Transgenic expression of proinsulin to inactivate insulin-specific
CD8+ T-cell responses in autoimmune diabetes
Peta Louise Susan Reeves
Bachelor of Science (Honours)
A thesis submitted for the degree of Doctor of Philosophy at
The University of Queensland in 2015
The University of Queensland Diamantina Institute
Abstract
Type 1 diabetes (T1D) results from autoimmune destruction of pancreatic β cells. The
current treatment for the fatal insulin-deficiency in T1D is injection of exogenous insulin. Despite
the dramatic increase in survival from the use of exogenous insulin, complications develop from
imperfect glycemic control and exogenous insulin does not rectify the underlying cause of disease,
the diabetogenic autoimmune response. Hence, there is a need for an effective therapy that impedes
the progress of the pathogenic autoimmune response. β cell-reactive CD8+ and CD4+ T cells are
important mediators of T1D development and progression, and CD8+ T cells directly destroy β
cells. For an effective immunotherapy, diabetogenic CD8+ T cells must therefore be inactivated.
Using the model antigen, ovalbumin (OVA), it has been shown that targeting OVA expression to
antigen-presenting cells (APC) inactivates cognate CD8+ T cells. It is not known whether this
inactivation would also occur for T cells recognising disease-relevant, β-cell antigens in
autoimmune-prone NOD (non-obese diabetic) mice where genetic immune tolerance defects exist.
In the NOD mouse model of T1D, insulin is the primary β-cell antigen that elicits disease. As such,
I have investigated T-cell tolerance to insulin in an adoptive transfer setting in which TCR
transgenic (G9) CD8+ T cells specific for insulin B15-23 (InsB15-23) were transferred to transgenic
mice over-expressing proinsulin in APC. Subsequently, I examined the effect of introducing
transgenic proinsulin expression on diabetes development in wild type NOD mice.
When transferred to mice transgenically expressing proinsulin, naïve G9 T cells proliferated
extensively prior to undergoing rapid deletion. In proinsulin transgenic recipients, the remaining
population of G9 T cells displayed reduced T-cell receptor (TCR) expression and bound less InsB1523-loaded,
H2-Kd-tetramer compared to G9 T cells that had been transferred to non-transgenic NOD
recipients. Interestingly, TCR down-regulation represented that which occurred for OVA-specific
CD8+ T cells transferred to mice expressing OVA in APC in a non-autoimmune prone strain.
Importantly, naïve G9 T cells did not expand or produce the important effector cytokine, interferon
(IFN)-γ, in response to InsB15-23 immunisation after transfer to proinsulin transgenic mice. These
data confirm naïve G9 T cells undergo rapid deletion and inactivation after transfer to mice
transgenically-expressing proinsulin, despite genetic tolerance defects in NOD mice. Memory CD8+
T cells (Tmem) perpetuate β-cell autoimmunity and pose a distinct barrier to immunotherapy. It was
pertinent to determine whether inactivation of insulin-specific CD8+ Tmem also occurs after
transfer to mice expressing proinsulin in APC. To test inactivation of insulin-specific CD8+ Tmem,
G9 Tmem were generated using an in vitro culture method. Cultured G9 Tmem were validated by
assessing phenotypic markers and cytokine production. Similarly to naïve G9 T cells, G9 Tmem
proliferated extensively and most G9 Tmem were rapidly deleted after transfer to mice overexpressing proinsulin in APC. The remaining, non-deleted population of G9 Tmem was infrequent
i
and exhibited reduced TCR expression. Furthermore, G9 Tmem did not expand or produce IFN-γ in
response to InsB15-23 immunisation, consistent with functional inactivation. These results
demonstrate transgenic proinsulin expression in APC overcomes genetic defects in NOD mice to
induce tolerance in insulin-specific CD8+ Tmem.
From the data described above, transgenic proinsulin expression is tolerogenic for insulinspecific CD8+ T cells. This has therapeutic potential to alter diabetes development by inactivating
potentially diabetogenic T cells. The next step was to investigate the effect on diabetes development
of introducing transgenic proinsulin expression to non-transgenic NOD mice using bone marrow
(BM)/haematopoietic stem and progenitor cell (HSPC) transfer facilitated by low-dose irradiation.
Expanding regulatory T cells (Treg) concurrently with antigen-encoding BM/HSPC transfer may be
beneficial in impeding diabetes progression. To determine the therapeutic window in which
inducing tolerance to insulin might protect NOD mice from diabetes development, proinsulin
transgenic BM/HSPC were transferred to NOD recipients at defined disease stages in conjunction
with administration of a Treg-boosting interleukin (IL)-2 complex. Proinsulin transgenic BM did
not significantly alter disease course when transferred to 10- and 16-week old NOD recipients.
Interestingly, an increase in overall diabetes incidence was observed after proinsulin transgenic BM
transfer to 6-week old NOD recipients, yet this increased diabetes incidence was not observed when
proinsulin-encoding HSPC were transferred. Finally, treatment with an IL-2 complex reduced
diabetes incidence in 3-week old recipients of NOD HSPC. The present study provides important
proof-of-concept regarding rapid inactivation of naïve and memory CD8+ T cells recognising a
diabetes-relevant, β-cell antigen that is, insulin, by transgenic expression of proinsulin in APC.
Furthermore, this study demonstrates transfer of proinsulin-encoding BM and treatment with a
Treg-boosting IL-2 complex alter diabetes development in an age-dependent manner.
ii
Declaration by author
This thesis is composed of my original work, and contains no material previously published or
written by another person except where due reference has been made in the text. I have clearly
stated the contribution by others to jointly-authored works that I have included in my thesis.
I have clearly stated the contribution of others to my thesis as a whole, including statistical
assistance, survey design, data analysis, significant technical procedures, professional editorial
advice, and any other original research work used or reported in my thesis. The content of my thesis
is the result of work I have carried out since the commencement of my research higher degree
candidature and does not include a substantial part of work that has been submitted to qualify for
the award of any other degree or diploma in any university or other tertiary institution. I have
clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.
I acknowledge that an electronic copy of my thesis must be lodged with the University Library and,
subject to the policy and procedures of The University of Queensland, the thesis be made available
for research and study in accordance with the Copyright Act 1968 unless a period of embargo has
been approved by the Dean of the Graduate School.
I acknowledge that copyright of all material contained in my thesis resides with the copyright
holder(s) of that material. Where appropriate I have obtained copyright permission from the
copyright holder to reproduce material in this thesis.
iii
Publications during candidature
Research papers
Peta L. S. Reeves, F. Susan Wong, Emma E. Hamilton-Williams, and Raymond J. Steptoe.
Transgenic expression of auto-antigen overcomes peripheral tolerance defects in NOD mice
(Manuscript in preparation).
Oral presentations

Visiting Scholar presentation, Institute of Molecular and Experimental Medicine, Cardiff
University School of Medicine (2015)
o ‘Reversing CD8+ T-cell responses in autoimmune diabetes.’

The Translational Research Institute Annual Poster Symposium (2014)
o ‘Transgenic expression of proinsulin overcomes tolerance defects in NOD mice.’

15th Annual Scientific Meeting, Brisbane Immunology Group (2014)
o ‘Transgenic expression of proinsulin overcomes tolerance defects in NOD mice.’
Poster presentations

Immunology of Diabetes Society 14th International Congress (2015)

Princess Alexandra Hospital Health Symposium (2014)

ASMR Post-graduate Student Conference (2014)

Immunology of Diabetes Society 13th International Congress (2013)

14th Annual Scientific Meeting, Brisbane Immunology Group (2013)

13th Annual Scientific Meeting, Brisbane Immunology Group (2012)
Publications included in this thesis
No publications included.
iv
Contributions by others to the thesis
Associate Professor Raymond Steptoe contributed to the design and analysis of experiments as well
as the written portion of this thesis. Dr Emma Hamilton-Williams contributed in the design of some
experiments and critiqued drafts of this thesis.
In Chapter 3, Ryan Galea performed the analysis of liver-infiltrating G9 T cells (Fig. 3.5). In
Chapter 5, Yoshihito Minoda performed intrahepatic injections of HSPC to neonatal NOD mice
(Fig. 5.17, 5.18).
Statement of parts of the thesis submitted to qualify for the award of another degree
None.
v
Acknowledgements
Without the following people, my PhD would have been by no means achievable, so thank you.
Firstly, I would like to extend my sincere gratitude to my principle supervisor, Associate Professor
Raymond Steptoe, for all the support and guidance throughout the past few years. I credit my
scientific development to your thorough supervision. Also, I wish to genuinely thank my cosupervisor, Dr Emma Hamilton-Williams, for providing an additional level of insight into this
project and for your constant support.
I have made life-long friends in my time at UQ Diamantina Institute and TRI, but in particular I
thank Jane AL-Kouba and Ryan Galea, both of whom were members of the Steptoe group for the
entirety of my PhD. Your friendship is truly appreciated.
To the members of my thesis committee, Dr James Wells, Dr Stephen Mattarollo and Dr Tony
Kenna, thank you for your time and constructive feedback over the course of my PhD.
For technical assistance, I thank staff from PA, PACE and TRI animal facilities for helping
procedures run smoothly and for sharing their immense murine knowledge. I also wish to
acknowledge staff within the flow cytometry facility, both in TRI and in R-Wing, for guiding me
within the FACS universe.
I thank Professor F. Susan Wong for sharing her vast knowledge of G9 mice with me over Skype
and in person.
Thank you to my family for your support, guidance, love and encouragement throughout this, at
times challenging, but always rewarding, endeavour. Finally, I sincerely thank my fiancé, Joseph,
for everything he has done for me during my candidature, it was made a much more pleasant
experience for all of your support and love.
vi
Keywords
diabetes, T cell, tolerance, memory, autoimmunity, insulin, NOD mouse, Treg
Australian and New Zealand Standard Research Classifications (ANZSRC)
ANZSRC code: 110703, Autoimmunity, 60%
ANZSRC code: 110704, Cellular Immunology, 20%
ANZSRC code: 110708, Transplantation Immunology, 20%
Fields of Research (FoR) Classification
FoR code: 1107, Immunology, 100%
vii
Table of contents
Abstract
i
Declaration by author
iii
Publications during candidature
iv
Contribution by others to the thesis
v
Acknowledgments
vi
Keywords
vii
Table of contents
viii
List of figures
xv
List of tables
xviii
List of abbreviations
xix
Chapter 1: Literature review
1
1.1 Introduction
2
1.2 T cells and the generation of immunological memory
2
1.3 T-cell tolerance
3
1.3.1 Central tolerance
3
1.3.2 Peripheral tolerance
3
1.3.2.1 Cell-intrinsic peripheral tolerance mechanisms
4
1.3.2.2 Cell-extrinsic peripheral tolerance mechanisms
6
1.4 Type 1 diabetes
6
1.5 T cells in autoimmune diabetes
8
1.6 Autoantigen involvement in autoimmune diabetes
9
1.6.1 Insulin
9
1.6.2 IGRP
10
viii
1.6.3 GAD65
11
1.6.4 Other autoantigens
12
1.7 Failure of immunological tolerance in autoimmune diabetes
12
1.7.1 Failure of immunological tolerance in the NOD mouse
12
1.7.2 Failure of immunological tolerance in human T1D
14
1.8 Therapies for autoimmune diabetes
15
1.8.1 Exogenous insulin/islet transplantation
15
1.8.2 Immune modulators
15
1.8.3 IL-2 to expand Treg
17
1.8.4 Insulin-specific immunotherapy
18
1.8.4.1 Insulin immunotherapy in NOD mice
18
1.8.4.2 Insulin immunotherapy in human T1D
19
1.9 Purpose of the study
22
1.10 Hypothesis and aims
22
1.10.1 Hypothesis
22
1.10.2 Aims
22
Chapter 2: Materials and methods
24
2.1 Mice
25
2.1.1 Inbred and Congenic strains
25
2.1.2 Transgenic strains
26
2.2 Reagents
26
2.2.1 Peptides
26
2.2.2 Buffers
27
2.2.3 Adjuvants
28
ix
2.2.4 Tissue culture media
29
2.2.5 Monoclonal antibodies
29
2.2.6 H2-Kd tetramers
32
2.3 Cell preparations
32
2.3.1 Lymph nodes and spleens
32
2.3.2 Bone marrow
33
2.3.3 Liver
33
2.3.4 Pancreas
33
2.3.5 CFSE-labelling
34
2.3.6 Purification of CD8+ T cells
34
2.3.7 Generation of G9 memory T cells
35
2.3.8 Purification hematopoietic stem and progenitor cells
35
2.4 Flow cytometry
36
2.4.1 Surface staining
36
2.4.2 Intracellular staining
36
2.4.3 Tetramer staining
36
2.5 Ex vivo assays
2.5.1 ELISPOT
37
37
2.6 Assessment of diabetes
37
2.7 Statistical analysis
37
Chapter 3: Transgenic expression of proinsulin induces peripheral tolerance
in insulin-specific CD8+ T cells
38
3.1 Introduction
39
3.2 Results
42
x
3.2.1 G9 T cells display a naïve T-cell phenotype despite expression of
cognate antigen in G9 mice
42
3.2.2 G9 T cells divide extensively after transfer to proinsulin transgenic
mice
44
3.2.3 G9 T cells are rapidly deleted after transfer to mice transgenically
expressing proinsulin
46
3.2.4 Transferred G9 T cells do not infiltrate the pancreas of NOD or
proinsulin transgenic mice
49
3.2.5 G9 T cells are not preferentially activated or retained in liver of
proinsulin transgenic recipients
51
3.2.6 Reduced recovery of G9 T cells from proinsulin transgenic
recipients is not due to allotype-mediated rejection
3.2.7 Transferred G9 T cells persist in the absence of cognate antigen
53
56
3.2.8 Down-regulation of TCR on G9 T cells after transfer to
proinsulin transgenic recipients
58
3.2.9 InsB15-23-loaded tetramer binding of G9 T cells is impaired after
transfer to proinsulin transgenic recipients
61
3.2.10 Transferred G9 T cells are predominantly CD44hiCD62L+ in
proinsulin transgenic recipients
3.2.11 Optimisation of InsB15-23 peptide immunisation
64
66
3.2.12 G9 T cells are unresponsive to InsB15-23 peptide immunisation
after transfer to proinsulin transgenic mice
68
3.2.13 Unresponsiveness of transferred G9 T cells in proinsulin transgenic
recipients is maintained
70
xi
3.2.14 Transferred G9 T cells do not contribute to diabetes development
72
3.3 Discussion
79
3.4 Future directions
86
Chapter 4: Transgenic expression of proinsulin rapidly inactivates
insulin-specific CD8+ memory T cells
87
4.1 Introduction
88
4.2 Results
91
4.2.1 G9 T cells are activated in vitro by InsB15-23
91
4.2.2 In vitro-cultured G9 Tmem display a phenotype resembling TCM and TEM 93
4.2.3 G9 Tmem produce IFN-γ
97
4.2.4 G9 Tmem proliferate extensively after transfer to proinsulin
transgenic recipients
4.2.5
99
G9 Tmem are rapidly deleted after transfer to proinsulin transgenic
recipients
101
4.2.6 G9 Tmem persist in the absence of their cognate antigen
103
4.2.7 G9 Tmem do not infiltrate the pancreas after transfer to recipients
105
4.2.8 Transferred G9 Tmem maintain a memory-like phenotype in
B16A and NOD recipients
4.2.9 G9 Tmem do not preferentially migrate to BM in recipients
107
109
4.2.10 TCR down-regulation on transferred G9 Tmem is observed in
proinsulin transgenic mice
112
4.2.11 Transferred G9 Tmem are unresponsive to InsB15-23 immunisation
in proinsulin transgenic recipients
114
4.2.12 G9 Tmem do not accumulate in LN of proinsulin transgenic recipients
xii
following InsB15-23 immunisation
116
4.2.13 G9 Tmem are unresponsive to InsB15-23 immunisation twenty one days
after transfer to proinsulin transgenic recipients
118
4.3 Discussion
121
4.4 Future directions
128
Chapter 5: Transfer of proinsulin-encoding BM and IL-2/αIL-2 alter diabetes
development in an age-dependent manner
129
5.1 Introduction
130
5.2 Results
133
5.2.1 BM from proinsulin transgenic mice is not rejected after transfer
to sublethally irradiated NOD mice
5.2.2 IL-2/αIL-2 complex expands Treg in spleen and LN of NOD mice
133
139
5.2.3 IL-2/αIL-2 complex enhances Treg expression of CD25, Foxp3 and
other markers indicative of suppressive function
141
5.2.4 IL-2/αIL-2 complex increases the frequency of non-Treg immune cell
populations
5.2.5
143
Transfer of proinsulin transgenic BM alters diabetes development in
an age-dependent manner
146
5.2.6 Donor-type leukocyte development is observed in peripheral blood
from 6-week old NOD recipients of proinsulin transgenic HSPC
149
5.2.7 Transfer of proinsulin transgenic HSPC does not increase diabetes
penetrance in 6-week old NOD recipients
154
5.2.8 Increased donor-type leukocyte development in 3-week old recipients
of proinsulin transgenic compared to NOD HSPC
156
xiii
5.2.9 IL-2/αIL-2 expands host-type CD25+CD4+ T cells in peripheral blood
from recipients of proinsulin transgenic HSPC
159
5.2.10 IL-2/αIL-2 treatment reduces diabetes incidence in 3-week old recipients
of NOD HSPC
161
5.2.11 Donor-type leukocytes develop after transfer of NOD and proinsulin
transgenic HSPC to 3-day old recipients
163
5.3 Discussion
167
5.4 Future directions
176
Chapter 6: Concluding remarks
177
6.1 Concluding remarks
178
Chapter 7: References
183
xiv
List of figures
Figure 3.1
G9 T cells display a typically naïve phenotype in the presence
of cognate antigen expression in G9 mice.
Figure 3.2
43
G9 T cells proliferate extensively after transfer to proinsulin
transgenic mice.
45
Figure 3.3
G9 T cells are deleted after transfer to proinsulin transgenic mice.
48
Figure 3.4
Low recovery of transferred G9 T cells in pancreas of NOD
and proinsulin transgenic recipients.
Figure 3.5
Low recovery of transferred G9 T cells in liver of NOD and proinsulin
transgenic recipients.
Figure 3.6
50
52
G9 T cells are not rejected due to an allotypic difference between NOD
and proinsulin transgenic recipients.
55
Figure 3.7
High recovery of transferred G9 T cells in B16A recipients.
57
Figure 3.8
G9 T cell TCR down-regulation after transfer to proinsulin transgenic
recipients.
Figure 3.9
Transferred G9 T cells bind less InsB15-23-loaded, H2-Kd-tetramer in
proinsulin transgenic compared to B16A and NOD recipients.
Figure 3.10
67
G9 T cells do not expand or produce IFN-γ following InsB15-23
immunisation in proinsulin transgenic recipients.
Figure 3.13
65
G9 T-cell expansion and IFN-γ production following InsB15-23 peptide
immunisation.
Figure 3.12
63
G9 T cells are predominantly CD44hiCD62L+ after transfer to proinsulin
transgenic recipients.
Figure 3.11
60
69
G9 T cells are unresponsive to InsB15-23 immunisation in proinsulin
transgenic recipients.
71
Figure 3.14
Transfer of G9 T cells does not contribute to diabetes development.
74
Figure 3.15
Transferred G9 T cells do not accumulate in pLN of recipients.
75
Figure 3.16
Transferred G9 T cells do not accumulate in spleen of recipients.
76
Figure 3.17
Transferred G9 T cells do not accumulate in sdLN of recipients.
77
Figure 3.18
Summary of main findings from Chapter 3
78
Figure 4.1
In vitro stimulation with InsB15-23 (23 Gly) activates G9 T cells in a
dose-dependent manner.
Figure 4.2
92
In vitro-culture of G9 T cells results in a mixed population of TEM
and TCM.
95
xv
Figure 4.3
G9 T cells up-regulate PD-1, CTLA-4 and LAG-3 following
in vitro-culture with InsB15-23 and IL-2.
96
Figure 4.4
G9 Tmem produce IFN-γ after transfer to B16A recipients.
98
Figure 4.5
G9 Tmem proliferate extensively in response to presentation of
transgene-derived, but not islet-derived, (pro)insulin.
Figure 4.6
100
G9 Tmem are rapidly deleted after transfer to proinsulin
transgenic recipients.
102
Figure 4.7
Transferred G9 Tmem persist in B16A recipients.
104
Figure 4.8
G9 Tmem do not traffic to the pancreas in immune-replete NOD and
proinsulin transgenic recipients.
Figure 4.9
Transferred G9 Tmem maintain a memory-like phenotype in B16A
and NOD recipients.
Figure 4.10
113
G9 Tmem recall response is ablated after transfer to proinsulin
transgenic mice.
Figure 4.14
111
TCR down-regulation by G9 Tmem after transfer to proinsulin
transgenic recipients.
Figure 4.13
110
Few tetramer-binding G9 Tmem detected in BM of NOD and
proinsulin transgenic recipients.
Figure 4.12
108
Increased recovery of CFSE-labelled G9 Tmem in BM of B16A
recipients compared to NOD and proinsulin transgenic recipients.
Figure 4.11
106
115
G9 Tmem do not accumulate in LN of proinsulin transgenic recipients
following InsB15-23 immunisation.
117
Figure 4.15
G9 Tmem are functionally inactivate in proinsulin transgenic recipients.
119
Figure 4.16
Summary of main findings from Chapter 4
120
Figure 5.1
No difference in donor-type leukocyte development in peripheral blood
from recipients of proinsulin transgenic and NOD BM.
Figure 5.2
136
Substantial donor-type leukocyte development in spleen from recipients
of proinsulin transgenic BM.
137
Figure 5.3
HSPC from NOD BM express variable levels of MHC class II.
138
Figure 5.4
Foxp3+ Treg expansion in spleen and LN of NOD mice following
IL-2/αIL-2 complex treatment.
Figure 5.5
Enhanced Treg expression of CD25, Foxp3, GITR, TNFRII and CD103
following IL-2/αIL-2 complex treatment.
Figure 5.6
140
142
Expansion of CD8+ TEM in spleen following IL-2/αIL-2
complex treatment.
144
xvi
Figure 5.7
Expansion of NKT cells, NK cells and DC in IL-2/αIL-2-treated
mice.
Figure 5.8
Proinsulin transgenic BM increases the penetrance of diabetes when
transferred to 6-week old recipients.
Figure 5.9
164
Similar donor-type leukocyte development in sublethally irradiated,
3-day old recipients of NOD and proinsulin transgenic HSPC.
Figure 5.19
162
Similar donor-type leukocyte development in sublethally irradiated,
3-day old recipients of NOD and proinsulin transgenic HSPC.
Figure 5.18
160
IL-2/αIL-2 complex reduces diabetes incidence in 3-week old recipients
of NOD HSPC.
Figure 5.17
158
Expansion of host-type CD25+CD4+ T cells in recipients of proinsulin
transgenic HSPC following IL-2/αIL-2 treatment.
Figure 5.16
157
Increased donor-type leukocyte development in 3-week old, PBS-treated
recipients of proinsulin transgenic compared to NOD HSPC.
Figure 5.15
155
Substantial donor-type leukocyte development in 3-week old recipients of
NOD and proinsulin transgenic HSPC.
Figure 5.14
153
Transfer of proinsulin-encoding HSPC does not increase the penetrance
of diabetes in 6-week old NOD recipients.
Figure 5.13
152
No difference in donor-type leukocyte development between sublethally
irradiated recipients of NOD and proinsulin transgenic HSPC.
Figure 5.12
151
No difference in donor-type leukocyte development between sublethally
irradiated recipients of NOD and proinsulin transgenic HSPC.
Figure 5.11
148
High purity of FACS-sorted HSPC from NOD and proinsulin
transgenic BM.
Figure 5.10
145
165
Summary of main findings from BM/HSPC transfer experiments
in Chapter 5
166
xvii
List of tables
Table 1.8.1
Summary of immunotherapy in mouse and man
21
Table 2.2.5.1 List of monoclonal antibodies for flow cytometry
29
Table 2.2.5.2 List of streptavidin conjugates for flow cytometry
31
Table 2.2.6.1 List of H2-Kd tetramers for flow cytometry
32
Table 3.1
Summary of Chapter 3
78
Table 4.1
Summary of Chapter 4
120
Table 5.1
Summary of BM/HSPC transfer experiments from Chapter 5
166
xviii
List of abbreviations
AIRE
Autoimmune regulator
APC
Antigen presenting cell(s)
APL
Altered peptide ligand
APS1
Autoimmune polyglandular syndrome type 1
BGL
Blood glucose level
BM
Bone marrow
BSA
Bovine serum albumin
CASAC
Combined Adjuvant for Synergistic Activation of Cellular Immunity
CCR
Chemokine receptor
CD
Cluster of differentiation
CFSE
Carboxyfluorescein succinimidyl ester
cGy
Centigray
ChgA
Chromogranin A
CTL
Cytotoxic T lymphocyte(s)
CTLA-4
Cytotoxic T lymphocyte antigen-4
DC
Dendritic cell(s)
EAE
Experimental autoimmune encephalomyelitis
EBV
Epstein-Barr virus
EDCI
1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide
Egr-2
Early growth response gene 2
ELISPOT
Enzyme-linked immunospot
ET
Enhanced transthyretin
FACS
Fluorescence-activated cell sorting
FCS
Foetal calf serum
Foxp3
Forkhead box P3
FSC
Forward scatter
GABA
γ-aminobutyric acid
GAD65
Glutamic acid decarboxylase 65
G-CSF
Granulocyte-colony stimulating factor
GFP
Green fluorescent protein
GITR
Glucocorticoid-induced TNF receptor
GVHD
Graft-versus-host disease
HA
Hemagglutinin
HLA
Human leukocyte antigen
xix
HPC
Haematopoietic progenitor cell(s)
HSC
Haematopoietic stem cell(s)
HSPC
Haematopoietic stem and progenitor cell(s)
IA2
Insulinoma antigen 2
Idd
Insulin-dependent diabetes
IDO
Indoleamine 2, 3-dioxygenase
IFA
Incomplete Freund’s adjuvant
IFN
Interferon
IGRP
Islet-specific glucose-6-phosphatase catalytic subunit-related protein
IL
Interleukin
IL-2R
Interleukin 2 receptor
InsB9-23
Insulin B9-23 peptide
InsB15-23
Insulin B15-23 peptide
i.p.
Intraperitoneal
IPEX
immune dysregulation, polyendocrinopathy, enteropathy, X-linked
iTreg
Induced regulatory T cell(s)
i.v.
Intravascular
KLRG1
Killer-cell lectin like receptor G1
LAG-3
Lymphocyte activation gene-3
LCMV
Lymphocytic choriomeningitis virus
Lin
Lineage
LLO
Listeriolysin O
LN
Lymph nodes
LT-HSC
Long-term repopulating haematopoietic stem cell(s)
LV
Lentiviral vector(s)
mAb
Monoclonal antibody
MFI
Mean fluorescence intensity
MHC
Major histocompatibility complex
MS
Multiple sclerosis
MT-PBS
Mouse-tonicity phosphate buffered saline
NFAT
Nuclear factor of activated T cells
NF-κB
Nuclear factor kappa-light-chain-enhancer of activated B cells
NK
Natural killer
NOD
Non-obese diabetic
NOD.SCID
Non-obese diabetes severe combined immunodeficiency
xx
nTreg
Natural regulatory T cell(s)
OVA
Ovalbumin
PBMC
Peripheral blood mononuclear cell(s)
PBS
Phosphate buffered saline
PD-1
Programmed death 1
PD-L1
Programmed death ligand 1
PI
Propidium iodide
pLN
Pancreatic lymph nodes
Poly I:C
Polyinosinic–polycytidylic acid potassium salt
proIns Tg
Proinsulin transgenic
QuilA
Quillaja saponaria A
RANKL
Receptor activator of NF-κB ligand
RBC
Red blood cell(s)
RIP
Rat insulin promoter
RPMI
Roswell Park Memorial Institute
R.T.
Room temperature
s.c.
Subcutaneous
sdLN
Skin-draining lymph nodes
SLE
Systemic lupus erythematosus
STZ
Streptozotocin
T1D
Type 1 diabetes
TCM
Central memory T cell(s)
TCR
T-cell receptor
TEM
Effector memory T cell(s)
Tg
Transgenic
TGF
Transforming growth factor
Th
T helper cell(s)
TIM-3
T cell immunoglobin and mucin domain-containing 3
Tmem
Memory T cell(s)
TNF
Tumour necrosis factor
TNFR
Tumour necrosis factor receptor
Tr1
Type1 regulatory T cell(s)
TRM
Tissue-resident effector memory T cells
Treg
Regulatory T cell(s)
TSDR
Treg-specific demethylation region
xxi
WT
Wild type
ZnT8
Zinc transporter 8
xxii
Chapter 1
Literature review
1
1.1 Introduction
Autoimmune disease occurs when an immune response is mounted against an individual’s
own cells or tissues, representing a breakdown of immune tolerance mechanisms that usually
moderate these undesirable immune responses. Autoimmune diseases are generally characterized as
systemic or organ-specific. Systemic autoimmune diseases affect multiple organs or target sites
within an individual such as the joints, skin and vasculature in systemic lupus erythematosus (SLE)
or the kidneys and lungs in systemic sclerosis. Organ-specific autoimmune diseases are the result of
a single organ or tissue being targeted, and classic examples are type 1 diabetes (T1D), in which
pancreatic β cells are targeted, or Graves disease, where follicular cells of the thyroid gland are
affected. The pathogenesis of organ-specific autoimmune disease usually involves T-cell attack and
destruction of target tissues. A thorough understating of how to interrupt pathogenic T-cell
responses could lead to effective therapies for T-cell mediated autoimmune diseases.
1.2 T cells and the generation of immunological memory
T cells are a subset of adaptive immune cells which develop in the thymus. In the periphery,
T cells can be characterised into two broad populations based on the presence of the surface
molecules CD4 and CD8. CD8+ T cells, as cytotoxic T lymphocytes (CTL), mediate cytotoxicity of
pathogen-infected and potentially tumourigenic cells, while CD4+ T cells provide ‘help’ to elicit
CD8+ T cell cytotoxicity as well as antibody production by B cells. The capacity to generate
immunological memory is an important characteristic of T cells and the adaptive immune system as
a whole. Naïve CD8+ T cells, by way of the T-cell receptor (TCR), recognise their cognate peptide
antigen presented in the context of multiple histocompatibility complex (MHC) class I molecules on
the surface of an antigen presenting cell (APC). Upon encounter with peptide/MHC class I
complexes in the presence of co-stimulation, CD8+ T cells expand and differentiate [reviewed in
(Williams & Bevan, 2007)]. Most CD8+ T cells differentiate into short-lived effector T cells,
capable of driving inflammation through the production of proinflammatory cytokines such as
interferon (IFN)-γ and tumour necrosis factor (TNF)-α. These short-lived CD8+ effector T cells are
also able to directly lyse target cells via cytotoxic mediators including granzyme B/perforin and
Fas/Fas ligand (FasL) (Li et al., 2014; Lowin et al., 1994). A small proportion (5-10%) of the initial
activated T-cell pool will differentiate into a long-lived memory population (Kaech et al., 2002),
capable of mounting a rapid response after secondary antigen challenge, and it is this rapid
secondary response of memory T cells (Tmem) which underlies vaccination. Human and murine
Tmem can be classified into two broad subsets based on phenotype and function: central memory
and effector memory T cells. Central memory T cells (TCM) express high levels of the lymph node
(LN)-homing receptor, CD62L, and chemokine receptor 7 (CCR7) and as such, generally reside in
2
LN. In contrast, effector memory T cells (TEM) express low levels of CD62L and do not express
CCR7, but possess cytotoxic activity and can migrate to non-lymphoid tissue (Sallusto et al., 1999;
Wherry et al., 2003).
1.3 T-cell tolerance
The immune system must not only prevent the individual from attack by pathogens, but also
protect self-tissue from misdirected immune responses. Tolerance mechanisms act as checkpoints to
balance the need to prevent deleterious autoimmune responses without impairing protective
immunity As a consequence, autoimmune disease results when immune tolerance mechanisms
break down.
1.3.1 Central tolerance
Central tolerance occurs in the thymus for T cells and during T-cell development, immature
CD4+CD8+ T cells are positively selected in the thymic cortex by finely-tuned levels of TCR
signalling. Thymocytes with very low levels of TCR signalling die by neglect and those with TCR
signalling above this threshold progress to negative selection (Starr et al., 2003). Negative selection
is deletion, via apoptosis, of thymocytes that strongly interact with self-antigen presented by MHC
(Klein et al., 2014). Many tissue-specific antigens are expressed in the thymus (Derbinski et al.,
2001) and the autoimmune regulator (Aire) gene has been implicated in controlling thymic
expression of self-antigens (Anderson et al., 2002). Humans with autoimmune polyglandular
syndrome type 1 (APS1) display multi-organ endocrinology and this syndrome has been attributed
to a mutation in AIRE (Peterson & Peltonen, 2005). Furthermore, Aire-deficient mice display
similar autoimmune phenotypes to APS1 patients (Anderson et al., 2002; Metzger & Anderson,
2011). Therefore, Aire-dependent thymic antigen expression is crucial for effective negative
selection of thymocytes and in the absence of Aire, an increase in self-reactive T cells is observed in
the periphery which may lead to development of autoimmune disease. Some thymocytes with an
intermediate to high self-reactivity are designated to develop into regulatory T cells (Treg)
(Stritesky et al., 2013) and thymocytes with low affinity for self are released into the periphery
(Stritesky et al., 2012).
1.3.2 Peripheral tolerance
While negative selection in the thymus greatly reduces the frequency of circulating,
potentially autoreactive mature T cells, this tolerance mechanism is incomplete as not all selfantigens are expressed in the thymus and low-affinity self-reactive T cells can escape central
3
tolerance, thus peripheral checkpoints are required to prevent autoimmune disease. Peripheral
tolerance mechanisms can be classified as cell-intrinsic or cell-extrinsic.
1.3.2.1 Cell-intrinsic peripheral tolerance mechanisms
Cell-intrinsic deletion or induction of functional unresponsiveness in a T cell requires an
antigen-specific interaction of the T cell with an APC. To generate immunity, T-cell activation
requires TCR recognition of peptide/MHC complex on APC and a second co-stimulatory signal
such as ligation of CD28 on the T cell and CD80/CD86 present on the APC. The result of an
‘unproductive’ encounter, in the absence of a co-stimulatory signal, between a T cell and an APC
presenting the T cell’s cognate antigen is deletion of that T cell, or the T cell existing in an
unresponsive state (Bonifaz et al., 2002; Hawiger et al., 2001; Steptoe et al., 2007). In studies
employing administration of peptide to induce tolerance, it has been shown that lower peptide doses
favour deletion, rather than unresponsiveness, of cognate T cells (Redmond et al., 2005) and
peptides with low TCR affinity also favour deletional T-cell tolerance (Smith et al., 2014). During
deletional tolerance, there is an initial period of T-cell proliferation but this ultimately leads to death
of the self-reactive T cells via apoptosis (Davey et al., 2002; Kenna et al., 2008; Kurts et al., 1997).
Proliferation of self-reactive CD8+ T cells preceding their deletion has been demonstrated upon
interaction with Aire-expressing cells that reside outside of the thymus, in secondary lymphoid
organs (Gardner et al., 2008). Using TCR transgenic mice in which CD8+ T cells recognise
hemagglutinin (HA), HA-specific CD8+ (clone 4) T cells proliferated before being deleted in
pancreatic LN (pLN) of transgenic mice expressing HA in pancreatic islets (Hernandez et al.,
2001). In mice expressing the model antigen, ovalbumin (OVA), in the pancreas only high levels of
OVA expression led to deletion of transferred OVA-specific CD8+ (OT-I) T cells, as opposed to
OT-I T-cell ignorance when low levels of OVA were expressed (Kurts et al., 1999).
The Bcl-2 family member, Bim, has been implicated in the apoptotic pathway of deletional
tolerance, in that Bim can activate the proapoptotic proteins, Bax and Bak, thereby initiating
mitochondrial cell death (Czabotar et al., 2009; Gavathiotis et al., 2008). For example, Bimdeficient CD8+ T cells recognising OVA were not deleted in response to cross-presented OVA,
unlike their wild type counterparts (Davey et al., 2002). Furthermore, inhibition of deletion in
response to cross-presented antigen has been demonstrated for Bim-deficient CD8+ T cells
recognising HA (Redmond et al., 2008). Another pathway of T-cell death is dependent on the death
receptor CD95 (Fas) and the role of CD95 in antigen-specific T-cell deletion is less clear, that is,
CD95-deficient OT-I T cells were not deleted when OVA was expressed in islets (Kurts et al.,
1998), while CD95-deficient clone 4 T cells were deleted when HA was expressed in islets
(Redmond et al., 2008). Transcriptional analyses have shown CD8+ T cells undergoing deletional
4
tolerance exhibited a distinct gene profile from T cells undergoing immunogenic activation, with
down-regulation of granzyme A, as well as rapid up-regulation of Bim and up-regulation of
programmed death 1 (PD-1) (Parish et al., 2009). Deletional tolerance of CD8+ T cells by targeting
antigen to steady-state dendritic cells (DC), was demonstrated to be independent of the PD-1/PDligand 1 (PD-L1) pathway (Mukherjee et al., 2013), and may represent different T-cell tolerance
outcomes depending on the antigen encounter. Targeting antigen to non-APC, such as, erythrocytes,
has been shown to induce T-cell tolerance by cross-presentation of antigen by DC and subsequent
deletion of cognate CD8+ and CD4+ T cells (Kontos et al., 2013). Therefore, T-cell deletion in the
periphery has been demonstrated in multiple model systems and is important in maintaining
tolerance to self, and in the absence of such deletional pathways, autoimmune disease can result.
Anergy occurs when T cells persist in an unresponsive state following antigen encounter in
the absence of a co-stimulatory signal and this state was initially characterised in vitro (Harding et
al., 1992; Jenkins & Schwartz, 1987; Lamb et al., 1983). The anergic state is characterized by
reduced TCR signalling and interleukin (IL)-2 expression (Schwartz, 2003). Altered signalling
pathways underlie T-cell anergy in that T cells lacking the E3 ubiquitin ligases, Itch or Cbl-b, are
resistant to anergy (Heissmeyer et al., 2004), probably due to the requirement for such ubiquitin
ligases to degrade TCR signalling proteins. Furthermore, microarray analyses revealed early growth
response gene 2 (Egr-2) and Egr-3 as negative regulators of TCR signalling, possibly through
association with Cbl-b (Safford et al., 2005). T-cell unresponsiveness can occur not only from a T
cell/APC interaction in the absence of co-stimulation, but also in the presence of co-inhibitory
molecules, for example PD-1 and cytotoxic T lymphocyte antigen-4 (CTLA-4). CTLA-4 has been
shown to compete with CD28 for binding of its ligand CD80 (van der Merwe et al., 1997), thereby
inducing unresponsiveness, rather than activation, of the interacting T cell. In a study using
administration of peptide to explore T-cell tolerance, CD8+ T-cell unresponsiveness was reversed
by PD-1 blockade (Tsushima et al., 2007). Despite studies implicating the PD-1/PD-L1 inhibitory
pathway in T-cell deletion and unresponsiveness, this pathway has been more extensively
investigated in, and thought to be a hallmark of, T-cell exhaustion in chronic viral infection (Day et
al., 2006; Trautmann et al., 2006) and tumour settings (Fourcade et al., 2012; Woo et al., 2012). The
transcription factor nuclear factor of activated T cells (NFAT) has recently been shown to bind to
genes associated with exhaustion which, for example, encode PD-1 and T cell immunoglobin and
mucin domain-containing 3 (TIM-3), thereby controlling CD8+ T-cell exhaustion and possibly other
forms of T-cell unresponsiveness dependent on co-inhibitory molecules (Martinez et al., 2015) The
importance of peripheral tolerance mechanisms is evident as autoreactive T cells often exist in
healthy individuals (Danke et al., 2004). It is this ability of APC to induce T cell-specific tolerance
5
that makes them an attractive tool for immunotherapy in autoimmune diseases where peripheral
tolerance mechanisms are perturbed.
1.3.2.2 Cell-extrinsic peripheral tolerance mechanisms
Cell-extrinsic peripheral tolerance mechanisms can be mediated by cells with regulatory
properties capable of suppressing effector T cells. There are a range of T-cell types with regulatory
abilities including conventional Treg, T helper 3 (Th3) and type 1 regulatory T cells (Tr1).
Conventional Treg, as characterised by CD4, CD25 and forkhead box P3 (Foxp3) transcription
factor expression are important in maintaining peripheral tolerance and thereby preventing
development of autoimmune disease (Hori et al., 2003; Sakaguchi et al., 1995). Treg can act
through contact-dependent or contact-independent mechanisms (Shalev et al., 2011). Contactdependent mechanisms include ligation of CTLA-4 on the surface of Treg with CD80/CD86 which
acts to inhibit co-stimulation and cytokine secretion of DC to impair effector T-cell activation
(Oderup et al., 2006). Treg, via CTLA—4, can induce production of indoleamine 2, 3-dioxygenase
(IDO) by DC which suppresses T-cell function by depletion of tryptophan (Fallarino et al., 2003).
Also, Treg can lyse effector T cells by release of perforin and granzymes (Cao et al., 2007;
Grossman et al., 2004; Oderup et al., 2006). Contact-independent mechanisms of Treg function
include secretion of regulatory cytokines interleukin (IL)-10 (Asseman et al., 1999), IL-35 (Collison
et al., 2007) and transforming growth factor (TGF)-β (Joetham et al., 2007) as well as limiting
expansion of effector T cells by IL-2 deprivation (McNally et al., 2011).
Two subsets of Foxp3+ Treg occur in mice and humans; thymic-derived natural Treg (nTreg) and
peripherally induced Treg (iTreg). nTreg develop in the thymus as a result of agonist self-peptide
selection (Jordan et al., 2001). During reduced immunogenic conditions, conventional T cells can
convert to iTreg in the presence of TGF-β and retinoic acid, with this process occurring
predominantly in the gut (Chen et al., 2003; Peng et al., 2004). Emphasising the importance of Treg
in autoimmune disease is ‘immune dysregulation, polyendocrinopathy, enteropathy, X-linked’
(IPEX) syndrome in which patients have mutations in the FOXP3 gene (Katoh et al., 2013),
resulting in defective Treg and development of multiple autoimmune diseases within the first few
years of life (Chatila et al., 2000; Powell et al., 1982).
1.4 Type 1 diabetes
T1D is an autoimmune disease mediated by T cells, resulting in destruction of insulinproducing β cells in the pancreatic islets of Langerhans (Chatenoud et al., 1994). Insulin is a
hormone secreted exclusively by pancreatic β cells and allows for cellular up-take of glucose from
the blood. As such, T1D is characterised by hyperglycemia as well as impaired lipid and protein
6
metabolism. The endocrine symptoms of T1D are due to the lack of insulin in T1D patients and
currently, injection of exogenous insulin is required for survival. However, injection of insulin is
not a cure and imperfect control of blood glucose levels renders patients susceptible to chronic
complications including retinopathy, neuropathy and nephropathy, as well as cerebrovascular and
cardiovascular disease (Bjornstad et al., 2015). These complications are the prime cause of
amputations and reduced lifespan in T1D patients.
T1D afflicts more than 90 000 individuals in Australian (2011) and is usually diagnosed in
childhood although this disease can also arise later in life (Naik et al., 2009). Genetic predisposition
and environmental factors are involved in T1D development, with the incidence increasing
dramatically over the last twenty years, especially in younger children (Harjutsalo et al., 2008),
highlighting the influence of environmental factors in disease progression. Some environmental
factors proposed to influence T1D development include lack of Vitamin D from decreased sun
exposure (Takiishi et al., 2010), reduced exposure to microbes leading to dysregulation of immune
responses (D'Angeli et al., 2010), viral infections such as Coxsackievirus-B (Hober et al., 2012) and
infant nutrition (Knip et al., 2010). In terms of genetic susceptibility, certain loci within the human
MHC, known as the human leukocyte antigen (HLA), region confer the greatest genetic risk for
T1D including DRB1 0401 and DQA1 0301 alleles (Erlich et al., 2008). Non-HLA candidate genes
associated with T1D include INS, PTPN22, CTLA4 [reviewed in (Bluestone et al., 2010; Concannon
et al., 2009)] and IL2RA (Lowe et al., 2007). The fact that the majority of T1D-associated
candidate-genes are involved in immune function emphasises the autoimmune nature of T1D. As
such, autoantibodies against various β-cell antigens, including insulin, glutamic acid decarboxylase
65 (GAD65), IA-2 (insulinoma antigen 2) and islet-specific glucose-6-phosphatase catalytic
subunit-related protein (IGRP), are present in the majority of T1D patients and detection of
autoantibodies is currently the most reliable predictor of high risk of developing T1D (Pihoker et
al., 2005).
Since T1D is a disease resulting from immune dysregulation, restoration of immune
regulation would provide an opportunity to therapeutically intervene in the course of disease.
Therefore, antigen-specific immunotherapy is an attractive approach for T1D. Due to the early age
of onset and generally non-fatal nature of T1D, the window for therapeutic manipulation is small
and for any immunotherapy there is emphasis on inactivating autoreactive T-cell responses in an
antigen-specific manner, while maintaining robust immunity to foreign antigens. At T1D diagnosis,
patients have developed a pool of well established effector T cell and Tmem responses as a result of
ongoing β-cell autoimmunity and for an immunotherapy to provide a curative effect, these Tmem
responses need to be inactivated. The host laboratory has demonstrated that transfer of antigenencoding haematopoietic stem and progenitor cells (HSPC) is an effective tool to instate potentially
7
tolerogenic antigen expression and thus inactivate cognate memory CD8+ T cells (Coleman et al.,
2012).
1.5 T cells in autoimmune diabetes
Although autoimmune diabetes is a T-cell mediated autoimmune disease, involvement of
other immune cell types, such as B cells and innate immune cells, are also important for diabetes
progression (Diana et al., 2013; Serreze et al., 1996). The primary role of both CD8+ and CD4+ T
cells in autoimmune diabetes of the non-obese diabetic (NOD) mouse is evident in the ability of
purified CD8+ and CD4+ T cells from NOD mice to transfer diabetes (Christianson et al., 1993;
Wong et al., 1996) and by inhibition of diabetes by therapies modulating T cells (Chatenoud et al.,
1994). Early histological studies identified both CD4+ and CD8+ T cells within the pancreatic islets
of diabetic patients (Hanninen et al., 1992; Itoh et al., 1993). Furthermore, IGRP-, GAD65- and
insulin-specific T cells have been detected in the peripheral blood of patients with T1D (Luce et al.,
2011; Yang et al., 2006; Yang et al., 2012a).
CD4+ T cells were initially suspected of being the cause of T1D as CD4+ T cells were highly
abundant in infiltrated islets of NOD mice, CD4+ T cell clones could transfer disease (Haskins et
al., 1988) and in humans, the most significant genetic risk for T1D mapped to MHC class II (Shiina
et al., 2004). Although current evidence suggests CD8+ T cells play a key role in β-cell destruction.
For instance, pancreatic β cells express only MHC class I (Thomas et al., 1998), and not MHC class
II (McInerney et al., 1991), suggesting only CD8+ T cells can directly recognise antigens presented
by β cells. However, a recent study suggests MHC class II expression is up-regulated by β cells in
response to IFN-γ in NOD mice (Zhao et al., 2015). Furthermore, NOD mice deficient in MHC
class I and CD8+ T cells are protected from insulitis as well as autoimmune diabetes (Katz et al.,
1993a; Serreze et al., 1994). While when only pancreatic β cells were deficient in MHC class I,
insulitis proceeded unhindered but diabetes development was impaired (Hamilton-Williams et al.,
2003), which may indicate a direct role for CD8+ T-cell interaction in the progression from insulitis
to overt diabetes . Using a humanised mouse model, it has been demonstrated that CD8+ T cells
cloned from peripheral blood of a new-onset T1D patient were able to induce destructive insulitis
(Unger et al., 2012). Also in human T1D, a strong association between the presence of autoreactive
CD8+ T cells and islet transplantation failure suggests CD8+ T cells are important for β-cell
destruction (Pinkse et al., 2005). Following lysis of β cells, further autoreactive T cells can be
primed as a result of antigen release from dead β cells (Turley et al., 2003), perpetuating
autoimmunity. Diabetogenic CD4+ T cells are primed in pLN (Hoglund et al., 1999), while
autoreactive CD8+ T cells may develop effector function in islets (Chee et al., 2014), suggesting
8
CD8+ T cells initiate islet autoimmunity through direct interaction with β cells and CD4+ T cells are
necessary to magnify and sustain this response through to the progression of insulitis.
NOD TCR transgenic mice in which T cells recognise β-cell antigens have been established
and have proven to be invaluable tools for dissecting the role of self-reactive T cells in autoimmune
diabetes. The TCRs utilised to generate these TCR transgenic mice are typically derived from T-cell
clones isolated from NOD mice at various disease stages. An example of an H2-Ag7-restricted TCR
transgenic mouse is the BDC2.5 mouse (Katz et al., 1993b), which expresses a transgenic TCR
derived from a CD4+ T-cell clone isolated from NOD mice (Haskins et al., 1988), recognising
chromogranin A (ChgA) (Stadinski et al., 2010). An example of an H2-Kd-restricted TCR
transgenic mouse is the G9 mouse (Wong et al., 2009), generated using the TCR chains from a
CD8+ T-cell clone isolated from young NOD mice (Wong et al., 1996) and which recognises insulin
B15-23 peptide (InsB15-23) (Wong et al., 1999). Another is the 8.3 mouse (Verdaguer et al., 1997), in
which the transgenic TCR was derived from an IGRP206-214-specific (Lieberman et al., 2003) CD8+
T cell clone isolated from NOD mice older than 10 weeks of age (Nagata & Yoon, 1992). Studies
utilising these T-cell clones and TCR transgenic mice have been instrumental in developing the
understanding of T-cell pathogenesis in autoimmune diabetes.
1.6 Autoantigen involvement in autoimmune diabetes
Since the identification of insulin as a target of the autoimmune response in T1D patients
(Palmer et al., 1983), over thirty β-cell antigens have been identified (DiLorenzo, 2011), including
insulin, IGRP and GAD65. The process by which immune responses are generated to many
antigens as a result of an initial response toward a primary antigen is termed epitope spreading.
Epitope spreading is evident in autoimmune disease models including the NOD mouse, where this
process is crucial to disease progression (Tian et al., 2006). Also, there is evidence to suggest
epitope spreading occurs in human T1D also (Ott et al., 2004). Therefore, it is important to
understand the contribution of β-cell antigens in the pathogenesis of autoimmune diabetes.
1.6.1 Insulin
Insulin is not only the hormone lacking in patients with T1D, it is also a key autoantigen
recognised by self-reactive T cells critical to autoimmune diabetes pathogenesis. Insulin was the
first target for autoantibodies in autoimmune diabetes to be characterised at the molecular level
(Palmer et al., 1983). However, these autoantibodies are not pathogenic (Bach, 1994), indicating a
prime role for autoreactive T cells in autoimmune diabetes. Several lines of evidence suggest
insulin is the primary autoantigen of autoimmune diabetes in the NOD mouse and T-cell responses
to insulin are required for diabetes development. The G9 T-cell clone, recognising InsB15-23 (Wong
9
et al., 1999), isolated from the islets of young NOD mice rapidly transfer disease (Wong et al.,
1996). Insulin-reactive CD4+ T-cells clones from NOD islets have also been identified, specifically
for InsB9-23 (Daniel et al., 1995). Moreover, insulin-specific T cells have been identified in human
T1D, both in the circulation (Luce et al., 2011; Sachdeva et al., 2015) and from pLN (Kent et al.,
2005). While the presence of insulin-reactive T cells in NOD mice and humans has indicated a role
for insulin-specific T-cell responses in diabetes, manipulation of insulin expression in mice has
been crucial in dissecting the role of insulin as an autoantigen in autoimmune diabetes. Mice
possess two insulin genes, Ins1 and Ins2, with greater expression of insulin II protein in the thymus
(Heath et al., 1998). NOD mice made deficient in the Ins2 gene show accelerated development of
autoimmune diabetes (Thebault-Baumont et al., 2003), this is in contrast to the protection from
diabetes observed for Ins1 knockout NOD mice (Moriyama et al., 2003). Proinsulin is the
prohormone precursor of insulin and transgenic expression of proinsulin in APC of NOD mice
protects from diabetes (French et al., 1997). Using this model, it has been shown that proinsulinspecific T-cell responses are required for expansion of T cells specific for other β-cell antigens
(Krishnamurthy et al., 2006; Krishnamurthy et al., 2008), providing strong evidence that
(pro)insulin is the primary autoantigen in NOD mice. Furthermore, altering the InsB9-23 sequence,
which contains overlapping CD8+ T-cell and CD4+ T-cell determinants (Daniel et al., 1995; Wong
et al., 1999), by a tyrosine to alanine substitution at position 16, protects NOD mice from insulitis
and diabetes development (Nakayama et al., 2005; Nakayama et al., 2007).
Genetically, the role of insulin in T1D is emphasised by the contribution of human insulin
(INS) gene polymorphisms to disease susceptibility (Concannon et al., 2009). Shorter repeat
polymorphisms in the promoter region of the INS gene confer risk, while longer repeats are
associated with protection from diabetes (Kockum et al., 1996). The longer repeat polymorphisms,
known as class III alleles, are thought to provide protection due to higher levels of thymic insulin
transcription compared to transcription from short class I polymorphisms (Vafiadis et al., 1997).
The protective effect of higher thymic insulin levels has been attributed to greater IL-10 production
from proinsulin-specific Tmem (Durinovic-Bello et al., 2005). While there is strong evidence for
the role of insulin as a prime autoantigen in autoimmune diabetes, there remain roles for other β-cell
antigens in the pathogenesis of T1D.
1.6.2 IGRP
IGRP-specific TCR transgenic mice show different diabetes kinetics dependent on whether
the IGRP-specific T cells are H2-Kd- or H2-Ag7-restricted. While monoclonal IGRP-specific CD4+
T cells are sufficient to cause diabetes, the reduced frequency and prolonged disease initiation
observed for mice with a monoclonal IGRP-specific CD8+ T-cell repertoire (Verdaguer et al.,
10
1997), suggests that IGRP-specific CTL require help from autoreactive CD4+ T cells to effectively
propagate autoimmune diabetes. Building on this, it was shown that proinsulin-specific T-cell
responses are necessary for IGRP-specific CD8+ T-cell expansion and transgenic expression of
IGRP in APC resulted in loss of IGRP-specific T cells in islets of transgenic mice but IGRP
expression had no effect on insulitis or diabetes development (Krishnamurthy et al., 2006).
Similarly, deletion of the gene encoding IGRP had no effect on diabetes progression in NOD mice,
despite the absence of IGRP-reactive T cells in islets of IGRP knockout mice (Oeser et al., 2011).
Altering the T-cell repertoire of NOD mice by specifically depleting IGRP-reactive CD8+ T cells
using toxin-coupled tetramer delayed the onset of, but did not prevent, autoimmune diabetes
(Vincent et al., 2010), which may indicate responses to IGRP are not crucial to initiate islet
autoimmunity, but rather act synergistically with other specificities to mediate diabetes
development. Furthermore, unlike insulin, IGRP expression is not detectable in thymus of NOD
mice (Gardener, 2008), suggesting IGRP-specific T cells in the periphery may be of a higher
affinity than insulin-specific T cells, which may underlie differences in T-cell responses in the
periphery and how these T-cell specificities mediate diabetes pathogenesis.
1.6.3 GAD65
GAD65 is an enzyme expressed in pancreatic β cells, as well as in neurons, and its role in
diabetes initiation and progression is conflicted in the literature. GAD65-specific T-cell responses
are detected in pre-diabetic NOD mice (Kaufman et al., 1993) and human T1D (Mannering et al.,
2004; Oling et al., 2005; Worsaae et al., 1995). Inducing tolerance to GAD65 in NOD mice has had
varying effects on diabetes development, in that, passive tolerance induction in young NOD mice
by i.v. injection of GAD65 prevented insulitis and autoimmune diabetes (Kaufman et al., 1993),
while oral administration of porcine GAD65 had no effect on disease progression in NOD mice
(Ramiya et al., 1997). In contrast, feeding transgenic plants expressing the other GAD isoform,
GAD67, reduced diabetes incidence in NOD mice (Ma et al., 1997). As another example, nasal
administration of various GAD65 peptides protected over 50% of treated NOD mice from diabetes
and T cells from treated mice displayed a Th2 phenotype, as opposed to the high Th1-bias present
in autoimmune diabetes (Tian et al., 1996). The efficacious therapies using GAD, and its peptides,
in NOD mice have not translated to human T1D. Injection with GAD65 and alum in recent-onset
T1D patients had no effect on C-peptide levels at four years follow-up (Ludvigsson et al., 2011),
but increased GAD65-specifc CD4+CD25hi T cells were observed in peripheral blood from GAD65treated relative to placebo-treated patients (Hjorth et al., 2011). Manipulation of GAD65 expression
in NOD mice suggests GAD65 is not required for autoimmune diabetes development. For example,
GAD65 knockout mice develop diabetes as per their wild type counterparts (Yamamoto et al.,
11
2004) and transgenic expression of GAD65 in APC does not protect mice from diabetes (Jaeckel et
al., 2003). Further analysis of young GAD65 transgenic mice displayed reduced β-cell specific Tcell responses (Tian et al., 2009), implying the involvement of GAD65 in diabetes development is
to promote activation and expansion of T cells specific for other β-cell antigens.
1.6.4 Other autoantigens
ChgA is a protein found in secretory granules of β cells and neuroendocrine tissues. It was
identified as a T1D autoantigen for its ability to activate the CD4+ BDC2.5 T-cell clone (Stadinski
et al., 2010). ChgA-specific CD4+ and CD8+ T-cell responses have been identified in T1D patients
(Gottlieb et al., 2014; Li et al., 2015), indicating ChgA may be an important autoantigen in human
T1D. Another more recently identified autoantigen is zinc transporter 8 (ZnT8) (Wenzlau et al.,
2007). Zinc is crucial for proper storage, secretion and function of insulin and ZnT8 is specifically
expressed in pancreatic β cells. The proportion of T1D patients positive for ZnT8 autoantibodies
ranges between 24%-80%, depending on the study (Kawasaki et al., 2011; Wenzlau et al., 2007).
The identification of novel T1D autoantigens is beneficial for screening at-risk individuals to
predict diabetes onset and for diagnostic applications.
Post-translational modification of protein, such as glycoslyation and methylation, can result
in the generation of unique self-antigens and post-translationally modified antigens have been
extensively investigated in autoimmune diseases including rheumatoid arthritis and multiple
sclerosis (MS) (Doyle & Mamula, 2001). T cells from T1D patients recognise post-translationally
modified insulin (Mannering et al., 2005) and GAD65 (McGinty et al., 2014) determinants, while a
post-translationally modified ChgA peptide more effectively activates BDC2.5 T cells than the
native ChgA peptide (Delong et al., 2012). Examination of post-translationally modified antigens in
T1D provides insight into the pathogenesis of diabetes and may explain how low-affinity antigens,
such as insulin, generate strongly diabetogenic responses.
1.7 Failure of immunological tolerance in autoimmune diabetes
1.7.1 Failure of immunological tolerance in the NOD mouse
The NOD mouse is the most well-established mouse model of autoimmune diabetes
(Makino et al., 1980). Diabetes occurs spontaneously in 60-80% of female and 20-30% of male
NOD mice, depending on housing conditions (Bach, 1994; Kikutani & Makino, 1992). The
presence of immune cell infiltrates that surround the islet, termed peri-insulitis, is observed by
approximately 3-4 weeks of age. Invasion of the islets by immune cells is termed insulitis and
occurs several weeks after the initiation of peri-insulitis in both male and female mice (Anderson &
Bluestone, 2005). Onset of overt diabetes begins to occur from 12-14 weeks of age in female NOD
12
mice. The presence of insulitis in both male and female NOD mice, but the disparity between
diabetes incidence, suggests late stage regulatory events are capable of controlling progression to
overt diabetes. NOD mice exhibit defects in multiple leukocyte subsets. Treg are important in
controlling autoimmunity and in the NOD mouse dysregulation of Treg has been described. For
example, the suppressive function of Treg decreases with age in NOD mice (Gregori et al., 2003)
and, compounding this decrease in Treg suppression, is the increased resistance of effector T cells
to Treg-mediated suppression in older NOD mice (You et al., 2005). Evidence suggests these
defects may not be Treg-intrinsic and that APC may be defective in maintaining Treg function. As
such, APC from autoimmune-resistant C57BL/6 mice were more effective in stimulating
CD4+CD25+ Treg than APC from NOD mice (Alard et al., 2006). Furthermore, CD11b+CD11cAPC from NOD mice expressing C57BL/6 resistant alleles prime more suppressive Treg than APC
from wild type NOD mice (Anderson et al., 2008). The maturation and function of macrophages has
been shown to be defective in NOD mice (Lee et al., 2012b; Padgett et al., 2015; Serreze et al.,
1993), as is the function of natural killer (NK) and NKT cells (Poulton et al., 2001). Such defects in
Treg, APC and other cell types may act in concert to bring about the autoimmune disease phenotype
observed in NOD mice.
NOD mice possess the unique H2-Ag7 MHC haplotype which is not only the most high risk
genetic factor, but is also necessary for diabetes development in NOD mice (Wicker et al., 1995).
There is a specific substitution on the I-A molecule, an aspartic acid to glutamine at position 57 on
the β chain (Acha-Orbea & McDevitt, 1987), that alters peptide binding (Kanagawa et al., 1998),
and this substitution is also present in the human susceptibility HLA DQ β chain (Todd et al.,
1987). Manipulations of MHC in NOD mice can provide protection from diabetes development
(Nishimoto et al., 1987; Uehira et al., 1989). Other loci, apart from MHC, which contribute to
diabetes susceptibility in the NOD mouse are termed Idd (insulin-dependent diabetes) loci (Todd &
Wicker, 2001). Ctla4 is contained in the Idd5 locus and polymorphisms in this gene can confer risk
of autoimmune diabetes in the NOD mouse and T1D in humans (Ueda et al., 2003). Il2 is the
strongest candidate-gene in the Idd3 region (Yamanouchi et al., 2007) and congenic NOD mice
which express diabetes-resistant Idd3 alleles have reduced diabetes incidence, due to an increase in
suppressive Treg in pLN and islets (Goudy et al., 2011).
NOD mice have been shown to possess defects in peripheral tolerance pathways, including
reduced deletion and increased accumulation of islet-specific CD8+ T cells (Hamilton-Williams et
al., 2012). These tolerance defects could be rescued with the introduction of diabetes-resistant Idd
alleles from non-autoimmune strains, such as C57BL/6. Expansion of HA-specific CD8+ T cells
was suppressed by CD25+ Treg in Idd3/5 mice expressing HA in islets (Martinez et al., 2005). This
is important when the candidate genes of Idd3 and Idd5, Il2 and Ctla4, respectively, are considered
13
with regard to their known function in Treg survival and suppressive activity (Barron et al., 2010;
Sojka et al., 2009). The presence of multiple susceptibility loci in the NOD mouse hints at the
complexity of genetic pathways involved in diabetes development and that numerous tolerance
pathways are perturbed in the NOD mouse. Since several of the genetic lesions in NOD mice are
similar to those described in human T1D, it is possible that similar peripheral tolerance defects
contribute to islet autoimmunity in human T1D.
1.7.2 Failure of immunological tolerance in human T1D
It has long been suggested that abnormalities within immune cell subsets exist in patients
with T1D. This includes differences, compared to healthy controls, in the proportion of T cells (alKassab & Raziuddin, 1990; Ilonen et al., 1991; Petersen et al., 1996) and also in their function
(Hehmke et al., 1995; Spatz et al., 2003). In particular, there has been much conjecture over the
frequency and function of Treg in T1D (Zhang et al., 2012). CD4+CD25hi, but not CD4+CD25-, T
cells from recent-onset T1D patients have shown increased propensity toward apoptosis than those
from healthy controls (Jailwala et al., 2009). This effect was replicated in CD4+CD25hi T cells from
healthy controls after depleting IL-2, suggesting reduced IL-2 signalling in T1D patients might be
responsible for an increase in apoptosis-prone Treg. Reduced IL-2 receptor (IL-2R) signalling in
T1D patients has been associated with impaired maintenance of Foxp3 expression by CD4+CD25+
T cells (Long et al., 2010). Also, reduced suppression by Treg from T1D patients compared to
healthy controls has been demonstrated (Brusko et al., 2005; Lindley et al., 2005). Augmenting
defects in the Treg population, is the reduced sensitivity of effector T cells and Tmem to Tregmediated inhibition in T1D patients (Afzali et al., 2011; Lawson et al., 2008; Schneider et al.,
2008).
While the majority of studies in humans have investigated immune cell subsets in the
peripheral blood of T1D patients, it is not known how these data represent what is occurring within
the target organ, that is, the pancreas, and the draining LN. In peripheral blood, there is no
difference in Treg number between T1D patients and non-diabetic controls (Brusko et al., 2005;
Lindley et al., 2005). Although, a significant decrease in Foxp3+ Treg frequency and function has
been observed in pLN from T1D patients relative to healthy controls (Ferraro et al., 2011). As
CD25 and Foxp3 expression can be influenced by environmental factors, Ferraro et al. employed a
Treg-specific demethylation region (TSDR) assay to identify epigenetically imprinted Treg. No
difference in Treg frequency was observed between T1D patients and healthy controls using this
method (Ferraro et al., 2011). Although, the suppressive function of CD25bright Treg isolated from
pLN of T1D patients was impaired, but those from peripheral blood were as suppressive as Treg
from healthy controls. This study highlights the potential for site-specific defects in immune cell
14
subsets and this may explain discrepancies in the literature regarding the function of regulatory cells
in peripheral blood of T1D patients.
With regard to the APC compartment, plasmacytoid DC are less frequent in T1D patients
than healthy controls (Chen et al., 2008a). Data from Treg suppression assays with cells from T1D
patients and healthy controls indicate APC dysfunction might be responsible for reduced Treg
suppression in T1D patients (Jin et al., 2009). Additionally, gene expression profiles of CD14+
monocytes from a subset of T1D patients were associated with cellular stress pathways and diabetes
progression (Irvine et al., 2012). It is the combination of multiple immunological defects that bring
about β-cell autoimmunity and subsequent progression to autoimmune diabetes in humans and
NOD mice. While there are differences between diabetes in the NOD mouse and human T1D, it is
important to investigate inflammatory sites, that is, the pancreas and pLN, from NOD mice which
can guide analyses of infrequently obtained human pancreatic samples.
1.8 Therapies for autoimmune diabetes
1.8.1 Exogenous insulin/islet transplantation
Exogenous insulin therapy is the standard treatment for T1D and while the use of insulin has
drastically reduced mortality, this treatment is not curative and complications from imperfect blood
glucose control still occur (Lachin et al., 2014; Steffes et al., 2003). In some cases where glucose
monitoring and subsequent insulin replacement therapy are inadequate, an alternative approach to
the management of T1D is islet, or to a lesser extent, whole pancreas transplantation [reviewed in
(Fotino et al., 2015; Watson, 2015). Inclusion criteria include recurrent, severe hypoglycemia,
marked hyperglycemia and ketoacidosis (Robertson et al., 2000; Robertson et al., 2006). Islet
transplantation is beneficial as β-cell replacement might reinstate some level of physiological
glucose homeostasis thereby helping to control long-term complications. However, there are risks
involved with islet or pancreas transplantation including chronic immunosuppression, limited
organ/islet availability and general risks of surgery. It should be noted that the β cell-specific
effector CD8+ T cells responsible for initial islet destruction and the subsequent pool of β cellspecific memory CD8+ T cells perpetuate islet autoimmunity and can contribute to the destruction
of transplanted, allogeneic islets (Okitsu et al., 2001). This highlights the importance of an
immunotherapy for T1D which terminates these pathogenic CD8+ Tmem responses.
1.8.2 Immune modulators
Since T1D was identified as an autoimmune disease (Bottazzo et al., 1974; MacCuish et al.,
1974), focus has shifted to immunotherapeutic approaches to address the underlying immune
dysregulation that results in β-cell destruction. Early studies focused on non-specific immune
15
modulators such as cyclosporine and anti-CD3 monoclonal antibody (mAb) therapy as a means to
suppress immune responses in autoimmune diabetes. While continuous cyclosporine treatment in
recent-onset T1D patients was effective in reducing insulin dependence, and in some patients
abolishing the need for exogenous insulin entirely, there were toxic side effects and eventually the
beneficial effects diminished leading to the decline in use of this therapy for T1D (Bougneres et al.,
1988; Stiller et al., 1984). Long-term remission of autoimmune diabetes was observed in recentonset NOD mice following anti-CD3 mAb treatment (Chatenoud et al., 1994). The presence of
insulitis and acceptance of syngeneic islet grafts in anti-CD3 mAb-cured mice (Mottram et al.,
2002) suggested restoration of immune regulation, and it was later demonstrated that administration
of anti-CD3 mAb induced T cells with regulatory properties (Belghith et al., 2003; Bresson et al.,
2006). These and other studies using anti-CD3 mAb in mice (Hayward & Shriber, 1992; Herold et
al., 1992) did not necessarily translate to human T1D. Single-dose anti-CD3 mAb treatment has
been shown to maintain insulin production for up to a year in T1D (Herold et al., 2002), however,
treatment may need to be repeated for long-term clinical outcomes. Dosing was a critical issue,
high-dose anti-CD3 mAb treatment has been associated with Epstein-Barr virus (EBV) reactivation
(Keymeulen et al., 2010a) and low-dose anti-CD3 mAb therapy had no effect on T1D progression
(Sherry et al., 2011). Despite β-cell destruction in T1D being T-cell mediated, anti-CD20 mAb
therapy, to deplete B cells, can reverse autoimmune diabetes in recent-onset NOD mice (Hu et al.,
2007). In human T1D, anti-CD20 mAb therapy has been shown to partially protect β-cell function
(Pescovitz et al., 2009), possibly by transient disruption of pathogenic T-cell recruitment (Herold et
al., 2011). It appears repeated or long-term dosing would be required for increased efficacy of antiCD20 mAb therapy, and neither has been tested for safety. It is also currently unknown whether
anti-CD20 mAb therapy would prevent or delay T1D onset in at-risk individuals.
Transplantation of bone marrow (BM) or HSPC has been shown to prevent autoimmune
diabetes in mouse models, possibly by ‘re-setting’ the immune system. Insulitis and diabetes have
been prevented in NOD mice following transfer of BM from diabetes-resistant mice (Ikehara et al.,
1985). Further study into the use of BM transfer for autoimmune diabetes highlighted the necessity
for sublethal conditioning for this approach to proceed clinically. Sublethal conditioning was shown
to be effective in achieving stable, allogeneic chimerism that prevented diabetes in pre-diabetic
NOD mice (Li et al., 1996). Transplantation of autologous HSC in combination with high-dose
immune suppression in recent-onset T1D patients has been shown to induce insulin independence,
however, this was transient (de Oliveira et al., 2012; Voltarelli et al., 2007). Targeting tolerogenic
antigen expression to HSPC may represent a means to overcome the transient efficacy observed for
studies employing HSPC transplantation, in that pathogenic T cells might be continually inactivated
by antigen expression in a long-lived population of APC (Coleman et al., 2012).
16
1.8.3 IL-2 to expand Treg
The notion of IL-2 and Treg induction as a therapeutic approach in T1D is logical
considering the genetic lesions of the NOD mouse and in human T1D. CTLA-4 is constitutively
expressed by Treg and has been shown to be important in Treg suppressive function (Sojka et al.,
2009). The Ctla4 gene is located in the Idd5 locus of the NOD mouse (Araki et al., 2009b) and
polymorphisms in CTLA4 can confer risk of T1D in humans (Ueda et al., 2003). IL-2 is a
pleiotropic cytokine important for effector T-cell function and essential for Treg survival [reviewed
in (Long et al., 2013)]. The Idd3 locus in NOD mice includes the Il2 gene and the IL-2Rα gene has
been associated with human T1D (Lowe et al., 2007). Furthermore, the candidate-gene PTPN2 has
been implicated in reduced IL-2R signaling in human T1D in that the risk allele conferred reduced
Foxp3 expression on activated CD4+ T cells in response to IL-2 (Long et al., 2011). These genetic
studies suggest therapies to expand Treg number and/or function may be beneficial in regulating the
inflammatory responses that drive autoimmune diabetes pathogenesis.
While daily administration of IL-2 did not prevent diabetes development in NOD mice, IL-2
in combination with rapamycin protected NOD mice from diabetes development (Rabinovitch et al.,
2002). Low-dose IL-2 administered for ten days was shown to reverse hyperglycemia in a
significant proportion of new-onset diabetic NOD mice when compared to PBS-treated controls
(Grinberg-Bleyer et al., 2010). Islet-infiltrating Treg in IL-2-treated mice displayed higher
expression of markers indicative of suppressive activity, including Foxp3, CD25, CTLA-4 and
GITR, compared to PBS-treated control mice. In this system, these data suggest IL-2 treatment
increases the number and function of Treg localised within the inflamed site, that is, the pancreas,
which are capable of reversing established diabetes. Therefore, at least in NOD mice, IL-2 may be
more beneficial after hyperglycemia onset.
With regard to human T1D, in a phase I/II clinical trial, low-dose IL-2 was shown to be
well-tolerated in T1D patients and increased the proportion of CD4+CD25+Foxp3+ Treg in
peripheral blood, however, no change in exogenous insulin requirement or C-peptide was observed
(Hartemann et al., 2013). Employing the same patient cohort as Hartemann et al., it was shown that
the expanded CD4+ Treg also displayed a suppressive phenotype indicated by increased CD25 and
CTLA-4 expression (Rosenzwajg et al., 2015). Furthermore, a reduction of β cell-specific effector
T-cell responses, compared to pre-treatment baseline values, was observed following IL-2
treatment. In addition to expansion of conventional CD4+ Treg, low-dose IL-2 treatment of patients
with T1D increased the proportion of CD8+CD25+Foxp3+ T cells (Churlaud et al., 2015). In a
combinatory approach, IL-2 in conjunction with rapamycin increased Treg in T1D patients,
however, the metabolic indicator, C-peptide, decreased transiently in all patients (Long et al., 2012).
The IL-2 dose used by Long et al. was higher than that used by Hartemann et al. which might
17
explain the undesired outcome. In NOD mice, IL-2 in combination with rapamycin reduced β-cell
division and glucose metabolism (Baeyens et al., 2013), despite previously being demonstrated to
protect NOD mice from diabetes (Rabinovitch et al., 2002), and in other mouse models rapamycin
reduced β-cell survival and function (Tanemura et al., 2012; Yang et al., 2012b), also providing a
possible explanation for the outcome in humans. Translating therapies from mice to humans is
therefore challenging, particularly with regard to IL-2 therapy in the context of T1D. The
complexity of translation is emphasised by the pleiotropic nature of IL-2 and the need to ensure the
dose remains effective in expanding Treg to mediate suppression of diabetogenic immune responses
but low enough to avoid expansion of potentially pathogenic IL-2-dependent immune cells such as
effector T cells and NK cells.
1.8.4 Insulin-specific immunotherapy
Since T1D arises as a result of β-cell destruction by immune cells, particularly T cells,
recognising β-cell antigens, antigen-specific immunotherapy is an attractive approach for the
treatment of T1D. Antigen-specific immunotherapy has the potential to restore tolerance to β-cell
antigens without the concern of general immune suppression. In the NOD mouse, there is strong
evidence to suggest insulin is the primary antigen involved in autoimmune diabetes, and while in
humans the evidence is not as clear, it does implicate a crucial role for insulin responses in the
pathogenesis of T1D. Therefore, there has been considerable focus on insulin-specific
immunotherapy studies in both mouse models and humans.
1.8.4.1 Insulin immunotherapy in NOD mice
Early immunotherapy studies in NOD mice found treatment with oral insulin delayed the
onset of autoimmune diabetes (Zhang et al., 1991). Transfer of T cells from oral insulin-treated
mice also protected untreated mice from diabetes, suggesting oral insulin promoted immune
regulation capable of controlling diabetes progression. Similarly, the induction of oral tolerance to
insulin, or its peptides, was characterised by an increase in IL-10, TGF-β and IL-4 production by
lymphocytes following in vitro re-stimulation with InsB (Polanski et al., 1997). Coupling oral
insulin to a non-toxic subunit of cholera toxin increased the efficacy of oral insulin immunotherapy,
which was thought to be mediated by the induction of insulin-specific Treg (Meng et al., 2011;
Petersen et al., 2003), but this protection was not absolute as some treated mice developed diabetes.
Nasal administration of insulin is another delivery route able to induce Treg capable of modulating
diabetes development in NOD mice (Daniel & Wegmann, 1996; Every et al., 2006).
The combination of
antigen delivery through nasal
insulin
therapy and the
immunomodulatory, Treg- boosting capacity of anti-CD3 mAb therapy has been shown to reverse
18
recent-onset autoimmune diabetes and reduce insulitis in both NOD mice and a mouse model of
virus-induced diabetes (Bresson et al., 2006). The increased efficacy of combination approaches is
evident from the reduction of autoimmune diabetes in mice treated with a TNF receptor family
member (OX40) agonist mAb, OX86, and InsB9-23 peptide (Bresson et al., 2011). The reduction in
diabetes development was due to insulin-specific Treg induction following combination treatment
compared to insulin peptide immunotherapy alone. Similarly, reversal of new-onset diabetes in the
NOD mouse has been achieved through oral dosing of a commensal bacterium, Lactococcus lactis,
genetically modified to secrete human proinsulin and IL-10 in combination with administration of
low-dose anti-CD3 mAb (Takiishi et al., 2012).
As opposed to the induction of mucosal tolerance, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCI)-fixed splenic APC pulsed with antigen can induce T-cell tolerance through Tcell unresponsiveness (Jenkins & Schwartz, 1987). Complete diabetes protection was observed in
young NOD mice injected with fixed APC coupled with whole insulin, while treatment with fixed
APC coupled with InsB9-23 protected approximately 80% of young NOD mice from diabetes
development (Prasad et al., 2012). The outcomes of insulin immunotherapy studies in mouse
models highlight the possible need to induce deletional tolerance or inactivation of β cell-reactive T
cells as regulation of effector T cells alone may be insufficient for complete protection from
diabetes.
Since mice in which proinsulin II is transgenically expressed by MHC class II+ APC are
protected from autoimmune diabetes (French et al., 1997), studies have subsequently instated this
tolerogenic antigen expression to non-transgenic NOD mice. Transfer of proinsulin-encoding HSPC
has been shown to reconstitute the hematopoietic compartment of young, myeloablated NOD mice
(Steptoe et al., 2003). Engraftment of proinsulin-encoding HSPC was protective as recipients of
proinsulin-encoding HSPC showed significantly reduced incidence of autoimmune diabetes
compared to recipients of NOD HSPC. In a potentially more translational manner, young NOD
mice were protected from insulitis by transfer of retrovirus-transduced HSC expressing proinsulin
(Chan et al., 2006). This effect was proinsulin-specific, as the incidence of sialitis did not differ
between recipients of proinsulin-expressing HSC and unmanipulated control mice. All of these
studies combined suggest (pro)insulin-specific immune responses are an attractive target for
immunomodulation that may have potential therapeutic implications in T1D treatment.
1.8.4.2 Insulin immunotherapy in human T1D
Translation of insulin-specific immunotherapy from animal models to human T1D patients
has been challenging, with routes of administration, dosing and timing of treatment being explored
in clinical trials. The purpose of the immunotherapy diabetes (IMDIAB) trial was to determine
19
whether oral insulin therapy at time of T1D diagnosis would induce immunological tolerance to
insulin, and its peptides, and thereby preserve remaining β-cell function (Pozzilli et al., 2000).
Pozzilli et al. proposed that if oral insulin therapy was effective in recent-onset T1D patients, it
might then be applied to high-risk individuals to potentially prevent diabetes development. The
results, however, showed no difference between patients receiving oral insulin or placebo in Cpeptide levels or exogenous insulin requirement at one year follow up. It was concluded that the
dose of 5mg of oral insulin daily did not alter diabetes progression. Similarly, another trial to
determine whether oral insulin therapy at T1D diagnosis could protect remaining β-cell function
found no effect of treatment on the decline of β-cell function (Chaillous et al., 2000). A phase I
clinical trial in which human InsB was administered in incomplete Freund’s adjuvant (IFA) to
recently-diagnosed T1D patients showed no effect on residual β-cell function (Orban et al., 2010).
As this was a pilot study, with the main objective being to test safety, and the number of
participants was small (n=12), this study was not powered to detect significant changes in residual
β-cell function. Data from the study by Orban et al., provide some evidence for a Treg response to
InsB immunisation, in that, peripheral blood cells from InsB-treated patients produced higher levels
of TGF-β after in vitro re-stimulation with InsB. Also, InsB-reactive CD4+ T cells from peripheral
blood of InsB-treated patients displayed Foxp3 expression. One study, in patients within five years
of T1D diagnosis, showed preservation of C-peptide in groups treated with defined doses of a DNA
plasmid encoding proinsulin (Roep et al., 2013). There was a reduction in the frequency of
proinsulin-specific CD8+ T cells in the treatment group compared to placebo. This effect was
antigen-specific, since the frequency of irrelevant islet-antigen or virus-reactive CD8+ T cells was
not altered. The immunological and metabolic effects observed in this study were transient,
potentially highlighting the importance of a long-lived source of tolerogenic antigen presentation.
Insulin immunotherapy has also been trialled in prevention studies, which aim to prevent or
delay diabetes development in at-risk individuals, and an example is the Diabetes Prevention Trial
(DPT-1) (Diabetes Prevention Trial--Type 1 Diabetes Study, 2002). DPT-1 involved an intervention
strategy consisting of twice daily, low-dose subcutaneous insulin injections or no treatment as a
control. No difference in the overall incidence of T1D was observed between insulin treated- and
untreated-individuals. Therefore, insulin therapy, as used in the DPT-1 study, did not delay or
prevent diabetes. A pilot study in patients with islet autoantibodies found nasal insulin therapy to be
safe, that is, acceleration of β-cell destruction was not observed, and provided evidence of reduced
T-cell responses to insulin, which is consistent with mucosal tolerance as a result of nasal antigen
delivery (Harrison et al., 2004). Another prevention trial investigated the effect of nasal insulin
therapy on children with high susceptibility HLA-DQB1 alleles and positivity for autoantibodies
(Nanto-Salonen et al., 2008). Despite the initiation of nasal insulin therapy soon after autoantibody
20
detection, there was no effect on diabetes incidence when compared to placebo. The second DPT-1
study evaluated the effect of oral insulin therapy in relatives at lower risk of developing T1D than
relatives in the first DPT-1 study, but still with a 10- to 20-fold higher risk than the general
population (Skyler et al., 2005). Overall, T1D incidence was similar in the oral insulin treated-group
and placebo group, suggesting there was no effect of oral insulin in delaying or preventing T1D in
at-risk individuals. There was, however, an indication of the benefit of oral insulin therapy in a
subgroup of patients with higher levels of insulin autoantibodies and further studies are required to
explore this effect.
Further stratification of patients by baseline immunologic, metabolic and genetic factors for
therapy trials might help to identify patients which may benefit most from the defined treatment.
The lack of efficacy observed for therapies which rely solely on induction of immune regulation,
suggest there is the need for an approach to induce robust tolerance of β cell-specific effector T
cells and Tmem through deletion or inactivation, the benefits of which might be enhanced with
concurrent expansion of Treg. If such a therapy was demonstrated to be efficacious using insulin as
the tolerogenic antigen, it might be possible to widen the window for therapeutic intervention by
targeting multiple β-cell antigens, which may prove capable of reversing established islet
autoimmunity.
Table 1.8.1 Summary of immunotherapy in mouse and man
NOD mice
Human T1D
αCD3
Diabetes remission, acceptance Maintenance
of
insulin
of syngeneic islet grafts
production (≥ 1 year), EBV
reactivation with high dose, no
effect with low-dose
αCD20
Reversal in new-onset NOD Transient protection of β-cell
mice
function
Diabetes
prevention
with Transient insulin independence
BM/HSPC transplantation
diabetes-resistant
and following autologous HSC
allogeneic BM
transplantation
Protection
with
IL-2/ No change with low-dose IL-2,
IL-2
rapamycin, reversal with low- decrease in C-peptide with ILdose IL-2
2/rapamycin
Delayed onset, transfer of T No change in C-peptide or
Oral insulin
cells from treated mice exogenous insulin requirement
protected untreated mice from
diabetes
Inducing T cell tolerance to Protection when proinsulin is C-peptide preservation after
over-expressed
in
APC, proinsulin-encoding
plasmid
proinsulin
protection when proinsulin- DNA treatment
expressing HSC transferred
21
1.9 Purpose of the study
For an effective T1D immunotherapy, pathogenic, islet-reactive T-cell responses must be
inactivated. Targeting antigen to APC using an MHC class II promoter rapidly inactivates cognate
CD8+ T cells (Kenna et al., 2010). There is strong evidence implicating insulin as the primary
autoantigen for autoimmune diabetes in the NOD mouse (French et al., 1997; Nakayama et al.,
2005) and studies suggest insulin is also an important autoantigen in human T1D (Kent et al., 2005;
Luce et al., 2011). Importantly, T1D in humans and autoimmune diabetes in NOD mice develop as
a consequence of a genetic background associated with failure of peripheral tolerance pathways.
Therefore, it is pertinent to understand if insulin-specific T-cell responses can be tolerised in a
diabetes-prone environment, where known genetic tolerance defects exist. This study aims to
determine whether tolerance induction occurs for naïve and memory CD8+ T cells recognising the
important β-cell antigen, insulin, in an autoimmune-prone background. Transfer of antigenencoding HSPC has been used as a tool to instate tolerogenic antigen expression in non-transgenic
recipient mice, specifically for the purpose of diabetes prevention (Steptoe et al., 2003). To date,
this procedure has been performed only under conditions where the host immune system is ablated
at the time of BM/HSPC transfer, which leads to a full ‘re-set’ of the immune system. Given the
potential capacity of APC-targeted proinsulin expression to inactivate naïve and memory insulinspecific CD8+ T cells, I propose immune-ablation may not be required for effective prevention of
diabetes by proinsulin-encoding BM/HSPC transfer. Due to the young age at diagnosis and nonfatal nature of T1D, safety afforded from reduced intensity, immune-preserving conditioning would
be favourable in a clinical setting. Therefore, this study seeks to extend current knowledge by
determining the effect transfer of proinsulin-encoding BM/HSPC has on diabetes development
when this procedure is performed under conditions where host immunity is not ablated.
1.10 Hypothesis and aims
1.10.1 Hypothesis
‘Transgenic expression of proinsulin in APC induces peripheral tolerance in naïve and memory
insulin-specific CD8+ T cells and transfer of proinsulin-encoding BM/HSPC will impede diabetes
development in non-transgenic NOD mice.’
1.10.2 Aims
Aim I
Determine whether transgenic proinsulin expression induces peripheral tolerance in
naïve insulin-specific CD8+ T cells.
Aim II
Determine whether transgenic proinsulin expression induces peripheral tolerance in
insulin-specific CD8+ memory T cells.
22
Aim III
Determine whether, and at what stage of disease, transfer of proinsulin-encoding
BM/HSPC impedes diabetes development in sublethally irradiated NOD mice.
23
Chapter 2
Materials and methods
24
2.1 Mice
Mice were maintained in a specific pathogen-free environment at the Princess Alexandra Hospital
Biological Resource Facility (PAH-BRF; Brisbane, Australia), the Australian Institute for
Bioengineering and Nanotechnology Animal Facility (AIBN; Brisbane, Australia), the Pharmacy
Australia Centre of Excellence Biological Resources (PACE-BR; Brisbane, Australia) and the
Translational Research Institute Biological Resources Facility (TRI-BRF; Brisbane, Australia). All
mice were euthanised by CO2 asphyxiation. Mice were sex-matched within individual experiments.
Animals were used between 6-15 weeks of age unless otherwise indicated and all experiments were
approved by the University of Queensland Animal Ethics Committee under ARC project 164/12.
2.1.1 Inbred and Congenic strains
NOD/Lt mice:
NOD/Lt mice (referred to as NOD) (Makino et al., 1980) are inbred mice in which approximately
80% of females spontaneously develop autoimmune diabetes. NOD mice carry the H-2d haplotype
and express the CD45.1 allelic variant of the CD45 pan-leukocyte molecule. All NOD mice were
purchased from the Animal Resource Centre (Perth, Australia) at 6-8 weeks of age and used at this
age unless otherwise indicated.
NOD.CD45.2 mice:
NOD.CD45.2 mice are congenic NOD mice that express the CD45.2 allelic variant of the CD45
pan-leukocyte molecule (Steptoe et al., 2004). NOD.CD45.2 mice were bred and maintained at
AIBN (Brisbane, Australia) and the TRI-BRF (Brisbane, Australia). For adoptive transfer
experiments, male NOD.CD45.2 mice at 8-10 weeks of age were used as recipients unless
otherwise stated.
NOD.CD45.2 x NOD mice:
NOD/Lt mice were mated with NOD.CD45.2 mice to generate F1 mice that express both CD45.1
and CD45.2. NOD.CD45.2 x NOD mice were bred and maintained at the PAH-BRF (Brisbane,
Australia), the PACE-BR (Brisbane, Australia) and the TRI-BRF (Brisbane, Australia). For
adoptive transfer experiments, male NOD.CD45.2xNOD mice at 8-10 weeks of age were used as
recipients unless otherwise stated.
25
2.1.2 Transgenic strains
Proinsulin transgenic mice:
Proinsulin transgenic mice are transgenic mice expressing mouse proinsulin II under the control of
an MHC class II (I-Eαk) promoter (French et al., 1997). Proinsulin transgenic mice are protected
from spontaneous autoimmune diabetes. Proinsulin transgenic mice were bred and maintained at the
PAH-BRF (Brisbane, Australia) or the TRI-BRF (Brisbane, Australia). Proinsulin transgenic mice
were a kind gift from Prof Leonard C. Harrison (The Walter and Eliza Hall Institute of Medical
Research, Parkville, Victoria, Australia).
NOD.CD45.2 x proinsulin transgenic mice:
Proinsulin transgenic mice were mated with NOD.CD45.2 to generate F1 mice that express both
CD45.1 and CD45.2. NOD.CD45.2 x proinsulin transgenic mice were bred and maintained the
PAH-BRF (Brisbane, Australia), the PACE-BR (Brisbane, Australia) or the TRI-BRF (Brisbane,
Australia). For adoptive transfer experiments, male NOD.CD45.2 x proinsulin transgenic mice at
8-12 weeks of age were used as recipients unless otherwise stated.
G9 mice:
G9 are transgenic mice expressing an H2-Kd-restricted TCR specific for InsB15-23. The G9 mice
used were offspring of F1 matings between TCRα transgenic Cα-/- (line 22) and TCRβ transgenic
Cα-/- (line 2) founder lines (Wong et al., 2009). G9 mice were bred and maintained at the PAH-BRF
(Brisbane, Australia) or the TRI-BRF (Brisbane, Australia). The TCRα and TCRβ founder lines
used to generate the G9 mice were kind gifts from Prof F. Susan Wong (Institute of Molecular and
Experimental Medicine, School of Medicine, Cardiff University, Cardiff, UK).
B16A mice:
B16A mice are double knockout mice lacking both insulin 1 and insulin 2 and expressing a mutated
preproinsulin (B16:A insulin) (Nakayama et al., 2005). B16A mice are protected from spontaneous
autoimmune diabetes. B16A mice were bred and maintained at the PAH-BRF (Brisbane, Australia)
or the TRI-BRF (Brisbane, Australia).
2.2 Reagents
2.2.1 Peptides
InsB15-23:
Insulin B15-23 (LYLVCGERG) is a H2-Kd-restricted determinant of insulin and together with H-2Kd
forms an MHC class I/peptide complex recognised by CD8+ T cells. Ins B15-23 (23 Gly) was
26
purchased from Auspep (Melbourne, Australia). Ins B15-23 was dissolved in pyrogen free H2O to a
concentration of 10mg/ml, aliquoted, and stored at -30oC.
InsB15-23 (23 Val):
InsB15-23 (23 Val) (LYLVCGERV) is an altered peptide ligand (APL) of native Ins B15-23 with
valine replacing glycine at p9. Ins B15-23 (23 Val) binds more strongly to H2-Kd than native Ins B1523
(Wong et al., 2002). Ins B15-23 (23 Val) was purchased from Auspep (Melbourne, Australia).
InsB15-23 (23 Val) was dissolved in pyrogen free H2O to a concentration of 10mg/ml, aliquoted, and
stored at -30oC.
2.2.2 Buffers
MT-PBS:
Mouse tonicity phosphate buffered saline (MT-PBS) was composed of Na2HPO4.2H20 (28.5g/L),
NaH2PO4.1H20 (5.52g/L) and NaCl (87g/L) dissolved in MilliQ water as a 10x stock and diluted for
use. MT-PBS was prepared in-house and sterilised by autoclaving prior to use.
MT-PBS / 2.5% FCS:
MT-PBS/2.5% FCS (F2.5) was used for organ collection and cell preparations. MT-PBS was
supplemented with 2.5% fetal calf serum (FCS) for cells to be adoptively transferred, or 2.5%
bovine serum for cells used in vitro.
ACK buffer:
ACK buffer was used to eliminate red blood cells (RBC) from spleens, bone marrow cell
suspensions and peripheral blood. ACK buffer was made in house and was composed of NH4Cl
(8.29g/L), KHCO3 (1g/L) add Na2EDTA (0.0372g/L) dissolved in MilliQ water. The pH of ACK
buffer was adjusted to 7.2-7.4, filter sterilised and stored at 4°C.
FACS wash:
FACS wash was made in house to 10x concentrations and was composed of 1% bovine serum
albumin (BSA) and 20mM EDTA in MT-PBS. 1x FACS wash was made by diluting 100mL of 10x
FACS wash in 1L of MT-PBS. 10x FACS wash was stored at -20°C and 1x FACS wash was stored
at 4°C.
27
FACS block:
FACS block was made in house from the supernatant of 2.4G2 hybridoma cells producing antiCD16/32 to allow blocking of mouse Fc-receptors, Fc-gamma II (CD32) and Fc-gamma III (CD16).
1% BSA was added to supernatant and filter-sterilised aliquots were stored at -20°C.
2.2.3 Adjuvants
Poly I:C:
Polyinosinic–polycytidylic acid potassium salt (Poly I:C) (Sigma-Aldrich, St. Louis, MA, USA)
was dissolved in sterile MT-PBS to a concentration of 1mg/ml, aliquoted and stored at -30°C.
100µg per mouse administered intraperitoneally (i.p.) was used for routine immunisations.
QuilA:
Quillaja saponaria A (QuilA) (Soperfos Biosector DK-Vedback, Denmark) was diluted in sterile
MT-PBS, aliquoted and stored at -30°C. 20µg per mouse administered subcutaneously was used for
routine immunisations.
CpG:
CpG 1663 oligonucleotide (Invitrogen, Carlsbad, CA, USA) was diluted in H2O, aliquoted and
stored at -30°C. 15.5nmol per mouse administered subcutaneously was used for routine
immunisations.
CASAC:
Combined Adjuvant for Synergistic Activation of Cellular Immunity (CASAC) (Wells et al., 2008)
is a combination adjuvant that induces a high level of CD8+ T cell activation in response to peptide
immunisation. CASAC comprises anti-mouse CD40 (clone: FGK4.5, isotype: Rat IgG2a. 25µg per
mouse, BioXCell, New Hampshire, USA), CpG 1826 oligonucleotides (50µg per mouse, Integrated
DNA Technologies, Iowa, USA), recombinant mouse IFN-γ (100ng per mouse, R&D Systems,
New South Wales, Australia) and adjuvant (Sigma Adjuvant System: 100µl per mouse, SigmaAldrich, Missouri, USA). CASAC was prepared immediately prior to injection. 200µl per mouse
administered intradermally was used for routine immunisations. CASAC was a kind gift from Dr
James W. Wells (The University of Queensland Diamantina Institute, Brisbane, Queensland,
Australia).
28
IL-2/αIL-2 complex:
Complexing recombinant mouse IL-2 with an anti-IL-2 mAb, JES6-1, has been shown to
specifically expand Treg in vivo (Webster et al., 2009). IL-2/αIL-2 complex was made fresh for
each use by diluting 2.5µl of JES6-1 (at 2mg/ml) in 195 µL of MT-PBS. 2.5µl of recombinant
mouse IL-2 (at 400µg/ml) was added and the solution mixed by gentle pipetting. The solution was
incubated at 37°C for 30 minutes, then stored on ice until use. The final complex (5µg of JES6-1;
1µg of IL-2) was administered to mice by injecting 200µl i.p. per day for 3 consecutive days.
2.2.4 Tissue culture media
Complete Roswell Park Memorial Institute (RPMI-1640) Media:
Complete RPMI-1640 consists of sterile RPMI-1640 (Gibco-Invitrogen, Carlsbad, CA, USA)
supplemented with 1mM sodium pyruvate (Gibco-Invitrogen, Carlsbad, CA, USA), 5x10-5M 2Mercaptoethanol (Sigma-Aldrich, St. Louis, MA, USA), 1x10-4M non-essential Amino Acids
(Gibco-Invitrogen, Carlsbad, CA, USA), 100units/ml Penicillin, 100units/ml Streptomycin and
2mM L-glutamine (Gibco-Invitrogen, Carlsbad, CA, USA).
2.2.5 Monoclonal antibodies, streptavidin conjugates and H2-Kd tetramers
Table 2.2.5.1- List of monoclonal antibodies for flow cytometry
Antigen
Conj.
Isotype
Diln (FACS)
Clone
Supplier/Cat#
CD3ε
FITC
hamster IgG
1/100
145-2C11
Biolegend/100306
CD3ε
PE
hamster IgG
1/400
145-2C11
Biolegend/100308
CD4
FITC
rat IgG2b k
1/200
GK1.5
Biolegend/100406
CD4
PE/Cy7
rat IgG2b k
1/100
GK1.5
Biolegend/100422
CD5
PE
rat IgG2a k
1/400
53-7.3
Biolegend/100608
CD8α
PerCP/Cy5.5
rat IgG2b k
1/200
53-6.7
Biolegend/100734
CD8α
APC/Cy7
rat IgG2b k
1/100
53-6.7
Biolegend/100414
CD8β
PE
rat IgG2a k
1/400
YTS156.7.7
Biolegend/126608
CD11b
FITC
rat IgG2bk
1/800
M1/70
Grown in house
CD11b
PerCP/Cy5.5
rat IgG2bk
1/400
M1/70
Biolegend/101228
CD11c
FITC
hamster IgG
1/200
N418
Biolegend/117306
29
CD25
PE
rat IgG1
1/400
PC61
Biolegend/102007
CD25
APC
rat IgG1
1/200
PC61
Biolegend/102012
CD44
PE
rat IgG2bk
1/400
IM7
Biolegend/103008
FITC
rat IgG2a k
1/800
PE
rat IgG2a k
1/200
CD45.1
PE/Cy7
mouse IgG2aκ
1/100
A20
Biolegend/110730
CD45.1
APC
mouse IgG2aκ
1/100
A20
Biolegend/110714
CD45.2
FITC
mouse IgG2aκ
1/200
104
Biolegend/109806
CD45.2
PE/Cy7
mouse IgG2aκ
1/100
104
Biolegend/109830
CD49b
FITC
rat IgM
1/100
DX5
Biolegend/108906
CD62L
PerCP/Cy5.5
rat IgG2a k
1/200
MEL-14
Biolegend/104432
CD62L
APC
rat IgG2a k
1/200
MEL-14
Biolegend/104412
CD69
PE
hamster IgG
1/200
H1.2F3
Biolegend/104507
CD103
PerCP/Cy5.5
hamster IgG
1/100
2E7
Biolegend/121416
CD117
PE
rat IgG2b k
1/800
2B8
Biolegend/105808
PE
hamster IgG
1/100
TR75-89
Biolegend/113406
PE
rat IgG2a k
1/150
A7R34
eBioscience/
CD45R
RA3-6B2
Biolegend/103206
(B220)
CD45R
RA3-6B2
Biolegend/103208
(B220)
(c-kit)
CD120b
(TNFRII)
CD127
(IL-7Rα)
CD152
12-1271-82
biotinylated
hamster IgG
1/100
biotinylated
rat IgG1 κ
1/100
PerCP/Cy5.5
rat IgG2a k
1/100
FITC
rat IgG2a k
1/200
VCIO-4B9
Biolegend/106304
(CTLA-4)
CD223
C9B7W
Biolegend/125206
(LAG-3)
CD279
29F.1A12
Biolegend/135208
(PD-1)
Foxp3
FJK-16s
eBioscience/
11-5773-82
Granzyme B
PE
rat IgG2b k
1/200
16G6
eBioscience/
12-8822
GITR
PE
hamster IgG
1/100
YGITR 765
Biolegend/120207
30
HELIOS
I-Ak,g7,r,f,s
I-Ak,g7,r,f,s
KLRG1
PE
hamster IgG
1/100
22F6
Biolegend/157216
FITC
mouse IgG2aκ
1/200
10.2.16
Grown in house
biotinylated
mouse IgG2aκ
1/400
10.2.16
Grown in house
APC
rat IgG2a k
1/100
2F1
eBioscience/
17-5893-82
Ly6-A/E
PerCP/Cy5.5
rat IgG2a
1/100
D7
Biolegend/108124
FITC
rat IgG2b k
1/800
RB6-8C5
Grown in house
Pan-TCRβ
APC
hamster IgG
1/200
H57-597
Biolegend/109212
Perforin
FITC
rat IgG2a k
1/100
MAK-D
eBioscience/
(Sca-1)
Ly-6G/C (Gr1)
11-9392-82
TER-119
FITC
rat IgG2b k
1/400
TER-119
Grown in house
APC
rat IgG2bk
1/400
RR4-7
eBioscience/
(Ly-76)
Vβ6
17-5795
Rat IgG2a
FITC
rat IgG2ak
1/200
RTK2758
Biolegend/400506
FITC
rat IgG2bk
1/200
RTK4530
Biolegend/400606
PE
hamster IgG
1/200
HTK888
Biolegend/400907
(isotype
control)
Rat IgG2b
(isotype
control)
Hamster IgG
(isotype
control)
Table 2.2.5.2 - List of streptavidin conjugates for flow cytometry
Antigen
Conj.
Diln (FACS)
Supplier/Cat#
Streptavidin
FITC
1/400
CALTAG/ SA1001
Streptavidin
PE
1/200
CALTAG/ SA1004-4
Streptavidin
PerCP/Cy5.5
1/200
eBioscience/45431782
Streptavidin
APC
1/200
CALTAG/ SA1005
Streptavidin
Brilliant Violet 421
1/400
Biolegend/405225
31
2.2.6 H2-Kd tetramers
InsB15-23- and IGRP206-214- loaded, H2-Kd tetramers were used to identify insulin- and IGRPspecific CD8+ T cells, respectively, in recipient mice. Listeriolysin O (LLO) is derived from
Listeria monocytogenes and a tetrameric complex of H2-Kd loaded with LLO91-99 was used as an
irrelevant peptide control in analyses using InsB15-23- and IGRP206-214- loaded, H2-Kd tetramers as
an indicator of the level of non-specific tetramer staining.
Table 2.2.6.1 - List of H2-Kd tetramers for flow cytometry
MHC class I allele
Peptide
Conj.
Diln (FACS)
Supplier/Cat#
H2-Kd
InsB15-23 (23 Val)
PE
1/1000
NIH Tetramer
(LYLVCGERV)
H2-Kd
IGRP206-214
Facility
PE
1/1000
(VYLKTNVFL)
H2-Kd
LLO91-99
NIH Tetramer
Facility
PE
(GYKDGNEYI)
1/1000
NIH Tetramer
Facility
2.3 Cell preparations
2.3.1 Lymph nodes and spleens
Using aseptic technique, axial, brachial, inguinal, mesenteric or pancreatic LN were removed and
pooled in a 15ml Falcon tube (BD Biosciences, San Jose, CA, USA) filled with cold F2.5.
Likewise, spleens were removed and placed in a 15ml Falcon tube with 5ml of cold F2.5. To
prepare single cell suspensions, LN or spleens were placed into separate 70µm nylon mesh cell
strainers (BD Biosciences, San Jose, CA, USA) placed on top of 50ml Falcon tubes (BD
Biosciences, San Jose, CA, USA) in a biological safety cabinet. The rubber end of a plunger from a
5ml syringe (BD Biosciences, San Jose, CA, USA) was used to manually disrupt the organ and the
cell strainer was regularly flushed with F2.5. Cells were then collected by centrifugation at 350g for
5 minutes at 4°C. Spleen cells underwent a RBC lysis step using ACK buffer, while LN cells were
washed a further time. LN and spleen cells were then re-suspended separately in F2.5 or culture
medium as required. Cells were counted using trypan blue exclusion and a haemocytometer. For
adoptive transfers, cell suspensions were further washed in 1x MT-PBS and transferred via
intravenous (i.v.) tail injection in 200µl of MT-PBS to each mouse. All adoptive transfers
performed within this thesis were between syngeneic NOD mice.
32
2.3.2 Bone marrow
Using aseptic technique, femurs and tibiae were harvested from donor mice by severing the femur
at the hip and the tibia above the ankle. Bulk tissue was removed from bones before being placed
into a 50ml Falcon tube with 25ml of cold F2.5. In a biological safety cabinet, BM was flushed
from individual bones. Forceps and scissors were used to cut the epiphyseal plate from the proximal
end of the tibia and a 23g needle (BD Biosciences, San Jose, CA, USA) inserted into the marrow
cavity. F2.5 was used to flush BM from individual bones. The distal end of the femur was cut and
marrow flushed. BM cell clumps were re-suspended by passaging several times through the syringe.
The cell suspension was then transferred to a sterile 50ml Falcon tube and centrifuged at 350g for 5
minutes at 4°C. The pellet was RBC lysed with ACK buffer for 2 minutes and washed with F2.5.
Cells were counted and re-suspended to the appropriate concentration for further use.
Recipient mice were gamma-irradiated (137Cs source) with a single dose of 300cGy. Following
irradiation, mice were reconstituted with 2x107 BM cells via i.v. injection, unless otherwise
indicated.
2.3.3 Liver
Mice were perfused with MT-PBS. Livers were removed by grasping the hilus with forceps and
severing the cruciate ligament attached to the diaphragm and then the portal vein. Livers were
placed in 50ml Falcon tubes (BD Biosciences, San Jose, CA, USA) with 10ml of MT-PBS. In a
biological safety cabinet, livers were gently disrupted using a cell scraper (Greiner Bio-One,
Frickenhausen, Germany). Liver cell suspensions were then passed through a 70µm nylon mesh cell
strainer and centrifuged at 75g for 1 minute at 4oC to sediment parenchymal cells. Supernatants
were transferred to a fresh 50ml Falcon tube and centrifuged at 350g for 5 minutes at 4oC. Cells
were re-suspended in 30-35% isotonic Percoll (Sigma-Aldrich, St. Louis, MA, USA) solution with
Lympholyte-M (CedarLane, Ontario, Canada) underlay. Cells were centrifuged at 750g for 20
minutes at R.T. and collected at the Percoll/Lympholyte-M interface. Cells were washed twice with
F2.5 and RBC lysed using ACK buffer as described above. Cells were re-suspended in F2.5 for
staining steps.
2.3.4 Pancreas
Digest medium was made fresh for each use and comprised collagenase type III (Worthington,
Lakewood, NJ, USA) at 1mg/ml and DNase I (Roche Diagnostics, Basel, Switzerland) at 0.1mg/ml
in RPMI 1640 supplemented with 2% FCS. Using aseptic technique, pancreata were excised from
recipient mice and placed in a 15ml Falcon tube with 5ml of cold F2.5. To obtain single cell
suspensions, pancreas tissue was gently disrupted using the end of a plunger from a 5ml syringe.
33
The pancreas tissue was then transferred using a transfer pipette to a 15mL Falcon tube filled with
4.5ml of digest medium and incubated at 37°C in water bath for 15 minutes. During this incubation
period, pancreas tissue was gently disrupted by pipetting every 5 minutes. 0.5ml of room
temperature 100mM EDTA was added per tube and pancreas tissue was continually disrupted by
pipetting for a further 5 minutes. The little remaining undigested material was placed on a 70µm
nylon mesh cell strainer positioned on a 50ml Falcon tube and further disrupted gently with a 5ml
syringe
plunger.
The
strainer
was
washed
through
with
10ml
ice-cold
RPMI/10%FCS/10mMEDTA. Cells were then collected by centrifugation 350g for 5 minutes at
4°C. Cells were washed further with 20ml ice-cold RPMI/10%FCS/10mMEDTA and collected by
centrifugation. Cells were re-suspended in RPMI/10%FCS/10mMEDTA as required for staining
steps.
2.3.5 CFSE-labelling
Pooled LN single cell suspensions were re-suspended in warm MT-PBS to approximately 107
cells/ml in a clean 50ml Falcon tube. 0.5µl/ml of carboxyfluorescein diacetate succinimidyl ester
(CFSE; 2.5µM final; Molecular Probes, Invitrogen, Carlsbad, CA, USA) was added to the side of
the tube and the cell suspension rapidly vortexed to ensure even labelling. The cell suspension was
incubated at 37°C in the dark for 10 minutes. 30ml of cold F2.5 was added to stop labelling. Cells
were then counted and re-suspended to appropriate volume in MT-PBS for adoptive transfer. 5x106
CFSE-labelled naïve G9 T cells or 2x106 CFSE-labelled G9 Tmem were transferred to recipients
unless otherwise stated.
2.3.6 Purification of CD8+ T cells
Following preparation of LN single cell suspension, cells were re-suspended in F2.5/20mM EDTA
at a concentration of 108 cells/ml and stained with an anti-CD8β mAb using standard FACS staining
procedures. Cells were washed and re-suspended to 5x107 cells/ml in F2.5/20mM EDTA and then
transferred to a polypropylene FACS tube (BD Biosciences, San Jose, CA, USA) by filtering
through a 35µm cell strainer (BD Biosciences, San Jose, CA, USA). Cells were maintained on ice,
with all centrifugation steps at 4°C. Sterile propidium iodide (PI) (Sigma-Aldrich, St. Louis, MA,
USA) was added to a working concentration of approximately 10µg/ml immediately prior to sorting
to identify and exclude dead cells. CD8β-positive cells were sorted to a purity of > 90% on a MoFlo
Astrios (Beckman Coutler, Fort Collins, USA). Sorted cells were collected into 10% FCS-coated
tubes containing 5ml of cold F2.5. After sorting, cells were counted and re-suspended in MT-PBS
for adoptive transfer. Aliquots of cells were kept for independent pre- and post-sort flow cytometry
analyses to determine purity and sort efficiency.
34
2.3.7 Generation of G9 memory T cells
LN and spleen single cell suspensions from G9 mice were prepared as described above. Both LN
and spleen cells were then resuspended in complete RPMI/10% FCS at 2x106 cells/ml. InsB15-23 (23
Gly) and recombinant human IL-2 (PeproTech, Rocky Hill, NJ) were added to cultures at 10µg/ml
and 10ng/ml, respectively. Cells were plated in 6 well plates (Greiner Bio-one, Frickenhausen,
Germany) at 3ml/well. Cells were incubated at 37°C, 5% CO2 for 3 days. On day 3, cells were
harvested and washed 3 times in warm RPMI. Cells were counted and re-suspended in complete
RPMI/10% FCS to 2x106/ml. Recombinant mouse IL-15 (PeproTech, Rocky Hill, NJ) was added at
10ng/ml and cells were plated and cultured at 37°C, 5% CO2 for 2 days. Tmem were washed 3
times in F2.5, counted and transferred via i.v. injection. 2x106 cells per mouse were transferred in
200µl of MT-PBS, unless indicated otherwise.
2.3.8 Purification of hematopoietic stem and progenitor cells
HSPC were isolated from BM by sorting on a MoFlo Astrios (Beckman Coutler, Fort Collins, USA)
or FACSAria II (BD Biosciences, San Jose, CA, USA). After RBC lysis, cells were re-suspended in
F2.5/20mM EDTA at 108 cells/ml and stained with anti-c-kit and anti-lineage (lin) mix (CD4, CD8,
CD45R (B220), CD11b, Gr-1 (Ly-6G), TER-119 (Ly-76)) mAb using standard FACS staining
procedures. Cells were washed and re-suspended to 5x107 cells/ml in F2.5/20mM EDTA. Cells
were passed through a 35µm cell strainer (BD Biosciences, San Jose, CA, USA) before being
transferred to a polypropylene FACS tube (BD Biosciences, San Jose, CA, USA). Cells were
maintained on ice with all centrifugation steps at 4°C. Sterile PI was added to cells immediately
prior to sorting to identify and exclude dead cells. c-kit-positive, lin-negative cells were sorted to a
purity of > 90%. Sorted cells were collected into 10% FCS-coated tubes with 5ml of cold F2.5.
After sorting, cells were counted and re-suspended in MT-PBS for adoptive transfer. Aliquots of
cells were kept for independent pre- and post-sort flow cytometry analyses to determine purity and
sort efficiency.
Recipient mice were γ-irradiated (137Cs source) with a single dose of 300cGy. Following irradiation,
mice were reconstituted with purified HSPC via i.v. injection or intrahepatic injection (for neonates;
(Pearson et al., 2008)). 2x105 HSPC were injected in 200µl MT-PBS for recipient mice at 6 weeks
of age or older (approximately 1x107 HSPC/kg). 1x105 HSPC were injected in 100µl MT-PBS for
recipient mice at 3 weeks of age. 2x104 HSPC were injected in 50µl MT-PBS for recipient mice at 3
days of age.
35
2.4 Flow cytometry
2.4.1 Surface staining
All staining procedures were performed on ice. For flow count analyses to determine total organ
cellularity, 1ml of spleen or LN samples were added to FACS tubes, centrifuged at 750g for 2
minutes at 4oC and supernatant was discarded to leave ~100µl. Samples were briefly vortexed to
resuspend cells. For all other flow analyses, 100µl of sample was added straight to each FACS tube.
All samples were treated with 10µl FACS block (anti CD16/32 [2.4G2]) for 5mins. Next, 10µl of
required antibodies (at 10x final working concentrations; Table 2.2.5.1) were added to each sample
and tubes were incubated in the dark for 30 minutes. Following staining, cells were washed with
cold FACS wash and cells collected. Supernatant was ‘tipped off’ to leave ~100µl and briefly
vortexed to resuspend cells. If required for flow count analysis, Flow-CountTM Fluorospheres
(Beckman Coulter, Miami Lakes, FL, USA) were added at 20µl per tube. Where appropriate, PI
was added to samples immediately prior to analysis to distinguish and exclude dead cells. Samples
were collected on a BD FACSCanto (BD Biosciences, San Jose, CA, USA). Diva (BD Biosciences,
San Jose, CA, USA) or FlowJo (TreeStar Inc, Ashland, OR, USA) software was used to analyze
data.
2.4.2 Intracellular staining
Cells were stained for surface markers as described above and kept on ice. Cells were fixed with
0.5ml of Fixation/Permeabilization solution (eBioscience, San Diego, CA, USA) and briefly
vortexed. Cells were incubated in the dark for 3 hours (for Foxp3) or 30 minutes (for other
intracellular markers) prior to 2 washes with Permeabilization buffer (eBioscience, San Diego, CA,
USA). Cells were blocked (10µl FACS-block and 10% normal rat serum), relevant antibodies added
and incubated in the dark for 30 minutes. Cells were washed twice with Permeabilization buffer and
re-suspended in an appropriate volume for flow cytometric analysis.
2.4.3 Tetramer staining
Cells were prepared as per ‘Surface staining’ method, however, the tetramer staining step was
performed at room temperature (R.T.). 10µl of required tetramer (at 10x final working
concentration; Table 2.2.6.1) was added to each sample and tubes were incubated at R.T. in the dark
for 15 minutes. Without washing, 10µl of required antibodies (at 10x final working concentrations;
Table 2.2.5.1) were added to each sample and tubes were incubated on ice in the dark for a further
15 minutes. Following staining, cells were washed with cold FACS wash and cells collected.
Supernatant was ‘tipped off’ to leave ~100µl and briefly vortexed to resuspend cells prior to flow
cytometric analysis.
36
2.5 Ex vivo assays
2.5.1 ELISPOT
Polyvinylidene difluoride enzyme-linked immunospot (ELISPOT) plates (Millipore) were pretreated with ethanol and washed with MT-PBS. The required capture antibody (Table 2.5.1.1) was
added and plates incubated overnight at 4°C wrapped in foil. The next day, plates were washed and
blocked with RPMI/5% FCS for 1-2 hours. Spleen cells were plated at 5x105 cells/well in 200µl
RPMI/5% FCS in triplicate. Necessary wells were left untreated or stimulated with InsB15-23 (23
Gly) (10µg/ml), InsB15-23 (23 Val) (0.5µg/ml), or anti-CD3 mAb (0.1µg/ml). Plates were incubated
for 2 days at 37°C, 5% CO2 in a damp chamber. After 2 days, plates were washed and required
detection antibody (Table 2.5.1.1) was added. Plates were then incubated overnight at 4°C wrapped
in foil. Avidin-HRP detection enzyme (Sigma-Aldrich, St. Louis, MA, USA) was added and plates
were incubated at R.T. for 1 hour. Plates were washed and AEC substrate (Calibochem, San Diego,
CA, USA) added. Development of colour change from the substrate reaction was observed under a
dissecting microscope for 5-15 minutes and plates washed with tap water as soon as background
development was observed. Plates were stored at R.T. in the dark until dry and then read using an
immuno-spot plate reader (AID GmbH, Strassburg, Germany).
Table 2.5.1.1- Monoclonal antibodies for IFN-γ ELISPOT
Cytokine
Capture antibody
Supplier/Cat#
Detection antibody
Supplier/Cat#
clone
(working conc.)
clone
(working conc.)
AN-18
eBioscience/
R4-6A2
eBioscience/
IFN-γ
14-7313-85
13-7312-85
(5µg/ml)
(2µg/ml)
2.6 Assessment of diabetes
Mice screened for glycosuria weekly using Diastix Reagent Strip (Bayer, Leverkusen, Germany).
Blood glucose was determined for glycosuric mice using an Accu-Chek Go meter (Roche, Basel,
Switzerland). Mice were deemed diabetic following two consecutive blood glucose readings
>12mM. Mice deemed diabetic were subsequently euthanised.
2.7 Statistical analysis
Graphpad Prism 6 was used to generate all graphs and statistical analyses. Means were compared by
student’s t-test for pairwise comparison, or one-way analysis of variance (ANOVA) with Tukey’s
multiple comparison post-test for multiple comparisons. Significance was assigned for the
following p-values: < 0.05 (*), < 0.01 (**), < 0.001 (***), < 0.0001 (****).
37
Chapter 3
Transgenic expression of proinsulin induces peripheral tolerance
in insulin-specific CD8+ T cells
38
3.1 Introduction
β cell-specific CD8+ T cells are important contributors to T1D pathogenesis, capable of
directly lysing β cells by the Fas/FasL pathway and by release of perforin/granzymes (Bulek et al.,
2012; Unger et al., 2012). As well as direct lysis, β cell-specific CD8+ T cells drive islet
inflammation through production of proinflammatory cytokines such as IFN-γ and TNF-α (Varanasi
et al., 2012; Yi et al., 2012). Peripheral tolerance induction in antigen-specific CD8+ T cells by
targeting antigen expression to APC has been well established using foreign model antigens in both
adoptive transfer (Kenna et al., 2010; Kurts et al., 1997; Steptoe et al., 2007) and inducible
Cre/LoxP systems (Probst et al., 2003). T-cell responses in islet autoimmunity are directed toward
β-cell antigens and robust evidence from the NOD mouse suggests insulin is the initiating antigen
from which islet autoimmunity arises. This is evidenced by firstly, B16A mice in which insulin 1
and insulin 2 are knocked out and replaced with a mutated, non-antigenic preproinsulin under
control of the rat insulin promoter (RIP) and B16A mice do not develop insulitis nor diabetes
(Nakayama et al., 2005). This indicates CD8+ and CD4+ T cells must recognize the
immunodominant region of insulin B chain, that is, InsB9-23, for diabetes development in NOD
mice. Secondly, transgenic expression of proinsulin under control of an MHC class II promoter
protects NOD mice from insulitis and autoimmune diabetes (French et al., 1997). Furthermore,
transfer of HSPC from proinsulin transgenic mice abrogates diabetes development in NOD
recipients (Steptoe et al., 2003). It was determined that transgenic expression of proinsulin targeted
to APC prevented development and expansion of diabetogenic T-cell responses to other β-cell
antigen specificities such as IGRP (Krishnamurthy et al., 2006). Krishnamurthy et al. also
demonstrated that while transgenic expression of IGRP resulted in deletion of IGRP-specific CD8+
T cells, this alone was not sufficient to prevent diabetes development. Such studies emphasise not
only the precedence of (pro)insulin in islet autoimmunity, but also, the potential of transgenic
antigen, particularly insulin, expression to alter diabetes development. As the studies described
demonstrate ‘developmental’ induction of antigen-specific tolerance, it remains unclear as to
whether transgenic proinsulin expression might induce peripheral tolerance in mature insulinspecific CD8+ T cells.
G9 TCR transgenic mice (Wong et al., 2009) express a H2-Kd-restricted InsB15-23-specific
TCR derived from a CD8+ T cell clone isolated from the islet infiltrate of young RIP.B7-1 NOD
mice (Wong et al., 1999; Wong et al., 1996). The original G9 T cell is highly diabetogenic and
effectively transfers diabetes to irradiated NOD and immunodeficient NOD.SCID recipients, even
in the absence of endogenous CD4+ T cells. Since the G9 T cell clone was isolated from young
NOD mice, prior to diabetes onset, it is possible that the InsB15-23 specificity is involved in early
islet autoimmunity. In keeping with this suggestion, InsB15-23-loaded, H2-Kd-tetramer binding (G939
like) T cells comprise a large proportion of islet-infiltrating T cells in 4-week old NOD mice and
the frequency of these G9-like T cells diminishes over time (Wong et al., 1999). G9 TCR transgenic
T cells (herein referred to as G9 T cells) may, therefore, be a good model of early CD8+ T-cell
responses in autoimmune diabetes development and they can be employed to examine the responses
of insulin-specific CD8+ T cells in an adoptive transfer setting and an example would be using G9 T
cells to investigate peripheral tolerance induction. Notably, unlike many model systems using
foreign model antigens to test tolerance induction, G9 T cells have undergone thymic selection in
the presence of their cognate antigen. This is an important consideration as autoreactive, β cellspecific T cells in T1D in human may have undergone thymic selection in the presence of their
cognate antigen. If different tolerance outcomes were observed in a study using G9 T cells
compared to studies using foreign model antigens, this may have implications in understanding
peripheral tolerance induction in T cells that have been subjected to thymic selection pressure. This
might impact the development of, for example, immunotherapies for T1D that induce T-cell
peripheral tolerance. Here, with these considerations in mind, I will test whether G9 T cells undergo
peripheral tolerance after transfer to proinsulin transgenic mice.
Transgenic mouse models have been used to understand the effect of the autoimmune-prone
NOD background on antigen-specific T cell tolerance. When HA is expressed transgenically in
autoimmune-resistant BALB/c or B10.D2 background mice, only low-avidity HA-specific T cells
are recovered from HA transgenic mice, indicating high-avidity HA-specific T cells have been
deleted (Lo et al., 1992). In addition, these mice did not develop diabetes after influenza
immunization. In contrast, in autoimmune-prone NOD HA-expressing mice, islets contained highavidity HA-specific CD8+ T cells, capable of transferring diabetes, suggesting that the NOD
background permits the survival and effector function of these cells (Kreuwel et al., 2001). The
different tolerance outcomes observed underlines the importance of peripheral tolerance in
controlling CD8+ T cell responses to prevent autoimmunity and suggests NOD mice are inherently
defective in pathway(s) associated with peripheral tolerance induction. Furthermore, transgenic
expression of protective Idd alleles restores peripheral tolerance in clone 4 T cells on the NOD
background (Martinez et al., 2005). Peripheral CD8+ T cell tolerance to the important β-cell antigen,
IGRP, is also restored upon introduction of protective Idd3/5 alleles to NOD mice (HamiltonWilliams et al., 2012). These studies using Idd congenic mice suggest genetic loci crucial to
autoimmune diabetes development in NOD mice have a direct influence on CD8+ T-cell peripheral
tolerance pathways.
In this chapter, I investigate whether targeting proinsulin expression to MHC class II+ APC
induces peripheral tolerance in insulin-specific CD8+ T cells. In an adoptive-transfer setting, the
response of G9 T cells to transgenic proinsulin expression and islet-derived insulin expression was
40
compared. G9 T cell proliferation and population dynamics in secondary lymphoid sites were
assessed in recipients, as well as G9 T cell TCR expression, phenotype and binding of InsB15-23loaded tetramer. The response of transferred G9 T cells to InsB15-23 immunisation and the
contribution of G9 T-cell transfer to diabetes development were assessed.
41
3.2 Results
3.2.1 G9 T cells display a naïve T-cell phenotype despite expression of cognate antigen in G9 mice
G9 TCR transgenic T cells have developed, undergone thymic selection and been
maintained peripherally in the presence of their cognate antigen, InsB15-23. It is possible that this
antigen exposure has activated G9 T cells in vivo. To investigate this, the phenotype of CD8+ T cells
expressing the transgenic TCR chain Vβ6+ from LN of G9 mice was compared to CD8+Vβ6+ T
cells from LN of non-transgenic NOD mice. As expected, CD8+Vβ6+ T cells from LN of TCR
transgenic G9 mice were consistently more frequent than from non-transgenic NOD mice (Fig.
3.1A, B). The level of Vβ6 expression was high on transgenic G9 T cells (MFI: 7529±2704,
mean±S.D., n=3) and was not significantly different to that of T cells from non-transgenic NOD
mice (MFI: 6815±3312, mean±S.D., n=3, p=0.8) (Fig. 3.1A). CD69 expression was low and similar
between G9 and NOD T cells (Fig. 3.1A). The proportion of CD8+Vβ6+ T cells expressing CD44
was slightly higher from NOD (% CD44+: 54.7±19.6, mean±S.D.) than G9 (% CD44+: 33.9±15.4,
mean±S.D.) (Fig. 3.1A). This difference might be due to the polyclonal nature of the T-cell
repertoire in non-transgenic NOD mice compared to that of the specificity-restricted TCR
transgenic G9. Finally, the majority of CD8+Vβ6+ T cells from both G9 and non-transgenic NOD
mice expressed high levels of CD62L (Fig. 3.1A). Since G9 TCR transgenic mice express insulin in
pancreatic β cells, it is possible that G9 T cells would be activated in pLN, the site of islet-derived
insulin presentation. To test this, the phenotype of CD8+Vβ6+ T cells from pLN of G9 mice was
also analysed. A small proportion of CD8+Vβ6+ T cells from pLN expressed CD69 (% CD69+:
0.4±0.2, mean±S.D.) (Fig. 3.1C). Similarly to CD8+Vβ6+ T cells from non-pancreatic LN of G9
mice (Fig. 3.1A), most CD8+Vβ6+ T cells expressed CD62L (% CD62L+: 85.8±3.3, mean±S.D.)
and low levels of CD44 (Fig. 3.1C). Some CD8+Vβ6+ T cells expressed CD44 (% CD44+: 20.5±9.1,
mean±S.D.), which may indicate activation (Fig. 3.1C). Despite some evidence of activation, the
majority of G9 T cells from pLN displayed a typically naïve T-cell phenotype. Despite expression
of insulin in pancreatic islets and thymus of G9 mice, these data confirm G9 T cells exhibit a
phenotype typical of naïve CD8+ T cells with similar expression patterns of CD69, CD44 and
CD62L to that of analogous, non-transgenic T cells.
42
A
NOD
SSC
CD8
Isotype
NOD
G9
Vβ6
G9
Vβ6
B
CD69
CD44
CD62L
CD8
FSC
Vβ6
Strain
CD8+Vβ6+ of live cells (%)
[mean±S.D.]
NOD (non-transgenic)
2.50±0.26
G9 (TCR transgenic)
36.83±8.16
C
Vβ6
CD44
Isotype
mAb
CD69
CD62L
Figure 3.1- G9 T cells display a typically naïve phenotype in the presence of cognate antigen
expression in G9 mice.
(A) Axillary, brachial, inguinal and mesenteric LN cells were harvested and pooled from G9 and NOD
mice. Surface expression of Vβ6, CD69, CD44 and CD62L on CD8+Vβ6+ T cells was determined
using flow cytometry. Histograms are representative of five G9 and three NOD mice. (B) The
proportion of CD8+Vβ6+ T cells from live cells was determined using flow cytometry. Proportions are
mean±S.D. of pooled male G9 LN cells from 11 independent experiments and pooled male NOD LN
cells from three independent experiments. (C) pLN were harvested from G9 mice. Surface expression
of CD69, Vβ6, CD44 and CD62L on CD8+Vβ6+ T cells was determined using flow cytometry. n=2.
The percentages in each quadrant refer to the mAb-stained population (blue) for each histogram. Dot
plots are representative of two independent experiments.
43
3.2.2 G9 T cells divide extensively after transfer to proinsulin transgenic mice
Since G9 T cells represent a population of naïve, insulin-specific CD8+ T cells, the next step
was to determine whether G9 T cells respond to presentation of InsB15-23 derived from transgenic
proinsulin. To test this, CFSE-labelled G9 LN cells were transferred to B16A, NOD and proinsulin
transgenic mice and G9 T cell CFSE dilution was determined three days later. As expected, due to
B16A mice lacking the G9 T-cell determinant, there was no CFSE dilution observed in G9 T cells
recovered from spleen, skin-draining LN (sdLN) or pLN of B16A recipients (Fig. 3.2A). A low
proportion of G9 T cells from pLN, but not spleen or sdLN, of NOD recipients showed evidence of
1-2 divisions, as indicated by CFSE dilution (Fig. 3.2A). In contrast, in proinsulin transgenic
recipients, widespread G9 T cell proliferation was observed in that the majority of G9 T cells from
spleen, sdLN and pLN had diluted CFSE (Fig. 3.2A). The proportion of CFSElo G9 T cells
recovered from proinsulin transgenic recipients was significantly higher than from B16A and NOD
recipients (Fig. 3.2B). In NOD recipients, the proportion of CFSElo G9 T cells was significantly
higher for pLN than for spleen or sdLN (Fig. 3.2B). When the proliferation index was calculated,
the extent of G9 T-cell division in pLN of NOD recipients was identical to that in spleen and sdLN
(Fig. 3.2C). However, the proliferation index of G9 T cells was significantly higher in spleen, sdLN
and pLN from proinsulin transgenic recipients than from B16A and NOD recipients (Fig. 3.2C).
Despite a low number of replicates analysed, the proportion of CFSElo G9 T cells recovered
from proinsulin transgenic recipients was similar at both two (Fig. 3.2D) and three days after
transfer (91.7% vs. 90.4%±6.9). The proportion of CFSElo G9 T cells in pLN of NOD recipients
appeared to be lower at two compared to three days after transfer (Fig. 3.2D). As early as two days
after transfer, there was a trend toward a much higher proportion of CFSElo G9 T cells in proinsulin
transgenic recipients compared to B16A and NOD recipients (Fig. 3.2D). The proliferation index of
G9 T cells from proinsulin transgenic recipients was lower at two than three days after transfer (2.5
vs 5.0±1.40; Fig. 3.2E), indicating ongoing division of G9 T cells between days 2 and 3 after
transfer to proinsulin transgenic recipients. The proliferation index of G9 T cells from proinsulin
transgenic recipients appeared to be higher than B16A and NOD recipients (Fig. 3.2E). These data
suggest functional presentation of Ins B15-23 from transgenic proinsulin occurs in proinsulin
transgenic mice and that G9 T cells are responsive to this presentation in vivo. Additionally, the
CFSE-based analyses suggest a difference in the proliferative response of G9 T cells to transgenederived antigen presentation compared to presentation of islet-derived insulin, in that, most G9 T
cells proliferate early after transfer to proinsulin transgenic recipients, whereas only a low
proportion of G9 T cells in pLN of NOD recipients proliferated and more slowly. A difference in
proliferative response to antigen presentation might indicate different T-cell tolerance outcomes
between recipient strains.
44
A
B16A
12%
2%
B
proIns Tg
NOD
89%
pLN
****
95%
2%
1%
G 9 T c e lls ( % )
100
****
60
40
CFSE
lo
sdLN
80
4%
20
0
87%
0%
s
p
l
s
Spleen
d
L
N
p
L
N
s
p
l
s
B16A
d
L
N
p
L
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
NOD
CFSE
C
D
3
15
40
20
B16A
NOD
p ro In s T g
NOD
p ro In s T g
B 16A
p
s
p
L
N
N
s
L
N
L
p
d
l
N
p
L
s
d
N
L
N
L
p
l
N
l
p
N
N
s
l
N
p
N
N
L
B16A
d
L
s
p
L
L
p
d
p
l
s
L
s
L
L
d
p
N
L
L
N
p
d
l
s
s
p
s
s
N
L
L
N
d
p
l
s
L
p
p
d
N
s
s
N
d
l
s
s
p
l
0
0
0
1
s
CFSE
n .s .
2
p
G 9 T c e lls ( % )
60
s
5
80
lo
10
P r o life r a tio n in d e x
100
****
P r o life r a tio n in d e x
E
NOD
p ro In s
Tg
Figure 3.2 – G9 T cells proliferate extensively after transfer to proinsulin transgenic mice.
(A-E) CFSE-labelled G9 LN cells (5x106) were transferred to B16A, NOD and proinsulin transgenic
recipients. Three (A-C) or two (D, E) days later, CFSE dilution and proliferation index was
determined in spleen, sdLN (axillary, brachial and inguinal) and pLN using flow cytometry. FACS
plots (A) display CFSE dilution of CFSE+ G9 T cells and are representative of at least six individual
mice. Data are pooled from four independent experiments (A-C) or from a single experiment (D, E).
Each point represents a single mouse. Error bars show S.D. One-way ANOVA followed by Tukey
post-test. (****p<0.0001).
45
3.2.3 G9 T cells are rapidly deleted after transfer to mice transgenically expressing proinsulin
Since transferred G9 T cells undergo several divisions in proinsulin transgenic recipients
within two to three days after transfer, it might be expected that transferred G9 T cells would
accumulate in secondary lymphoid organs. To determine the population dynamics of G9 T cells
after transfer, a congenic system was established to identify G9 T cells in recipients. NOD.CD45.2
mice (CD45.2+CD45.1-) were mated to proinsulin transgenic (CD45.1+CD45.2-) and NOD
(CD45.1+CD45.2-) mice. F1 offspring (CD45.1+CD45.2+) were used as recipients for G9 LN cells
(CD45.1+CD45.2-) and, at defined time points, G9 T cells (CD45.1+CD45.2-CD8+Vβ6+) were
enumerated (Fig. 3.3A). To control for background mAb staining, mice of each strain that were not
injected with G9 T cells were analysed alongside test mice in each experiment. Mice that were not
injected with G9 T cells consistently displayed negligible CD45.1+CD45.2-Vβ6+CD8+-gated events.
Also, since similar trends were observed at the day 3 time point when bulk G9 LN cells or purified
CD8+ G9 T cells were transferred, the results obtained from when both bulk G9 LN cells and
purified CD8+ G9 T cells were used as donor cells were pooled for the day 3 time point represented
in the time course analyses. Despite extensive proliferation of G9 T cells after transfer to proinsulin
transgenic recipients (Fig. 3.2A), the total number of G9 T cells was similar between NOD and
proinsulin transgenic recipients in spleen, sdLN or pLN at one, two or three days after transfer (Fig.
3.3B). At both five and seven days after transfer, fewer G9 T cells were detected in spleens of
proinsulin transgenic recipients than NOD recipients (Fig.3.3B). The total number of G9 T cells
present in spleen of proinsulin transgenic recipients decreased over time (day 2 vs day 5, p<0.05),
whereas the total number of G9 T cells in spleen of NOD recipients remained unchanged over this
period (Fig. 3.3B). When the proportion of G9 T cells in the total CD8+ T cell population was
analysed, there was a decrease in the proportion of G9 T cells in spleen of proinsulin transgenic
recipients over time (day 2 vs day 5, p<0.05; day 3 vs day 5, p<0.05) (Fig. 3.3C). Likewise, there
was a decrease in the proportion of G9 T cells across time in sdLN of proinsulin transgenic
recipients (day 1 vs day 3, p<0.05; day 3 vs day 5, p<0.05) (Fig. 3.3C). In contrast, the proportion
of G9 T cells of total CD8+ T cells in spleen and sdLN of NOD recipients did not differ over time
(Fig. 3.3C). At five and seven days after transfer, the proportion of G9 T cells was significantly
lower in proinsulin transgenic recipients than in NOD recipients for spleen, sdLN and pLN (Fig.
3.3C).
Since bulk LN cells from G9 mice comprise CD8+Vβ6+ G9 T cells and a population of
uninvolved CD8-Vβ6- cells (referred to herein as non-G9 cells), analysis of the non-G9 cell
population might be an indicator of the antigen-specificity of G9 T-cell population dynamics
observed after transfer to NOD and proinsulin transgenic recipients. When the ratio of G9 T cells
(CD45.1+CD45.2-CD8+Vβ6+) to non-G9 cells (CD45.1+CD45.2-CD8-) was analysed, the ratio
46
remained similar in NOD and proinsulin transgenic recipients (Fig.3.3D). An increased ratio was
observed in pLN of NOD recipients (Fig. 3.3D), suggesting expansion of G9 T cells in response to
islet-derived proinsulin was specific to transferred G9 T cells. Data derived from timecourse studies
demonstrate a decrease in the total number and proportion of G9 T cells over time after transfer to
proinsulin transgenic recipients. In contrast, the number of G9 T cells was relatively stable after
transfer to NOD recipients. Therefore, G9 T cells appear to be deleted after transfer to proinsulin
transgenic recipients.
47
A
CD8
CD45.2
G9 T cells
Vβ6
40
C
****
30
N O D spl
20
p r o In s T g s p l
10
N O D sdLN
0 .5
****
****
****
N O D pLN
0 .4
+
T o t a l G 9 T c e lls ( x 1 0
-3
)
****
p r o In s T g s d L N
6
N O D pLN
4
p r o In s T g p L N
2
0
0
1
2
3
4
5
6
7
G 9 T c e lls /t o t a l C D 8
B
c e lls ( % )
CD45.1
N O D spl
0 .3
N O D sdLN
p r o In s T g s p l
0 .2
p r o In s T g s d L N
p r o In s T g p L N
0 .1
0 .0
0
1
2
3
4
5
6
7
D a y s (p o s t tra n s fe r)
D a y s (p o s t tra n s fe r)
D
R a t io ( G 9 : n o n G 9 )
3
****
N O D pLN
p r o In s T g p L N
N O D sdLN
2
p r o In s T g s d L N
N O D spl
p r o In s T g s p l
1
0
0
2
4
6
8
D a y s (p o s t tra n s fe r)
Figure 3.3- G9 T cells are deleted after transfer to proinsulin transgenic mice.
G9 LN cells (5x106; day 1, 2, 3, 5, 7; n=4-5 day 3) or purified CD8+ G9 T cells (2x106; day 3, n=3-4)
were transferred to NOD and proinsulin transgenic recipients. (A) Gating strategy for analysis of G9 T
cells transferred to congenic (CD45.1+CD45.2+) recipients. At times shown, the total number (B) and
proportion of G9 T cells (CD45.1+CD8+Vβ6+) of total CD8+ T cells (C) in spleen, sdLN and pLN was
determined using flow cytometry. (D) Ratio of G9 (CD45.1+CD8+Vβ6+) to non-G9 cells
(CD45.1+CD8-). Data are pooled from two (day 1, 2, 5) or four (day 3) independent experiments. n=4
(day 1, 2, 5 and 7); n= 7-9 (day 3). Error bars show S.D. One-way ANOVA followed by Tukey posttest. (B) (****p<0.0001: NOD spl vs proIns Tg spl) (C) (**** p<0.0001: NOD pLN vs proIns Tg
pLN) (D) (**** p<0.0001: NOD pLN vs NOD spl).
48
3.2.4 Transferred G9 T cells do not infiltrate the pancreas of NOD or proinsulin transgenic
recipients
Since low numbers of G9 T cells were recovered from secondary lymphoid organs of
proinsulin transgenic recipients, it is possible that G9 T cells might traffic to the pancreas after
transfer thereby reducing the G9 T cell number observed in spleen and LN. To test this, G9 LN cells
were transferred to NOD and proinsulin transgenic mice and pancreas-infiltrating lymphocytes were
enumerated three days later. Few host (CD45.1+CD45.2+) leukocytes were observed in pancreas of
NOD and proinsulin transgenic recipients (Fig. 3.4A). Transferred leukocytes (CD45.1+CD45.2-) as
a proportion of host leukocytes (CD45.1+CD45.2+) did not differ between NOD and proinsulin
transgenic recipients (Fig. 3.4B). Despite a low number of biological replicates analysed, there
appeared to be no difference in the proportion of transferred leukocytes in NOD and proinsulin
transgenic mice that were not injected with G9 T cells compared to either NOD or proinsulin
transgenic recipients of G9 T cells (Fig. 3.4B). Similarly, there was no significant difference in the
proportion of transferred G9 T cells (CD45.1+CD45.2-CD8+) of host CD8+ T cells between NOD
and proinsulin transgenic recipients (Fig. 3.4C). With the caveat that a low number of untransferred
control mice were analysed, there appeared to be no difference in the proportion of G9 T cells of
host CD8+ T cells between untransferred mice and recipients of G9 T cells (Fig. 3.4C). It should be
noted that it is possible that transferred G9 T cells may adhere to the collagen matrix within the
pancreas, precluding their identification using flow cytometry. Therefore, histology on pancreas
samples from recipients of G9 T cells could be examined to validate these data. From the available
data, G9 T cells do not appear to preferentially traffic to pancreas in NOD or proinsulin transgenic
recipients.
49
PI
SSC
A
FSC
FSC
proIns Tg
CD45.1
CD45.1
NOD
CD45.2
CD45.2
C
n .s .
+
2 .0
1 .5
1 .0
0 .5
0 .0
NOD
p r o In s T g
NOD
p r o In s T g
G 9 T c e lls /t o t a l C D 8
o f t o t a l le u k o c y t e s ( % )
T r a n s f e r r e d le u k o c y t e s
2 .5
T c e lls ( % )
B
n .s .
0 .3
0 .2
0 .1
0 .0
NOD
p r o In s T g
NOD
p r o In s T g
u n tra n s fe rr e d
G 9 T c e ll
u n tra n s fe rr e d
G 9 T c e ll
c o n tr o ls
r e c ip ie n t s
c o n tr o ls
r e c ip ie n t s
Figure 3.4- Low recovery of transferred G9 T cells in pancreas of NOD and proinsulin
transgenic recipients.
(A-C) G9 LN cells (5x106) or purified CD8+ G9 T cells (2x106) were transferred to NOD and
proinsulin transgenic recipients (males, at 8-10 weeks of age). Three days later, the proportion of
pancreas-infiltrating transferred leukocytes (CD45.1+/CD45.2-) of total leukocytes (B) and the
proportion of G9 T cells (CD45.1+/CD45.2-/CD8+) of total CD8+ T cells (C) were determined using
flow cytometry. Representative gating strategy to analyse infiltrating leukocytes in pancreas samples
(A). Data are pooled from three independent experiments. Each point represents a single mouse. Error
bars show S.D. One-way ANOVA followed by Tukey post-test.
50
3.2.5 G9 T cells are not preferentially activated and retained in the liver of proinsulin transgenic
recipients
Since the liver is a site of CD8+ T cell activation and retention or clearance (Benseler et al.,
2011; Bertolino et al., 2001; Huang et al., 1994), it is possible that G9 T cells may become activated
and retained in liver after transfer to proinsulin transgenic mice. This could account for the reduced
number of G9 T cells recovered from secondary lymphoid organs in proinsulin transgenic
recipients. To test this, G9 LN cells were transferred to NOD and proinsulin transgenic recipients
and the frequency and CD69 expression by G9 T cells in liver were assessed three hours later.
Although only two mice were analysed, the percentage of G9 T cells was higher in spleen compared
to liver for both strains, while in liver the percentage of G9 T cells was similar between strains (Fig.
3.5A). In proinsulin transgenic recipients, the proportion of CD69+ G9 T cells was higher in spleen
than in liver but in NOD recipients the inverse was observed (Fig. 3.5B). These data suggest G9 T
cells are not preferentially activated and retained in liver after transfer to proinsulin transgenic
recipients.
51
2 .5
2 .0
G 9 T c e lls /t o t a l C D 8
+
c e lls ( % )
A
1 .5
1 .0
0 .5
0 .0
L iv e r
S p le e n
NOD
L iv e r
S p le e n
p ro In s T g
B
40
30
20
C D 69
+
G 9 T c e lls ( % )
50
10
0
L iv e r
S p le e n
NOD
L iv e r
S p le e n
p ro In s T g
Figure 3.5- Low recovery of transferred G9 T cells in liver of NOD and proinsulin transgenic
recipients.
G9 LN cells (5x106) were transferred to NOD and proinsulin transgenic recipients. Three hours later,
the proportion of G9 T cells (CD45.1+CD45.2-CD8+Vβ6+) of total CD8+ T cells (A) and the proportion
of CD69+ G9 T cells (B) in liver and spleen were determined using flow cytometry. Data are from a
single experiment. Each point represents a single mouse.
52
3.2.6 Reduced recovery of G9 T cells from proinsulin transgenic recipients is not due to allotypemediated rejection
Overall, the number of G9 T cells recovered from lymphoid tissues of proinsulin transgenic
recipients was lower than that typically recovered in other adoptive transfer studies either in
C57BL/6 (Nasreen et al., 2010; Steptoe et al., 2007) or NOD mice (Krishnamurthy et al., 2012).
While the data considered to this point are consistent with deletion of G9 T cells in proinsulin
transgenic recipients, it is possible that the low number of G9 T cells recovered was due to a
suboptimal number of cells being transferred. Therefore, a pilot experiment was performed where
graded doses of G9 LN, up to 20x106, cells were transferred. The total number of G9 T cells
recovered from both NOD and proinsulin transgenic recipients remained relatively low three days
after transfer, particularly in proinsulin transgenic recipients, even at the highest G9 T cell number
transferred (Fig. 3.6A). Importantly, the frequency of G9 T cells was similar relative to the number
of G9 T cells transferred for both strains (Fig. 3.6B). These data suggest that increasing the number
of G9 LN cells transferred does not alter the pattern of G9 T-cell recovery from proinsulin
transgenic recipients and this remains low relative other models (Nasreen et al., 2010; Steptoe et al.,
2007). Furthermore, the data indicate that the number of G9 T cells transferred in experiments
shown to date (that is, 5x106 G9 LN cells, representing approximately 2x106 CD8+Vβ6+ G9 T cells)
provides a valid insight into G9 T-cell dynamics.
The lower than expected recovery of G9 T cells from NOD and proinsulin transgenic
recipients prompted the question of whether unexpected immune-mediated rejection of transferred
G9 T cells due to genetic differences between donor and recipient strains might be occurring.
Therefore a pilot experiment was performed where CFSE-labelled G9 LN cells were transferred to
syngeneic G9 and non-transgenic NOD mice. When analysed three days after transfer, the
proportion of G9 T cells within the total CD8+ T cell pool was only slightly higher in syngeneic G9
compared to NOD recipients (Fig. 3.6C). However, T cell-mediated rejection of transferred G9 T
cells is unlikely to be ruled out by these data as three days is insufficient time for T cell-mediated
rejection to occur, although NK cell-mediated rejection can occur within this timeframe (Liu et al.,
2012). To examine the possibility of immune rejection further, CFSE-labelled G9 LN cells were
transferred
to
proinsulin
transgenic
(CD45.1+CD45.2-),
F1
proinsulin
transgenic
(CD45.1+CD45.2+), NOD (CD45.1+CD45.2-), F1 NOD (CD45.1+CD45.2+), NOD.CD45.2 (CD45.1CD45.2+), B16A (CD45.1+CD45.2-) and G9 (CD45.1+CD45.2-) mice and G9 T cells analysed
twenty one days later. These combinations of recipient mice allowed not only the interaction
between recipient strains, but also CD45 congenic status, to be examined. The percentage of G9 T
cells was low in spleen, sdLN and pLN of proinsulin transgenic recipients, regardless of the
congenic background (Fig. 3.6D-F). G9 T cells were somewhat increased in NOD recipients than
53
proinsulin transgenic recipients, but remained similar between the groups of NOD recipients,
regardless of CD45 congenic status (Fig. 3.6D-F). The frequency of G9 T cells was similar between
B16A and G9 recipients but was higher than NOD and proinsulin transgenic recipients (Fig. 3.6DF). From this it can be concluded that G9-T cell recovery from proinsulin transgenic or NOD
recipients is not influenced by CD45 congenic status. Similarly, reduced recovery of G9 T cells
from proinsulin transgenic relative to NOD is likewise not influenced by CD45 congenic status. If,
to account for reduced G9 T cell recovery in proinsulin transgenic recipients, immune-mediated
rejection of G9 T cells was occurring in this strain and not NOD recipients, then it might be
expected that the non-G9 cell population described in Figure 3.3 would, in addition to G9 T cells,
decrease at a faster rate than in NOD recipients. When analysed across all time points, non-G9 cells
displayed similar dynamics in both NOD and proinsulin transgenic recipients for spleen, sdLN and
pLN (Fig. 3.6G-I). This suggests the rapid reduction of G9 T cells in proinsulin transgenic
recipients is not due to non-specific immune rejection but most likely due to antigen-specific effects
of transgenic proinsulin expression.
54
20
spl
5 10 20
5 10 20
5 10 20
5 10 20
5 10 20
spl
sdLN
pLN
pLN
sdLN
+
0 .5
0 .0
5 10 20
spl
5 10 20
5 10 20
5 10 20
5 10 20
5 10 20
spl
sdLN
pLN
pLN
sdLN
0 .2
0 .1
0 .0
T
g
ro
In
s
T
g
F
1
N
O
D
F
N
O
1
D
N
O
D
.C
D
4
5
.2
B
1
6
A
G
9
-3
NOD
proIns Tg
400
200
0
5
10
0 .0
ro
In
s
T
g
ro
In
s
T
g
F
15
Days (post transfer)
20
1
N
O
D
F
N
O
1
D
N
O
D
.C
D
4
5
.2
B
1
6
A
G
9
spl
sdLN
pLN
spl
sdLN
pLN
G9
pLN
0 .8
0 .6
30
20
10
0
10
0 .0
ro
In
T
s
g
ro
In
s
T
g
F
1
N
15
Days (post transfer)
20
O
D
F
N
O
1
D
N
O
D
.C
D
4
5
.2
B
1
6
A
G
9
pLN
I
NOD
proIns Tg
5
0 .2
p
40
0
0 .4
p
sdLN
H
600
0
0 .2
p
Spleen
G
0 .4
p
Total non-G9/sdLN (x10-3)
In
s
p
Total non-G9/spleen (x10 )
0 .6
G 9 T c e lls /t o t a l C D 8
0 .3
G 9 T c e lls /t o t a l C D 8
0 .4
ro
0 .8
+
+
0 .5
0 .0
F
sdLN
c e lls ( % )
c e lls ( % )
0 .6
0 .5
NOD
E
Spleen
1 .0
+
D
1 .5
p ro In s T g
NOD
p r o In s T g
NOD
c e lls ( % )
1 .0
2 .0
20
Total non-G9/pLN (x10-3)
5 10 20
p
G 9 T c e lls /t o t a l C D 8
40
0
G 9 T c e lls /t o t a l C D 8
C
1 .5
+
T o t a l G 9 T c e lls ( x 1 0
-3
)
60
c e lls ( % )
B
G 9 T c e lls /t o t a l C D 8 c e lls ( % )
A
NOD
proIns Tg
15
10
5
0
0
5
10
15
20
Days (post transfer)
Figure 3.6 - G9 T cells are not rejected due to an allotypic difference between NOD and proinsulin
transgenic recipients.
(A, B) G9 LN cells (5x106, 10x106 or 20x106) were transferred to NOD and proinsulin transgenic
recipients. Three days later, the total number of G9 T cells (CD45.1+CD45.2-CD8+Vβ6+) and G9 T cells as
proportion of total CD8+ T cells was determined in spleen, sdLN and pLN using flow cytometry. (C)
CFSE-labelled G9 LN cells (5x106) were transferred to G9 and NOD recipients. Three days later, G9 T
cells (CFSE+CD8+Vβ6+) as a proportion of total CD8+ T cells was determined using flow cytometry. (D-F)
CFSE-labelled G9 LN cells (5x106) were transferred to G9 (CD45.1+CD45.2-), B16A (CD45.1+CD45.2-),
NOD.CD45.2 (CD45.1-CD45.2+), F1 NOD (CD45.1+CD45.2+), NOD (CD45.1+CD45.2-), F1 proinsulin
transgenic (CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2-) mice. Twenty one days later,
G9 T cells (CFSE+CD8+Vβ6+) as a proportion of total CD8+ T cells was determined in spleen (D), sdLN,
(E) and pLN (F) using flow cytometry. (G-I) G9 LN cells (5x106) were transferred to NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. At times shown, the ratio of G9 T
cells (CD45.1+CD8+Vβ6+) to non-G9 cells (CD45.1+CD8-) was determined using flow cytometry. Data are
from one (A-F) or pooled from at least two (G-I) independent experiments. n=3-4 at each time point (G-I).
Each point represents a single mouse. Error bars show S.D.
55
3.2.7 Transferred G9 T cells persist in the absence of cognate antigen
Using a congenic marker-based adoptive transfer setting, it was demonstrated that the
number of G9 T cells decreased rapidly after transfer to proinsulin transgenic recipients, while the
number of G9 T cells remained relatively stable after transfer to NOD recipients (Fig. 3.3).
Inclusion of B16A mice, which do not express the G9 T-cell determinant would be informative, but
these timecourse studies were not conducive to this as B16A mice express the same CD45 allotype
as G9 mice. If the assumption is made that G9 T cells are not expected to proliferate after transfer to
B16A recipients, CFSE-labelling might be useful to visualise transferred G9 T cells in B16A
recipients. To explore this, CFSE-labelled G9 LN cells were transferred to B16A, NOD and
proinsulin transgenic mice, and G9 T cells were enumerated at both three and twenty one days after
transfer. Three days after transfer, G9 T-cell recovery was similar between B16A, NOD and
proinsulin transgenic recipients in spleen and sdLN (Fig. 3.7A, B). The proportion of G9 T cells of
total CD8+ T cells was significantly lower in pLN from proinsulin transgenic compared to NOD
recipients (Fig. 3.7B). At twenty one days after transfer, despite a low number of replicates, G9 Tcell number appeared to be higher in spleen and sdLN of B16A recipients compared to NOD and
proinsulin transgenic recipients (Fig. 3.7C). When the proportion of G9 T cells was analysed, this
was significantly lower from both NOD and proinsulin transgenic recipients compared to B16A
recipients in spleen, sdLN and pLN (Fig. 3.7D). There is the possibility that this reduced G9 T-cell
frequency observed for NOD and proinsulin transgenic recipients may be due to loss of CFSElabelled G9 T cells as a result of proliferation. Although, G9 T cells were enumerated using a
congenic marker twenty one days after transfer to NOD and proinsulin transgenic recipients and
low recovery of G9 T cells was observed in both strains also (data not shown). A small population
of G9 T cells was detected in spleen, sdLN and pLN of NOD recipients, which was not present in
proinsulin transgenic recipients (Fig. 3.7D). Taken together, these data indicate G9 T cells persist in
B16A recipients, in the absence of their cognate antigen. On the other hand, expression of isletderived insulin in NOD mice appears to lead to the gradual deletion of transferred G9 T cells.
Notably, there were no G9 T cells detected in proinsulin transgenic recipients, indicating the rapid
deletion observed in timecourse studies (Fig. 3.3) is maintained.
56
B
40
*
+
30
1 .5
20
10
0
s
p
l
s
d
L
N
p
L
N
s
p
l
s
B16A
d
L
N
p
L
N
s
p
1 .0
G 9 T c e lls /t o t a l C D 8
T o t a l G 9 c e lls /o r g a n ( x 1 0
-3
)
A
c e lls ( % )
3 days after transfer
l
s
d
L
N
p
L
0 .5
0 .0
N
s
l
s
d
L
N
p
L
N
s
p
l
s
B16A
p ro In s T g
NOD
p
d
L
N
p
L
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
NOD
21 days after transfer
D
***
N
p
L
N
0 .0
NOD
p ro In s T g
s
s
B16A
B16A
N
L
L
s
d
p
l
l
p
N
s
p
N
L
L
d
p
s
N
s
L
N
s
d
L
l
p
p
l
s
N
N
L
L
d
p
N
N
p
L
d
s
d
0 .1
s
l
0 .2
L
p
0 .3
p
s
0 .4
l
0
0 .5
N
5
0 .6
p
10
0 .7
L
+
15
***
s
c e lls ( % )
20
G 9 T c e lls /t o t a l C D 8
T o t a l G 9 c e lls /o r g a n ( x 1 0
-3
)
C
NOD
p ro In s T g
Figure 3.7- High recovery of transferred G9 T cells in B16A recipients.
CFSE-labelled G9 LN cells (5x106) were transferred to B16A, NOD and proinsulin transgenic
recipients. Three (A, B) or twenty one (C, D) days later, the total number (A, C) of G9 T cells
(CFSE+CD8+Vβ6+) and G9 T cells as a proportion of total CD8+ T cells (B, D) in spleen, sdLN and
pLN were determined using flow cytometry. Data are pooled from four (B) or two (A, D) independent
or a single (C) experiment(s). Each point represents a single mouse. Error bars show S.D. One-way
ANOVA followed by Tukey post-test. (*p<0.05, ***p<0.001).
57
3.2.8 Down-regulation of TCR on G9 T cells after transfer to proinsulin transgenic recipients
It has been previously shown that rapid inactivation of OT-I T cells after transfer to mice in
which OVA expression is driven by an MHC class II promoter (MII.OVA) was due to TCR downregulation (Kenna et al., 2010). Since, in proinsulin transgenic mice, proinsulin expression is under
control of an MHC class II promoter, it is possible that TCR expression by G9 T cells is reduced
after transfer to proinsulin transgenic recipients. Initially, a pan-TCRβ mAb, recognising all
rearranged TCRβ chains, was used to determine TCR expression on G9 T cells. However, the
relative level of anti-TCRβ mAb staining was higher on G9 CD8+TCRβ+ T cells compared to nontransgenic NOD CD8+TCRβ+ T cells (Fig. 3.8A). When an anti-Vβ6 mAb was used, the level of
staining was not significantly different for G9 CD8+Vβ6+ T cells (MFI: 7529±2704 mean ±SD,
n=3) compared to non-transgenic NOD CD8+Vβ6+ T cells (MFI: 6815±3312, mean ±SD; n=3,
p=0.8) (Fig. 3.8A). The difference in mAb staining intensity might indicate the pan-TCRβ mAb had
an increased specificity for Vβ6 arrangements and the higher frequency of Vβ6-expressing CD8+ T
cells in G9 mice resulted in an overall increased staining intensity compared to non-transgenic NOD
T cells (Fig. 3.8B). To determine TCR expression by G9 T cells after transfer, CFSE-labelled G9
LN cells were transferred to B16A, NOD and proinsulin transgenic recipients and TCR expression
analysed three days later. G9 T cell TCR expression was determined relative to TCR expression of
host CD8+Vβ6+ T cells as an internal control for the staining technique. Three days after transfer,
there appeared to be an increase in TCRβ expression on G9 T cells relative to host CD8 +Vβ6+ T
cells from B16A and NOD recipients (Fig. 3.8C), however, this is probably due to the higher
staining intensity of transferred G9 T cells relative to the analogous host population (Fig. 3.8A, B).
Considering this, in proinsulin transgenic recipients, G9 T cells expressed significantly lower TCRβ
compared to B16A and NOD recipients (Fig. 3.8C). There was no significant difference in G9 T
cell TCR expression between B16A and NOD recipients (Fig. 3.8C).
Having accounted for the different levels of staining between anti-TCRβ and anti-Vβ6 mAb,
the anti-Vβ6 mAb was used in future experiments to determine TCR expression of transferred G9 T
cells. To determine the dynamics of TCR expression by G9 T cells, G9 LN cells were transferred to
congenic NOD and proinsulin transgenic recipients and TCR expression analysed at defined time
points. In proinsulin transgenic recipients, transferred G9 T cells displayed significantly lower Vβ6
expression than those from NOD recipients as early as one day after transfer and Vβ6 expression
continued to decrease over time on G9 T cells in proinsulin transgenic recipients (Fig. 3.8D, E).
Despite proliferation of G9 T cells in pLN of NOD recipients (Fig. 3.2), the level of Vβ6 expression
on G9 T cells from pLN of NOD recipients remained similar to that of the host CD8+Vβ6+ T-cell
population and was also similar to G9 T cells from spleen and sdLN of NOD recipients (Fig. 3.8E).
58
Since only minimal G9 T cell proliferation was observed in pLN of NOD recipients, it might
be possible that TCR down-regulation is occurring for proliferating G9 T cells but that this would
not be evident when non-proliferating G9 T cells were included in the analysis. To test this, TCRβ
expression of proliferating (CFSElo) G9 T cells was analysed relative to both the non-proliferating
(CFSEhi) G9 T cells and host CD8+ T cells. There was no significant difference between TCRβ
expression on proliferating and non-proliferating G9 T cells, or when this expression was compared
to host CD8+ T cells (Fig. 3.8F). This suggests rapid TCR down-regulation occurs only for G9 T
cells in response to transgenic proinsulin expression.
These data indicate TCR down-regulation occurs on G9 T cells after transfer to proinsulin
transgenic recipients. In contrast, TCR down-regulation was not observed for G9 T cells in response
to islet-derived insulin presentation in pLN of NOD recipients, suggesting rapid TCR downregulation is not merely a consequence of G9 T-cell activation.
59
A
NOD
CD8
Isotype
NOD
G9
SSC
G9
Vβ6
Vβ6
CD8
PanTCRβ
FSC
C
Pan-TCRβ
(clone: H57-957)
Anti-Vβ6
(clone:RR4-7)
D
Relative MFI
(G9/NOD)
1.51±0.01
1 .0
0 .5
0 .0
sp
l
sd
L
N
p
L
N
sp
B16A
E
1 .2
**
***
****
****
l
sd
L
N
p
L
NOD
N
sp
l
sd
L
N
p
L
N
p ro In s T g
****
N O D spl
N O D sdLN
0 .8
N O D pLN
p r o In s T g s p l
p r o In s T g s d L N
0 .4
p r o In s T g p L N
0 .0
0
1
2
3
4
5
6
7
8
D a y s (p o s t tra n s fe r)
F
Divided G9 T cells
relative to non-divided
G9
1 .5
1.15±0.13
NOD
proIns Tg
unstained
Vβ6
Vβ6
V 6 expression (relative to host)
mAb
TCR expression
(relative to host)
B
***
2 .0
Divided G9 T cells
relative to host CD8+ T
cells
Non-divided G9 T cells
relative to host CD8+ T
cells
1.08±0.29
1.36±0.24
1.30±0.30
Figure 3.8 – G9 T cell TCR down-regulation after transfer to proinsulin transgenic recipients.
(A) Vβ6 and pan-TCRβ expression on G9 or NOD LN cells was analysed using flow cytometry. Gating
strategy and relative histograms of CD8+Vβ6+ T cells. (B) Ratio of MFI of TCRβ staining on CD8+Vβ6+ T cells
or Vβ6 staining CD8+TCRβ+ T cells from G9 relative to NOD (mean±S.D.; n=3). (C) CFSE-labelled G9 LN
cells (5x106) were transferred to B16A, NOD and proinsulin transgenic mice. Three days later, TCRβ
expression on G9 T cells (CFSE+CD8+Vβ6+) relative to host CD8+Vβ6+ T cells in spleen, sdLN and pLN was
determined using flow cytometry. (D-E) G9 LN cells (5x106) or purified CD8+ G9 T cells (2x106) were
transferred to NOD and proinsulin transgenic mice. (D) Representative histogram of Vβ6 expression on G9 T
cells seven days after transfer. (E) At times shown, Vβ6 expression on G9 T cells (CD45.1+CD8+Vβ6+) relative
to host CD8+Vβ6+ T cells in spleen, sdLN and pLN was determined using flow cytometry. Data are pooled
from two (E), three (B) and four (C) independent experiments. Histograms are representative of at least two
independent experiments. Each point represents a single mouse. (E) n=4 (day 1, 2, 5, 7); n= 7-9 (day 3). (F)
Ratio of TCRβ expression by dividing (CFSElo), non-dividing (CFSEhi) G9 T cells and host CD8+ T cells
(CD8+CFSE-), n=6. Error bars show S.D. One-way ANOVA followed by Tukey post-test. (**p<0.01,
***p<0.001, ****p<0.0001: NOD pLN vs proIns Tg pLN).
60
3.2.9 InsB15-23-loaded tetramer binding of G9 T cells is impaired after transfer to proinsulin
transgenic recipients
Since the remaining, non-deleted G9 T cells in proinsulin transgenic recipients display
evidence of TCR down-regulation, it is plausible that these G9 T cells are functionally inactivated.
A reduction in binding of cognate peptide/MHC tetrameric complex (tetramer) by CD8+ T cells has
been associated with reduced effector function (Drake et al., 2005). To investigate this, InsB15-23loaded, H2-Kd-tetramer binding to transferred G9 T cells was analysed in B16A, NOD and
proinsulin transgenic recipients three days after transfer. Tetramer-binding G9 T cells as a
proportion of total CD8+ T cells were generally more frequent in B16A recipients (Fig. 3.9A).
Tetramer-binding G9 T cells were also relatively abundant in pLN of NOD recipients (Fig. 3.9A),
and this was similar to the proportion of CFSE-labelled G9 T cells of total CD8+ T cells in the same
site (Fig. 3.7B). Strikingly, few tetramer-binding G9 T cells were detected at any lymphoid site of
proinsulin transgenic recipients (Fig. 3.9A). In fact, this frequency was lower than the mean
proportion of control peptide (LLO91-99)-loaded tetramer-binding CD8+ T cells of total CD8+ T
cells, as indicated in Figure 3.9 by the dashed line. When the proportion of G9 T cells binding
tetramer was analysed, approximately 40-75% (depending on the site of collection) of G9 T cells
from B16A recipients bound tetramer (Fig. 3.9B). The proportion of G9 T cells that bound tetramer
from NOD recipients was slightly lower, but this was not statistically significant (Fig. 3.9B).
Interestingly, the proportion of tetramer-binding G9 T cells was higher in pLN than that of spleen or
sdLN in NOD recipients (Fig. 3.9B). In contrast, in proinsulin transgenic recipients, the proportion
of tetramer-binding G9 T cells was low and in some cases, below the level of control tetramer
binding (Fig. 3.9B). This indicates substantial loss of tetramer-binding capacity by G9 T cells after
transfer to proinsulin transgenic recipients, and when considered in relation to the TCR downregulation reported above (Fig. 3.8), this explains the low recovery of tetramer-binding G9 T cells
in proinsulin transgenic recipients.
The relationship between reduced tetramer binding and TCR expression was particularly
evident when the level of G9 T cell tetramer binding was shown relative to the level of G9 T cell
TCR expression. As such, G9 T cells in any lymphoid site from proinsulin transgenic recipients
generally clustered together, co-expressing low levels of Vβ6 and tetramer (Fig. 3.9C), although
one sample from pLN displayed a high level of tetramer expression despite low Vβ6 expression. In
contrast, G9 T cells from spleen, sdLN and pLN of B16A and NOD recipients clustered with high
expression of Vβ6 and tetramer (Fig. 3.9C) and were not distinguishable from each other.
Interestingly, some G9 T cells from pLN of B16A and NOD recipients displayed higher levels of
tetramer binding, despite similar Vβ6 expression compared to G9 T cells from spleen and sdLN of
B16A and NOD recipients (Fig. 3.9C). Overall, these data suggest G9 T cells transferred to
61
proinsulin transgenic mice are less able to recognise their cognate antigen in the context of H2-Kd,
which is probably due to TCR down-regulation, and may be indicative of functional inactivation.
62
**
*
1 .0
0 .8
G 9 T c e lls /t o t a l C D 8
+
c e lls ( % )
A
0 .6
0 .4
0 .2
0 .0
s
p
l
s
d
L
N
p
L
N
s
p
l
s
B16A
d
L
N
p
L
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
NOD
T e t r a m e r - b in d in g G 9 T c e lls ( % )
B
***
100
*
**
80
60
40
20
0
s
p
l
s
d
L
N
p
L
N
s
B16A
p
l
s
d
L
N
p
NOD
L
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
C
20000
p r o In s T g s p l
N O D spl
B 16A spl
M F I K : In s
15000
p r o In s T g s d L N
d
N O D sdLN
10000
B 16A sdLN
p r o In s T g p L N
5000
N O D pLN
B 16A pLN
0
0
1000
2000
3000
MFI V 6
Figure 3.9 – Transferred G9 T cells bind less InsB15-23-loaded, H2-Kd-tetramer in proinsulin
transgenic compared to B16A and NOD recipients.
CFSE-labelled G9 LN cells (5x106) were transferred to B16A, NOD and proinsulin transgenic
recipients. Three days later, the proportion of CFSE+Kd:Ins+CD8+Vβ6+ T cells of total CD8+ T cells
(A), the proportion of Kd:Ins-binding CFSE+CD8+Vβ6+ T cells (B) and the MFI of Vβ6 relative to the
MFI of Kd:Ins on CFSE+CD8+Vβ6+ T cells was determined using flow cytometry. Data are pooled
from two independent experiments. Each point represents a single mouse. Error bars show S.D. The
dashed line represents the limit of detection for the Kd:Ins tetramer and shows the mean of the
proportion of control (LLO91-99) tetramer-binding CD8+ T cells of total CD8+ T cells (A) or the
proportion of control tetramer-binding CFSE+CD8+Vβ6+ T cells (B). One-way ANOVA followed by
Tukey post-test. (*p<0.05, **p<0.01, ***p<0.001).
63
3.2.10 Transferred G9 T cells are predominantly CD44hiCD62L+ in proinsulin transgenic recipients
Since G9 T cells appear to undergo peripheral tolerance induction after transfer to proinsulin
transgenic mice, it is possible that these G9 T cells will have a different phenotype than G9 T cells
transferred to NOD recipients. To test this, G9 LN cells were transferred to NOD and proinsulin
transgenic recipients and the phenotype of G9 T cells analysed at defined time points. The
proportion of G9 T cells expressing the early activation marker, CD69, was higher in proinsulin
transgenic recipients than NOD recipients for spleen and sdLN as early as one day after transfer
(Fig. 3.10A). In NOD recipients, the proportion of CD69-expressing G9 T cells was higher in pLN
than spleen and sdLN (Fig. 3.10A). The proportion of G9 T cells expressing high levels of CD44
was significantly higher from proinsulin transgenic recipients than NOD recipients at one day after
transfer (Fig. 3.10B). In NOD recipients, G9 T cells expressed high levels of CD44 by three days
after transfer in pLN, while G9 T cells from spleen and sdLN remained consistently CD44 lo (Fig.
3.10B). The proportion of G9 T cells expressing CD44 and low levels of CD62L was relatively low
from both proinsulin transgenic and NOD recipients (Fig. 3.10C). At five and seven days after
transfer, the proportion of CD44hiCD62L- G9 T cells was higher in pLN compared to spleen of
NOD recipients (Fig. 3.10C). Most G9 T cells from proinsulin transgenic recipients were
CD44hiCD62L+ (Fig. 3.10D) The proportion of host CD8+Vβ6+ T cells expressing CD69 and CD44
remained low, while CD62L expression remained high over time in proinsulin transgenic and NOD
recipients (data not shown). At seven days after transfer, the total number of CD44hiCD62L- G9 T
cells in pLN of NOD recipients was significantly higher than in proinsulin transgenic recipients
(Fig. 3.10E). An increase in the number of CD44hiCD62L- G9 T cells was not observed in spleen of
NOD recipients (Fig. 3.10E). These data suggest G9 T cells are rapidly activated after transfer to
proinsulin transgenic recipients, and are predominantly CD44hiCD62L+ seven days after transfer. In
contrast, G9 T cells in spleen and sdLN from NOD recipients did not display an activated
phenotype. In pLN from NOD recipients, there was an accumulation of CD44hiCD62L- G9 T cells,
which might indicate development of effector function in response to islet-derived insulin
presentation.
64
A
B
**
**
***
****
**
G 9 T c e lls ( % )
G 9 T c e lls ( % )
100
80
60
C D 44
C D 69
+
hi
40
20
0
**
p r o In s T g s p l
80
p r o In s T g s d L N
p r o In s T g p L N
60
N O D spl
N O D sdLN
40
N O D pLN
20
1
2
3
4
5
6
7
0
1
D a y s (p o s t tra n s fe r)
2
3
4
5
6
7
D a y s (p o s t tra n s fe r)
D
100
80
**
*
*
100
***
****
*
**
p r o In s T g s p l
80
p r o In s T g s d L N
60
p r o In s T g p L N
N O D spl
C D 62L
+
-
60
G 9 T c e lls ( % )
C
40
hi
C D 44
20
0
0
1
2
3
4
5
6
7
C D 44
C D 6 2 L G 9 T c e lls ( % )
**
0
0
hi
****
***
100
40
N O D sdLN
N O D pLN
20
0
0
1
D a y s (p o s t tra n s fe r)
2
3
4
5
6
7
D a y s (p o s t tra n s fe r)
E
)
-3
1 .0
0 .5
0 .0
NOD
p r o In s T g
T o t a l G 9 T c e lls /s p le e n ( x 1 0
T o t a l G 9 T c e lls /p L N ( x 1 0
-3
)
*
1 .5
8
hi
-
hi
+
CD44 CD62L
6C
D44 CD62L
hi
-
hi
+
CD44 CD62L
n .s .
CD44 CD62L
4
2
0
NOD
p r o In s T g
Figure 3.10 – G9 T cells are predominantly CD44hiCD62L+ after transfer to proinsulin
transgenic recipients.
G9 LN cells (5x106) were transferred to NOD and proinsulin transgenic mice. At times shown, the
proportion of CD69+ (A), CD44+ (B), CD44hiCD62L- (C) and CD44hiCD62L+ (D) G9 T cells
(CD45.1+CD45.2-CD8+Vβ6+) in spleen, sdLN and pLN were determined using flow cytometry. (E)
Total number of CD44hiCD62L- and CD44hiCD62L+ G9 T cells in pLN (left) and spleen (right) seven
days after transfer, n=4. Data are pooled from two (day 1, 2, 5, 7 (A); (E)) or four (day 3 (A))
independent experiments. n=4 (day 1, 2, 5 and 7); n= 6-8 (day 3). Error bars show S.D. One-way
ANOVA followed by Tukey post-test. (A) (**p<0.01, ***p<0.001, ****p<0.0001: NOD spl vs proIns
Tg spl) (B) (**p<0.01, ***p<0.001, ****p<0.0001: NOD spl vs proIns Tg spl) (C) (*p<0.05,
**p<0.01: NOD spl vs NOD pLN) (D) (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001: NOD spl vs
proIns Tg spl) (E) (*p<0.05: CD44hiCD62L- NOD vs proIns Tg).
65
3.2.11 Optimisation of InsB15-23 peptide immunisation
As the data above are consistent with deletion of a portion of transferred G9 T cells in
proinsulin transgenic recipients and the retention of a small population of, possibly inactivated, G9
T cells, I next wanted to test the functional response of G9 T cells. To do so, it was first necessary
to develop an appropriate peptide immunisation regimen. To determine the optimal
antigen/adjuvant combination to induce an IFN-γ response from transferred G9 T cells, a pilot
experiment was performed. G9 LN cells were transferred to NOD mice and one day later, recipients
were immunised with defined InsB15-23/adjuvant combinations and IFN-γ production assessed five
days later using an ELISPOT assay. The highest number of IFN-γ spots was observed for NOD
recipients immunised with poly I:C/InsB15-23, with the least variability between replicates (Fig.
3.11A). Poly I:C was therefore employed for future peptide immunisations. Native InsB15-23 binds
poorly to H2-Kd (Wong et al., 2002) and it is possible that poor binding limits presentation to G9 T
cells. Increasing peptide/H2-Kd binding affinity might result in an increased G9 T cell response to
immunisation. An APL of InsB15-23, where glycine is replaced by valine at p9, InsB15-23 (23 Val),
has been shown to increase the cytotoxicity of the G9 T cell clone (Wong et al., 2002). To test
whether InsB15-23 (23 Val) immunisation increases the G9 T cell recall response, G9 LN cells were
transferred to NOD mice. One day later, recipients were immunised with defined doses of InsB15-23
(23 Val) or InsB15-23 (23 Gly) in combination with poly I:C and five days later, G9 T cells were
enumerated by flow cytometry and IFN-γ production was determined by ELISPOT. A trend toward
an increase in the number of IFN-γ spots was observed following InsB15-23 (23 Val) immunisation
compared to InsB15-23 (23 Gly) immunisation, however, more variability between replicates was
observed in NOD recipients immunised with InsB15-23 (23 Val) (Fig. 3.11B, left). The total number
of G9 T cells was increased only marginally after immunisation with 10µg InsB15-23 (23 Val)
compared to InsB15-23 (23 Gly) and 50µg InsB15-23 (23 Val), (Fig. 3.11B, middle). The proportion of
G9 T cells of total CD8+ T cells in recipients immunised with InsB15-23 (23 Gly) was similar
compared to 10 µg InsB15-23 (23 Val) (Fig. 3.11B, right). Given the reduced variability of the G9 Tcell recall response elicited by immunisation with 50µg InsB15-23 (23 Gly), future peptide
immunisations contained 50µg InsB15-23 (23 Gly).
66
 IFN-  spots (per 5x105 cells)
A
30
20
10
0
u
n
In
s
B
1
In
s
B
1
5
-2
3
(2
3
V
l)
5
0
u
g
0
u
n
In
im
s
B
m
1
5
.
-2
3
(2
3
G
ly
)
In
5
s
0
B
u
1
g
5
-2
3
(2
3
V
a
l)
In
1
s
0
B
u
1
5
g
-2
3
(2
3
V
a
l)
5
0
u
g
0 .0
g
a
u
g
0
u
g
0
u
1
g
l)
0
3
a
5
(2
V
l)
3
g
a
-2
u
V
5
0
u
5
1
3
)
l)
(2
ly
a
3
G
3
B
-2
m
(2
s
5
ti
3
In
1
s
0
3
l)
5
(2
a
-2
3
V
0 .2
V
-2
m
20
0 .4
3
B
5
ti
40
0 .6
5
s
1
s
0 .8
(2
In
)
.
60
1
.,
ly
.
1 .0
)
0
80
3
20
100
-2
40
3
o
-3
60
G
p
ly
+
)
+
G
B
im
3
p
80
(2
m
-2
(C
5
B
5
C
)
1
s
3
1
+
C
s
im
In
-2
B
3
I:
In
n
.,
5
-2
ly
T o t a l G 9 T c e lls /s p l ( x 1 0
 IFN-  spots (per 5x105 cells)
u
m
1
s
5
C
B
In
B
1
A
s
s
In
B
S
In
+
A
.
3
C
G
-2
I:
3
5
ly
m
s
1
o
(2
In
B
p
im
-2
5
A
n
1
il
3
B
u
u
s
Q
-2
In
+
3
5
-2
G
1
5
p
B
1
C
s
B
+
In
s
3
c e lls ( % )
In
.
m
+
im
n
G 9 T c e lls /t o t a l C D 8
u
Figure 3.11 – G9 T-cell expansion and IFN-γ production following InsB15-23 peptide
immunisation.
(A) G9 LN cells (5x106, CD45.1+CD45.2-) were transferred to NOD (CD45.1-CD45.2+) mice. One day
later, recipients were immunised with InsB15-23 (50µg) combined with CpG (15.5nM, s.c.), QuilA
(20µg, s.c.), poly I:C (100µg, i.p.), CASAC (200µl, s.c. flank), poly I:C+CpG (100µg+15.5nM, s.c.) or
left unimmunised. Five days later, IFN-γ production was determined in spleen using ELISPOT. (B) G9
LN cells (20x106, CD45.1+CD45.2-) were transferred to NOD (CD45.1-CD45.2+) mice. One day later,
recipients were immunised with 50µg InsB15-23 (23 Gly), 10µg InsB15-23 (23 Val) or 50µg InsB15-23 (23
Val) combined with 100µg poly I:C i.p. or left unimmunised. Five days later, IFN-γ production was
determined in spleen using ELISPOT (left panel), and the total number of G9 T cells
(CD45.1+CD45.2-CD8+Vβ6+) (middle) and G9 T cells as a proportion (right) of total CD8+ T cells
were determined in spleen using flow cytometry. Mean number of IFN-γ+ spots determined by
subtracting unstimulated wells from peptide-stimulated wells (B, left panel). Data are from one
individual experiment. Each point represents a single mouse. Error bars show S.D.
67
3.2.12 G9 T cells are unresponsive to InsB15-23 immunisation after transfer to proinsulin transgenic
mice
Since immunisation with InsB15-23 (23 Gly)/poly I:C expanded and induced IFN-γ
production by transferred G9 T cells in NOD recipients, I wanted to use this immunisation regimen
to investigate whether G9 T cells in proinsulin transgenic recipients were functionally inactive. To
do so, G9 LN cells were transferred to B16A, NOD and proinsulin transgenic mice and seven days
later recipients were immunised. Five days later, G9 T cells were enumerated using flow cytometry
and IFN-γ production analysed by ELISPOT. Since B16A mice express the same CD45.1 allotype
as transferred G9 T cells, InsB15-23-loaded, H2-Kd-tetramer was employed to visualise G9 T cells
after transfer to B16A mice. In B16A recipients, the total number of tetramer-binding G9 T cells
and their proportion of total CD8+ T cells remained unchanged following immunisation (Fig. 3.12A,
B). The total number of G9 T cells detected with a congenic marker was increased following
immunisation in NOD recipients (Fig. 3.12A). In contrast, the total number of G9 T cells and their
proportion of total CD8+ T cells was low in proinsulin transgenic recipients and this was not altered
by immunisation when detected with a congenic marker (Fig. 3.12A, B). To allow for direct
comparison between the three strains, tetramer-binding G9 T cells were also enumerated in
congenic NOD and proinsulin transgenic recipients. The pattern of the response was similar when
transferred G9 T cells were identified using tetramer (Fig. 3.12C, D) or a congenic marker (Fig.
3.12A, B). By ELISPOT, there was no significant difference in the number of IFN-γ-producing
cells following immunisation in neither B16A nor NOD recipients, however, a significant level of
IFN-γ production was observed in both B16A and NOD recipients (Fig. 3.12E, F). In contrast, IFNγ production was low in proinsulin transgenic recipients regardless of immunisation (Fig. 3.12F).
There were minimal IFN-γ-producing cells detected from mice which were not injected with G9 T
cells, despite immunisation (Fig. 3.12F), which indicates that in recipients of G9 T cells, IFN-γ
production can be attributed to the transferred G9 T cells. These data indicate G9 T cells are
frequent and produce IFN-γ in B16A and NOD recipients. This response, however, is subverted
after transfer to proinsulin transgenic mice, suggesting transferred G9 T cells are functionally
inactivate in proinsulin transgenic recipients.
68
u
u
/t
u
n
im
im
m
+
u
/t
u
B16A
NOD
im
+
/t
.
u
p ro In s T g
n
im
m
.
im
0 .0
.
m
NOD
B16A
p ro In s T g
D
c e lls ( % )
0 .4
0 .3
+
15
u
u
+
/t
im
m
.
u
m
im
.
m
im
.
u
/t
u
/t
+
im
m
.
u
n
im
m
.
im
m
.
u
/t
u
/t
im
+
m
.
u
NOD
n
im
m
.
im
m
.
p ro In s T g
**
50
n .s .
40
****
n .s .
30
20
n .s .
10
0
.
/t
proIns Tg
n
B16A
F
NOD
/t
p ro In s T g
E
B16A
0 .0
m
u
n
im
u
+
/t
/t
u
u
B16A
.
.
m
m
.
im
m
.
im
m
.
u
n
im
m
.
im
m
n
/t
+
im
u
u
NOD
/t
/t
u
u
.
im
m
+
im
/t
.
u
m
.
u
n
im
.
.
m
m
m
/t
+
im
im
u
B16A
/t
n
u
u
.
Im
m
/t
im
u
.
+
m
.
u
n
im
.
.
m
m
im
/t
+
im
m
u
/t
im
u
0 .1
.
0
0 .2
m
5
im
G 9 T c e lls /t o t a l C D 8
10
 IFN-  spots (per 5x105 cells)
T o t a l G 9 T c e lls /s p le e n ( x 1 0
-3
)
C
m
.
0 .0
.
.
.
.
m
m
m
.
m
m
im
+
im
/t
/t
im
u
u
.
n
im
m
im
.
+
n
m
u
u
im
.
.
m
/t
+
/t
im
.
/t
m
u
.
0
m
.
/t
m
m
im
u
.
n .s .
im
n
m
0 .1
u
u
im
0 .1
im
+
.
0 .2
n
/t
m
n .s .
****
n .s .
0 .2
u
u
im
5
0 .3
im
/t
0 .3
+
u
10
0 .4
/t
0
n .s .
u
5
*
15
0 .4
G 9 T c e lls /t o t a l C D 8
10
c e lls ( % )
n .s .
****
+
15
20
G 9 T c e lls /t o t a l C D 8
-3
T o t a l G 9 T c e lls /s p le e n ( x 1 0
)
-3
T o t a l G 9 T c e lls /s p le e n ( x 1 0
20
c e lls ( % )
B
)
A
NOD
p ro In s T g
Figure 3.12 - G9 T cells do not expand or produce IFN-γ following InsB15-23 immunisation in
proinsulin transgenic recipients.
G9 LN cells (5x106, CD45.1+CD45.2-) were transferred to B16A (CD45.1+CD45.2-), NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. Seven days later, recipients
were immunised with 50µg InsB15-23 and 100µg poly I:C i.p. or left unimmunised. Five days later, the
total number (A) or proportion (B) of Kd:Ins+CD8+Vβ6+ (for B16A recipients, left) or
CD45.1+CD45.2-CD8+Vβ6+ (for NOD and proinsulin transgenic recipients, right) G9 T cells and the
total number (C) or proportion (D) of Kd:Ins+CD8+Vβ6+ G9 T cells in spleen was determined using
flow cytometry. IFN-γ production in spleen was determined using ELISPOT. Representative well
images of InsB15-23-stimulated wells from immunised B16A (left), NOD (middle) and proinsulin
transgenic (right) recipients (E). Mean number of IFN-γ+ spots determined by subtracting unstimulated
wells from InsB15-23-stimulated wells (F). Data are pooled from three independent experiments. Each
point represents a single mouse. Error bars show S.D. The dashed line represents the detection limit for
Kd:Ins tetramer binding and shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T
cells. One-way ANOVA followed by Tukey post-test. (*p<0.05, **p<0.01, ****p<0.0001).
69
3.2.13 Unresponsiveness of transferred G9 T cells in proinsulin transgenic recipients is maintained
G9 T cells did not respond to InsB15-23 peptide immunisation seven days after transfer to
proinsulin transgenic recipients. It is possible that G9 T-cell inactivation observed in proinsulin
transgenic recipients is transient and may not be maintained. To test the longevity of G9 T-cell
inactivation, G9 LN cells were transferred to B16A, NOD and proinsulin transgenic mice and
twenty one days later recipients were immunised. Five days after immunisation, G9 T cells were
enumerated using flow cytometry and IFN-γ production analysed by ELISPOT. In B16A recipients,
the total number of tetramer-binding G9 T cells and their proportion of total CD8+ T cells were not
significantly altered following immunisation (Fig. 3.13A, B). The total number of G9 T cells and
their proportion of total CD8+ T cells were low and unchanged by immunisation when detected by a
congenic marker in NOD and proinsulin transgenic recipients (Fig. 3.13A, B). When transferred G9
T cells were detected using tetramer in all three strains, the pattern of the G9 T cell response was
similar as to when G9 T cells were detected with a congenic marker (Fig. 3.13A, B), in that the total
number and proportion of G9 T cells was significantly higher in B16A recipients than NOD and
proinsulin transgenic recipients (Fig. 3.13C, D). Although the total number of tetramer-binding G9
T cells (Fig. 3.13C) appeared marginally higher than CD45.1-gated G9 T cells (Fig. 3.13A), most
tetramer-binding G9 T cells in NOD and proinsulin transgenic recipients fell below the limit of
detection, which represents the level of control tetramer binding.
Systemic immunisation might increase the cellularity of spleens and since the total G9 T cell
number is based on the calculated spleen cellularity, a non-specific increase could skew the G9 T
cell total number. There was no significant difference in the cellularity of spleens from B16A, NOD
and proinsulin transgenic recipients following immunisation (Fig. 3.13E). Next, production of IFNγ was assessed as an indicator of effector function of transferred G9 T cells following
immunisation. There were a higher number of IFN-γ-producing spots in InsB15-23 peptide-stimulated
wells from immunised B16A recipients compared to immunised NOD and immunised proinsulin
transgenic recipients (Fig. 3.13F). IFN-γ production was higher in immunised B16A recipients
compared to both immunised NOD and immunised proinsulin transgenic recipients (Fig. 3.13G).
IFN-γ production was not significantly different when unimmunised and immunised mice were
compared within each recipient strain (Fig. 3.13G). These data confirm inactivation of G9 T cells
after transfer to proinsulin transgenic recipients is maintained. Interestingly, in NOD recipients, G9
T cells were unresponsive to InsB15-23 immunisation at this late time point despite IFN-γ production
observed when InsB15-23/poly I:C was administered seven days after transfer, suggesting G9 T cells
become unresponsive slowly over time in response to islet-derived insulin.
70
0
u /t
u n im m .
n .s .
0
im m .
u /t
u n im m .
)
T o t a l s p le e n c e lls ( x 1 0
im m .
c e lls ( % )
0 .3
+
c e lls ( % )
0 .4
0 .1
0 .0
u /t
u n im m .
0 .2
0 .1
n .s .
n .s .
0 .0
u /t
im m .
u n im m .
im m .
u /t
u n im m .
im m .
p ro In s T g
NOD
B16A
D
****
0 .5
**
0 .4
+
10
n .s .
n .s .
5
G 9 T c e lls /t o t a l C D 8
-3
T o t a l G 9 T c e lls /s p le e n ( x 1 0
u n im m .
0 .2
p ro In s T g
0
u /t
u n im m . im m .
u /t
u n im m . im m .
u n im m . im m .
p ro In s T g
NOD
n .s .
150
u /t
n .s .
n .s .
n .s .
0 .3
0 .2
n .s .
n .s .
0 .1
0 .0
u /t b 1 6 u n imbm
16
. im m .
B16A
F
n .s .
-6
)
u /t
n .s .
B16A
E
im m .
NOD
**
15
0 .3
+
n .s .
5
B16A
C
n .s .
0 .4
G 9 T c e lls /t o t a l C D 8
5
10
G 9 T c e lls /t o t a l C D 8
10
B
15
c e lls ( % )
)
-3
n .s .
15
T o ta l G 9 c e lls /s p le e n ( x 1 0
T o ta l G 9 c e lls /s p le e n ( x 1 0
-3
)
A
B16A
u /t
u n im m . im m .
NOD
NOD
u /t
u n im m . im m .
p ro In s T g
proIns Tg
100
50
0
u /t u n im m .im m .
u /t u n im m . im m .
u /t u n im m . im m .
B16A
NOD
p ro In s T g
G
 IFN- spots (per 5x105 cells)
****
40
***
30
20
n .s .
10
n .s .
0
u /t
u n im m .im m .
B16A
u /t u n im m . im m .
NOD
u /t u n im m . im m .
p ro In s T g
Figure 3.13 - G9 T cells are unresponsive to InsB15-23 immunisation in proinsulin transgenic
recipients.
G9 LN cells (5x106, CD45.1+CD45.2-) were transferred to B16A (CD45.1+CD45.2-), NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. Twenty one days later,
recipients were immunised with 50µg InsB15-23 and 100µg poly I:C i.p. or left unimmunised. Five days
later, the total number (A) and proportion (B) of Kd:Ins+CD8+Vβ6+ (B16A, left) or CD45.1+CD45.2CD8+Vβ6+ (NOD and proinsulin transgenic, right) G9 T cells of total CD8+ T cells, the total number
(C) and proportion (D) of Kd:Ins+CD8+Vβ6+ G9 T cells of total CD8+ T cells and the total number of
spleen cells (E) in spleen were determined using flow cytometry. IFN-γ production was determined
using ELISPOT. Representative well images of InsB15-23-stimulated wells from immunised B16A
(left), NOD (middle) and proinsulin transgenic (right) recipients (F). Mean number of IFN-γ+ spots
determined by subtracting unstimulated wells from InsB15-23-stimulated wells (G). Data are pooled
from two independent experiments. Each point represents a single mouse. Error bars show S.D. The
dashed line shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells. One-way
ANOVA followed by Tukey post-test. (** p<0.01, *** p<0.001, **** p<0.0001).
71
3.2.14 Transferred G9 T cells do not contribute to diabetes development
Male mice have been used as recipients for the adoptive transfer studies described so far. As
males exhibit reduced islet inflammation and diabetes incidence compared to female NOD mice,
this should minimise the effects of islet inflammation on tolerance outcomes in this adoptive
transfer setting. Therefore, transferring G9 T cells to male and female recipients may provide
insight into the effect of islet inflammation on this setting by directly comparing the population
dynamics of transferred G9 T cells in both male and female NOD and proinsulin transgenic
recipients. Considering the population dynamics of G9 T cells after transfer to NOD and proinsulin
transgenic recipients, the previous analyses have focused on early time points after transfer. To
investigate the influence of islet inflammation and to determine the involvement of transferred G9 T
cells in islet autoimmunity, G9 LN cells were transferred to both male and female congenic NOD
and proinsulin transgenic mice and analysis was performed eight weeks later, at the stage of severe
insulitis in NOD mice. G9 T cells were enumerated based on CD45 expression while host CD8+ T
cells recognising insulin and IGRP were analysed, as indicators of islet autoimmunity, using InsB1523-
and IGRP206-214- loaded , H2-Kd-tetramers, respectively, in pancreas and secondary lymphoid
organs of recipients by flow cytometry.
There was a trend toward a lower total number of pancreas-infiltrating leukocytes in
proinsulin transgenic compared to NOD recipients (Fig. 3.14A), as expected (French et al., 1997)
and this pattern was similar for the number of pancreas-infltrating CD8+ T cells, in that, this was
significantly higher in male NOD recipients than in male proinsulin transgenic recipients (Fig.
3.14B). The proportion of pancreas-infiltrating leukocytes of live-gated cells was higher in female
NOD compared to female proinsulin transgenic recipients, while this appeared to be higher in male
NOD compared to male proinsulin transgenic recipients, it was not significant (Fig. 3.14C). The
proportion of pancreas-infiltrating CD8+ T cells of total leukocytes was higher in female NOD
recipients than from female proinsulin transgenic recipients (Fig. 3.14D). The number of pancreasinfiltrating insulin-specific CD8+ T cells was low in all recipients (Fig. 3.14E, F), as indicated by
the level of control tetramer binding, represented here by the dashed line. This is consistent with the
insulin response waning with age (Wong et al., 1999), given the recipients were approximately 16weeks old at the time of analysis. More pancreas-infiltrating IGRP-specific CD8+ T cells were
observed in recipients than insulin-specific CD8+ T cells, but there were no significant differences
observed between male and female NOD or proinsulin transgenic recipients (Fig. 3.14G, H). It
should be noted that all recipients were negative for glycosuria at the time of cull for these analyses.
Next, the lymphoid site draining the pancreas was assessed to determine the involvement of
G9 T cell transfer on islet autoimmunity. The total number and proportion of G9 T cells of total
CD8+ T cells in pLN was low and there were no differences between male and female recipients of
72
either strain (Fig. 3.15A, B). The total number of insulin specific CD8+ T cells was lower from pLN
of male compared to female counterparts (Fig. 3.15C). The number of IGRP-specific CD8+ T cells
was significantly higher in female NOD recipients compared to male proinsulin transgenic
recipients (Fig. 3.15D). When tetramer-binding CD8+ T cells were expressed as a proportion of
total CD8+ T cells, there were no significant differences in insulin- or IGRP-specific CD8+ T cells
observed between groups (Fig. 3.15E, F). There were no significant differences in the proportion of
Vβ6+ insulin- or IGRP-specific CD8+ T cells in pLN from female or male proinsulin transgenic and
NOD recipients (Fig. 3.15G, H). In spleen, the number and proportion of G9 T cells was low and
similar from proinsulin transgenic and NOD recipients, for both male and female (Fig. 3.16A, B).
The total number and proportion of insulin-specific CD8+ T cells was similar between groups (Fig.
3.16C, E). The number of IGRP-specific CD8+ T cells was significantly higher in female NOD
recipients compared to male NOD and proinsulin transgenic recipients (Fig. 3.16D). The proportion
of IGRP-specific CD8+ T cells was significantly higher from female NOD recipients compared to
male proinsulin transgenic recipients (Fig. 3.16F). No significant differences were observed in the
proportion of Vβ6+ insulin- or IGRP-specific CD8+ T cells between groups (Fig. 3.16G, H). Finally,
sdLN were analysed in order to ascertain cell population dynamics in non-pancreatic draining LN
tissue following G9 T cell transfer. The total number and proportion of G9 T cells was low and
remained similar in proinsulin transgenic and NOD recipients, between both males and females
(Fig. 3.17A, B). The number and proportion of insulin- and IGRP-specific CD8+ T cells was similar
between groups (Fig. 3.17C-F). In peripheral LN, no significant differences were observed in the
proportion of Vβ6+ insulin- or IGRP-specific CD8+ T cells between proinsulin transgenic and NOD
recipients (Fig. 3.17G, H).
These data indicate that the rapid inactivation of G9 T cells after transfer to proinsulin
transgenic recipients is maintained over time and are in support of the finding of reduced G9 T-cell
frequency twenty one days after transfer to NOD recipients. From these findings, transferred G9 T
cells do not contribute to diabetes development by accumulating in recipients. It might be expected
that naïve G9 T cells would not contribute to diabetes development in an adoptive transfer setting as
G9 T cells require activation to become diabetogenic (Wong et al., 2009). Also, the full
diabetogenic potential of G9 T cells may not be observed where G9 T cells are transferred to
immune-replete NOD recipients, in which competition from host β cell-reactive T cells may
influence the diabetogenicity of transferred G9 T cells. To more accurately determine G9 T-cell
involvement in diabetes development, it would be important to include age-matched NOD mice not
injected with G9 T cells to attribute the observed outcomes directly to transferred G9 T cells rather
than predisposed autoimmune events which occur in wild type NOD mice.
73
A
B
)
-3
3
m a le
20
100
0
f e m a le
m a le
p ro In s T g
D
*
f e m a le
NOD
T c e lls ( % o f le u k o c y t e s )
f e m a le
NOD
15
10
*
f e m a le
p ro In s T g
*
15
10
5
+
5
m a le
20
CD8
In f ilt r a t in g le u k o c y t e s ( % )
200
T o ta l C D 8
1
m a le
0
m a le
f e m a le
NOD
m a le
f e m a le
0
m a le
p ro In s T g
f e m a le
NOD
m a le
f e m a le
p ro In s T g
F
6
+
K : In s
d
d
0
f e m a le
NOD
m a le
T c e lls )
m a le
p ro In s T g
30
+
K : IG R P
d
20
10
0
T c e lls )
T c e lls
40
+
50
f e m a le
NOD
H
)
-3
T c e lls ( x 1 0
5
f e m a le
+
d
T o t a l K : IG R P
10
0
m a le
(% o f to ta l C D 8
T o t a l K : In s
2
+
4
(% o f to ta l C D 8
T c e lls
15
+
T c e lls ( x 1 0
-3
)
E
G
300
+
2
0
C
*
400
T c e lls ( x 1 0
6
T o t a l le u k o c y t e s ( x 1 0 )
4
m a le
f e m a le
p ro In s T g
60
40
20
0
m a le
f e m a le
NOD
m a le
f e m a le
p r o In s T g
m a le
f e m a le
NOD
m a le
f e m a le
p ro In s T g
Figure 3.14 – Transfer of G9 T cells does not contribute to diabetes development.
G9 LN cells (5x106, CD45.1+CD45.2-) were transferred to male and female NOD and proinsulin
transgenic (CD45.1+CD45.2+) mice. Eight weeks after transfer, pancreata were collected and the total
number of infiltrating leukocytes (A), total number of infiltrating CD8+ T cells (B), leukocytes as a
proportion of live cells (C), CD8+ T cells as a proportion of total leukocytes (D), total number of
Kd:Ins+CD8+ T cells (E), total number of Kd:IGRP+CD8+ T cells (F), Kd:Ins+CD8+ T cells as a
proportion of total CD8+ T cells (G) and Kd:IGRP+CD8+ T cells as a proportion of total CD8+ T cells
(H) were analysed using flow cytometry. Data are pooled from two independent experiments. Each
point represents a single mouse. Error bars show S.D. The dashed line represents the detection limit for
Kd:Ins and Kd:IGRP tetramer binding and shows the mean frequency of Kd:LLO (control) tetramerbinding CD8+ T cells. One-way ANOVA followed by Tukey post-test. (*p<0.05).
74
T c e lls ( % )
B
)
A
T o t a l G 9 T c e lls /p L N ( x 1 0
-3
0 .1 0
+
0 .0 8
G 9 T c e lls /t o t a l C D 8
0 .0 6
0 .0 4
0 .0 2
0 .0 0
m a le
f e m a le
NOD
m a le
f e m a le
0 .0 1 0
0 .0 0 5
0 .0 0 0
m a le
p ro In s T g
D
*
2 .5
**
2
)
(x 1 0
d
1 .0
f e m a le
p ro In s T g
-3
T o t a l C D 8 K : IG R P
)
+
-3
1 .5
1
+
+
d
(x 1 0
m a le
3
+
2 .0
T o t a l C D 8 K : In s
f e m a le
NOD
T c e lls /p L N
C
T c e lls /p L N
0 .0 1 5
0 .5
0 .0
m a le
f e m a le
NOD
m a le
f e m a le
T c e lls )
m a le
f e m a le
p r o In s T g
0 .6
+
(% o f C D 8
T c e lls )
0 .3
+
0 .4
+
0 .2
C D 8 K : IG R P
(% o f C D 8
f e m a le
NOD
F
+
m a le
p ro In s T g
E
0 .1
0 .2
+
+
d
d
C D 8 K : In s
0
0 .0
m a le
f e m a le
NOD
m a le
f e m a le
0 .0
m a le
G
f e m a le
m a le
f e m a le
p ro In s T g
NOD
p ro In s T g
H
T c e lls )
+
0
+
+
20
75
50
d
d
+
+
C D 8 K : IG R P V b 6
40
( % o f K : IG R P
60
+
+
100
d
d
+
( % o f K : In s
C D 8 K : In s V b 6
T c e lls )
80
25
0
m a le
f e m a le
NOD
m a le
f e m a le
p ro In s T g
m a le
f e m a le
NOD
m a le
f e m a le
p ro In s T g
Figure 3.15 – Transferred G9 T cells do not accumulate in pLN of recipients.
G9 LN cells (5x106, CD45.1+CD45.2-) were transferred to male and female NOD and proinsulin
transgenic (CD45.1+CD45.2+) mice. Eight weeks later, the total number of G9 T cells
(CD45.1+CD45.2-CD8+Vβ6+) (A), G9 T cells as a proportion of total CD8+ T cells (B), total number of
Kd:Ins+CD8+ T cells (C), total number of Kd:IGRP+CD8+ T cells (D), Kd:Ins+CD8+ T cells as a
proportion of total CD8+ T cells (E), Kd:IGRP+CD8+ T cells as a proportion of total CD8+ T cells (F),
Vβ6+ Kd:Ins+CD8+ T cells as a proportion of total Kd:Ins+CD8+ T cells (G) and Vβ6+Kd:IGRP+CD8+ T
cells as a proportion of total Kd:IGRP+CD8+ T cells (H) in pLN were analysed using flow cytometry.
Data are pooled from two independent experiments. Each point represents a single mouse. Error bars
show S.D. The dashed line represents the detection limit for Kd:Ins and Kd:IGRP tetramer binding and
shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells. One-way ANOVA
followed by Tukey post-test. (* p<0.05, ** p<0.01).
75
B
1 .5
0 .0
m a le
f e m a le
T c e lls /s p l
)
50
0
m a le
+
K : IG R P
d
0 .2
(% o f to ta l C D 8
T c e lls
T c e lls )
F
1 .0
0 .5
m a le
H
d
+
5
0
T c e lls )
+
+
+
10
m a le
f e m a le
p ro In s T g
25
d
15
f e m a le
NOD
p ro In s T g
C D 8 K : IG R P V b 6
+
T c e lls )
*
1 .5
f e m a le
+
+
m a le
d
d
+
f e m a le
p r o In s T g
2 .0
( % o f K : IG R P
f e m a le
NOD
C D 8 K : In s V b 6
m a le
0 .0
m a le
( % o f K : In s
f e m a le
NOD
p ro In s T g
0 .0
G
p ro In s T g
100
f e m a le
0 .6
f e m a le
150
d
0
m a le
m a le
-3
+
)
10
f e m a le
f e m a le
(x 1 0
T c e lls /s p l
T c e lls )
m a le
T o t a l K : IG R P
20
0 .4
+
T c e lls
+
d
K : In s
0 .0 0 0
NOD
30
NOD
(% o f to ta l C D 8
0 .0 0 5
D
m a le
E
0 .0 1 0
p ro In s T g
40
-3
+
f e m a le
+
d
m a le
50
(x 1 0
T o t a l C D 8 K : In s
G 9 T c e lls /t o t a l C D 8
0 .5
NOD
C
0 .0 1 5
+
1 .0
0 .0 2 0
+
T o t a l G 9 T c e lls /s p le e n ( x 1 0
-3
)
T c e lls ( % )
A
20
15
10
5
0
m a le
f e m a le
NOD
m a le
f e m a le
p ro In s T g
m a le
f e m a le
NOD
m a le
f e m a le
p ro In s T g
Figure 3.16 – Transferred G9 T cells do not accumulate in spleen of recipients.
G9 LN cells (5x106, CD45.1+CD45.2-) were transferred to male and female NOD and proinsulin
transgenic (CD45.1+CD45.2+) mice. Eight weeks later, the total number of G9 T cells
(CD45.1+CD45.2-CD8+Vβ6+) (A), G9 T cells as a proportion of total CD8+ T cells (B), total number of
Kd:Ins+CD8+ T cells (C), total number of Kd:IGRP+CD8+ T cells (D), Kd:Ins+CD8+ T cells as a
proportion of total CD8+ T cells (E), Kd:IGRP+CD8+ T cells as a proportion of total CD8+ T cells (F),
Vβ6+ Kd:Ins+CD8+ T cells as a proportion of total Kd:Ins+CD8+ T cells (G) and Vβ6+ Kd:IGRP+CD8+
T cells as a proportion of total Kd:IGRP+CD8+ T cells (H) in spleen were analysed using flow
cytometry. Data are pooled from two independent experiments. Each point represents a single mouse.
Error bars show S.D. The dashed line represents the detection limit for Kd:Ins and Kd:IGRP tetramer
binding and shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells. One-way
ANOVA followed by Tukey post-test. (* p<0.05).
76
0 .1 0
0 .0 6
0 .0 4
0 .0 2
0 .0 0
f e m a le
m a le
4
0
f e m a le
NOD
m a le
T c e lls /s d L N
)
5
0
m a le
0 .0 5
T c e lls )
0 .4
0 .3
0 .2
0 .1
m a le
f e m a le
H
0
f e m a le
p ro In s T g
T c e lls )
+
60
40
d
d
+
5
( % o f K : IG R P
+
+
10
m a le
80
C D 8 K : IG R P V b 6
15
f e m a le
NOD
p ro In s T g
d
+
m a le
+
T c e lls )
9
6
3
0 .0
f e m a le
20
( % o f K : In s
+
f e m a le
p ro In s T g
+
T c e lls
+
d
K : IG R P
0 .1 0
(% o f to ta l C D 8
T c e lls )
+
T c e lls
+
d
(% o f to ta l C D 8
K : In s
0 .1 5
NOD
d
m a le
F
0 .2 0
m a le
+
f e m a le
NOD
p ro In s T g
0 .0 0
C D 8 K : In s V b 6
10
f e m a le
E
G
f e m a le
p ro In s T g
15
d
2
m a le
100
+
6
m a le
f e m a le
200
T o t a l K : IG R P
)
+
D
8
-3
0 .0 0 0
NOD
+
d
0 .0 0 1
p ro In s T g
10
(x 1 0
T o t a l C D 8 K : In s
f e m a le
0 .0 0 2
-3
T c e lls /s d L N
NOD
m a le
0 .0 0 3
(x 1 0
m a le
C
G 9 T c e lls /t o t a l C D 8
+
0 .0 8
T c e lls ( % )
B
)
T o t a l G 9 T c e lls /s d L N ( x 1 0
-3
A
20
0
m a le
f e m a le
NOD
m a le
f e m a le
p ro In s T g
m a le
f e m a le
NOD
m a le
f e m a le
p ro In s T g
Figure 3.17 – Transferred G9 T cells do not accumulate in sdLN of recipients.
G9 LN cells (5x106, CD45.1+CD45.2-) were transferred to male and female NOD and proinsulin
transgenic mice (CD45.1+CD45.2+). Eight weeks later, the total number of G9 T cells
(CD45.1+CD45.2-CD8+Vβ6+) (A), G9 T cells as a proportion of total CD8+ T cells (B), total number of
Kd:Ins+CD8+ T cells (C), total number of Kd:IGRP+CD8+ T cells (D), Kd:Ins+CD8+ T cells as a
proportion of total CD8+ T cells (E), Kd:IGRP+CD8+ T cells as a proportion of total CD8+ T cells (F),
Vβ6+ Kd:Ins+CD8+ T cells as a proportion of total Kd:Ins+CD8+ T cells (G) and Vβ6+ Kd:IGRP+CD8+
T cells as a proportion of total Kd:IGRP+CD8+ T cells (H) in sdLN were analysed using flow
cytometry. Data are pooled from two independent experiments. Each point represents a single mouse.
Error bars show S.D. The dashed line represents the detection limit for Kd:Ins tetramer binding and
shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells. One-way ANOVA
followed by Tukey post-test.
77
Table 3.1 Summary of Chapter 3
B16A
NOD
Proliferation
No
pLN only
Absolute number
n/a
Proportion (of total
CD8+ cells)
Unchanged
TCR downregulation
Tetramer binding
IFN-γ production
No
Slow decrease (by 21
days after transfer)
Accumulation in pLN
(at day 3) followed by
slow decrease in all
sites analysed (by day
21)
No
High
Yes
High
Yes (at day 7
following transfer)
Proinsulin
transgenic
Yes, in all sites
analysed
Rapid decrease (by 7
days after transfer)
Rapid decrease (by day
7)
Yes
Low
No
Figure 3.18 – Summary of main findings from Chapter 3.
This table represents summarised data from Chapter 3. ‘Proliferation’ refers to Figure 3.2, ‘Absolute
number’ and ‘proportion’ refer to Figure 3.3 and Figure 3.7, ‘TCR down-regulation’ refers to Figure
3.8, ‘Tetramer binding’ refers to Figure 3.9 and ‘IFN-γ production’ refers to Figure 3.12 and Figure
3.13.
78
3.3 Discussion
Targeted, transgenic antigen expression inactivates antigen-specific CD8+ T cells. Here, the
ability of transgenic proinsulin expression to inactivate insulin-specific CD8+ T cells on the
autoimmune-prone NOD background has been tested. The data suggest G9 T cells are deleted after
transfer to mice transgenically expressing proinsulin. Remaining, non-deleted G9 T cells were
infrequent, displayed both reduced TCR expression and InsB15-23-loaded tetramer binding, and
finally, were no longer able to respond to InsB15-23 immunisation, consistent with peripheral
tolerance. The typically naïve resting phenotype of TCR transgenic G9 T cells, even in pLN, is
consistent with a previous report showing CD8+ T cells from spleen of unimmunised G9 mice
display low CD44 and high CD62L expression (Wong et al., 2009). This might indicate lowavidity G9 T cells are ignorant of cognate antigen presentation in the periphery. Transgenic mice
expressing lymphocytic choriomeningitis virus (LCMV) glycoprotein under control of the RIP
develop autoimmune diabetes only after infection with LCMV (Ohashi et al., 1991; Oldstone et al.,
1991). Likewise, G9 mice rapidly develop autoimmune diabetes in an insulin-specific manner
following immunisation with InsB15-23 and CpG (Wong et al., 2009). When crossed to RIP.B7-1
mice, G9 mice also develop diabetes (Thomas et al., 2007), indicating a role of co-stimulation in G9
T cell activation. This is distinct from some other β cell-specific H2-Kd-restricted TCR transgenic
NOD mice. For instance, 8.3 TCR transgenic NOD mice, in which CD8+ T cells recognise IGRP206214,
spontaneously develop diabetes (Verdaguer et al., 1997) and AI4 TCR transgenic NOD mice, in
which CD8+ T cells recognise InsA14-20, demonstrate accelerated diabetes incidence (Graser et al.,
2000; Lamont et al., 2014). The low-affinity binding of InsB15-23/MHC class I might allow for
escape of G9-like T cells from thymic tolerance mechanisms into the periphery where activation
might occur. Due to their apparent primacy in islet autoimmune responses, it might be possible that
insulin-specific T cells are primed by non-self ligands (Bulek et al., 2012), and subsequent
specificities are primed directly as a result of insulin-specific T cell-mediated events such as β-cell
lysis and proinflammatory cytokine production. Therefore, G9 T cells might participate in early
diabetogenic responses in autoimmune diabetes. CD8+ T cells directed towards low-affinity MHC
class I-binding preproinsulin peptides are not only identifiable in human T1D (Abreu et al., 2012)
but are capable of lysing β cells with basal MHC class I expression (Bulek et al., 2012).
Proliferation of G9 T cells exclusively in pancreatic LN from NOD recipients observed here
is consistent with a previous study (Wong et al., 2009). Since B16A mice do not express the G9
determinant, incorporation of B16A recipients in adoptive transfer experiments in the present study
allows for the involvement of islet-derived antigen presentation to be dissected. Islet-derived insulin
presentation appeared to be localised to pLN of non-transgenic NOD mice as indicated by CFSE
dilution by G9 T cells only in that site (Fig. 3.2). It is important to establish the effect of islet79
derived insulin presentation in an adoptive transfer setting as this will be a confounding factor in
developing a T1D immunotherapy since varying levels of insulin presentation will occur in patients
depending on diabetes duration. Despite the fact that the majority of G9 T cells proliferated
extensively after transfer to proinsulin transgenic mice, the number of G9 T cells decreased over
time. Extensive proliferation preceding deletional tolerance has been demonstrated for 8.3 T cells in
response to IGRP expression by extrathymic Aire-expressing cells (Gardner et al., 2008) and for
OT-I T cells in MII.OVA mice (Kenna et al., 2010). In contrast, G9 T cells accumulated in pLN of
NOD recipients, even though the level of proliferation was not as substantial as observed in
proinsulin transgenic recipients. These different proliferative responses suggests G9 T cells become
activated and accumulate in response to islet-derived insulin presentation, however, in response to
transgenic proinsulin presentation, G9 T cells become activated but are rapidly deleted, emphasising
the tolerogenic nature of transgenic antigen expression.
The liver has been shown to be a reservoir for apoptotic T cells (Bertolino et al., 2001;
Huang et al., 1994). Kenna et al. did not observe a difference in the rate of apoptosis between OT-I
T cells transferred to mice expressing OVA in MHC class II+ APC or CD11chi DC, despite
significantly reduced OT-I T cell expansion in MII.OVA mice (Kenna et al., 2010). These data are
in agreeance with those from a pilot experiment in the present study suggesting transferred G9 T
cells did not accumulate to a greater extent in the liver of proinsulin transgenic recipients compared
to non-transgenic NOD recipients (Fig. 3.5). It is possible that the robustly tolerogenic signals
provided by antigen expression under control of an MHC class II promoter instruct the cognate
CD8+ T cell to undergo highly efficient apoptosis, precluding their identification in the liver.
Another possibility is that examining the liver of recipients of G9 T cells at a later time point may
show evidence of apoptotic G9 T cells. Further mechanistic studies in which known apoptotic
pathways are blocked in transferred G9 T cells are required to elucidate how G9 T cells are deleted
following transfer to proinsulin transgenic recipients.
OVA expression under control of an MHC class II promoter protected OVA expressing
islets from OT-I T cell-induced insulitis and autoimmune destruction (Kenna et al., 2010). In
contrast, in mice in which OVA expression is driven from a CD11c promoter, transferred OT-I T
cells were present in the islets and β-cell destruction was evident. Targeting tolerogenic antigen
expression to diverse MHC class II-expressing APC might be preferred in an immunotherapeutic
setting in order to prevent damage to antigen-expressing tissue by cognate T cells undergoing
tolerance induction. In the context of T1D, increased β-cell function is associated with a reduced
risk of complication development, therefore, bystander damage to β cells during T-cell tolerance
induction would be undesirable. Despite different antigens in each setting, deletion and induction of
unresponsiveness in G9 T cells after transfer to proinsulin transgenic recipients was similar to that
80
reported for OT-I T cells transferred to MII.OVA mice. Most G9 T cells from proinsulin transgenic
recipients were CD62Lhi, despite displaying an otherwise activated phenotype indicated by high
CD69 and CD44 expression. In contrast, in other studies, G9 T cells activated by InsB15-23/CpG
immunisation develop diabetogenicity and down-regulate CD62L expression (Wong et al., 2009).
In the present study, a higher proportion of CD62Llo G9 T cells were observed from pLN of NOD
recipients, suggesting islet-derived insulin presentation might elicit G9 T cell effector function.
Therefore, the tolerogenic nature of targeting transgenic antigen expression with an MHC class II
promoter is not restricted to high-affinity T cells recognising foreign antigens such as OVA but
translates to self-antigens such as insulin.
With regard to an alternative β-cell antigen, IGRP, expression under control of an MHC
class II promoter protects TCR transgenic 8.3 mice from autoimmune diabetes (Krishnamurthy et
al., 2012). Despite 8.3 T cells in the periphery becoming activated and proliferating in response to
transgenic IGRP expression, these 8.3 T cells produced less IFN-γ and TNF-α after IGRP206-214
stimulation in vitro. Significantly fewer 8.3 T cells were recovered from IGRP transgenic recipients
compared to NOD recipients, providing evidence for the ability of transgenic antigen expression to
delete diabetogenic CD8+ T cells. However, transgenic expression of IGRP in NOD mice does not
protect from autoimmune diabetes (Krishnamurthy et al., 2006), whereas transgenic expression of
proinsulin does protect NOD mice from diabetes development (French et al., 1997). Transgenic
IGRP expression is capable of protecting from diabetes development by peripheral T-cell deletion
when the diabetogenic CD8+ T cell population is comprised solely of 8.3-like T cells. Nevertheless,
deletion of 8.3-like T cells is insufficient to protect from diabetes development when insulin or
other specificities are present, such as in wild type NOD mice. Furthermore, where 8.3 T cells are
intermediate-avidity T cells, G9 T cells are a population of lower avidity T cells and low-avidity
autoreactive T cells escape peripheral tolerance mechanisms (Kawamura et al., 2008; Zehn &
Bevan, 2006), representing a population potentially diabetogenic T cells circulating in the periphery
which will inevitably encounter cognate antigen. Demonstration of tolerance induction in lowavidity autoreactive T cells is important physiologically as it might be expected that diabetogenic T
cells in the periphery would be low-avidity and tolerance outcomes may differ for low-avidity
compared to higher avidity T cells. Data from the present study suggest low-avidity β cell-specific
CD8+ T cells undergo peripheral tolerance induction after transfer to recipients expressing antigen
under the control of an MHC class II promoter.
Deletion of autoreactive AI4 T cells has been demonstrated using a superagonist mimotope,
MimA2, coupled to DC via the DEC-205 endocytic receptor (Mukhopadhaya et al., 2008). In
addition to a reduction in the proportion of AI4 T cells, the remaining AI4 T cells were unable to
produce IFN-γ after transfer of MimA2-coupled DC. Unresponsiveness was reversed when DC
81
maturation stimuli, poly I:C and an anti-CD40 mAb, were administered. Therefore, robust,
infallible antigen expression in which inflammatory stimuli would not alter tolerogenicity, would be
clinically beneficial in terms of T1D immunotherapies where islet-related and unrelated
inflammation would occur. In addition, the superagonist mimotope, MimA2, used in the DEC-205
model might impact the tolerance outcome of AI4 T cells by skewing toward higher affinity MHC
class I binding. Similarly, deletion of endogenous IGRP-specific CD8+ T cells was observed in
NOD mice treated with an IGRP mimotope, NRP-V7, coupled to DC (Mukherjee et al., 2013) .
These studies indicate tolerance induction is possible using mimotopes to target diabetogenic CD8+
T cells, although it would be pertinent to ensure mimotope presentation could not become
immunostimulatory, as this might increase the risk of autoimmune damage to antigen-expressing
tissue by activated T cells. Targeting antigen expression using cell-specific promoters, such as an
MHC class II promoter, may represent a means to ensure strictly tolerogenic antigen presentation,
and this could be tested by administration of inflammatory stimulus such as poly I:C and an antiCD40 mAb to determine whether this influenced G9 T-cell responses in proinsulin transgenic and
NOD recipients.
Persistent antigen exposure leads to deletion of antigen-specific CD8+ T cells (Redmond et
al., 2003) and transgenically expressed antigen is a robust method to ensure constant presentation of
potentially tolerogenic antigen. Deletion of autoreactive CD8+ T cells by cell-intrinsic mechanisms
involving the proapoptotic molecule, Bim, has been demonstrated in transgenic models with
expression of experimental antigens (Davey et al., 2002; Kenna et al., 2008). Clone 4 T cells
accumulate in pLN of mice expressing HA in islets equally in the presence or absence of cellextrinsic death molecules CD95 and TNF receptor (TNFR) 1 or TNFR2 (Redmond et al., 2008).
The absence of Bim, or the presence of Bcl-2, resulted in accumulation of clone 4 T cells.
Although, over-expression of Bcl-2 in clone 4 T cells led to more substantial insulitis in recipients,
suggesting different outcomes of CD8+ T-cell peripheral tolerance depending on whether a driver of
apoptosis (Bim) was absent or an apoptosis inhibitor (Bcl-2) was present. The different tolerance
outcome, in that the clone 4 T cells had increased effector function when Bcl-2 was over-expressed,
may provide insight into the rapid inactivation of antigen-specific CD8+ T cells in mice expressing
antigen driven by an MHC class II promoter. Since the absence of Bim had little effect on the
contraction of OT-I cells in MII.OVA recipients (Kenna et al., 2010), it is possible that Bcl-2sensitive pathways may be involved in the similar inactivation of G9 T cells after transfer to
proinsulin transgenic mice.
The present study and others have investigated T-cell tolerance using constitutive expression
transgenic systems targeting antigen expression to self tissue or APC and transfer of monoclonal
TCR transgenic CD8+ T cell repertoires. It is possible that the effect of transgenic antigen
82
expression might be different for endogenous self-reactive CD8+ T-cell populations compared to
transferred, monoclonal T cells. Peripheral tolerance induction has been investigated in a
polyclonal, endogenous CD8+ T-cell population using a tetracycline inducible and reversible
antigen expression system in which OVA expression is restricted to DC using a CD11c promoter
(Jellison et al., 2012). Naïve OVA-specific CD8+ T cells expanded following antigen induction and
finally became unresponsive, with high PD-1 expression. Cognate T-cell expansion preceding
tolerance induction is similar to the response of OT-I T cells transferred to mice expressing OVA
under control of a CD11c promoter (Kenna et al., 2008). These studies emphasise differing
population dynamics of antigen-specific CD8+ T cells depending on whether antigen is targeted to
DC or MHC class II+ APC. Deletion of OVA-specific CD8+ T cells was not evident in the inducible
antigen system used by Jellison et al., but this might have been due to recent thymus emigrants
becoming involved in the process thereby concealing a reduced number of OVA-specific CD8+ T
cells in the periphery.
Down-regulation of surface TCR acts to prevent hyper-responsiveness of activated T cells
by reducing receptor available to respond to antigen stimulation (Naramura et al., 2002; Shamim et
al., 2007). TCR down-regulation observed on G9 T cells after transfer to proinsulin transgenic mice
is less substantial than observed for OT-I T cells transferred to MII.OVA mice (Kenna et al., 2010).
TCR down-regulation observed in the present study might be associated with peripheral tolerance
or merely the result of recent activation of G9 T cells. In support of the former, G9 T cells from
pLN of NOD recipients expressed equivalent TCR relative to an analogous host T-cell population,
despite an activated phenotype of the transferred G9 T cells. Therefore, G9 T-cell TCR downregulation was observed only after encounter with the determinant derived from transgenic
proinsulin rather than from islet-derived insulin, suggesting G9 T-cell TCR down-regulation is
consistent with peripheral tolerance induction. A study of peptide-mediated tolerance involving
clone 4 T cells and HA peptide showed slight surface TCR down-regulation on clone 4 T cells after
administration of high-dose HA peptide (Redmond et al., 2005). Using a tumour antigen-specific
TCR transgenic model, it has been demonstrated that high-affinity MHC binding peptides induced
more substantial surface TCR down-regulation than lower affinity peptides (Cai et al., 1997). As G9
T cells recognise the low-affinity H-2Kd-binding InsB15-23, the level of TCR down-regulation
observed in the present study may be expected. In the absence of G9 determinant, in B16A
recipients, TCR expression on transferred G9 T cells remained high and similar to an analogous
host population (Fig. 3.8). This validates TCR expression observed for G9 T cells transferred to
NOD mice and suggests islet-derived insulin expression is contained within pLN as G9 T cells from
non-pancreatic draining sites of NOD recipients expressed a similar and high level of TCR
83
compared to G9 T cells from B16A recipients. Finally, these data indicate TCR down-regulation by
G9 T cells is a unique response to transgenically expressed proinsulin.
The mechanism of inactivation of autoreactive CD8+ T cells in the periphery may involve
inhibitory receptors such as PD-1. OT-I T cells lacking PD-1 are not tolerised, as indicated by OT-I
T cell accumulation and enhanced IFN-γ production, after transfer to recipients in which pancreatic
islets express high levels of OVA (Keir et al., 2007). Furthermore, PD-1 knockout OT-I T cells
effectively transfer autoimmune diabetes to recipients expressing OVA in islets, whereas, wild type
OT-I T cells do not under the experimental conditions used. This suggests, in this OT-I/RIP-OVA
tolerance system, PD-1 plays an important role in maintaining peripheral T-cell tolerance to tissuetargeted antigen expression. Likewise, accelerated onset and complete penetrance of diabetes by 15
weeks of age was observed for PD-1-deficient NOD mice, and islet-infiltrating CD8+ T cells
displayed a strong IFN-γ bias (Wang et al., 2005). The PD-1/PD-L1 inhibitory pathway has also
been implicated in a model of induced autoimmune diabetes in which RIP.B7-1 mice were
administered preproinsulin-encoding plasmid DNA to induce preproinsulin-specific CD8+ T cells
(Schuster et al., 2013). The relative affinity of epitopes derived from the preproinsulin-encoding
DNA influenced co-stimulation dependence in that, CD8+ T cells specific for the lower affinity
MHC-binding epitope, preproinsulin A12-21, became diabetogenic in the absence of transgenic B7-1
and PD-1 expression. This suggests differential maintenance of autoreactive T cells in response to
peptides of varying MHC-binding affinities. In the present study, G9 T cell expansion after transfer
to proinsulin transgenic mice is limited, which might indicate the PD-1 pathway is involved.
Although, targeting antigen expression using an MHC class II promoter leads to rapid inactivation
of cognate CD8+ T cells and this differs compared to tolerance in mice expressing antigen in islets
as in the study by Schuster et al. Blocking PD-1 or PD-L1 in proinsulin transgenic mice would
allow the involvement of this pathway in peripheral tolerance of G9 T cells to be elucidated,
however, unpublished data from the host laboratory using a blocking approach suggests this
inhibitory pathway is not involved in rapid inactivation of OT-I T cells in MII.OVA mice.
Taken together, the data in this chapter suggest transgenic expression of proinsulin induces
peripheral tolerance in insulin-specific CD8+ T cells in autoimmune-prone NOD mice. G9 T cells
underwent extensive proliferation prior to deletion after transfer to proinsulin transgenic recipients.
The remaining, non-deleted G9 T cells displayed TCR down-regulation and no longer effectively
recognised cognate peptide presented by H2-Kd. Finally, residual G9 T cells were unresponsive to
InsB15-23 immunisation. The response of insulin-specific CD8+ T cells to transgenic proinsulin
expression was consistent with other studies investigating CD8+ T cell responses to antigen
expressed under an MHC class II promoter, suggesting the robust mechanism of peripheral
tolerance in this setting is translatable from foreign, model antigens to self-antigens. While the
84
mechanism by which rapid inactivation of antigen-specific CD8+ T cells occurs is still unclear,
these findings suggest an immunotherapeutic approach based on such an antigen targeting strategy
may be feasible.
85
3.4. Future directions
An interesting observation in this chapter was the reduced frequency of G9 T cells from NOD
recipients twenty one days after transfer. Determining G9 T cell population dynamics at time points
in line with this finding may help elucidate whether this slow deletion occurs solely through G9 T
cell traffic to pLN in response to islet-derived insulin presentation or whether the insulin
determinant is presented elsewhere. Further dissection of the mechanism of this response to isletderived insulin compared to that which occurs in response to transgenic proinsulin expression will
provide novel insight into different tolerance pathways of insulin-specific CD8+ T cells. The
relative contributions of molecules shown to be important for T-cell tolerance in other models,
including Bim/Bcl-2, PD-1/PD-L1 and CD95, could be tested in G9 T-cell tolerance using knockout
mice and blocking antibodies.
86
Chapter 4
Transgenic expression of proinsulin rapidly inactivates
insulin-specific CD8+ memory T cells
87
4.1 Introduction
The previous chapter demonstrated peripheral tolerance induction in naïve insulin-specific
CD8+ T cells. It is important to demonstrate the inactivation of naïve CD8 + T cells in the context of
autoimmune diabetes as defects in central tolerance mechanisms allow for continual escape of naïve
islet-reactive T cells from the thymus. Antigen-experienced CD8+ effector T cells and Tmem in the
periphery must also be inactivated for a T1D immunotherapy to be effective. β-cell antigen-specific
CD8+ Tmem perpetuate islet autoimmunity and in the case of islet transplantation, can mediate graft
rejection (Alkemade et al., 2013; Mottram et al., 2002). Since G9 T cells are diabetogenic only after
activation, it is therefore pertinent to address tolerance induction in activated G9 T cells, which is
the focus of this chapter.
It had long been thought Tmem were impervious to cell-intrinsic peripheral tolerance
mechanisms due to their terminally differentiated nature and reduced requirement for co-stimulation
(Croft et al., 1994; Dengler & Pober, 2000; London et al., 2000). The resistance of Tmem to costimulation blockade in the context of transplantation tolerance has been widely explored (Chen et
al., 2004; Valujskikh et al., 2002). In humans, CD8+ T cells can display reduced expression of the
co-stimulatory molecule, CD28, after differentiation to a memory phenotype (Wills et al., 2002).
CD28- CD8+ Tmem proliferated more extensively in the presence of IL-15 and an anti-CTLA-4
mAb, compared to CD28+ CD8+ Tmem (Traitanon et al., 2014). Also, autoreactive Tmem in
experimental autoimmune encephalomyelitis (EAE), a mouse model of MS, were less susceptible to
apoptosis possibly due to expression of the anti-apoptotic molecule, Bcl-2, and as a result of
increased IFN-γ production these Tmem were more pathogenic that their effector counterparts
(Elyaman et al., 2008). These points lend credence to the argument that CD8+ Tmem may less
readily undergo peripheral tolerance compared to naïve CD8+ T cells. Additionally, naïve and
effector T cells are dependent on IL-2, and IL-2 production is modulated by immunosuppressants
such as rapamycin (Feuerstein et al., 1995; Gonzalez et al., 2001b). In contrast, CD8+ Tmem are
less dependent on IL-2 for homeostasis and survival (Araki et al., 2009a). Moreover, CD8+ Tmem
proliferation is less effectively controlled by CD4+Foxp3+ Treg compared to that for naïve and
effector T cells (Afzali et al., 2011; Yang et al., 2007). Defective IL-2 signalling has been
associated with reduced maintenance of Foxp3 expression by Treg in T1D patients (Long et al.,
2010) and in mice, Foxp3 instability for CD4+ T cells resulted in their development to diabetogenic
memory T cells (Zhou et al., 2009). It is possible that genetic defects associated with predisposition
to autoimmune diabetes in humans and NOD mice may result in otherwise suppressive Treg
developing pathogenicity in the context of islet inflammation. β cell-specific Tmem, therefore,
represent a barrier to effective T1D immunotherapy. To overcome this barrier, transgenic
expression of antigen represents a means to potentially inactivate Tmem responses.
88
In transgenic mice expressing HA under control of the RIP , transferred clone 4 Tmem underwent
deletional tolerance (Kreuwel et al., 2002). It was suggested that the observed tolerance occurred in
response to cross-presented HA determinants in the pLN of HA transgenic mice. Using a cell
subset-specific promoter, that is, a CD11c promoter, it has been shown that targeting OVA
expression to CD11chi DC resulted in peripheral tolerance induction in transferred OT-I Tmem
(Kenna et al., 2008). In this setting, a transient phase of effector function was exhibited by the OT-I
Tmem prior to tolerance induction. However, when OVA expression was enforced under an MHC
class II promoter, OT-I Tmem were rapidly inactivated (Kenna et al., 2010). Similarly, tolerogenic
DC have been shown to delete CD8+ Tmem. Expansion of CD8+ Tmem was observed after a single
stimulation with tolerogenic DC and further stimulation was required for subsequent deletion of the
CD8+ Tmem (Kleijwegt et al., 2013). In light of the tolerance outcomes achieved in previous
studies, promoter (or APC) choice will therefore require consideration for effective termination of β
cell-specific CD8+ Tmem responses. Since naïve and memory CD8+ T cells differ in their response
to antigen, it is important to address how G9 Tmem respond to islet-derived insulin expression as
well as transgenically-expressed proinsulin. It is possible that G9 Tmem might undergo more robust
expansion and develop effector function in the pancreatic LN of NOD recipients compared to naïve
G9 T cells.
While most studies have focused on model antigens such as OVA and HA, here CD8+ T-cell
responses to the important β-cell antigen, insulin, are investigated. In the NOD mouse, immune
responses directed toward insulin are required for diabetes development (French et al., 1997;
Nakayama et al., 2005). Insulin-reactive Tmem have been identified in the periphery of patients
with T1D (Luce et al., 2011) and in the NOD mouse (Chee et al., 2014; Diz et al., 2012). More
importantly, insulin-specific CD8+ T cells have been detected in the pLN and islets in human T1D
(Coppieters et al., 2012). Considering the patients’ disease duration and the site of analysis in the
study by Coppieters et al., it would be expected that the islet-infiltrating CD8+ T cells were of an
effector/memory phenotype. Homeostatic proliferation following immunosuppressive conditioning
and islet transplantation has been demonstrated in T1D patients to increase the frequency of
GAD65-specific CD8+ Tmem (Monti et al., 2008). While T1D patients and healthy controls alike
where shown to possess GAD- and insulin-reactive naïve T cells, only in T1D patients were GADand insulin- memory T cell responses observed (Monti et al., 2007).Tolerogenic DC have been
shown to induce tolerance in insulin-specific effector/memory CD4+ T cells from patients with
T1D, and the tolerised T cells were characterised by high IL-10 and low IFN-γ expression (TorresAguilar et al., 2010).
In this chapter, I examine whether transgenic expression of proinsulin induces peripheral tolerance
in antigen-experienced, insulin-specific memory CD8+ T cells. The response of G9 Tmem to
89
transgenic proinsulin and islet-derived insulin presentation was explored in an adoptive transfer
setting. G9 T cells were cultured to generate a mixed population of insulin-specific TCM-like and
TEM-like CD8+ T cells. In recipient mice presenting transgene-derived, islet-derived or no antigen,
the proliferation and population dynamics of G9 Tmem were assessed as well as the level of TCR
expression. G9 Tmem expansion and IFN-γ production were determined following InsB15-23
immunisation at an early and a late time point after transfer.
90
4.2 Results
4.2.1 G9 T cells are activated in vitro by InsB15-23
To determine the optimal InsB15-23 (23 Gly) concentration for in vitro activation of G9 T cells
and to explore the affinity of G9 T cells relative to another TCR transgenic CD8+ T cell, G9 and
OT-I LN cells were stimulated overnight with titrated concentrations of InsB15-23 (23 Gly), InsB15-23
(23 Val) or OVA257-264 peptide. CD69 and TCR expression on G9 and OT-I T cells were
determined using flow cytometry. CD69 is a marker of T cell activation and is up-regulated within
hours of activation (Sancho et al., 2005), while TCR down-regulation is a consequence of antigen
stimulation of T cells (Luton et al., 1994; Minami et al., 1987; Valitutti et al., 1995). Unstimulated
G9 T cells (CD8+Vβ6+) did not express CD69 (Fig. 4.1A). Upon addition of 1x10-5M InsB15-23 (23
Gly) and overnight culture, the majority of G9 T cells expressed CD69 (Fig. 4.1A). When titrated,
there was a dose-dependent response to peptide stimulation, with the highest proportion of CD69expressing G9 T cells after stimulation with 1x10-5M InsB15-23 (23 Gly) (Fig. 4.1B). Stimulation
with a two-log-fold lower concentration of InsB15-23 (23 Val) resulted in the maximum proportion of
CD69-expressing G9 T cells (Fig. 4.1B). Most OT-I T cells expressed CD69 with a five-log-fold
lower OVA257-264 concentration compared to InsB15-23 (23 Gly) (Fig. 4.1B). The proportion of
CD69-expressing OT-I T cells remained low after stimulation with InsB15-23 (23 Gly) or InsB15-23
(23 Val) (Fig. 4.1B). Down-regulation of Vβ6 was observed on G9 T cells at a log-fold higher
concentration of InsB15-23 (23 Gly) than was required for maximal CD69 up-regulation (Fig. 4.1C).
Reduced Vα2 expression observed on OT-I T cells after stimulation with 1x10-9M or higher
OVA257-264 (Fig. 4.1C). The concentration of OVA257-264 at which reduced TCR expression on OT-I
T cells was observed was five log-fold lower than the concentration of InsB15-23 (23 Gly) required
for G9 T cell TCR down-regulation (Fig. 4.1C). In this in vitro model of T-cell activation, a higher
concentration of cognate peptide is required to maximally activate G9 T cells compared to OT-I T
cells. Stimulation with an InsB15-23 APL, which is known to bind more strongly to MHC class I,
resulted in maximal G9-T cell activation at a lower peptide concentration than the native InsB15-23.
The lower InsB15-23 APL concentration requirement was not due to non-specific T-cell stimulation,
as OT-I T cells were not activated upon stimulation with InsB15-23 (23 Val). For TCR downregulation, this occurred at a peptide concentration approximately one-log-fold higher than required
for maximal CD69 up-regulation, suggesting a threshold of T-cell activation must be reached prior
to TCR down-regulation.
91
CD8α
PI
Unstimulated
SSC
A
Vβ6
FSC
isotype
CD8α
FSC
CD8α
PI
SSC
Peptidestimulated
FSC
CD69
T C R e x p r e s s io n
80
60
40
20
0
-2 4
-2 2
-2 0
-1 8
-1 6
-1 4
-1 2
-1 0
n
[ P e p t id e ] ( M x 1 0 )
-8
-6
-4
-2
( r e la t iv e t o u n s t im u la t e d )
100
+
T c e lls )
isotype
C
+
(o f C D 8
Vβ6
FSC
B
% CD69
CD69
1 .5
G 9 + In s B 1 5 -2 3 ( 2 3 G ly )
G 9 + In s B 1 5 -2 3 ( 2 3 V a l)
O T -I + O V A 2 5 7 -2 6 4
1 .0
O T - I + In s B 1 5 -2 3 ( 2 3 V a l)
O T - I + In s B 1 5 -2 3 ( 2 3 G ly )
0 .5
0 .0
-2 4
-2 2
-2 0
-1 8
-1 6
-1 4
-1 2
-1 0
-8
-6
-4
-2
n
[ P e p t id e ] ( M x 1 0 )
Figure 4.1 –In vitro stimulation with InsB15-23 (23 Gly) activates G9 T cells in a dose-dependent
manner.
LN cells from G9 or OT-I mice (2.5x105) were cultured with titrated concentrations of InsB15-23 (23
Gly), InsB15-23 (23 Val) or OVA257-264 peptide. The following day, cells were harvested and CD69 and
TCR expression on G9 (CD8+Vβ6+) or OT-I (CD8+Vα2+) T cells was determined using flow
cytometry. (A) Representative FACS plots of unstimulated G9 T cells (upper) and InsB15-23 (23 Gly) at
1x10-5M stimulated G9 T cells (lower). (B) Proportion of CD69+ T cells (from total CD8+Vβ6+ or
CD8+Vα2+ cells) and (C) relative TCR expression. Data are pooled from three independent
experiments, n=3. Error bars show S. D.
92
4.2.2 In vitro-cultured G9 Tmem display a phenotype resembling TCM and TEM
When OT-I T cells are cultured in vitro in the presence of OVA257-264 peptide, a mixed
population of cells phenotypically resembling TCM and TEM are generated (Kenna et al., 2008;
Kenna et al., 2010). To determine the phenotype of cultured G9 T cells, G9 LN and spleen cells
were cultured with InsB15-23 (23 Gly) and IL-2 for 3 days. On day 3, cells were washed and cultured
in IL-15 for a further 2 days as performed for OT-I T cell in the cited studies. The optimal
conditions required for G9 Tmem generation were found to be 10µg/ml of InsB15-23 (23 Gly) with
10% FCS (data not shown). Prior to culture, G9 T cells did not express CD69 and were
predominantly CD44lo and CD62Lhi (Fig. 4.2A), consistent with a naïve T-cell phenotype. One day
after culture, G9 T cells were predominantly CD69hi and CD44hi and a proportion of G9 T cells
were CD62Llo (Fig. 4.2A). At day 3, G9 T cells were predominantly CD69loCD44hi and there were
two populations of CD62Lhi and CD62Llo G9 T cells (Fig. 4.2A). On the final day of culture, after 2
days exposure to IL-15 and absence of InsB15-23, G9 T cells were predominantly CD69-CD44hi with
two populations of CD62Lhi and CD62Llo G9 T cells (Fig. 4.2A). Over the culture period, G9 T
cells enriched to comprise approximately 90% of live cells (Fig. 4.2B). At day 5, there was a higher
proportion of TCM-like G9 T cells (approximately 60%) compared to TEM-like G9 T cells
(approximately 30%, Fig. 4.2B). The forward scatter (FSC) profile of InsB15-23-stimulated G9 T
cells at day 3 was higher than naïve G9 T cells at day 0 (Fig. 4.2C). At day 5 of culture, G9 T cells
display an intermediate FSC profile (Fig 4.2C), indicative of T-cell activation. At day 3, a small
proportion of G9 T cells displayed increased granzyme B and perforin expression compared to
naïve G9 T cells (Fig. 4.3A). By day 5, only a small proportion of G9 T cells expressed granzyme B
(Fig. 4.3A) A small proportion of G9 T cells expressed IL-7Rα and/or killer-cell lectin like receptor
G1 (KLRG1) at day 3 compared to naïve G9 T cells (Fig. 4.3B). By day 5, a small proportion of G9
T cells expressed of IL-7Rα but there were almost no KLRG1-expressing G9 T cells (Fig. 4.3B).
The majority of G9 T cells expressed lymphocyte activation gene 3 (LAG-3) at day 3 and a
proportion of LAG-3+ G9 T cells also expressed CTLA-4 (Fig. 4.3C). By day 5, the proportion of
G9 T cells expressing CTLA-4 was similar to that observed for naïve G9 T cells, although a
proportion of G9 T cells expressing LAG-3 remained (Fig. 4.3C). CD5 expression was observed for
naïve G9 T cells relative to isotype control staining and at day 3 a small proportion of G9 T cells
appeared to further up-regulate CD5 (Fig. 4.3D), however, this might be an artifact of increased Tcell size as a result of activation. By day 5, CD5 expression was similar to naïve G9 T cells (Fig.
3D). Unlike CD5, naïve G9 T cells did not express PD-1, yet, the majority of G9 T cells expressed
PD-1 by day 3 and even at day 5, a proportion of G9 T cells displayed PD-1 expression (Fig. 4.3D).
Taken together, these data suggest G9 T cells undergo activation and effector differentiation
during the first 3 days of culture, as indicated by up-regulation of T-cell activation and co-inhibitory
93
markers including CD69, CD44, PD-1 and LAG-3 as well as an increase in FSC profile. An
additional 2 days of culture in the absence of InsB15-23 and in the presence of IL-15, resulted in a
mixed population of G9 TEM and TCM. Therefore, in vitro-culture is a valid method for generating a
mixed population of insulin-specific CD8+ T cells that phenotypically represent TEM and TCM.
94
A
B
CD62L
Isotype
Relevant
mAb
CD8α
Day 0
CD44
SSC
CD69
Day 5
Vβ6
FSC
isotype
Day 3
CD44
CD8α
Day 2
isotype
Day 1
CD69
+
Enrichment
(%)
CD69
(%)
89.95±5.17
3.67±2.43
+
CD44 CD62L
(%)
33.5±14.45
C
CD62L
lo
+
CD44 CD62Lhi
(%)
61.67±14.97
Day 0
Day 3
Day 5
Day 4
Day 5
FSC
Figure 4.2 – In vitro-culture of G9 T cells results in a mixed population of TEM and TCM.
Pooled LN and spleen cells from G9 mice were cultured for 3 days with 10µg/ml InsB15-23 and IL-2.
On day 3, cells were washed and re-cultured with IL-15 for a further 2 days. A sample of cells was
taken each day throughout the culture period and expression of surface markers was determined using
flow cytometry. (A) CD69, CD44 and CD62L expression (dashed line) on CD8+Vβ6+ T cells
compared to isotype controls (solid line). (B) Gating strategy of G9 Tmem at day 5 of culture.
Enrichment (CD8+Vβ6+ T cells as a proportion of total leukocytes) of and expression of CD69, CD44
and CD62L on G9 T cells at day 5 of culture. Table shows mean±S.D. Data are pooled from six
independent experiments. (C) Histogram shows FSC profile of lymphocyte gate. FACS plots are
representative of at least two independent experiments.
95
A
B
Day 3
Day 5
Day 0
Day 3
Day 5
IL-7Rα
Granzyme B
Day 0
KLRG1
isotype
isotype
Perforin
isotype
isotype
D
CD5
CTLA-4
C
PD-1
isotype
isotype
LAG-3
isotype
isotype
Figure 4.3 – G9 T cells up-regulate PD-1, CTLA-4 and LAG-3 following in vitro-culture with
InsB15-23 and IL-2.
Pooled LN and spleen cells from G9 mice were cultured for 3 days with 10µg/ml InsB15-23 and IL-2.
On day 3, cells were washed and re-cultured with IL-15 for a further 2 days. A sample of cells was
taken at defined time points throughout the culture period and expression of surface and intracellular
markers were determined using flow cytometry. Expression of indicated markers and relevant isotype
controls for fixed and permeabilised (A, C) or unfixed (B, D) G9 T cells. FACS plots are
representative of two independent experiments.
96
4.2.3 G9 Tmem produce IFN-γ
In vitro-cultured G9 Tmem display a phenotype typical of TCM and TEM and it might be
expected that G9 Tmem exert effector function. To test this, G9 Tmem were transferred to B16A
recipients and IFN-γ production was analysed five days later. The concentration of InsB15-23 (23
Gly) and InsB15-23 (23 Val) used to re-stimulate recipient splenoctyes in these ELISPOT assays was
titrated to determine the optimal concentration for detection of IFN-γ in future assays. As expected,
there were no IFN-γ-producing cells in spleen of B16A mice that were not injected with G9 Tmem
(Fig. 4.4A), indicating IFN-γ production is attributable to transferred G9 Tmem. In the absence of
in vitro InsB15-23 (23 Gly) re-stimulation, few IFN-γ-producing cells were observed in ELISPOT
cultures established from B16A recipients of G9 Tmem and IFN-γ production was increased
following re-stimulation with InsB15-23 (23 Gly) (Fig. 4.4A). In B16A recipients of G9 Tmem, the
number of IFN-γ-producing cells from spleen was significantly increased by re-stimulation with
1µg/ml and 10µg/ml of InsB15-23 (23 Gly) (Fig. 4.4B). Re-stimulation with any defined
concentration of InsB15-23 (23 Val) resulted in similar numbers of IFN-γ-producing cells in B16A
recipients of G9 Tmem (Fig. 4.4B). Re-stimulation of splenocytes from B16A G9 Tmem recipients
with 10µg/ml of InsB15-23 (23 Gly) resulted in a high number of IFN-γ-producing cells, with less
variability observed between replicates than for other groups (Fig. 4.4B). Therefore, 10µg/ml of
InsB15-23 (23 Gly) was used to re-stimulate splenocytes in future ELISPSOT assays. The high
number of G9 Tmem transferred in these experiments ensured adequate resolution of the G9 Tmem
IFN-γ response. In later experiments, the number of G9 Tmem transferred was reduced to 2x106 G9
Tmem, in line with the total number of naïve G9 T cells transferred in adoptive transfer studies in
Chapter 3. Overall, these data demonstrate G9 Tmem produce IFN-γ after transfer to B16A mice,
indicating effector function and considering transferred G9 Tmem were not primed in vivo by
immunisation, G9 Tmem have been primed during culture and, at this time point, produce basal
levels of IFN-γ.
97
A
No G9 Tmem
+ InsB15-23 stim.
400
 IFN-  spots (per 5x105 cells)
B
0
.1
u
g
/m
l
1
B
u
G9 Tmem
unstim.
G9 Tmem
+ InsB15-23 stim.
**
300
n .s .
200
100
0
1
g
5
-2
/m
1
3
l
0
(2
B
u
3
1
g
5
G
-2
/m
ly
3
l
)
(2
B
0
G
3
1
.1
5
-2
u
g
)
ly
3
(2
/m
0
3
l
.5
G
B
u
1
g
ly
5
)
-2
/m
3
l
1
(2
B
u
3
1
g
5
V
-2
/m
1
a
3
l
0
l)
(2
B
u
3
1
g
5
V
-2
/m
a
3
l
l)
(2
B
1
0
5
-2
.1
l)
a
V
3
3
u
(2
g
V
3
/m
l
1
a
B
u
n o G 9 T m e m tra n s fe r
1
g
l)
5
-2
/m
1
3
l
0
(2
B
u
3
1
g
5
G
-2
/m
ly
3
l
)
(2
B
0
G
3
1
.1
5
-2
u
g
)
ly
3
(2
/m
0
3
l
.5
G
B
u
1
g
ly
5
)
-2
/m
3
l
1
(2
B
u
3
1
g
5
V
-2
/m
1
a
3
l
0
l)
(2
B
u
3
1
g
5
V
-2
/m
a
3
l
l)
(2
B
3
1
5
V
-2
a
3
l)
(2
3
V
a
l)
G 9 T m e m tra n s fe r
Figure 4.4 - G9 Tmem produce IFN-γ after transfer to B16A recipients.
G9 Tmem (10x106) were transferred to B16A mice and five days later, IFN-γ production in spleen was
determined using ELISPOT. (A) Representative well images showing IFN-γ+ spots from a mouse not
injected with G9 Tmem (left), a mouse injected with G9 Tmem without re-stimulation (middle) and
the same recipient with re-stimulation (10µg/ml InsB15-23 (23 Gly); right). (B) The mean number of
spots in peptide-stimulated wells minus the mean number of spots in unstimulated wells (delta value)
per 5x105 spleen cells. The concentration of peptide used for re-stimulation is shown on the x-axis.
Data are pooled from two independent experiments. Each point represents a single mouse. Error bars
show S.D. One-way ANOVA followed by Tukey post-test. (**p<0.01).
98
4.2.4 G9 Tmem proliferate extensively after transfer to proinsulin transgenic recipients
When transferred to proinsulin transgenic mice, naïve G9 T cells proliferated in response to
transgenic proinsulin expression and this proliferation preceded tolerance induction. To determine
whether G9 Tmem also proliferate in response to transgenic proinsulin expression, CFSE-labelled
G9 Tmem were transferred to B16A, NOD and proinsulin transgenic and G9 Tmem CFSE dilution
was assessed three days later. As expected, there was no CFSE dilution observed for G9 Tmem
from spleen, sdLN or pLN of B16A recipients (Fig. 4.5A). Similarly, CFSE dilution was not
observed for G9 Tmem from spleen or sdLN of NOD recipients (Fig. 4.5A). A small proportion of
G9 Tmem from pLN of NOD recipients showed evidence of proliferation (Fig. 4.5A). In contrast,
the majority of G9 Tmem transferred to proinsulin transgenic recipients exhibited extensive CFSE
dilution in spleen, sdLN and pLN samples (Fig. 4.5A). The proportion of CFSElo G9 T cells was
significantly higher from proinsulin transgenic compared to B16A and NOD recipients for spleen,
sdLN and pLN (Fig. 4.5B). The proportion of CFSElo G9 T cells from pLN was not significantly
different to spleen or sdLN in NOD recipients (Fig. 4.5B). The proportion of proliferating naïve G9
T cells observed in pLN from NOD recipients was higher than that observed for G9 Tmem. When
proliferation index was calculated, the extent of G9 Tmem division from proinsulin transgenic
recipients was significantly higher than from NOD recipients in sdLN and pLN (Fig. 4.5C). Taken
together, these data indicate G9 Tmem, similarly to naïve G9 T cells, proliferate extensively after
transfer to proinsulin transgenic recipients. Interestingly, G9 Tmem proliferated to a lesser extent
compared to naïve G9 T cells in response to islet-derived insulin presentation in NOD recipients.
Despite differences in the response of naïve and memory G9 T cells to islet-derived insulin,
extensive proliferation of G9 Tmem in response to transgenic proinsulin expression might precede
tolerance induction.
99
A
B16A
proIns Tg
NOD
pLN
sdLN
Spleen
CFSE
B
C
P r o life r a tio n in d e x
5
100
50
CFSE
lo
G 9 T c e lls ( % )
****
*
4
3
2
1
0
0
s
p
l
s
d
L
N
p
L
B16A
N
s
p
l
s
d
L
N
p
L
NOD
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
s
p
l
s
d
L
N
p
L
B16A
N
s
p
l
s
d
L
N
p
L
NOD
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
Figure 4.5 - G9 Tmem proliferate extensively in response to presentation of transgene-derived,
but not islet-derived, (pro)insulin.
CFSE-labelled G9 Tmem (2x106) were transferred to B16A, NOD and proinsulin transgenic mice.
Three days later, G9 Tmem (CFSE+CD8+Vβ6+) CFSE dilution was determined using flow cytometry.
(A) FACS plots are representative of at least five individual mice. (B) CFSElo G9 Tmem as a
proportion of total G9 Tmem. (C) Proliferation index of G9 Tmem. Data are pooled from two
independent experiments. Each point represents a single mouse. Error bars show S. D. One-way
ANOVA followed by Tukey post-test. (*p<0.05, ****p<0.0001).
100
4.2.5 G9 Tmem are rapidly deleted after transfer to proinsulin transgenic recipients
Since G9 Tmem proliferate extensively in proinsulin transgenic recipients, it was next
important to investigate the population dynamics of transferred G9 Tmem. Firstly, CFSE-labelled
G9 Tmem were transferred to B16A, NOD and proinsulin transgenic recipients and G9 Tmem were
enumerated three days later. The proportion of G9 Tmem (CFSE+Vβ6+CD8+) of total CD8+ T cells
was significantly higher from B16A compared to proinsulin transgenic recipients in spleen and pLN
(Fig. 4.6A). No difference in G9 Tmem as a proportion of total CD8+ T cells was observed from
proinsulin transgenic and NOD recipients in pLN (Fig. 4.6A), this lack of G9 Tmem accumulation
is consistent with minimal G9 Tmem proliferation observed in pLN from NOD recipients (Fig. 4.5).
The proportion of G9 Tmem was higher from B16A compared to NOD recipients in spleen (Fig.
4.6A).
Secondly, to track the population dynamics of transferred G9 Tmem over time, G9 Tmem
(CD45.1+CD45.2-) were transferred to NOD (CD45.1+CD45.2+) and proinsulin transgenic
(CD45.1+CD45.2+) recipients and G9 Tmem were enumerated at defined time points. To control for
background mAb staining, mice of each genotype that were not injected with G9 Tmem were
analysed alongside test mice in each experiment. Mice that were not injected with G9 Tmem
consistently displayed negligible CD45.1+CD45.2-Vβ6+CD8+-gated events. Despite proliferation of
G9 Tmem in proinsulin transgenic recipients (Fig. 4.5), the total number of G9 Tmem in spleens of
these recipients decreased over time (Fig. 4.6B). The total number of G9 Tmem from proinsulin
transgenic recipients was significantly lower than NOD recipients in spleen at five and seven days
after transfer (Fig. 4.6B). The total number of G9 Tmem remained low across time in both sdLN
and pLN from proinsulin transgenic recipients (Fig. 4.6B). When the number of G9 Tmem was
represented as a proportion of total CD8+ T cells, there was a reduction in the proportion of G9
Tmem of total CD8+ T cells over time in spleen of both proinsulin transgenic and NOD recipients
(Fig. 4.6C). This reduction in the proportion of G9 T cells of total CD8+ T cells in spleen was not
observed when naïve G9 T cells were transferred (Fig. 3.3) and may be due to different trafficking
properties of Tmem as a result of differential expression of CD62L. Seven days after transfer the
proportion of G9 Tmem of total CD8+ T cells from proinsulin transgenic recipients was
significantly lower than NOD recipients in spleen, sdLN and pLN (Fig. 4.6C). The proportion of G9
Tmem of total CD8+ T cells remained low in sdLN and pLN of proinsulin transgenic recipients
(Fig. 4.6C). Similarly to when CFSE was used to track G9 Tmem, no increase in the proportion of
G9 Tmem of total CD8+ T cells was observed in pLN of NOD recipients at any time point when G9
Tmem were identified based on differential CD45 expression (Fig. 4.6C). The population dynamics
of G9 Tmem after transfer to proinsulin transgenic and NOD recipients indicate G9 Tmem are
rapidly deleted in mice transgenically expressing proinsulin.
101
1 .4
****
1 .2
1 .0
G 9 T m e m /t o t a l C D 8
+
T c e lls ( % )
A
0 .8
0 .6
**
0 .4
n .s .
0 .2
0 .0
s
p
l
s
d
L
N
p
L
N
s
p
l
s
B16A
d
L
N
p
L
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
NOD
B
T o ta l G 9 T m e m (x 1 0
-3
)
150
****
****
100
N O D spl
50
p r o In s T g s p l
N O D sdLN
10
p r o In s T g s d L N
N O D pLN
5
p r o In s T g p L N
0
0
2
4
6
8
G 9 T m e m /t o t a l C D 8
+
C
T c e lls ( % )
D a y s (p o s t tra n s fe r)
2 .0
1 .5
****
****
1 .0
N O D spl
p r o In s T g s p l
0 .5
0 .5
N O D sdLN
0 .4
p r o In s T g s d L N
0 .3
N O D pLN
0 .2
p r o In s T g p L N
0 .1
0 .0
0
2
4
6
8
D a y s (p o s t tra n s fe r)
Figure 4.6- G9 Tmem are rapidly deleted after transfer to proinsulin transgenic recipients.
(A) CFSE-labelled G9 Tmem (2x106) were transferred to B16A, NOD and proinsulin transgenic mice.
Three days later, G9 Tmem (CFSE+CD8+Vβ6+) as a proportion of total CD8+ T cells in spleen, sdLN
and pLN was determined using flow cytometry. (B, C) G9 Tmem (2x106, CD45.1+CD45.2-) were
transferred to NOD (CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. At times
shown, total number of G9 Tmem (CD45.1+CD45.2-CD8+Vβ6+) (B) and G9 Tmem as a proportion of
total CD8+ T cells (C) in spleen, sdLN and pLN were determined using flow cytometry. Data are
pooled from at least two independent experiments at each time point. Each point represents a single
mouse. (B, C) n=5 (day 1 and 5); n=6 (day 2 and 3); n= 9 (day 7). Error bars show S.D. One-way
ANOVA followed by Tukey post-test. (**p<0.01, ****p<0.0001: NOD spl vs proIns Tg spl (B, C)).
102
4.2.6 G9 Tmem persist in the absence of their cognate antigen
In Chapter 3 it was demonstrated that the frequency of naïve G9 T cells slowly decreased
over time after transfer to NOD recipients, possibly reflecting deletion induced by islet-derived
insulin. It is possible that G9 Tmem, due to their more differentiated phenotype, may be less
responsive to islet-derived insulin presentation and as such the frequency of G9 Tmem might
remain stable after transfer to NOD recipients. To test this, CFSE-labelled G9 Tmem were
transferred to B16A, NOD and proinsulin transgenic recipients and G9 Tmem were enumerated
three weeks later. The total number of G9 Tmem from B16A recipients was significantly higher
than NOD and proinsulin transgenic recipients for both spleen and sdLN (Fig. 4.7A). The total
number of G9 Tmem was similar between B16A, NOD and proinsulin transgenic recipients in pLN
(Fig. 4.7A). There was no significant difference in the total number of G9 Tmem between NOD and
proinsulin transgenic recipients in spleen, sdLN and pLN (Fig 4.7A). When the proportion of G9
Tmem of total CD8+ T cells was analysed, there was a higher proportion of G9 Tmem from B16A
compared to NOD and proinsulin transgenic recipients in spleen, sdLN and pLN (Fig. 4.7B). There
was a trend toward an increase in the proportion of G9 Tmem recovered from NOD compared to
proinsulin transgenic recipients, although, this was not significant (Fig. 4.7B). The total cellularity
of spleens from B16A recipients was significantly lower than both NOD and proinsulin transgenic
recipients (Fig. 4.7C). There was also a trend toward lower cellularity of sdLN from B16A than
NOD and proinsulin transgenic recipients (Fig. 4.7C), yet, this was not significant. This highlights
the higher number of G9 Tmem in B16A recipients compared to NOD and proinsulin transgenic
recipients despite reduced cellularity of organs in B16A recipients. These data indicate G9 Tmem
are slowly deleted after transfer to NOD recipients. Since the frequency of G9 Tmem remained
stable in B16A recipients, the slow deletion of G9 Tmem in NOD recipients might be due to isletderived insulin presentation. An important observation here is the complete absence of G9 Tmem
three weeks after transfer to proinsulin transgenic recipients, further supporting the notion of rapid
deletion of G9 Tmem in response to transgenic proinsulin expression.
103
A
B
T c e lls ( % )
****
15
+
10
5
0
s
p
l
s
d
L
N
p
L
N
s
p
l
s
B16A
d
L
N
p
L
N
s
p
l
s
d
L
N
p
N
L
****
0 .6
0 .4
0 .2
0 .0
s
p
l
s
p ro In s T g
NOD
G 9 T m e m /t o t a l C D 8
T o t a l G 9 T m e m /o r g a n ( x 1 0
-3
)
****
d
L
N
p
B16A
L
N
s
p
l
s
d
L
N
p
L
NOD
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
100
**
T o t a l c e ll n u m b e r ( x 1 0
-6
)
C
80
60
40
n .s .
20
n .s .
T
g
D
A
O
s
In
ro
p
6
1
B
s
In
sdLN
N
T
g
D
A
ro
6
O
N
1
B
s
In
ro
p
s p le e n
p
g
D
T
O
N
B
1
6
A
0
pLN
Figure 4.7- Transferred G9 Tmem persist in B16A recipients.
CFSE-labelled G9 Tmem (2x106) were transferred to B16A, NOD and proinsulin transgenic mice.
Three weeks later, the total number of G9 Tmem (CFSE+CD8+Vβ6+) (A), G9 Tmem as a proportion of
total CD8+ T cells (B) and the total number of cells (C) in spleen, sdLN and pLN were determined
using flow cytometry. Data are pooled from two independent experiments. Each point represents a
single mouse. Error bars show S.D. One-way ANOVA followed by Tukey post-test. (**p<0.01,
****p<0.0001).
104
4.2.7 G9 Tmem do not infiltrate the pancreas after transfer to recipients
Since G9 Tmem are post-activated CD8+ T cells which recognise an insulin-derived
determinant, it is possible that G9 Tmem might preferentially migrate to the pancreas after transfer
and this might underlie the decrease in number of G9 Tmem in secondary lymphoid organs of
proinsulin transgenic recipients. To test this, G9 Tmem (CD45.1+CD45.2-) were transferred to NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) recipients and pancreasinfiltrating leukocytes were analysed seven days later. The level of CD45.1 mAb background
staining in pancreas was determined for NOD and proinsulin transgenic mice that were not injected
with G9 Tmem (Fig. 4.8A). Seven days after transfer, no donor (G9) leukocytes were observed in
pancreas of proinsulin transgenic or NOD recipients (Fig. 4.8A). A G9 mouse was included as a
technical control for staining of CD45.1+CD45.2- cells in pancreas (Fig. 4.8A). There was no
difference observed for the proportion of host-type pancreas-infiltrating leukocytes between NOD
and proinsulin transgenic recipients (Fig. 4.8B). This might be due to 8-10 week old male NOD
mice being used as recipients and these mice may be expected to have milder insulitis than older
female NOD mice and as a result, fewer endogenous leukocytes might infiltrate the pancreas. The
proportion of donor leukocytes of host-type leukocytes was not significantly different between
NOD and proinsulin transgenic recipients in pancreas (Fig. 4.8C). Therefore the decrease in G9
Tmem number after transfer to proinsulin transonic recipients is not due to G9 Tmem trafficking to
pancreas. It is possible that islet inflammation, which is expected to be mild in the NOD recipients
used, is a requirement for attracting transferred G9 Tmem to pancreas.
105
PI
SSC
A
FSC
proIns Tg
NOD (untrf.)
proIns Tg (untrf.)
G9 (untrf.)
CD45.2
NOD
FSC
CD45.1
C
-
+
10
C D 4 5 .1 C D 4 5 .2 c e lls
( % o f liv e )
+
n .s .
15
5
( % o f t o t a l le u k o c y t e s )
0 .8
20
+
H o s t ( C D 4 5 .1 C D 4 5 .2 ) c e lls
B
n .s .
0 .6
0 .4
0 .2
0 .0
0
NOD
p r o In s T g
no G 9 Tm em
NOD
p r o In s T g
+ G 9 Tm em
NOD
p r o In s T g
no G 9 Tm em
NOD
p r o In s T g
+ G 9 Tm em
Figure 4.8 - G9 Tmem do not traffic to the pancreas in immune-replete NOD or proinsulin
transgenic recipients.
G9 Tmem (2x106, CD45.1+CD45.2-) were transferred to NOD (CD45.1+CD45.2+) and proinsulin
transgenic (CD45.1+CD45.2+) mice (A). Seven days later, host (CD45.1+CD45.2+) leukocytes as
proportion of live cells (B) and transferred (CD45.1+CD45.2-) leukocytes as a proportion of total
leukocytes (C) in pancreas were determined using flow cytometry. Data are pooled from two
independent experiments. FACS plots are representative of at least two individual mice. Each point
represents a single mouse. Error bars show S.D. One-way ANOVA followed by Tukey post-test.
106
4.2.8 Transferred G9 Tmem maintain a memory-like phenotype in B16A and NOD recipients
G9 Tmem persist in the absence of cognate antigen, but it is not known whether G9 Tmem
retain a memory phenotype and whether this phenotype might differ between NOD (islet-derived
insulin) and B16A (no antigen) recipients. To test this, CFSE-labelled G9 Tmem were transferred to
B16A, NOD and proinsulin transgenic recipients and G9 Tmem phenotype was determined three
weeks later. Prior to transfer, G9 Tmem comprise a mixed population of CD44hiCD62Lhi TCM and
CD44hi CD62Llo TEM. Transferred G9 Tmem from sdLN of B16A recipients were predominantly
CD44hi, in contrast to the host CD8+Vβ6+ T-cell population which was predominantly CD44lo (Fig.
4.9A). The level of relevant isotype control mAb staining was low on both G9 Tmem and host
CD8+Vβ6+ T cells (Fig. 4.9A), although the increased fluorescence observed for G9 Tmem might
be expected due to CFSE labelling. Three weeks after transfer, G9 Tmem from both B16A and
NOD recipients were predominantly CD44hi in spleen, sdLN and pLN (Fig. 4.9B). The phenotype
of G9 Tmem after transfer to proinsulin transgenic recipients was not determined (Fig. 4.9B-D) as
there were no G9 Tmem observed in spleen, sdLN or pLN three weeks after transfer, consistent
with data described above (Fig. 4.7). The proportion of G9 Tmem expressing high levels of CD44
and CD62L was similar between B16A and NOD recipients for spleen, sdLN and pLN (Fig. 4.9B).
Similarly, the proportion of CD44hiCD62L- G9 Tmem in B16A recipients did not differ to NOD
recipients (Fig. 4.9C). Therefore, the proportion of TEM- and TCM-like G9 Tmem after transfer to
B16A and NOD recipients was similar to that observed prior to transfer, indicating in vitro-cultured
G9 Tmem maintain a memory phenotype in vivo, consistent with a long-lived Tmem population.
107
Host CD8+Vβ6+
T cells
Host CD8+Vβ6+
T cells
G9 Tmem
isotype
G9 Tmem
CD44
A
CD62L
isotype
C
G 9 T m e m (% )
100
80
100
80
60
+
60
C D 62L
40
hi
20
n .d . n .d . n .d .
0
s
p
l
s
d
L
N
p
L
N
s
p
l
s
B16A
d
L
N
p
L
N
s
p
l
s
d
L
N
p
L
40
20
n .d . n .d . n .d .
0
N
s
p
l
s
p ro In s T g
NOD
C D 44
T o ta l C D 4 4
hi
G 9 T m e m (% )
B
d
L
N
p
L
B16A
N
s
p
l
s
d
L
N
p
NOD
L
N
s
p
l
s
d
L
N
p
L
N
p ro In s T g
100
75
50
25
s
p
L
N
N
p
L
s
N
L
p
d
l
N
p
L
s
d
s
p
L
N
N
p
L
s
d
s
B16A
l
n .d . n .d . n .d .
0
l
C D 44
hi
-
C D 6 2 L G 9 T m e m (% )
D
NOD
p ro In s T g
Figure 4.9 - Transferred G9 Tmem maintain a memory-like phenotype in B16A and NOD
recipients.
CFSE-labelled G9 Tmem (2x106) were transferred to B16A, NOD and proinsulin transgenic mice.
Three weeks later, the proportion of CD44hi (B), CD44hiCD62L+ (C) and CD44hiCD62L- (D) G9
Tmem (CFSE+CD8+Vβ6+) were determined using flow cytometry. Data are pooled from two
independent experiments. Representative FACS plots showing CD44 and CD62L expression on
transferred G9 Tmem and host CD8+Vβ6+ T cells in sdLN from B16A recipients (A). Each point
represents a single mouse. Error bars show S.D. n.d.= not determined. One-way ANOVA followed by
Tukey post-test.
108
4.2.9 G9 Tmem do not preferentially migrate to BM in recipients
Since BM is an important site of TCM accumulation (Becker et al., 2005; Mazo et al., 2005),
it is possible that transferred G9 Tmem, of which approximately 50% display a TCM-like phenotype,
may preferentially home to this compartment in recipients. If so, this might explain the reduced
number of G9 Tmem recovered from secondary lymphoid organs after transfer to proinsulin
transgenic recipients (Fig. 4.6, 4.7). To test this, CFSE-labelled G9 Tmem were transferred to
B16A, NOD and proinsulin transgenic recipients and G9 Tmem enumerated from BM three weeks
later. The total number of G9 Tmem from BM of proinsulin transgenic recipients did not differ
compared to NOD recipients (Fig. 4.10A). A higher total number and proportion of G9 Tmem
within total CD8+ T cells was observed for BM of B16A recipients compared to both NOD and
proinsulin transgenic recipients (Fig. 4.10A, B), consistent with the increased proportion of G9
Tmem from other sites of B16A recipients at this time point (Fig. 4.7). Since B16A mice show
reduced spleen cellularity than NOD and proinsulin transgenic mice (Fig. 4.7), it is possible that the
cellularity of BM from B16A mice may be lower compared to that of wild type NOD mice. The
total cellularity of BM from B16A recipients was similar to NOD and proinsulin transgenic
recipients (Fig. 4.10C). This validates the increase in the total number of G9 Tmem in BM of B16A
compared to NOD recipients, as this was not an artefact of differing BM cellularity. Since CFSE
was used to track transferred G9 Tmem, it was possible that loss of G9 Tmem-gated events due to
CFSE dilution could account for the reduced number of G9 Tmem from BM of NOD and proinsulin
transgenic recipients (Fig. 4.10A, B).
To confirm data in Figure 4.10, G9 Tmem were transferred to congenic NOD and proinsulin
transgenic recipients (CD45.1+CD45.2+) as well as B16A recipients (CD45.1+CD45.2-). Three
weeks later, G9 Tmem in BM were identified using both congenic markers and InsB15-23-loaded,
H2-Kd-tetramer for NOD and proinsulin transgenic recipients and tetramer alone for B16A
recipients. Similarly to data in Figure 4.10, the recovery of G9 Tmem was lower from BM of NOD
and proinsulin transgenic recipients compared to B16A recipients, when G9 Tmem were identified
with congenic markers (Fig. 4.11A, C) or tetramer (Fig. 4.11B, D). Despite variability, the
proportion of G9 Tmem within total CD8+ T cells from BM of B16A recipients as identified with
tetramer (Fig. 4.11B), was similar as identified with CFSE (Fig. 4.10B). Also, no difference in BM
cellularity was observed between congenic NOD, congenic proinsulin transgenic and B16A
recipients (Fig. 4.11C). These data exclude BM as a preferential site of deposition or reservoir for
G9 Tmem after transfer to recipients, and this finding helps to reinforce the notion of G9 Tmem
undergoing rapid deletion after transfer to proinsulin transgenic recipients.
109
T o t a l G 9 T m e m /B M ( x 1 0
-3
)
A
0 .5
**
0 .4
0 .3
0 .2
0 .1
0 .0
NOD
p r o In s T g
NOD
p r o In s T g
NOD
p r o In s T g
0 .4
**
0 .3
G 9 T m e m /t o t a l C D 8
+
B
c e lls ( % )
B 16A
C
0 .2
0 .1
0 .0
B 16A
B M c e ll n u m b e r (x 1 0
-6
)
30
20
10
0
B 16A
Figure 4.10- Increased recovery of CFSE-labelled G9 Tmem in BM of B16A recipients compared
to NOD and proinsulin transgenic recipients.
CFSE-labelled G9 Tmem (2x106) were transferred to B16A, NOD and proinsulin transgenic mice.
Three weeks later, the total number of G9 Tmem (CFSE+CD8+Vβ6+) (A), G9 Tmem as a proportion of
total CD8+ T cells (B) and the total number of cells (C) in BM were determined using flow cytometry.
Data are pooled from two independent experiments. Each point represents a single mouse. Error bars
show S.D. One-way ANOVA followed by Tukey post-test. (**p<0.01).
110
B
0 .1
0 .0
0 .1
0 .0
NOD
c e lls ( % )
c e lls ( % )
)
0 .2
0 .4
0 .3
G 9 T m e m /t o t a l C D 8
0 .2
0 .1
0 .0
B 16A
E
0 .3
0 .2
0 .1
0 .0
B 16A
D
0 .4
0 .3
*
0 .4
p r o In s T g
+
+
G 9 T m e m /t o t a l C D 8
-3
0 .3
B 16A
C
T o t a l G 9 T m e m /B M ( x 1 0
0 .2
0 .4
c e lls ( % )
0 .3
n .s .
0 .5
0 .4
NOD
p r o In s T g
*
0 .3
+
0 .4
0 .5
0 .2
0 .1
0 .0
NOD
p r o In s T g
G 9 T m e m /t o t a l C D 8
-3
)
0 .5
T o t a l G 9 T m e m /B M ( x 1 0
T o t a l G 9 T m e m /B M ( x 1 0
-3
)
A
0 .2
0 .1
0 .0
B 16A
NOD
p r o In s T g
n .s .
T o t a l B M c e lls ( x 1 0
-6
)
40
30
20
10
0
B 16A
NOD
p r o In s T g
Figure 4.11- Few tetramer-binding G9 Tmem detected in BM of NOD and proinsulin transgenic
recipients.
G9 Tmem (2x106, CD45.1+CD45.2-) were transferred to B16A (CD45.1+CD45.2-), NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. Three weeks later, the total
number of G9 Tmem (CD45.1+CD8+Kd:Ins+Vβ6+ in B16A or CD45.1+CD45.2-CD8+Vβ6+ in NOD and
proinsulin transgenic) (A), total number of G9 Tmem (CD45.1+CD8+Kd:Ins+Vβ6+) (B), G9 Tmem
(CD45.1+CD8+Kd:Ins+Vβ6+ in B16A or CD45.1+CD45.2-CD8+Vβ6+ in NOD and proinsulin
transgenic) as a proportion of total CD8+ T cells (C), G9 Tmem (CD45.1+CD8+Kd:Ins+Vβ6+) as a
proportion of total CD8+ T cells (D) and the total number of cells (E) in BM were determined using
flow cytometry. Data are pooled from two independent experiments. Each point represents a single
mouse. Error bars show S.D. The dashed line represents the detection limit for Kd:Ins tetramer binding
and shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells. One-way ANOVA
followed by Tukey post-test. (*p<0.05).
111
4.2.10 TCR down-regulation on transferred G9 Tmem is observed in proinsulin transgenic
recipients
Since naïve G9 T cells down-regulate TCR expression after transfer to proinsulin transgenic
recipients, it is possible that G9 Tmem down-regulate TCR in a similar manner. To test this, G9
Tmem were transferred to NOD and proinsulin transgenic recipients and TCR expression on G9
Tmem was analysed at defined time points. G9 Tmem from proinsulin transgenic recipients
displayed reduced Vβ6 expression compared to G9 Tmem from NOD recipients (Fig. 4.12A). As
early as one day after transfer, the level of Vβ6 expression was lower for G9 Tmem from pLN in
proinsulin transgenic recipients compared to NOD recipients (Fig. 4.12B). This reduction of Vβ6
expression was maintained, and at seven days after transfer, G9 Tmem Vβ6 expression was
significantly lower from proinsulin transgenic recipients compared to NOD recipients in spleen,
sdLN and pLN (Fig. 4.12B). After transfer to NOD recipients, G9 Tmem Vβ6 expression was lower
relative to the host CD8+Vβ6+ T-cell population as a result of in vitro activation of G9 Tmem
during culture (Fig. 4.12B). To determine G9 Tmem Vβ6 expression after transfer to B16A
recipients, CFSE-labelled G9 Tmem were transferred to B16A, NOD and proinsulin transgenic
recipients and TCR expression on G9 Tmem analysed three days later. G9 Tmem Vβ6 expression
was similar from B16A compared to NOD recipients (Fig. 4.12C). From proinsulin transgenic
recipients, G9 Tmem displayed significantly lower Vβ6 expression compared to those from NOD
and B16A recipients in spleen, sdLN and pLN (Fig. 4.12C). These data confirm G9 Tmem rapidly
down-regulate TCR after transfer to proinsulin transgenic recipients. TCR down-regulation might
indicate G9 Tmem are functionally inactivate after transfer to mice transgenically expressing
proinsulin.
112
B
V 6 expression (relative to hos t )
A
NOD
proIns Tg
unstained
*
*
****
****
****
N O D spl
1 .0
p r o In s T g s p l
0 .8
N O D sdLN
p r o In s T g s d L N
0 .6
N O D pLN
0 .4
p r o In s T g p L N
0 .2
0 .0
0
1
2
3
4
5
6
7
8
D a y s (p o s t tra n s fe r)
Vβ6
***
n .s .
1 .0
0 .8
0 .6
0 .4
0 .2
p
L
N
N
l
p
L
s
d
s
p
L
N
N
p
L
d
s
s
p
L
N
N
p
L
s
d
s
B16A
l
0 .0
l
V 6 expression (relative to host)
C
NOD
p ro In s T g
Figure 4.12 – TCR down-regulation by G9 Tmem after transfer to proinsulin transgenic
recipients.
G9 Tmem (2x106, CD45.1+CD45.2-) were transferred to NOD (CD45.1+CD45.2+) and proinsulin
transgenic (CD45.1+CD45.2+) mice. (A) Vβ6 expression on G9 Tmem (CD45.1+CD45.2-CD8+Vβ6+)
in sdLN from NOD and proinsulin transgenic recipients seven days after transfer. (B) At times shown,
Vβ6
expression
on
G9
Tmem
(CD45.1+CD45.2-CD8+Vβ6+)
relative
to
host
+
+
+
+
(CD45.1 CD45.2 CD8 Vβ6 ) T cells in spleen, sdLN and pLN was determined using flow cytometry.
(C) CFSE-labelled G9 Tmem (2x106) were transferred to B16A, NOD and proinsulin transgenic mice.
Three days later, Vβ6 expression on G9 Tmem (CFSE+CD8+Vβ6+) in spleen, sdLN and pLN was
determined using flow cytometry. Histograms are representative of nine mice from three independent
experiments. Data are pooled from at least two independent experiments at each time point (B) and
two independent experiments (C). n=5 (day 1 and 5); n=6 (day 2 and 3); n= 12 (day 7). Each point
represents a single mouse. Error bars show S.D. One-way ANOVA followed by Tukey post-test.
(*p<0.05, ****p<0.0001: NOD pLN vs proIns Tg pLN (B)). (***p<0.001 (C)).
113
4.2.11 Transferred G9 Tmem are unresponsive to InsB15-23 immunisation in proinsulin transgenic
recipients
Since the data presented to this point suggest G9 Tmem are deleted after transfer to
proinsulin transgenic recipients and residual G9 Tmem exhibit reduced TCR expression, it is
possible that the residual population of G9 Tmem in proinsulin transgenic recipients are
unresponsive. To test this, G9 Tmem were transferred to B16A, NOD and proinsulin transgenic
mice and seven days later recipients immunised with InsB15-23 (23 Gly)/poly I:C. Five days after
immunisation, G9 Tmem were enumerated and IFN-γ production was determined using ELISPOT.
Since B16A mice express the same CD45.1 allotype as G9 Tmem, InsB15-23-loaded, H2-Kd-tetramer
was used to detect transferred G9 Tmem in recipients. There was a trend toward an increase in the
total number of G9 Tmem in immunised B16A recipients compared to unimmunised B16A
recipients, however, this was not significant (Fig. 4.13A). The total number of G9 Tmem and their
proportion of total CD8+ T cells remained unchanged in NOD recipients after immunisation (Fig.
4.13A, B). In proinsulin transgenic recipients, the total number of G9 Tmem and their proportion of
total CD8+ T cells was low and did not differ following immunisation (Fig. 4.13A, B). The
proportion of G9 Tmem within CD8+ T cells from B16A recipients was increased following
immunisation (Fig. 4.13B). G9 Tmem were also identifiable in NOD and proinsulin transgenic
recipients based on expression of CD45.1 and CD45.2. A similar pattern was observed when G9
Tmem were identified using tetramer and congenic markers (Fig. 4.13C, D).
Immunisation did not significantly alter the total cellularity of spleens for any strain (Fig.
4.13E). The total cellularity of spleens from immunised B16A mice was significantly lower than
immunised proinsulin transgenic mice (Fig. 4.13E). Immunised NOD recipients were
normoglycemic at the time of cull and BGL were similar between immunised and unimmunised
NOD recipients and age-matched NOD mice that were not injected with G9 Tmem (Fig. 4.13F). By
ELISPOT, there appeared to be fewer IFN-γ-producing cells from proinsulin transgenic than NOD
and B16A recipients (Fig. 4.13G). The number of IFN-γ-producing cells was increased in B16A
recipients by immunisation but remained unchanged in NOD recipients following immunisation
(Fig. 4.13H). In proinsulin transgenic recipients, there were very few IFN-γ-producing cells
observed and this was unchanged following immunisation (Fig. 4.13H). The number of IFN-γproducing cells appeared to be lower from immunised proinsulin transgenic than immunised NOD
recipients yet this was not significantly different (Fig. 4.13H). These data suggest G9 Tmem are
responsive to InsB15-23 immunisation after transfer to B16A recipients, that is, in the absence of
cognate antigen. In contrast to naïve G9 T cells, G9 Tmem did not respond to InsB15-23
immunisation after transfer to NOD recipients. Importantly though, the response of G9 Tmem to
immunisation was ablated after transfer to mice expressing transgenic proinsulin.
114
B
30
n .s .
n .s .
0
0 .6
u /t
u n im m . im m .
u /t
NOD
u n im m . im m .
10
0
u /t
u n im m .
im m .
n .s .
0 .0
u /t
u n im m . im m .
c e lls ( % )
30
20
**
n .s .
n .s .
10
0
u /t
u n im m .
im m .
u /t
NOD
u n im m .
-6
n .s .
0 .8
0 .6
0 .4
0 .2
0 .0
u /t
im m .
u n im m . im m .
u /t
NOD
u n im m . im m .
p ro In s T g
0 .8
0 .6
u n im m .
im m .
**
0 .4
n .s .
n .s .
0 .2
0 .0
u /t
B16A
p ro In s T g
u n im m . im m .
NOD
u /t
u n im m . im m .
p ro In s T g
F
*
100
u /t
B16A
D
+
-3
0 .2
p ro In s T g
)
20
T o ta l G 9 T m e m /s p le e n (x 1 0
n .s .
30
n .s .
c e lls ( % )
u n im m . im m .
0 .4
+
u /t
)
-3
T o ta l G 9 T m e m /s p le e n (x 1 0
*
G 9 T m e m /t o t a l C D 8
10
B16A
)
G 9 T m e m /t o t a l C D 8
n .s .
B16A
T o t a l s p le e n c e lls ( x 1 0
****
0 .8
+
20
G 9 T m e m /t o t a l C D 8
T o ta l G 9 T m e m /s p le e n (x 1 0
-3
)
*
C
E
T c e lls ( % )
A
n .s .
n .s .
80
60
NOD
imm.
8.4±1.2
n .s .
40
NOD
unimm.
9.1±2.6
NOD
u/t
7.3±0.8
20
0
B16A
u /t u n im m . im m .
NOD
u /t u n im m . im m .
****
p ro In s T g
G
H
B16A imm.
NOD imm.
proIns Tg imm.
 IFN- spots (per 5x105 cells)
u /t u n im m . im m .
250
****
***
200
150
n .s .
100
n .s .
50
n .s .
0
u /t u n im m . im m .
B16A
u /t u n im m . im m .
NOD
u /t u n im m . im m .
p ro In s T g
Figure 4.13- G9 Tmem recall response is ablated after transfer to proinsulin transgenic mice.
G9 Tmem (2x106, CD45.1+CD45.2-) were transferred to B16A (CD45.1+CD45.2-), NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. Seven days later, mice were
injected with InsB15-23 (50µg) and poly I:C (100µg) i.p. or left unimmunised. Five days later, the total
number of G9 Tmem (CD8+Kd:Ins+Vβ6+) (A), G9 Tmem (CD8+Kd:Ins+Vβ6+) as a proportion of total
CD8+ T cells (B), the total number of G9 Tmem (CD45.1+CD8+Kd:Ins+Vβ6+ in B16A;
CD45.1+CD45.2-CD8+Vβ6+ in NOD and proinsulin transgenic) (C), G9 Tmem
(CD45.1+CD8+Kd:Ins+Vβ6+ in B16A; CD45.1+CD45.2-CD8+Vβ6+ in NOD and proinsulin transgenic)
as a proportion of total CD8+ T cells (D) and the total number of spleen cells (E) in spleen were
determined using flow cytometry. The non-fasting BGL (mM) of NOD mice was determined at time
of cull and table shows mean±S.D (F). IFN-γ production in spleen was determined using ELISPOT
(G, H). (G) Representative images from InsB15-23-stimulated wells. Data are from two independent
experiments. Each point represents a single mouse. Error bars shown S.D. The dashed line shows the
mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells. One-way ANOVA followed by
Tukey post-test. (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001).
115
4.2.12 G9 Tmem do not accumulate in LN of proinsulin transgenic recipients following InsB15-23
immunisation
Since InsB15-23/poly I:C immunisation did not increase the number of G9 Tmem in spleen of
NOD recipients, it is possible that G9 Tmem might have accumulated in LN which drain the site of
immunisation instead. In light of this, as InsB15-23/poly I:C immunisation was administered i.p., an
increase in G9 Tmem may have occurred in inguinal LN or pLN of immunised recipients which
both drain the peritoneal cavity. To test this, G9 Tmem were transferred to B16A, NOD and
proinsulin transgenic mice and seven days later recipients immunised with InsB15-23/poly I:C. Five
days after immunisation, G9 Tmem were enumerated in spleen, brachial, inguinal and pancreatic
LN as described above. There was no significant difference in the total number of G9 Tmem or
their proportion of total CD8+ T cells in brachial LN of NOD recipients after immunisation (Fig.
4.14B), consistent with the response observed in spleen (Fig. 4.14A). Despite the small number of
mice analysed, the number of G9 Tmem remained unchanged in inguinal LN (Fig. 4.14C) and pLN
(Fig. 4.14D) of NOD recipients following immunisation. No significant difference in the total
number of G9 Tmem or their proportion of total CD8+ T cells was observed in brachial LN from
B16A recipients (Fig. 4.14B). Importantly, in immunisation-draining and non-draining sites alike,
the number and proportion of G9 Tmem was low and this remained unchanged following
immunisation in proinsulin transgenic recipients (Fig. 4.14A-D). These findings suggest the pattern
of G9 Tmem dynamics observed for spleen was similar in LN tissue. An important observation was
that G9 Tmem did not expand in any lymphoid site in response to InsB15-23 immunisation.
116
10
5
0
u n im m .
0 .6
+
10
0
u /t
c e lls ( % )
n .s .
15
***
0 .8
im m .
u /t u n im m . im m .
u /t
0 .4
0 .2
0 .0
u /t
u n im m . im m .
0 .3
u n im m .
im m .
**
n .s .
+
n .s .
G 9 T m e m /t o t a l C D 8
20
c e lls ( % )
)
**
-3
T o ta l G 9 T m e m (x 1 0
Spleen
T o ta l G 9 T m e m /s p le e n (x 1 0
-3
)
20
n .s .
30
n .s .
0 .2
G 9 T m e m /t o t a l C D 8
A
0 .1
0 .0
u /t u n im m . im m .
u /t
u n im m . im m .
B16A
B16A
NOD
NOD
p ro In s T g
p ro In s T g
1
2
n .s .
1
n .s .
0
0
u /t
u n im m .
u /t
im m .
u n im m . im m .
NOD
-3
)
1 .0
0 .4
0 .2
u n im m . im m .
u /t
u n im m . im m .
u /t
u n im m .
im m .
*
n .s .
0 .2
n .s .
0 .0
u /t
u n im m . im m .
NOD
u /t
u n im m . im m .
p ro In s T g
0 .1 0
0 .0 8
0 .0 6
0 .0 4
0 .0 2
0 .0 0
u /t u n im m . im m .
p ro In s T g
-3
)
0 .8
u n im m . im m .
p ro In s T g
0 .2 0
0 .1 5
+
0 .6
u /t
NOD
c e lls ( % )
NOD
T o ta l G 9 T m e m (x 1 0
0 .0
0 .4
B16A
G 9 T m e m /t o t a l C D 8
0 .6
u /t
pLN
0 .2
0 .6
+
0 .8
0 .0
D
u n im m . im m .
0 .4
p ro In s T g
0 .4
0 .2
0 .0
u /t u n im m . im m .
NOD
u /t u n im m . im m .
p ro In s T g
G 9 T m e m /t o t a l C D 8
T o ta l G 9 T m e m (x 1 0
Inguinal
LN
u /t
c e lls ( % )
B16A
C
+
n .s .
G 9 T m e m /t o t a l C D 8
T o ta l G 9 T m e m (x 1 0
2
n .s .
0 .6
+
)
-3
)
-3
T o ta l G 9 T m e m (x 1 0
Brachial
LN
c e lls ( % )
3
G 9 T m e m /t o t a l C D 8
n .s .
3
c e lls ( % )
B
0 .1 0
0 .0 5
0 .0 0
u /t u n im m . im m .
NOD
u /t
u n im m . im m .
p ro In s T g
Figure 4.14- G9 Tmem do not accumulate in LN of proinsulin transgenic recipients following
InsB15-23 immunisation.
G9 Tmem (2x106, CD45.1+CD45.2-) were transferred to B16A (CD45.1+CD45.2-), NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. Seven days later, mice were
injected with InsB15-23 (50µg) and poly I:C (100µg) i.p. or left unimmunised. Five days later, the total
number of G9 Tmem (CD45.1+CD8+Kd:Ins+Vβ6+in B16A or CD45.1+CD45.2-CD8+Vβ6+ in NOD and
proinsulin transgenic) (left) and G9 Tmem as a proportion of total CD8+ T cells (right) in spleen (A),
brachical LN (B), inguinal LN (C) and pancreatic LN (D) were determined using flow cytometry. Data
are from a single (pLN and inguinal LN, B16A recipients) or two (spleen and brachial LN)
independent experiments. Each point represents a single mouse. Error bars shown S.D. The dashed
line shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells. One-way ANOVA
followed by Tukey post-test. (*p<0.05, **p<0.01, ***p<0.001).
117
4.2.13 G9 Tmem are unresponsive to InsB15-23 immunisation twenty one days after transfer to
proinsulin transgenic recipients
Since G9 Tmem are unresponsive to InsB15-23/poly I:C immunisation in proinsulin
transgenic recipients seven days after transfer, it is possible that this functional inactivation is not
maintained and G9 Tmem responsiveness is re-instated. To test the longevity of G9 Tmem
unresponsiveness, G9 Tmem were transferred to B16A, NOD and proinsulin transgenic mice and
twenty one days later recipients immunised with InsB15-23/poly I:C. Five days after immunisation,
G9 Tmem were enumerated using flow cytometry and IFN-γ production was determined using
ELISPOT. Neither the total number of G9 Tmem nor their proportion of total CD8+ T cells was
increased by immunisation in B16A, NOD or proinsulin transgenic recipients when G9 Tmem were
identified using tetramer (Fig. 4.15A, B) and expression of congenic markers (Fig. 4.15C, D).
Recovery of G9 Tmem was higher in B16A recipients than NOD and proinsulin transgenic
recipients (Fig. 4.15A, B). Immunisation did not affect the total number of spleen cells for any
strain but the total number of spleen cells was lower from immunised B16A recipients compared to
immunised NOD and immunised proinsulin transgenic recipients (Fig. 4.15E). NOD recipients and
age-matched NOD mice not injected with G9 Tmem were normoglycemic at the time of cull (Fig.
4.15F). After immunisation, there appeared to be fewer IFN-γ-producing cells from NOD and
proinsulin transgenic recipients compared to B16A recipients (Fig. 4.15E). The number of IFN-γproducing cells was not significantly altered by immunisation in B16A, NOD or proinsulin
transgenic recipients (Fig. 4.15F). There was a trend toward an increased number of IFN-γproducing cells from immunised B16A recipients compared to immunised NOD and immunised
proinsulin transgenic recipients, however, this was not significant (Fig. 4.15F). These data suggest
G9 Tmem persist and remain functional in the absence of cognate antigen, as indicated by IFN-γ
production, yet do not exhibit an increased response following InsB15-23 immunisation in B16A
recipients. A second InsB15-23 immunisation may elicit a stronger response from transferred G9
Tmem in B16A recipients. G9 Tmem, like naïve G9 T cells, are also unresponsive to InsB15-23
immunisation twenty one days after transfer to NOD recipients. Particularly significant here is the
complete functional unresponsiveness of G9 Tmem after transfer to proinsulin transgenic recipients,
suggesting peripheral tolerance induction in G9 Tmem is maintained in mice transgenically
expressing proinsulin.
118
A
B
****
0 .5
**
0 .4
+
n .s .
10
G 9 T c e lls /t o t a l C D 8
T o ta l G 9 T m e m /s p le e n (x 1 0
-3
)
*
n .s .
15
T c e lls ( % )
n .s .
n .s .
5
0
u /t
u n im m .
im m .
u /t
u n im m .
B16A
im m .
u /t
NOD
u n im m .
im m .
0 .3
0 .2
0 .0
u /t
u n im m .
im m .
u /t
B16A
p ro In s T g
u n im m .
im m .
NOD
u /t
u n im m .
im m .
p ro In s T g
5
0
u /t
u n im m .
im m .
5
0
u /t
u n im m .
B16A
u /t
u n im m .
im m .
0 .4
0 .3
0 .2
0 .1
0 .0
u /t
p ro In s T g
0 .5
0 .4
u n im m .
im m .
0 .3
0 .2
0 .1
0 .0
u /t
u n im m . im m .
u /t
NOD
B16A
u n im m . im m .
p ro In s T g
F
**
*
-6
)
E
T o t a l s p le e n c e lls ( x 1 0
im m .
NOD
100
**
+
10
0 .5
+
15
G 9 T m e m /t o t a l C D 8
10
c e lls ( % )
)
-3
15
T o ta l G 9 T m e m /s p le e n (x 1 0
T o t a l G 9 T c e lls /s p le e n ( x 1 0
-3
)
*
T c e lls ( % )
D
G 9 T c e lls /t o t a l C D 8
C
n .s .
n .s .
0 .1
n .s .
n .s .
NOD
imm.
7.4±0.4
80
n .s .
60
40
NOD
unimm.
7.2±0.4
NOD
u/t
8.2±0.3
20
0
u n im m .
B16A
B16A imm.
im m .
u /t
u n im m . im m .
NOD
NOD imm.
u /t
u n im m . im m .
p ro In s T g
proIns Tg imm.
H
 IFN-  spots (per 5x105 cells)
u /t
G
100
80
60
40
20
0
im m . u n im m .
B16A
u /t
im m . u n im m .
NOD
u /t
im m . u n im m .
u /t
p r o In s T g
Figure 4.15- G9 Tmem are functionally inactivate in proinsulin transgenic recipients.
G9 Tmem (2x106, CD45.1+CD45.2-) were transferred to B16A (CD45.1+CD45.2-), NOD
(CD45.1+CD45.2+) and proinsulin transgenic (CD45.1+CD45.2+) mice. Twenty-one days later, mice
were injected with InsB15-23 (50µg) and poly I:C (100µg) i.p. or left unimmunised. Five days later, the
total number of G9 Tmem (CD45.1+CD8+Kd:Ins+Vβ6+) (A), G9 Tmem (CD45.1+CD8+Kd:Ins+Vβ6+) as
a proportion of total CD8+ T cells (B), the total number of G9 Tmem (CD45.1+CD45.2-CD8+Vβ6+ in
NOD and proinsulin transgenic) (C), G9 Tmem (CD45.1+CD45.2-CD8+Vβ6+ in NOD and proinsulin
transgenic) as a proportion of total CD8+ T cells (D) and the total number of spleen cells (E) were
determined using flow cytometry. The non-fasting BGL (mM) of NOD mice was determined at time
of cull and the table shows mean±S.D (F). The number of IFN-γ-producing cells in spleen of recipients
was determined using ELISPOT (G, H). (G) Representative images from InsB15-23-stimulated wells.
Data are from two independent experiments. Each point represents a single mouse. Error bars shown
S.D. The dashed line shows the mean frequency of Kd:LLO (control) tetramer-binding CD8+ T cells.
One-way ANOVA followed by Tukey post-test. (*p<0.05, **p<0.01, ****p<0.0001).
119
Table 4.1 Summary of Chapter 4
B16A
NOD
Proliferation
No
pLN only
Absolute number
n/a
Proportion (of total
CD8+ cells)
Unchanged
TCR downregulation
IFN-γ production
No
Slow decrease (by 21
days after transfer)
Accumulation in pLN
(at day 3) followed by
slow decrease in all
sites analysed (by day
21)
No
Recall response
Yes
Yes
Yes (at day 7
following transfer)
No increase in G9
Tmem number
following
immunisation
Proinsulin
transgenic
Yes, in all sites
analysed
Rapid decrease (by 7
days after transfer)
Rapid decrease (by day
7)
Yes
No
No
Figure 4.16 – Summary of main findings from Chapter 4.
This table represents summarised data from Chapter 3. ‘Proliferation’ refers to Figure 4.5, ‘Absolute
number’ and ‘proportion’ refer to Figure 4.6 and Figure 4.7, ‘TCR down-regulation’ refers to Figure
4.12, ‘IFN-γ production’ and ‘Recall response’ refer to Figure 4.13 and Figure 4.15.
120
4.3 Discussion
Induction of peripheral tolerance in naïve CD8+ T cells has been studied widely, whereas
this has not been investigated as extensively for CD8+ Tmem. To an even lesser extent have studies
examined tolerance induction in CD8+ Tmem specific for self-antigens in the context of
autoimmune disease. In this chapter, transgenic expression of proinsulin in MHC class II+ APC as a
means to induce peripheral tolerance in insulin-specific CD8+ Tmem was investigated. After
activation with InsB15-23, G9 T cells rapidly transfer autoimmune diabetes to NOD.SCID recipients
(Wong et al., 2009). Naïve G9 T cells are not diabetogenic and as such may be more amenable to
tolerogenic signals in the periphery, due to being maintained in a naïve state (Fig. 3.1) in G9 mice
despite expression of their cognate antigen, InsB15-23, in the thymus and the periphery. In the present
study, G9 T cells were cultured in the presence of InsB15-23 and IL-2 for three days and IL-15 alone
for a further two days. The resulting CD8+ T cells displayed a memory-like phenotype and secreted
IFN-γ, consistent with TEM differentiation. It could therefore be hypothesized that G9 Tmem are
less likely, compared to their naïve counterparts, to undergo peripheral tolerance induction after
transfer to proinsulin transgenic mice. The data in this chapter, however, suggest transgenic
expression of proinsulin provides robust tolerogenic signals to induce tolerance in a mixed
population of insulin-specific TEM and TCM.
G9 T cells were activated by InsB15-23 (23 Gly) in vitro, as indicated by CD69 up-regulation.
Synchronously, an increase in FSC fluorescence was observed with addition of InsB15-23 (23 Gly),
indicating blastogenesis, further validating activation of G9 T cells. Whilst there was no difference
in the proportion of activated G9 T cells, a reduced concentration of the higher affinity APL InsB1523
(23 Val) was required for maximal G9 T-cell activation compared to stimulation with native
InsB15-23 (23 Gly), consistent with a previous study (Wong et al., 2002). As a high-affinity
TCR/peptide interaction (Hogquist et al., 1994), OT-I T cells and the agonist peptide OVA257-264
was assessed in the present study. The kinetics of OT-I T cells in response to OVA257-264 stimulation
was similar to previously published findings (Salmond et al., 2014), although the peptide
concentration required for maximal T-cell activation was lower in the present study. This could be
due to different stimulation periods, 18 hours in the present study compared to 4 hours in the study
by Salmond et al. Overall, the outcomes of the peptide titration studies confirmed G9 T cells exhibit
a low affinity for their cognate determinant from insulin.
The method of generating a mixed population of insulin-specific CD8+ TEM and TCM
employed here was based on widely-accepted protocols and has been validated for investigating
antigen-specific effector/memory T-cell responses (Carrio et al., 2004; Klebanoff et al., 2005;
Manjunath et al., 2001; Weninger et al., 2001). In vitro-cultured OT-I Tmem displayed similar
kinetics compared to long-lived in vivo-generated OT-I Tmem in response to transgenic OVA
121
expression targeted to CD11chi DC (Kenna et al., 2008). G9 Tmem in the present study displayed a
phenotype similar to that of OT-I T cells cultured under equivalent conditions (Kenna et al., 2008).
Despite G9 Tmem displaying a similar phenotype to OT-I Tmem, it is possible that there would be
differences in tolerance outcomes between the in vitro-generated G9 Tmem used in the present
study and in vivo-generated G9 Tmem. To test whether this was the case, G9 T cells would be
transferred to congenic B16A recipients, the B16A recipients immunised with InB15-23/poly I:C and
six weeks later TEM and TCM sorted from recipients. These in vivo-generated G9 Tmem could then
be transferred to recipients alongside with in vitro-generated G9 Tmem and analyses performed
accordingly. It is possible that the more implicit signals available to G9 T cells during culture skew
the ‘memory’ phenotype toward one more readily amenable to peripheral tolerance induction. The
aforementioned experiment may address this. With regard to effector differentiation, G9 T cells
displayed an activated, effector-like phenotype after stimulation with InsB15-23 and IL-2, expressing
similar levels of CD44 and CD62L at day 3 of culture compared to OT-I T cells stimulated with
OVA257-264 and IL-2 (Salmond et al., 2014). Salmond et al. reported little KLRG1 and IL-7Rα
expression on the surface of OT-I CTL which was similar with the phenotype of G9 T cells at day 3
of culture in the present study. Salmond et al. also reported intracellular expression of granzyme B
by effector differentiated OT-I T cells. Unstimulated OT-I T cells were not included in the analysis
by Salmond et al., therefore, background staining on naïve T cells, as observed for G9 T cells in
Figure 4.3, cannot be conclusively ruled out. To skew effector G9 T cells present at day 3 toward a
memory-like phenotype, T cells were cultured for a further two days in the presence of IL-15, since
IL-15 has been shown to maintain the memory phenotype of CD8+ T cells in the absence of antigen
(Liu et al., 2002). At the end of the culture period, in the absence of InsB15-23 and the presence of
IL-15, G9 Tmem displayed reduced expression of co-inhibitory markers PD-1, LAG-3 and CTLA4, consistent with development from post-activated effector T cells to Tmem. Tmem represent a
heterogenous population of TEM and TCM (Olson et al., 2013), therefore investigating the response
of a mixed population of insulin-specific TEM and TCM to transgenic proinsulin expression was valid
and physiologically relevant.
It might be possible that a recently identified subset of TEM, termed tissue-resident effector
memory T cells (TRM), play an important role in diabetes pathogenesis. TRM express CD69 and
CD103, do not recirculate and have been identified in the intestinal epithelium and the epidermis of
the skin (Carbone, 2015). One study indicated expression of CD69 and CD103 on TEM within the
pancreas, potentially indicating the presence of TRM in this site (Casey et al., 2012). Since TRM
represent a population of potentially long-lived TEM with potent effector function, pancreas-residing
TRM may prove another barrier to an immunotherapeutic treatment for T1D. Another study
demonstrated CD103 expression on CD8+ T cells was necessary for rejection of islet grafts (Feng et
122
al., 2002). Additional studies are required to further characterise and quantify potential pancreasresiding TRM, then it may be possible to investigate the role of these TRM in the context of
autoimmune diabetes and how this could be manipulated in a therapeutic approach.
G9 Tmem proliferated to a lesser extent than naïve G9 T cells in pLN of NOD recipients.
This is seemingly in contrast to the long-standing dogma that Tmem are more readily activated and
responsive to antigen than their naïve counterparts (Curtsinger et al., 1998; Tanchot et al., 1997).
There is recent evidence, however, to suggest this may not always hold true. Transferred naïve OT-I
T cells proliferated to a greater extent than OT-I TCM in pLN of mice expressing OVA in the islets
(Mehlhop-Williams & Bevan, 2014). This observation is analogous to the reduced proliferation of
G9 Tmem, compared to naïve G9 T cells, in pLN of NOD recipients in the present study. Also, G9
Tmem comprise approximately 60% TCM, which when considering TCM might proliferate less than
naïve T cells in response to cross-presented antigen, may account for the reduced proliferation of
G9 Tmem compared to naïve G9 T cells in response to islet-derived insulin. The evidence
suggesting naïve T cells proliferate more so in response to limited amounts of antigen may hold true
in the present study as young, male NOD were used as recipients and it might be expected that islet
inflammation would be mild in these mice, and therefore a reduced co-stimulatory environment
with presentation of insulin determinants compared to that of older female NOD mice. It might be
possible that G9 Tmem proliferate to a greater extent in pLN of older, female NOD recipients
compared to the young, male NOD recipients used. This is true for IGRP-reactive CD8+ T cells, as
8.3 T-cell proliferation in pLN has been shown to be increased with the age of NOD recipients
(Krishnamurthy et al., 2006). Work from the host laboratory suggests transfer of G9 Tmem to prediabetic female NOD mice triggers accelerated β-cell damage and subsequent hyperglycaemia.
Also, G9 Tmem secrete IFN-γ following in vitro InsB15-23 re-stimulation, which indicates rapid
production of cytotoxic mediators with less antigen requirement, that is, in the absence of InsB15-23
immunisation, than naïve G9 T cells. Programming of G9 thymocytes during thymic selection in the
presence of thymic insulin presentation may account for the reduced G9 Tmem response observed
in response to islet-derived insulin presentation. Despite the extensive use of TCR transgenic CD8+
T-cell models to investigate peripheral T-cell tolerance, a pertinent issue that is not addressed is the
effect of thymic selection in the presence of cognate antigen on the CD8+ T cell. Data from the
present study indicate targeting proinsulin expression to MHC class II+ APC induces rapid deletion
and unresponsiveness in insulin-specific CD8+ Tmem, and this was despite thymic selection of G9
thymocytes in the presence of insulin antigen.
The reduced recall ability of G9 Tmem (Fig. 4.13) compared to naïve G9 T cells (Fig. 3.12)
following InsB15-23 immunisation was somewhat unexpected when the robust IFN-γ response of G9
Tmem in B16A recipients (Fig. 4.4) was considered. Naïve G9 T cells mounted, albeit a weak,
123
recall response in immunised NOD, but not B16A, recipients (Fig. 3.12). G9 Tmem, however,
mounted a robust recall response indicated by increased production of IFN-γ in B16A recipients
following InsB15-23 immunisation. In contrast, no increase in IFN-γ production was observed for
NOD recipients following immunisation. The robust recall response of G9 Tmem, compared to
naïve G9 T cells, in B16A recipients could be explained by priming of G9 T cells during culture,
suggesting G9 T cells require more than a single round of priming to respond to InsB15-23
immunisation. The absence of G9 Tmem expansion in NOD recipients after immunisation might
have been a result of low G9 Tmem proliferation in response to islet-derived insulin. It is possible
that the increased PD-1 expression by G9 Tmem, compared to naïve G9 T cells, inhibited the G9
Tmem recall response in the present study, and this could be elucidated by blocking PD-1 in the G9
Tmem adoptive transfer setting. An important point to be taken from defining the G9 Tmem
dynamics in response to InsB15-23 immunisation is the absence of G9 Tmem expansion after transfer
to proinsulin transgenic recipients in any site investigated, indicating the residual population of G9
Tmem in proinsulin transgenic recipients are functionally inactive.
Differences in the recall ability of Tmem subsets might explain variation in the recall ability
of naïve G9 T cells compared to G9 Tmem in the present study. Due to the increased recovery of
transferred G9 Tmem in spleen, compared to LN, this was the site of analysis for recall response
experiments and TEM are predominantly found in spleen compared to LN (Brinkman et al., 2013).
TCM expand in number in response to immunisation while TEM proliferate little but produce
cytolytic mediators (Olson et al., 2013). While applying such analyses as in the present study to a
mixed population of Tmem is physiologically relevant in the context of autoimmunity, there is the
possibility that interpretation of Tmem responses to immunisation may be skewed, that is,
expansion in TCM, for example, might be concealed by the reduction of TEM. At twenty one days
after transfer, there was no difference in the proportion of TEM-like and TCM-like G9 Tmem in
B16A and NOD recipients, it is unknown whether this would be the same at earlier time points. It is
plausible that in spleens of B16A recipients, G9 TEM and TCM might comprise a different proportion
compared to that in spleens of NOD recipients due to islet-derived insulin presentation modulating
G9 Tmem trafficking into and out of the spleen in NOD recipients. This may account for the
different responses of G9 Tmem to InsB15-23 immunisation observed for NOD and B16A recipients.
Studies employing soluble peptide administration as a means of T-cell tolerance induction
have shown different outcomes depending on the dose and affinity of peptide used. Using clone 4 T
cells, it was shown that administration of a Kd-HA super agonist resulted in unresponsiveness of
transferred clone 4 T cells (Smith et al., 2014). Unresponsive clone 4 T cells expressed high levels
of PD-1 and the number of clone 4 T cells in spleen was similar to PBS-treated controls. In contrast,
clone 4 T cells were deleted in recipient mice after administration of a weak agonist APL of HA.
124
After treatment with a weak agonist APL, clone 4 T cells expressed lower levels of PD-1 and were
less frequent in spleen than those treated with native peptide, which may be indicative of T-cell
deletion. However, a lower dose of strong agonist peptide was used as a comparison to the weak
agonist peptide, that is, 1µg and 10µg, respectively, in the study by Smith et al. and the difference
in peptide dose may have affected the T-cell tolerance outcomes observed. Therefore, this study
suggests different T-cell tolerance outcomes are possible depending on the strength of TCR/peptideMHC interactions. Clone 1 T cells differ from clone 4 T cells in that they express a transgenic HAspecific TCR cloned from CD8+ T cells recovered from mice expressing HA in the islets and
thymus (Lyman et al., 2005) and thus express a low-affinity TCR. Clone 1 T cells are similar to G9
T cells in that both T cells have undergone thymic selection in the presence of their cognate antigen.
Most clone 1 T cells were deleted following HA-peptide administration recipients, yet a population
of clone 1 T cells expressing PD-1 persisted after peptide administration, potentially indicating
unresponsiveness and this was in contrast to the complete deletion of clone 4 T cells following
weak agonist peptide administration. While PD-1 has been examined on T cells with regard to
unresponsiveness, PD-1 is increased on CD8+ T cells after activation and PD-1 expression has been
implicated in shaping the Tmem population. It has been shown that PD-1 expression may inhibit
conversion of TCM to TEM (Charlton et al., 2013), and in the absence of PD-1, homeostatic
expansion of CD4+ T cells resulted in a skew toward TEM-like phenotype (Charlton et al., 2015).
Overall, the data from Smith et al. suggest T cells with low-affinity TCR may undergo inefficient
tolerance induction in response to peptide administration. The response of clone 1 T cells to peptide
administration can be contrasted to the robust, effective tolerance observed for low-affinity G9
Tmem in recipient mice over-expressing proinsulin, emphasising the different tolerance outcomes
between peptide antigen administration and when antigen is targeted to MHC class II-expressing
APC.
IL-7 is a crucial cytokine for CD8+ Tmem survival (Tan et al., 2002) and in some settings
IL-7 can promote autoreactive T-cell expansion (Calzascia et al., 2008). Blocking IL-7Rα was
shown to not only prevent but also reverse hyperglycemia in NOD mice (Lee et al., 2012a). IL-7Rα
blockade favoured PD-1 up-regulation on effector T cells using a BDC2.5 T-cell adoptive transfer
model and PD-1 involvement was further implicated as PD-1 blockade reversed normoglycemia in
IL-7Rα blockade-cured NOD mice. Similarly, Penaranda et al. demonstrated reversal of
hyperglycemia in NOD mice by IL-7Rα blockade and PD-1 was implicated in this reversal also
(Penaranda et al., 2012). The number of Tmem within pLN was not reduced after anti-IL-7Rα
treatment, but the ability of Tmem to produce IFN-γ and to transfer diabetes to NOD.SCID mice
was impaired, consistent with inhibition afforded by PD-1 expression rather than a deletional
mechanism. In the context of islet transplantation, blocking IL-7Rα protected allogeneic islet grafts
125
from rejection in streptozotocin (STZ)-induced diabetic mice (Mai et al., 2014). Yet again,
protection was associated with reduced T-cell IFN-γ production and PD-1 expression. Therefore,
inhibiting Tmem responses is beneficial in impeding diabetes development in NOD mice. From the
perspective of immunotherapy development, however, a more targeted, antigen-specific approach
would be favourable and as such, the present study has demonstrated insulin-specific CD8+ Tmem
are rapidly inactivated when proinsulin is expressed under control of an MHC class II promoter.
A clinical trial has tested whether depletion of TEM in recent-onset T1D patients had any
effect on disease progression (Rigby et al., 2013). In this study, a dimeric fusion protein, alefacept,
was used to target CD2-expressing TEM and was shown to deplete TEM more readily than TCM.
Rigby et al. rationalised that this was a more selective method of pathogenic T-cell depletion, with
less widespread immunosuppression than general immune suppressive agents such as cyclosporine.
At one year follow-up, alefacept-treated patients showed a reduction in CD4+ TEM and TCM
populations and reduced CD8+ TCM compared to placebo-treated patients, while CD4+Foxp3+ Treg
were unaltered. Most encouragingly, insulin use had decreased as a result of alefacept treatment. An
issue not thoroughly addressed in the study by Rigby et al. was maintenance of protective
immunity, and assessment of patient responses to common viral pathogens may have elucidated
this, at least to some extent. These immunological and metabolic outcomes suggest targeting TEM
and TCM populations using a therapy which maintains Treg may be a useful approach for treatment
of T1D. Therefore, taking the notion of ‘selective therapy’ one step further and deleting
diabetogenic TEM and TCM in a β cell antigen-specific manner might allow for interruption of
pathogenic islet autoimmunity without compromising protective immunity.
Overall, this chapter investigated peripheral tolerance induction in insulin-specific CD8+
Tmem. A method of generating G9 Tmem was established using in vitro culture and verified by
analysis of phenotypic markers. G9 Tmem were shown to proliferate extensively in proinsulin
transgenic recipients while the number of G9 Tmem recovered from secondary lymphoid organs
rapidly decreased. The remaining population of G9 Tmem in proinsulin transgenic recipients was
minimal and displayed reduced TCR expression than those in NOD recipients. Similarly to naïve
G9 T cells, G9 Tmem did not expand or produce IFN-γ in response to InsB15-23 immunisation in
proinsulin transgenic recipients. These observations in proinsulin transgenic recipients are
consistent with the rapid in activation of OT-I Tmem in MII.OVA mice. In NOD recipients, G9
Tmem waned slowly overtime in secondary lymphoid organs and prior to this contraction, G9
Tmem produced IFN-γ without InsB15-23 immunisation. In the absence of cognate antigen, that is, in
B16A recipients, the number of G9 Tmem was maintained and G9 Tmem produced IFN-γ twenty
one days after transfer. Therefore, insulin-specific CD8+ TCM and TEM are rapidly inactivated after
transfer to mice which transgenically express proinsulin under control of an MHC class II promoter,
126
consistent with the induction of peripheral tolerance. Since transgenic expression of proinsulin
induces rapid inactivation in naïve and memory G9 T cells, it would be logical to then determine the
effect on diabetes development of introducing tolerogenic proinsulin expression in non-transgenic
NOD mice. The following chapter will examine the effect of proinsulin-encoding HSPC/BM
transfer on diabetes development in sub-lethally conditioned NOD recipients.
127
4.4 Future directions
This chapter has addressed tolerance induction in CD8+ Tmem specific for an important islet
antigen, insulin. Since male NOD mice develop insulitis and subsequent autoimmune diabetes to a
reduced extent than female, it may be possible that reduced islet inflammation in male NOD
recipients would alter the response of G9 Tmem to islet-derived insulin presentation. It is important
to address the effect of increased islet inflammation, as seen in pre-diabetic female NOD mice, on
the slow deletion of G9 Tmem. To test this, G9 Tmem could be transferred to male and female
NOD recipients at ages, that represent different, defined stages of diabetes development and then
effector function assessed. Furthermore, G9 Tmem proliferation in pLN of NOD mice was reduced
compared to naïve G9 T cells. Since a proportion of G9 Tmem expressed PD-1 prior to transfer, this
may explain the reduced proliferation of G9 Tmem to islet-derived insulin presentation and the
involvement of PD-1 could be examined by blocking PD-1 or PD-L1 in NOD recipients. G9 Tmem
expression of PD-1 may also explain the reduced recall response to immunisation of G9 Tmem
compared to naïve G9 T cells. PD-1/PD-L1 blockade and immunisation might elucidate whether
this is the case. Finally, G9 Tmem displayed effector function as demonstrated by IFN-γ
production, therefore, G9 Tmem may contribute to diabetes progression in NOD mice. To test this,
G9 Tmem would be transferred to male and female NOD and proinsulin transgenic recipients and
6-8 weeks later, just prior to diabetes onset, pancreas and secondary lymphoid organs could be
analysed for β-cell antigen-specific CD8+ T cells as indicators of islet autoimmunity.
128
Chapter 5
Transfer of proinsulin-encoding BM and IL-2/αIL-2 complex alter
diabetes development in an age-dependent manner
129
5.1 Introduction
The previous chapter investigated tolerance induction in G9 Tmem. Since the potentially
diabetogenic T cells were rapidly inactivated after transfer to mice expressing transgenic proinsulin,
the question is raised whether introducing tolerogenic proinsulin expression and, as a consequence,
tolerance to insulin would affect diabetes development in non-transgenic NOD mice. OT-I T cells
are inactivated in non-transgenic recipients after OVA-encoding BM transfer (Coleman et al.,
2013). BM transfer is, therefore, a means to possibly instate the tolerogenicity of transgenic
proinsulin expression in non-transgenic NOD recipients. Antigen-encoding BM/HSPC transfer
might represent a clinical approach to developing an immunotherapy which requires cognate T-cell
inactivation.
Re-placing the haematopoietic compartment of NOD mice by transfer of allogeneic BM
from non-autoimmune-prone BALB/c mice prevents autoimmune diabetes (Ikehara et al., 1985). In
addition, transfer of purified, allogeneic HSC protects NOD mice from diabetes after lethal
irradiation (Beilhack et al., 2003). Therefore, diabetes protection by means of allogeneic HSC
transplantation suggests an immunological re-set from a diabetogenic repertoire to one tolerant to βcell antigens is possible. Transplantation of allogenic BM/HSPC poses substantial risks such as
highly toxic conditioning regimens, graft rejection and graft-versus-host disease (GVHD) [reviewed
in (Leventhal et al., 2013; Ramadan & Paczesny, 2015)]. Transplantation of autologous HSPC/BM,
however, might negate at least some risks associated with such toxic procedures. Transfer of
autologous, mobilised HSPC after sublethal, yet relatively high dose, irradiation (800cGy) protected
6-week old NOD recipients from autoimmune diabetes and this protection was attributed to
expansion of Foxp3+ Treg (Kared et al., 2006). Re-setting of the diabetogenic immune compartment
and subsequent protection from diabetes can be achieved even when autoimmune-prone NOD
HSPC are transferred, suggesting the conditioning regimen itself affords protection from diabetes
progression. In humans, transplantation of autologous HSC has proved to be one of the few
efficacious therapies, in terms of metabolic outcome, for recent-onset T1D patients. Voltarelli et al.
demonstrated that transplantation of autologous, mobilised HSC to T1D patients conditioned with
high-dose cyclophosphamide and anti-thymocyte globulin resulted in insulin independence for a
proportion of patients (Voltarelli et al., 2007). Immunologically, there was a reduction in the
humoral response to GAD while metabolically, an increase in C-peptide was observed in HSCtransplanted patients. Similarly, Li et al. demonstrated reduced exogenous insulin usage and
increased C-peptide in a cohort of recent-onset Chinese T1D patients (Li et al., 2012). These
promising human trials exemplify transplantation of autologous HSC as a means to overcome
established β-cell autoimmunity. However, results from these trials suggest relapse is mostly
inevitable, which might indicate a requirement for long-lasting immune tolerance to completely
130
reverse T1D. Ideally, such immune tolerance would be β-cell antigen-specific, directly impeding
diabetogenic autoimmunity without compromising protective, anti-pathogen immune responses.
Antigen-specific immunotherapy would be preferred to non-specific immunosuppression, such as
anti-CD3 mAb therapy, that has been previously trialed in mice (Chatenoud et al., 1994) and
humans (Keymeulen et al., 2010b), due to safety considerations such as EBV re-activation and Tcell cytokine release, of particular consideration in the context of T1D due to the young age of
onset.
Since antigen-specific therapy represents a means to inactivate diabetogenic T-cell
responses without compromising protective immunity, the question is raised as to which β-cell
antigen might be targeted for tolerance induction. The notion of insulin as the primary antigen in
autoimmune diabetes in NOD mice is exemplified by the fact that NOD mice in which proinsulin
expression is under control of an MHC class II promoter are protected from insulitis and diabetes
development (French et al., 1997). Whereas transgenic over-expression of other important β-cell
antigens, IGRP (Krishnamurthy et al., 2006) and GAD65 (Jaeckel et al., 2003), does not protect
NOD mice from diabetes. When proinsulin is transgenically expressed, the inherently diabetogenic
repertoire of the NOD mouse is most likely regulated centrally and peripherally from birth, perhaps
even in utero, instigating robust protection from autoimmune events that underlie diabetes
development. Transfer of proinsulin-encoding HSPC to immune-ablated NOD mice protected from
diabetes (Chan et al., 2006; Steptoe et al., 2003). The use of a conditioning regimen, required to
create ‘space’ for HSPC reconstitution, that ablated the diabetogenic immune repertoire of NOD
mice meant the transferred HSPC completely re-established the immune compartment, resulting in
protection from diabetes. In an immune-ablated recipient, the pathogenic T cells have been cleared
by the irradiation, therefore antigen-expressing APC, derived from transferred HSPC, are not
required to actively induce T-cell tolerance in this setting. Considering T1D, when correctly
managed, is generally not fatal, it is especially pertinent to ensure patient safety before any
immunotherapy might be adopted clinically. Therefore, given the potential tolerogenic capacity of
transgenic proinsulin expression to inactivate naïve and memory insulin-specific CD8+ T cells, it is
advantageous to determine the efficacy of proinsulin-encoding BM/HSPC transfer on impeding
diabetes development when performed under reduced intensity, immune-preserving conditions.
Using Treg to impede β-cell autoimmunity in NOD mice has been achieved through various
approaches including generation of an immunosuppressed, Treg-favoured environment (Battaglia et
al., 2006), modulation of APC to promote Treg function (Richer et al., 2012) and recruitment of
Treg to islets (Montane et al., 2011). A complex of IL-2 and an anti-IL-2 mAb, JES6-1, specifically
expands Treg (Webster et al., 2009) and ameliorates hyperglycemia in diabetic NOD mice
(Grinberg-Bleyer et al., 2010). It is plausible that expanding Treg synchronously with transfer of
131
proinsulin-encoding BM/HSPC might induce or expand (pro)insulin-specific Treg, mediating
insulin-specific regulation as well as bystander suppression toward other β-cell specificities. An
example of a combination therapy based on similar principles employed the T-cell modulator γaminobutyric acid (GABA) and proinsulin administration which showed synergistic efficacy to
restore glycemic control in recent-onset diabetic NOD mice (Tian et al., 2014). This combination
therapy led to generation of antigen-specific Treg, inhibition of pathogenic T cells and regeneration
of β cells, while there was minimal effect of either therapy individually. A similar effect was
observed with combination therapy involving GAD65 and GABA which increased survival of islet
grafts in diabetic NOD mice (Tian et al., 2011). These studies emphasise the potential of
combinatory approaches which boost Treg and supply antigen, to synergise and induce tolerance
which results in a disease-relevant outcome. Therefore, this chapter assesses whether coupling Treg
expansion with proinsulin-encoding BM/HSPC transfer will enhance the tolerogenicity of
BM/HSPC transfer to instate tolerance to β-cell antigens and inhibit diabetes development.
In this chapter, I investigate the effect of introducing transgenic proinsulin expression on
diabetes development in wild type NOD recipients. Firstly, donor-type leukocyte development was
assessed in recipients of proinsulin transgenic BM after low-dose irradiation. Secondly, the efficacy
of an IL-2/αIL-2 complex to expand Treg was assessed in NOD mice. Finally, proinsulin transgenic
BM/HSPC were transferred to NOD mice at defined ages and diabetes incidence was assessed to
determine the disease stage at which proinsulin-encoding BM/HSPC transfer altered diabetes
development. IL-2/αIL-2 complex was administered to BM/HSPC recipients to determine whether
this therapy synergises with tolerogenic BM/HSPC transfer.
132
5.2 Results
5.2.1 BM from proinsulin transgenic mice is not rejected after transfer to sublethally irradiated
NOD mice
It has been shown that BM encoding a neo-antigen is susceptible to immune-mediated
rejection if the neo-antigen is expressed ubiquitously or in components of BM, such as HSC, that
are required for engraftment under immune-preserving conditions (Eixarch et al., 2009; Rosenzweig
et al., 2001). One means to prevent rejection of antigen-encoding BM is to restrict neo-antigen
expression to differentiated APC using APC-specific promoters which protects the engrafting
components, that is, HSPC, from antigen-specific immune destruction. In C57BL/6 mice, HSC
transiently express MHC class II and MHC class II-promoter controlled OVA expression in HSC
lead to their immune rejection after transfer (Coleman et al., 2013). Such immune rejection might
also occur after transfer of proinsulin transgenic BM, if the proinsulin expressed from the MHC
class II promoter used becomes a target for immune attack. To test this in NOD mice where
proinsulin-specific T-cell responses are likely present, CD45 congenic NOD mice (CD45.1CD45.2+) were irradiated (300cGy) and proinsulin transgenic (CD45.1+CD45.2-) or NOD
(CD45.1+CD45.2-) BM transferred. Development of donor-type leukocytes was assessed in
peripheral blood at defined time points and in spleen six weeks after transfer as an indicator of
donor BM engraftment.
The proportion of donor-type cells in peripheral blood increased in a time-dependent manner
after transfer of BM (Fig. 5.1A). Donor-type leukocyte development was marginally higher in
recipients of NOD compared to proinsulin transgenic BM at two and four weeks after transfer but
this difference was not observed at six weeks after transfer of BM (Fig. 5.1A). This suggests,
perhaps, a more rapid development of NOD compared to proinsulin transgenic-derived leukocytes,
but that engraftment, once established, was overall similar. Donor-type DC, B cells and myeloid
cells all increased in a time-dependent manner in recipients, and this did not differ between
recipients of NOD and proinsulin transgenic BM (Fig. 5.1B-D). Notably, donor-type CD8+ and
CD4+ T cells increased more slowly than other cell subsets but was similar for recipients of
proinsulin transgenic and NOD BM (Fig. 5.1E, F). To validate observations from peripheral blood,
the frequency of donor-type leukocytes was compared between peripheral blood and spleen. Six
weeks after transfer of BM, the proportion of donor-type leukocytes was somewhat lower in spleen
than peripheral blood for both recipients of proinsulin transgenic and NOD BM, but overall there
was no difference in the frequency of donor-type leukocytes in spleen between recipient groups
(Fig. 5.2A). The proportion of donor-type myeloid cells was marginally lower in spleen than blood
of recipients of proinsulin transgenic BM, yet this was not different in recipients of NOD BM (Fig.
5.2D). There was no significant difference between spleen and peripheral blood in the proportion of
133
other donor-derived leukocyte subsets at this time point (Fig. 5.2B, C, E, F) nor was there a
difference between recipients of proinsulin transgenic and NOD BM for each cell subset analysed
(Fig. 5.2B-F). There was a trend toward a decreased donor-type B cells and myeloid cells in spleen
compared to peripheral blood of proinsulin transgenic and NOD BM recipients (Fig. 5.2C, D).
While not significant, this might account for the reduced proportion of donor-type leukocytes in
spleen compared to peripheral blood of BM recipients (Fig. 5.2A). However, a substantial level
(greater than 50%) of donor chimerism was still observed in spleen of recipients of proinsulin
transgenic and NOD BM (Fig. 5.2A).
The engraftment studies described above indicate that BM expressing MHC class II
promoter-controlled proinsulin was not rejected in recipients which differs from that observed for
BM expressing OVA under control of an MHC class II promoter. HSC from C57BL/6 mice
transiently express MHC class II which resulted in priming of OVA transgene-specific T-cell
responses and subsequent rejection of OVA-encoding BM when OVA was expressed under an
MHC class II promoter (Coleman et al., 2013). It is possible that HSC from NOD mice do not
express MHC class II, thereby permitting the observed engraftment of proinsulin transgenic BM in
wild type NOD recipients. To test this, surface expression of MHC class II, I-Ag7, was determined
on various mature and progenitor cell subsets within non-transgenic NOD BM. A small proportion
of lin-c-kit+ progenitor cells expressed I-Ag7 (Fig. 5.3). A similar trend for I-Ag7expression was
observed on the less differentiated c-kit+sca-1+ and c-kit+sca-1+CD135- populations (Fig. 5.3). With
the caveat of minimal cells in the lin-c-kit+sca-1+CD135-CD48-CD150+ long-term repopulating
HSC (LT-HSC) population to analyse, there appeared to be a small shift in the level of IAg7expression relative to isotype control (Fig. 5.3). These results indicate HSC in NOD mice may
express MHC class II, particularly in the LT-HSC population, which might lead to expression of the
proinsulin II transgene by LT-HSC. In this experiment the only control for I-Ag7 mAb binding was
the level of relevant isotype control staining. To more stringently control for non-specific binding of
the anti-I-Ag7 mAb, and to allow for more accurate interpretation of these data, I-Ag7 expression
was assessed on known MHC class II-expressing and MHC class II-negative mature cell
populations. CD11chi DC and CD19+B220+ B cells served as positive controls for MHC class II
expression as both DC and B cells express MHC class II in NOD mice (De Riva et al., 2013). I-Ag7
expression was observed on a substantial proportion of B cells and DC within NOD BM (Fig. 5.3).
On the other hand, T cells do not express MHC class II and were used to assess the level of nonspecific binding of the anti-I-Ag7 mAb. As expected, both CD3+CD4+ and CD3+CD8+ T cells
displayed nominal I-Ag7expression compared to B cells and DC (Fig. 5.3).With regard to progenitor
cell subsets in experiment 2, low I-Ag7 expression by lin-c-kit+ progenitors cells was observed when
compared to the isotype control (Fig. 5.3). Similarly, only a very small proportion of cells from the
134
less differentiated subsets displayed I-Ag7 expression (Fig. 5.3). When LT-HSC were analysed, a
marginal increase in I-Ag7 expression was observed relative to the isotype control (Fig. 5.3). Due to
variation observed between two individual experiments, further investigation is required regarding
I-Ag7 expression by HSPC from NOD mice. However, these data are to some extent consistent with
the study by Coleman et al. demonstrating MHC class II expression on LT-HSC in C57BL/6 mice.
It should also be noted that directly measuring expression of the proinsulin transgene using PCR
could be used in conjunction with antibody-mediated detection of MHC class II to ascertain
whether proinsulin transgene expression by HSC occurs in this setting and if so, why this proinsulin
transgene expression precludes immune rejection of transferred BM/HSPC facilitated by low-dose
irradiation.
Transfer of proinsulin transgenic and NOD BM after low-dose irradiation allows for a high
level of donor-type leukocyte development, particularly in APC subsets, presumably as a result of
effective engraftment. The similarity in the frequency of donor-type leukocytes observed in blood
and spleen validates using peripheral blood to assess donor engraftment. The evidence suggests
HSPC from NOD mice might express MHC class II but this needs to be investigated further. These
data demonstrate BM encoding proinsulin under control of an MHC class II promoter is not rejected
in wild type NOD recipients after low-dose irradiation and thus BM transfer provides a means to
instate the potentially tolerogenic capacity of transgenic proinsulin expression to wild type NOD
recipients.
135
Overall
A
DC
B
100
**
1 0s0 T g
p r o In
**
nonTg
80
% donor
% donor
80
60
40
60
40
20
20
0
0
p r o In s T g
NOD
0
10
20
30
40
50
0
10
D a y s (p o s t tra n s fe r)
B cells
C
50
p r o In s T g
nonTg
80
% donor
60
40
60
40
20
20
0
0
0
10
20
30
40
50
0
10
20
30
40
50
D a y s (p o s t tra n s fe r)
D a y s (p o s t tra n s fe r)
CD8+ T cells
CD4+ T cells
F
100
100
p r o In s T g
p r o In s T g
nonTg 80
nonTg
% donor
80
% donor
40
100
80
E
30
Myeloid cells
D
100
% donor
20
D a y s (p o s t tra n s fe r)
60
40
60
40
20
20
0
0
0
10
20
30
D a y s (p o s t tra n s fe r)
40
0
10
20
30
40
50
D a y s (p o s t tra n s fe r)
Figure 5.1 – No difference in donor-type leukocyte development in peripheral blood from
recipients of proinsulin transgenic and NOD BM.
NOD.CD45.2 (CD45.1-CD45.2+) mice were irradiated (300cGy) prior to transfer of 2x107 bulk BM
cells from NOD (CD45.1+CD45.2-, blue) or proinsulin transgenic (CD45.1+CD45.2-, red) mice. At
indicated time points, the proportion of donor-type (CD45.1+CD45.2-) total donor cells (A), DC
(CD11chi) (B), B cells (B220+) (C), myeloid cells (CD11b+) (D), CD8+ T cells (E) and CD4+ T cells
(F) were determined in peripheral blood using flow cytometry. Data are pooled from two independent
experiments. n=4-5. Error bars show S.D. Two-way ANOVA followed by Bonferroni post-test.
(**p<0.01: proIns Tg vs NOD).
136
Overall
A
100
DC
B
n .s .
*
n .s .
100
***
80
**
% donor
% donor
80
60
40
60
40
20
20
0
0
NOD
p r o In s T g
NOD
b lo o d
NOD
p r o In s T g
s p le e n
B cells
C
p r o In s T g
NOD
b lo o d
s p le e n
Myeloid cells
D
100
p r o In s T g
*
100
n .s .
80
% donor
% donor
80
60
40
20
60
40
20
0
0
NOD
p r o In s T g
NOD
b lo o d
p r o In s T g
p r o In s T g
p r o In s T g
s p le e n
CD4+ T cells
F
100
100
80
80
n .s .
% donor
n .s .
% donor
NOD
b lo o d
s p le e n
CD8+ T cells
E
NOD
60
40
20
60
40
20
0
0
NOD
p r o In s T g
b lo o d
NOD
p r o In s T g
s p le e n
NOD
p r o In s T g
b lo o d
NOD
p r o In s T g
s p le e n
Figure 5.2 – Substantial donor-type leukocyte development in spleen from recipients of NOD
and proinsulin transgenic BM.
NOD.CD45.2 (CD45.1-CD45.2+) mice were irradiated (300cGy) prior to transfer of 2x107 bulk BM
cells from NOD (CD45.1+CD45.2-) or proinsulin transgenic (CD45.1+CD45.2-) mice. Six weeks later,
the proportion of donor-type (CD45.1+CD45.2-) total donor cells (A), DC (CD11chi) (B), B cells
(B220+) (C), myeloid cells (CD11b+) (D), CD8+ T cells (E) and CD4+ T cells (F) were determined in
peripheral blood and spleen using flow cytometry. Data are pooled from two independent experiments.
Each point represents a single mouse. One-way ANOVA followed by Tukey post-test. Error bars show
S.D. (*p<0.05, **p<0.01, ***p<0.001).
137
A
lin-c-kit+sca-1+
CD135-CD48CD150+
lin+c-kit-
lin-c-kit+
c-kit
sca-1
lin-c-kit+sca-1+
CD135-CD48+
CD48
Experiment 1
B
CD150
lin
c-kit
CD135
lin-c-kit+sca-1+
CD48
Experiment 2
Isotype
I-Ag7
CD4+ T cells
lin+c-kit-
CD8+ T cells
lin-c-kit+
lin-c-kit+sca-1+
DC
lin-c-kit+sca-1+
CD135-CD48+
B cells
I-Ag7
lin-c-kit+sca-1+
CD135-CD48CD150+
I-Ag7
Figure 5.3 – HSPC from NOD BM express variable levels of MHC class II.
BM from NOD mice was harvested and I-Ag7 expression on mature, lin-positive leukocyte subsets and
HSPC was determined using flow cytometry. (A) Gating strategy of HSPC populations in descending
order of reduced differentiation and (B) I-Ag7 (solid line) and isotype (dashed line) expression on
CD3+CD4+ T cells, CD3+CD8+ T cells, B220+CD19+ B cells, CD11c+CD19- DC, lin+c-kit- mature BM
cells, lin-c-kit+ HPC, lin-c-kit+sca-1+ HSC, lin-c-kit+sca-1+CD135- HSC and lin-c-kit+sca-1+CD135CD48-CD150+ LT-HSC. Data are from two independent experiments.
138
5.2.2 IL-2/αIL-2 complex expands Treg in spleen and LN of NOD mice
Transfer of proinsulin-encoding BM as an approach to induce insulin-specific CD8+ T-cell
tolerance in NOD mice might be benefited by concurrently enhancing Treg activity. Treatment with
a cytokine/antibody complex comprised of IL-2 and the anti-IL-2 mAb, JES-61 (IL-2/αIL-2),
expands Treg in C57BL/6 mice (Webster et al., 2009). To determine whether Treg are expanded
and the dynamics of this in NOD mice, IL-2/αIL-2 complex or PBS was administered to NOD mice
for three consecutive days and the total number, frequency and phenotype of Treg analysed in
spleen and LN at defined time points. Little change was observed in the total number of
CD4+FoxP3+ Treg or their proportion of total CD4+ T cells in spleen after one or two days of IL2/αIL-2 treatment (Fig. 5.4A, B). However, three days after the third and final administration of IL2/αIL-2, the number of CD4+FoxP3+ Treg and their proportion of CD4+ T cells in spleen was
increased almost two-fold over PBS-treated mice (Fig. 5.4A, B). The total number and proportion
of Treg of total CD4+ T cells peaked at this time and remained elevated for at least four days after
treatment (Fig. 5.4A, B). As expected, neither the number nor proportion of Treg were altered in
PBS-treated mice (Fig. 5.4A, B).
In LN, little change was observed in the total number and proportion of Treg after two days
of IL-2/αIL-2 treatment (Fig. 5.4C, D). An increase in the total number of Treg was observed on the
final day of IL-2/αIL-2 treatment and the number of Treg peaked two days after cessation of
treatment (Fig. 5.4C). The increase in Treg number was transient, and by four days after cessation
of treatment the number of Treg in LN of IL-2/αIL-2-treated mice returned to the level observed in
PBS-treated mice (Fig. 5.4C). The number and proportion of Treg in LN was not altered in PBStreated mice (Fig. 5.4C, D). In LN of IL-2/αIL-2-treated mice, the proportion of Treg of total CD4+
T cells was increased at one day after cessation of treatment and this increase was maintained until
at least four days after treatment (Fig. 5.4D). The pattern of Treg expansion, defined as a proportion
of total CD4+ T cells, was similar in LN and spleen, although the proportion of Treg in the total
CD4+ T cell population was lower in LN than in spleen (Fig. 5.4B, D). These data confirm Treg
expansion in spleen and LN of NOD mice after treatment with IL-2/αIL-2 and this may represent a
therapeutic option in NOD mice to expand Treg, which might also act synergistically with transfer
of proinsulin-encoding BM to protect autoimmune-prone NOD mice from diabetes development.
139
B
***
***
+
Foxp3
2000
+
T re g
6000
4000
***
30
( % o f t o t a l C D 4 T c e lls )
***
8000
Foxp3
+
T r e g /s p le e n ( x 1 0
-3
)
A
***
Spleen
10
IL - 2 / Il- 2
PBS
0
1
2
3
4
5
6
0
1
2
D ay
3
4
5
6
***
***
***
D ay
Treatment
Treatment
D
*
**
*
-3
)
2000
+
T re g
1500
+
Foxp3
1000
500
0
( % o f t o t a l C D 4 + T c e lls )
C
T r e g /L N ( x 1 0
***
20
0
0
Foxp3
***
*
15
10
LN
5
0
0
1
2
3
4
5
6
0
1
2
D ay
Treatment
3
4
5
6
D ay
Treatment
Figure 5.4 - Foxp3+ Treg expansion in spleen and LN of NOD mice following IL-2/αIL-2 complex
treatment.
NOD mice were injected with IL-2/αIL-2 or PBS i.p. for three consecutive days. At indicated time
points, the total number of CD4+FoxP3+ Treg (A, C) and their proportion of total CD4+ T cells (B, D)
in spleen (A, B) and LN (C, D) were determined using flow cytometry. Data are from 1-2 experiments
at each time point, n=4-7. Error bars show S.D. Two-way ANOVA followed by Bonferroni post-test.
(*p<0.05, **p<0.01, ***p<0.001 between IL-2/αIL-2 and PBS).
140
5.2.3 IL-2/αIL-2 complex enhances Treg expression of CD25, Foxp3 and other markers indicative
of suppressive function
Next, the phenotype of IL-2/αIL-2-expanded Treg was assessed as an indicator of Treg
function. Increased CD25 and Foxp3 expression by Treg is observed following IL-2/αIL-2
treatment (Webster et al., 2009) and increased expression of these markers has been associated with
increased suppressive function (Wan & Flavell, 2007). Similarly, increased expression of TNFRII,
glucocorticoid-induced TNFR-related receptor (GITR), LAG-3, and CD103 have been associated
with suppressive function Treg (Chen et al., 2008b; Huang et al., 2004; McHugh et al., 2002;
Zabransky et al., 2012). Since IL-2/αIL-2 complex expands Treg, it might be expected that this
increase would be due to induction of peripheral iTreg, rather than an increase in emergence of
thymic-derived nTreg. Expression of the transcription factor, helios, is a marker of nTreg (Singh et
al., 2015; Thornton et al., 2010). To determine the phenotype of IL-2/αIL-2-expanded Treg, NOD
mice were treated with IL-2/αIL-2 or PBS for three consecutive days and Treg phenotype was
analysed two days later.
CD25 expression by CD4+Foxp3+ Treg from spleen and LN was significantly higher in IL2/αIL-2-treated mice than in PBS-treated controls (Fig. 5.5A). Foxp3 expression by Treg was
increased in spleen following IL-2/αIL-2 treatment, but Foxp3 expression was not significantly
altered in LN (Fig. 5.5B). As expected, most CD4+Foxp3+ Treg in spleen and LN expressed GITR,
and the proportion of GITR-expressing Treg was not altered by IL-2/αIL-2 treatment (Fig. 5.5C).
However, the level of GITR expression by Treg from spleen and LN was significantly increased in
response to IL-2/αIL-2 treatment (Fig. 5.5D). The proportion of Treg from spleen and LN
expressing LAG-3 remained unchanged after IL-2/αIL-2 treatment (Fig. 5.5E). The proportion of
TNFRII-expressing Treg was increased in IL-2/αIL-2-treated mice compared to PBS-treated mice
for spleen, yet remained unchanged in LN (Fig. 5.5F). The proportion of Treg from spleen and LN
expressing CD103 was increased after IL-2/αIL-2 treatment (Fig. 5.5G). Finally, the proportion of
Treg expressing helios in spleen and LN did not differ between IL-2/αIL-2-treated and PBS-treated
mice (Fig. 5.5H).
These data are in agreement with increased expression of CD25 and Foxp3 following IL2/αIL-2 treatment and constitutive GITR expression by Treg observed in other studies (McHugh et
al., 2002; Webster et al., 2009). The fact that IL-2/αIL-2 treatment did not alter the proportion of
helios-expressing Treg might indicate IL-2/αIL-2-expanded Treg are iTreg, rather than nTreg,
although there is conjecture as to whether helios is an exclusive marker of nTreg (Gottschalk et al.,
2012). The increase in CD25, Foxp3, GITR, CD103 and TNFRII expression by Treg in IL-2/αIL-2treated mice may indicate an increase in suppressive function.
141
CD25
A
***
***
20000
15000
10000
5000
2000
1000
0
0
IL - 2 /  IL - 2
PBS
IL - 2 /  IL - 2
PBS
spl
G IT R e x p r e s s io n ( M F I)
+
25
**
****
15000
50
LN
GITR
D
75
IL - 2 /  IL - 2
PBS
spl
100
T re g (% )
IL - 2 /  IL - 2
PBS
LN
GITR
C
G IT R
***
3000
F o x p 3 e x p r e s s io n ( M F I)
C D 2 5 e x p r e s s io n (M F I)
25000
Foxp3
B
10000
5000
0
0
IL - 2 /  IL - 2
PBS
IL - 2 /  IL - 2
PBS
spl
E
IL - 2 /  IL - 2
PBS
LN
F
LAG-3
IL - 2 /  IL - 2
PBS
spl
LN
TNFRII
80
15
T re g (% )
T re g (% )
***
40
T N F R II
L A G -3
+
+
10
60
5
20
0
0
IL - 2 /  IL - 2
PBS
IL - 2 /  IL - 2
PBS
spl
G
T re g (% )
10
IL - 2 /  IL - 2
LN
Helios
40
+
h e lio s
C D 103
+
T re g (% )
15
PBS
60
*
**
IL - 2 /  IL - 2
spl
H
CD103
20
PBS
LN
5
0
20
0
PBS
IL - 2 /  IL - 2
spl
PBS
IL - 2 /  IL - 2
LN
PBS
IL - 2 /  IL - 2
spl
PBS
IL - 2 /  IL - 2
LN
Figure 5.5 – Enhanced Treg expression of CD25, Foxp3, GITR, TNFRII and CD103 following
IL-2/αIL-2 complex treatment.
NOD mice were injected with IL-2/αIL-2 or PBS i.p. for three consecutive days. On day 4 of
treatment, the phenotype of CD4+Foxp3+ Treg in spleen and LN was determined using flow cytometry.
(A) CD25 MFI, (B) Foxp3 MFI, (C) GITR+ Treg (%) (D) GITR MFI, (E) helios+ Treg (%), (F) LAG3+ Treg (%), (G) TNRFII+ Treg (%) and (H) CD103+ Treg (%). Data are pooled from two independent
experiments. Each point represents single mouse. Error bars show S.D. One-way ANOVA followed by
Tukey post-test. (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001).
142
5.2.4 IL-2/αIL-2 complex increases the frequency of non-Treg immune cell populations
The IL-2/αIL-2 complex generated with the anti-IL-2 mAb, JES6-1, has been shown to be
selective for Treg in C57BL/6 mice (Webster et al., 2009) due to enhanced signalling through the
high-affinity IL-2Rα, CD25. In contrast, the anti-IL-2 mAb, S4B6, promotes signalling though IL2Rβ (CD122) leading to expansion of CD8+ Tmem and NK cells (Boyman et al., 2006). It is
possible that the selectively of IL-2/JES6-1 complex for Treg is different in NOD mice and other
non-Treg cell populations might be expanded also. Expansion of potentially pathogenic, IL-2responsive cell types such as CD8+ T cells, NK and NKT cells after IL-2/αIL-2 treatment might be
detrimental in the context of proinsulin transgenic BM transfer. To test the effect of IL-2/JES61complex on non-Treg cell subsets, NOD mice were injected with IL-2/αIL-2 or PBS daily for three
consecutive days and lymphocyte populations were analysed two or three days after the final IL2/αIL-2 administration.
The total number and proportion of CD8+ T cells of total leukocytes was not altered by IL2/αIL-2 treatment in spleen or LN (Fig. 5.6A, B). The total number and proportion of CD8+ TCM of
total CD8+ T cells remained unchanged after IL-2/αIL-2 treatment in spleen and LN (Fig. 5.6C, D).
However, the total number and proportion of CD8+ TEM of total CD8+ T cells was significantly
higher in IL-2/αIL-2-treated mice compared to PBS-treated mice in spleen, yet remained unchanged
in LN (Fig. 5.6E, F). NK and NKT cells were defined in the present study by expression of CD49b,
as NOD mice do not express NK1.1 (Baxter et al., 1997). As such, the total number and proportion
of NKT cells of total leukocytes was increased following IL-2/αIL-2 treatment in spleen, but not in
LN (Fig. 5.7A, B). Similarly, the total number and proportion of NK cells of total leukocytes was
higher in spleen of IL-2/αIL-2-treated mice compared to PBS-treated mice, but not in LN (Fig.
5.7C, D). An increase in the total number and proportion of CD11chi DC in spleen was observed in
response to IL-2/αIL-2 treatment, yet these parameters were not affected in LN (Fig. 5.7E, F).
Finally, the total number of B cells in spleen and LN did not differ significantly between IL-2/αIL2-treated and PBS-treated mice (Fig. 5.7G). Interestingly, the proportion of B cells from spleen was
significantly lower after IL-2/αIL-2 treatment, but remained unchanged in LN (Fig. 5.7H). It is
possible that IL-2/αIL-2 treatment increased the cellularity of spleen, however, there was no
significant difference between spleen cellularity in IL-2/αIL-2-treated (57.0x106±12.7) and PBStreated (40.0x106±9.9) mice. These data suggest treatment with IL-2/JES6-1 complex not only
expands Treg, but also non-Treg cell populations including CD8+ TEM, NK- and NKT-like cells as
well as DC in NOD mice.
143
B
30
15
T c e lls ( % )
10
+
5
CD8
T o ta l C D 8
+
T c e lls ( x 1 0
-6
)
A
IL - 2 /  IL - 2
PBS
spl
PBS
spl
IL - 2 /  IL - 2
LN
D
60
T c e lls )
T C M (% o f C D 8
+
-6
)
3
T C M (x 1 0
IL - 2 /  IL - 2
pbs
LN
4
2
1
40
CD8+ TCM
20
0
0
IL - 2 /  IL - 2
PBS
PBS
spl
IL - 2 /  IL - 2
IL - 2 /  IL - 2
PBS
PBS
spl
LN
E
IL - 2 /  IL - 2
LN
F
**
15
1 .5
T c e lls )
*
+
1 .0
T E M (% o f C D 8
)
10
IL - 2 /  IL - 2
C
-6
Total CD8+
T cells
0
0
PBS
T E M (x 1 0
20
0 .5
10
CD8+ TEM
5
0
0 .0
PBS
IL - 2 /  IL - 2
spl
PBS
IL - 2 /  IL - 2
LN
PBS
IL - 2 /  IL - 2
spl
PBS
IL - 2 /  IL - 2
LN
Figure 5.6 – Expansion of CD8+ TEM in spleen following IL-2/αIL-2 complex treatment.
NOD mice were injected with IL-2/αIL-2 or PBS i.p. for three consecutive days. On day 5 or 6 of
treatment, the total number of CD8+ T cells (A), CD8+ T cells as a proportion of total leukocytes (B),
the total number of CD44hiCD62Lhi CD8+ TCM (C), CD8+ TCM as a proportion of total CD8+ T cells
(D), the total number of CD44hiCD62Llo CD8+ TEM (E) and CD8+ TEM as a proportion of total CD8+ T
cells (F) were determined in spleen and LN using flow cytometry. Data are pooled from two
independent experiments. Each point represents a single mouse. Error bars show S.D. One-way
ANOVA followed by Tukey post-test. (*p<0.05, **p<0.01).
144
B
A
2 .0
N K T c e lls ( % )
)
-6
N K T c e lls ( x 1 0
4
**
2 .5
1 .5
1 .0
NKT cells
2
1
0
IL - 2 /  IL - 2
PBS
PBS
spl
IL - 2 /  IL - 2
IL - 2 /  IL - 2
PBS
PBS
spl
LN
IL - 2 /  IL - 2
LN
D
C
4
2 .5
****
***
N K c e lls ( % )
2 .0
-6
)
3
0 .5
0 .0
N K c e lls ( x 1 0
**
1 .5
1 .0
3
NK cells
2
1
0 .5
0
0 .0
IL - 2 /  IL - 2
PBS
PBS
spl
IL - 2 /  IL - 2
IL - 2 /  IL - 2
PBS
PBS
spl
LN
IL - 2 /  IL - 2
LN
F
E
*
3
*
2 .0
D C (x 1 0
-6
D C (% )
)
1 .5
1 .0
2
DC
1
0 .5
0
0 .0
PBS
G
IL - 2 /  IL - 2
PBS
spl
IL - 2 /  IL - 2
LN
H
20
PBS
spl
IL - 2 /  IL - 2
LN
*
40
25
B c e lls ( % )
)
30
-6
B c e lls ( x 1 0
IL - 2 /  IL - 2
PBS
15
10
B cells
20
10
5
0
0
PBS
IL - 2 /  IL - 2
spl
PBS
IL - 2 /  IL - 2
LN
PBS
IL - 2 /  IL - 2
spl
PBS
IL - 2 /  IL - 2
LN
Figure 5.7 – Expansion of NKT cells, NK cells and DC in IL-2/αIL-2-treated mice.
NOD mice were injected with IL-2/αIL-2 or PBS i.p. for three consecutive days. On day 5 or 6 of
treatment, the total number (left panel) and their proportion of total leukocytes (right panel) of
TCRβ+CD49b+ NKT cells (A, B), CD3-CD49b+ NK cells (C, D), CD11chi DC (E, F) and B220+ B
cells (G, H) were determined in spleen and LN using flow cytometry. Data are pooled from two
independent experiments. Each point represents a single mouse. Error bars show S.D. One-way
ANOVA followed by Tukey post-test. (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001).
145
5.2.5 Transfer of proinsulin transgenic BM alters diabetes development in an age-dependent manner
Since transferred G9 T cells are inactivated in proinsulin transgenic recipients, it is possible
that transfer of proinsulin-encoding BM to NOD recipients might inactivate host diabetogenic
insulin-specific CD8+ T cells, preventing further β-cell damage and subsequently protecting NOD
recipients from diabetes. To identify the therapeutic window in which instating tolerogenic
proinsulin expression protects from diabetes development, NOD mice at defined disease stages
were used as recipients of BM (In't Veld, 2014). In 16-week old, pre-diabetic NOD mice β-cell
autoimmunity is well-established and destructive, and these mice would possess the minimum
functional β-cell mass to maintain normoglycemia. It is possible that transfer of proinsulin
transgenic BM at this disease stage might prevent further β-cell damage, thereby inhibiting
progression to hyperglycemia. In 10-week old NOD mice substantial insulitis, that is, immune cell
infiltration in the islets, would be expected and it might be possible that proinsulin-encoding BM
transfer at this disease stage would prevent recipients developing overt diabetes. NOD mice at 6weeks of age show evidence of immune cells surrounding and infiltrating the pancreatic islets, and
it may be possible to prevent destructive insulitis and diabetes by intervening at this disease stage.
Also, expanding Treg concurrently might help to promptly suppress inflammation to allow for
mature proinsulin-encoding APC derived from transferred BM to induce T-cell tolerance. To test
the effect of proinsulin-encoding BM transfer and Treg expansion on diabetes development at
different stages of islet autoimmunity, 16-, 10- and 6-week old NOD mice were irradiated (300cGy)
prior to transfer of NOD or proinsulin transgenic BM with or without IL-2/αIL-2 administration.
Diabetes incidence was assessed commencing one week after BM transfer, for 16-week old
recipients, or at approximately 80 days of age for 6- and 10-week old recipients.
Development of diabetes did not differ between 16-week old recipients of proinsulin
transgenic BM and NOD BM (Fig. 5.8A). While minor changes were apparent, no significant
differences were observed between groups (Fig. 5.8A) While low-dose irradiation without BM
transfer appeared to reduce diabetes incidence compared to untreated NOD mice, statistical
significance was not reached due to low number of replicates in the 300cGy/no BM group (Fig.
5.8A). It should be noted that the untreated mice represent a cumulative cohort of female NOD mice
run concurrently with BM transfer experiments. As such, the onset of diabetes development in this
untreated cohort was prior to BM transfer in experimental mice (indicated by the red arrow), and all
experimental mice were negative for glycosuria prior to irradiation and BM transfer. Whilst there
were no significant differences observed between treatment or control groups based on analysis
using a log-rank test, there was a very noticeable trend toward an increase in diabetes incidence for
10-week old recipients of proinsulin transgenic BM compared to NOD BM (Fig. 5.8B). IL-2/αIL-2
treatment did not alter diabetes incidence for recipients of proinsulin transgenic BM and NOD BM
146
(Fig. 5.8B). Similarly to recipients at 16 weeks of age, low-dose irradiation without BM transfer
appeared to reduce diabetes incidence in 10-week old NOD mice compared to untreated NOD mice,
yet this was not significant (Fig. 5.8B). Diabetes incidence was significantly increased in 6-week
old recipients of proinsulin transgenic BM compared to recipients of NOD BM (Fig. 5.8C). Despite
the increased penetrance of diabetes in recipients of proinsulin transgenic BM, the time of diabetes
onset in this group was relatively similar to that of recipients of NOD BM and untreated NOD mice
(Fig. 5.8C). A noticeable trend was observed in that diabetes incidence appeared to be increased for
IL-2/αIL-2-treated recipients of NOD BM compared to PBS-treated recipients of NOD BM treated,
however, this was not significant (Fig. 5.8C). No such trend was observed for IL-2/αIL-2-treated
recipients of proinsulin transgenic BM (Fig. 5.8C). The diabetes incidence was similar between
untreated and 300cGy/no BM groups (Fig. 5.8C).
In recipients of proinsulin transgenic BM, diabetes development was not altered at 16
weeks, while there was a noticeable trend towards an increase in diabetes incidence at 10 weeks and
this increased diabetes incidence was significant at 6 weeks. It is possible that the increased diabetes
penetrance is due to priming of an immune response by the proinsulin transgene. It might be
expected that this would be more prominent in younger NOD recipients due to the predominance of
insulin specificities at this stage (Wong et al., 1999). On the other hand, transfer of NOD BM
appeared to impede diabetes development more so in younger NOD recipients, which is consistent
with an increased repertoire of potentially diabetogenic immune cells present in transferred BM of
older NOD mice. Low-dose irradiation without transfer of BM might be more effective in inhibiting
diabetes development when performed in older NOD mice, suggesting established diabetogenic
responses evident at the pre-diabetic stage may be cleared by irradiation resulting in reduced
diabetes incidence. Therefore, transfer of proinsulin transgenic BM facilitated by low-dose
irradiation affects diabetes development in an age-dependent manner.
147
16 weeks
A
100
u n tr e a te d ( n = 2 5 )
N O D B M (n = 1 1 )
P e r c e n t d ia b e t ic
80
n .s .
60
N O D B M + IL - 2 / IL - 2 ( n = 1 5 )
p r o In s T g B M ( n = 1 5 )
p r o In s T g B M + IL - 2 / IL - 2 ( n = 1 5 )
40
3 0 0 c G y , n o B M (n = 6 )
20
0
0
100
200
Irradiation +A g e
BM transfer
300
(d a y s )
10 weeks
B
100
u n tr e a te d ( n = 2 5 )
P e r c e n t d ia b e t ic
80
n .s .
60
p r o In s T g B M + IL - 2 / IL - 2 ( n = 8 )
p r o In s T g B M ( n = 9 )
N O D B M + IL - 2 / IL - 2 ( n = 1 1 )
N O D B M (n = 8 )
40
3 0 0 c G y , n o B M (n = 6 )
20
0
0
100
200
300
A g e (d a y s )
Irradiation +
BM transfer
C
6 weeks
100
P e r c e n t d ia b e t ic
u n tr e a te d ( n = 2 5 )
80
**
N O D B M (n = 8 )
N O D B M + IL - 2 / IL - 2 ( n = 7 )
60
p r o In s T g B M ( n = 1 2 )
p r o In s T g B M + IL - 2 / IL - 2 ( n = 1 1 )
40
3 0 0 c G y ., n o B M ( n = 3 )
20
0
0
100
200
300
A g e (d a y s )
Irradiation +
BM transfer
Figure 5.8 – Proinsulin transgenic BM increases the penetrance of diabetes when transferred to
6-week old NOD recipients.
Sixteen- (A), ten- (B) and six-week old (C), non-diabetic female NOD mice were irradiated (300cGy)
and 2x107 NOD or proinsulin transgenic BM cells transferred. One group were irradiated without BM
transfer and another group left untreated. Recipients of NOD and proinsulin transgenic BM were
injected with IL-2/αIL-2 or PBS i.p. on day 0, 1 and 2 after transfer. Mice were screened for glycosuria
weekly commencing one week after transfer (A) or at approximately 80 days of age (B, C). Diabetes
was confirmed by two consecutive weekly readings of non-fasting BGL>12mM. Survival curves
showing overall diabetes incidence (A-C). Data are pooled from at least three independent
experiments. Log-rank (Mantel-Cox) test. (**p<0.01: NOD BM vs proIns Tg BM).
148
5.2.6 Donor-type leukocyte development is observed in peripheral blood from 6-week old NOD
recipients of proinsulin transgenic HSPC
Clinically, transfer of HSPC is now more widely used due to the ease of harvesting blood
derived CD34+ cells in patients compared to bulk BM aspirate (Deotare et al., 2015). Furthermore,
transfer of HSPC negates the effect of transferring potentially diabetogenic immune cells residing
within bulk BM, which might have been a confounding factor in interpretation of experiments using
transfer of BM. Also, transfer of HSPC, rather than BM, meant age-mismatched NOD donors could
be used in future experiments as harvesting BM from NOD mice younger than 6 weeks of age could
be technically challenging. It is possible, however, that effects of transfer might differ depending on
whether HSPC or BM was transferred. To test this, 6-week old NOD mice were irradiated (300cGy)
prior to transfer of NOD or proinsulin transgenic HSPC. Development of donor-type leukocytes
was assessed in peripheral blood six weeks after transfer.
Lin-c-kit+ HSPC (CD45.1+CD45.2+), that were congenically distinct from recipients
(CD45.1+CD45.2), were sorted to a high purity using FACS (Fig. 5.9A, B). Six weeks after
irradiation and transfer of HSPC, there were fewer peripheral blood leukocytes in recipients of both
NOD and proinsulin transgenic HSPC compared to unirradiated NOD mice but the total number of
leukcoytes was similar for recipients of both NOD and proinsulin transgenic HSPC (Fig. 5.10A).
This indicates recipients remain lymphopenic for at least six weeks following low-dose irradiation.
The total number of donor-type DC, B cells, myeloid and T cells in peripheral blood did not differ
between recipients of NOD and proinsulin transgenic HSPC (Fig. 5.10B-F). Similarly, there was no
significant difference in the overall proportion of donor-type leukocytes between recipients of NOD
and proinsulin transgenic HSPC (Fig. 5.11A). At six weeks after transfer, the overall proportion of
donor-type leukocytes was lower in recipients of HSPC (approximately 40%, Fig. 5.11A) than in
recipients of bulk BM (approximately 60%, Fig. 5.1A). No significant differences were observed
for the proportion of donor-type cell subsets between recipients of NOD and proinsulin transgenic
HSPC (Fig. 5.11B-F). The proportion of donor-type myeloid and T cells appeared to be higher in
recipients of bulk BM (Fig. 5.1D-F) compared to recipients of HSPC (Fig. 5.11D-F), which may
account for the reduced donor-type development in recipients of HSPC compared to BM.
Importantly, development of donor-type APC subsets, DC and B cells, was similar between
recipients of HSPC (Fig. 5.11B, C) and BM (Fig. 5.1B, C).
These data demonstrate substantial development of mature leukocyte populations derived
from both NOD and proinsulin HSPC in peripheral blood of recipients after low-dose irradiation. It
should be noted that due to the nature of these long-term diabetes incidence studies, donor-type
leukocyte development was not confirmed in lymphoid tissue, however, data in Figure 5.2
demonstrated donor-type leukocyte development in peripheral blood was similar to spleen under
149
like conditions. This finding confirms HSPC transfer as a valid means to introduce transgenic
proinsulin expression to wild type NOD recipients without the confounding factor of transferring
mature immune cells as for transfer of bulk BM.
150
A
FSC
c-kit
proIns Tg
lin
FSC
NOD
lin
PI
SSC
Pre-sort
c-kit
FSC
c-kit
proIns Tg
lin
FSC
NOD
lin
PI
SSC
Post-sort
B
c-kit
NOD
proIns Tg
94.5±4.0
93.7±2.6
Figure 5.9 – High purity of FACS-sorted HSPC from NOD and proinsulin transgenic BM.
HSPC (lin-c-kit+) from BM of NOD and proinsulin transgenic were sorted using FACS. (A)
Representative FACS plots of lin- and c-kit-stained bulk BM (‘Pre-sort’, top panel) and sorted HSPC
(‘Post-sort, bottom panel) from NOD and proinsulin transgenic mice, (B) Final proportion of lin-c-kit+
HSPC of live (PI-) cells from NOD and proinsulin transgenic mice, mean ± S.D. Dot plots (A) are
representative of seven independent experiments. Data in table (B) are pooled from seven independent
experiments.
151
Overall
DC
n .s .
A
B
n .s .
***
0 .0 1 0
-3
D o n o r D C /u l ( x 1 0
T o t a l c e lls /u l ( x 1 0
-3
)
)
8
6
4
2
0 .0 0 8
0 .0 0 6
0 .0 0 4
0 .0 0 2
0 .0 0 0
0
u n ir r . N O D
C
NOD
NOD
p r o In s T g
D
Myeloid cells
-3
)
B cells
p r o In s T g
D o n o r B c e lls /u l ( x 1 0
-3
)
D o n o r m y e lo id c e lls /u l ( x 1 0
n .s .
0 .8
0 .6
0 .4
0 .2
0 .0
NOD
-3
c e lls /u l (x 1 0
n .s .
0 .1 5
0 .1 0
0 .2
0 .0
NOD
p r o In s T g
n .s .
0 .8
0 .6
0 .4
+
Donor CD4
Donor CD8
0 .4
CD4+ T cells
)
0 .2 0
0 .6
F
CD8+ T cells
n .s .
0 .8
p r o In s T g
+
c e lls /u l (x 1 0
-3
)
E
1 .0
0 .0 5
0 .0 0
NOD
p r o In s T g
0 .2
0 .0
NOD
p r o In s T g
Figure 5.10 – No difference in donor-type leukocyte development between sublethally irradiated
recipients of NOD and proinsulin transgenic HSPC.
Five- to six-week old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 2x105 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HSPC transferred. Six weeks
later, the total number of total leukocytes (A) and the total number of donor-type (CD45.1+CD45.2+)
DC (B), B cells (C), myeloid cells (D), CD8+ T cells (E) and CD4+ T cells (F) in peripheral blood
were determined using flow cytometry. Data are pooled from two independent experiments. Each
point represents a single mouse. Error bars shows S.D. One-way ANOVA followed by Tukey post-test
(A) or student’s t-test (B-F). (***p<0.001).
152
Overall
A
100
80
100
D o n o r D C (% )
T o t a l d o n o r c e lls ( % )
DC
B
n .s .
60
40
n .s .
80
60
40
20
20
0
0
NOD
NOD
p r o In s T g
C
p r o In s T g
D
B cells
Myeloid cells
100
D o n o r m y e lo id c e lls ( % )
D o n o r B c e lls ( % )
100
n .s .
80
60
40
20
0
E
40
20
NOD
p r o In s T g
CD8+ T cells
F
p r o In s T g
CD4+ T cells
100
c e lls ( % )
100
80
80
60
n .s .
+
60
+
n .s .
Donor CD4
c e lls ( % )
60
0
NOD
Donor CD8
n .s .
80
40
20
40
20
0
0
NOD
p r o In s T g
NOD
p r o In s T g
Figure 5.11 – No difference in donor-type leukocyte development between sublethally irradiated
recipients of NOD and proinsulin transgenic HSPC.
Five- to six-week old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 2x105 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HSPC transferred. Six weeks
later, the proportion of donor-type (CD45.1+CD45.2+) total cells (A), DC (B), B cells (C), myeloid
cells (D), CD8+ T cells (E) and CD4+ T cells (F) in peripheral blood were determined using flow
cytometry. Data are pooled from two independent experiments. Each point represents a single mouse.
Error bars shows S.D. Student’s t-test.
153
5.2.7 Transfer of proinsulin transgenic HSPC does not increase diabetes penetrance in 6-week old
NOD recipients
In order to compare incidence studies using BM transfer with incidence studies using HSPC
transfer, diabetes incidence was assessed for 6-week old NOD recipients of NOD and proinsulin
transgenic HSPC. Similarly to transfer of bulk BM, the age diabetes onset was not affected by
transfer of proinsulin transgenic or NOD HSPC when compared to untreated NOD mice (Fig.
5.12A). The overall incidence was similar between recipients of NOD and proinsulin transgenic
HSPC compared to untreated NOD mice (Fig. 5.12A). Unlike transfer of proinsulin transgenic BM
(Fig. 5.8C), transfer of proinsulin transgenic HSPC did not alter diabetes development in 6-week
old NOD recipients (Fig. 5.12A). Quantifying the level of donor-type leukocytes in peripheral blood
from recipients of HSPC allowed for exploration of the relationship between the level of donor-type
leukocyte development and the age of diabetes onset. There was no correlation between donor-type
leukocyte development and the age of diabetes onset for recipients of NOD HSPC (Fig. 5.12B).
Interestingly, there was a negative correlation between the proportion of donor-type leukocytes and
the age of diabetes onset for recipients of proinsulin transgenic HSPC (Fig. 5.12B). Since the
overall diabetes incidence was similar in HSPC recipients at 6 weeks of age and untreated NOD
mice, diabetes incidence following transfer of BM and HSPC can be compared between NOD
recipients at different ages. However, as diabetes incidence was not increased following proinsulin
transgenic HSPC transfer, the increased penetrance of diabetes following transfer of proinsulinencoding BM to 6-week old NOD recipients was BM-dependent.
154
A
100
p r o In s T g H P C ( n = 1 1 )
P e r c e n t d ia b e t ic
N O D H P C (n = 9 )
80
u n tr e a te d ( n = 1 7 )
60
40
20
0
0
50
100
150
200
A g e (d a y s )
Irradiation +
HPC transfer
B
T o t a l d o n o r c e lls ( % )
55
p r o In s T g H P C *
N O D HPC
50
r = - 0 .7 0
45
r = 0 .0 2
40
35
30
0
50
100
150
200
A ge of onset
Figure 5.12 – Transfer of proinsulin-encoding HSPC does not increase the penetrance of
diabetes in 6-week old NOD recipients.
Five- to six-week old female NOD mice were irradiated (300cGy) and 2x105 NOD or proinsulin
transgenic purified HSPC transferred. One group was left untreated. Mice were screened weekly for
glycosuria commencing at approximately 80 days of age. Diabetes was confirmed by two consecutive
weeks of non-fasting BGL>12mM. (A) Survival curve showing diabetes incidence. (B) Correlation of
the proportion of donor-type leukocytes in peripheral blood 6 weeks after HPC transfer and the age of
diabetes onset shown with linear regression. Data are pooled from two independent experiments. Each
point represents a single mouse. Log-rank (Mantel-Cox) test (A) and Pearson correlation (B).
(*p<0.05: proIns Tg HPC (B)).
155
5.2.8 Increased donor-type leukocyte development in 3-week old recipients of proinsulin transgenic
compared to NOD HSPC
Since transfer of proinsulin-encoding HSPC did not alter diabetes incidence when
performed in 6-week old NOD recipients, it is possible that transferring proinsulin transgenic HSPC
to NOD recipients prior to the onset of insulitis may protect from diabetes development. To test
this, 3-week old NOD mice were irradiated (300cGy) and NOD or proinsulin transgenic HSPC
transferred. Development of donor-type leukocytes in peripheral blood was assessed six weeks after
transfer. The total number of leukocytes in peripheral blood of recipients was significantly lower
than unirradiated NOD mice but was not different between recipients of NOD and proinsulin
transgenic HSPC, with or without IL-2/αIL-2 treatment (Fig. 5.13A). The total number of donortype DC, B cells, myeloid and T cells did not differ significantly between recipients of proinsulin
transgenic and NOD HSPC (Fig. 5.13B-F). The proportion of donor-type leukocytes was higher for
PBS-treated recipients of proinsulin transgenic compared to NOD HSPC (Fig. 5.14A). The
proportion of donor-type DC, CD8+ and CD4+ T cells was higher in PBS-treated recipients of
proinsulin transgenic compared to NOD HSPC, while IL-2/αIL-2 treatment reduced development of
these cell subsets for recipients of proinsulin transgenic HSPC (Fig. 5.14B, E, F). In recipients of
proinsulin transgenic HSPC, the proportion of donor-type B cells was higher compared to recipients
of NOD HSPC and this remained unchanged by IL-2/αIL-2 treatment (Fig. 5.14C). The proportion
of donor-type myeloid cells was higher in recipients of proinsulin transgenic compared to NOD
HSPC and this was unchanged by IL-2/αIL-2 treatment (Fig. 5.14D). These data demonstrate
development of donor-type leukocytes following transfer of HSPC to 3-week old NOD recipients.
Interestingly, a higher level of donor-type leukocyte development was observed for recipients of
proinsulin transgenic than NOD HSPC and may indicate increased expression of potentially
tolerogenic proinsulin expression. Also in recipients of proinsulin transgenic HSPC, IL-2/αIL-2
treatment inhibited development of donor-type DC and T cells.
156
Overall
A
0 .0 2 5
-3
D o n o r D C /u l ( x 1 0
6
4
2
0
u n ir r . N O D P B S
PBS
PBS
D
-3
D o n o r m y e lo id c e lls /u l ( x 1 0
)
-3
D o n o r B c e lls /u l ( x 1 0
1 .5
1 .0
0 .5
IL - 2 /  IL - 2
PBS
IL - 2 /  IL - 2
CD8+ T cells
Myeloid cells
1 .5
1 .0
0 .5
0 .0
PBS
IL - 2 /  IL - 2
PBS
IL - 2 /  IL - 2
p r o In s T g
CD4+ T cells
)
2 .0
c e ll/u l ( x 1 0
-3
)
-3
1 .5
1 .0
Donor CD4
+
0 .4
+
c e ll/u l ( x 1 0
p ro In s T g
NOD
F
0 .6
IL - 2 /  IL - 2
PBS
2 .0
p r o In s T g
NOD
IL - 2 /  IL - 2
NOD
0 .0
Donor CD8
0 .0 0 5
p ro In s T g
B cells
E
0 .0 1 0
IL - 2 /  IL - 2
2 .0
PBS
0 .0 1 5
0 .0 0 0
IL - 2 /  IL - 2
NOD
C
0 .0 2 0
)
C e lls /u l ( x 1 0
-3
)
)
8
DC
B
***
0 .2
0 .0
PBS
IL - 2 /  IL - 2
NOD
PBS
IL - 2 /  IL - 2
p r o In s T g
0 .5
0 .0
PBS
IL - 2 /  IL - 2
NOD
PBS
IL - 2 /  IL - 2
p r o In s T g
Figure 5.13 – Substantial donor-type leukocyte development in 3-week old recipients of NOD
and proinsulin transgenic HSPC.
Three-week old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 1x105 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HSPC transferred. HSPC
recipients were injected with IL-2/αIL-2 complex or PBS i.p. on day 0, 1 and 2 of transfer. Six weeks
later, the total number of total leukocytes (A) and the total number of donor-type (CD45.1+CD45.2+)
DC (B), B cells (C), myeloid cells (D), CD8+ T cells (E) and CD4+ T cells (F) in peripheral blood
were determined using flow cytometry. Data are pooled from three independent experiments. Each
point represents a single mouse. Error bars shows S.D. One-way ANOVA followed by Tukey posttest. (***p<0.001).
157
Overall
A
DC
B
*
100
80
***
D o n o r D C (% )
T o t a l d o n o r c e lls ( % )
100
60
40
IL - 2 /  IL - 2
PBS
PBS
40
IL - 2 /  IL - 2
PBS
p ro In s T g
NOD
C
IL - 2 /  IL - 2
Myeloid cells
100
D o n o r m y e lo id c e lls ( % )
100
**
80
IL - 2 /  IL - 2
PBS
p ro In s T g
NOD
D
B cells
D o n o r B c e lls ( % )
60
0
0
**
60
40
20
*
80
60
40
20
0
0
IL - 2 /  IL - 2
PBS
E
PBS
IL - 2 /  IL - 2
PBS
IL - 2 /  IL - 2
F
CD8+ T cells
IL - 2 /  IL - 2
PBS
p ro In s T g
NOD
p ro In s T g
NOD
CD4+ T cells
100
c e lls ( % )
100
80
**
80
**
****
60
+
60
+
***
Donor CD4
c e lls ( % )
80
20
20
Donor CD8
***
40
20
40
20
0
0
PBS
IL - 2 /  IL - 2
NOD
PBS
IL - 2 /  IL - 2
p ro In s T g
PBS
IL - 2 /  IL - 2
NOD
PBS
IL - 2 /  IL - 2
p ro In s T g
Figure 5.14 – Increased donor-type leukocyte development in 3-week old, PBS-treated recipients
of proinsulin transgenic HSPC compared to NOD HSPC.
Three-week old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 1x105 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HPC transferred. HPC
recipients were injected with IL-2/αIL-2 complex or PBS i.p. on day 0, 1 and 2 of transfer. Six weeks
later, the proportion of donor-type (CD45.1+CD45.2+) total cells (A), DC (B), B cells (C), myeloid
cells (D), CD8+ T cells (E) and CD4+ T cells (F) in peripheral blood were determined using flow
cytometry. Data are pooled from three independent experiments. Each point represents a single mouse.
Error bars shows S.D. One-way ANOVA followed by Tukey post-test. (*p<0.05, **p<0.01, ***p<0.001,
****p<0.0001).
158
5.2.9 IL-2/αIL-2 expands host-type CD25+CD4+ T cells in peripheral blood from recipients of
proinsulin transgenic HSPC
The data presented in Figure 5.4 indicated CD4+CD25+Foxp3+ Treg expanded 1-2 days after
IL-2/αIL-2 treatment in NOD mice. It is not known whether IL-2/aIL-2 treatment expands Treg
following low-dose irradiation, and whether Treg expansion, if evident, is maintained. To test this,
recipients of HSPC were treated with IL-2/αIL-2 or PBS for three consecutive days starting at the
time of HSPC transfer. Six weeks later, CD25-expressing CD4+ and CD8+ T cells in peripheral
blood were enumerated. The total number of CD25-expressing donor-type CD4+ T cells and their
proportion of host CD4+ T cells remained unchanged following IL-2/αIL-2 treatment in recipients
of NOD and proinsulin transgenic HSPC (Fig. 5.15A, C). The number of host-type CD25expressing CD4+ T cells was higher in recipients of NOD HSPC than recipients of proinsulin
transgenic HSPC, although, this did not differ significantly with IL-2/αIL-2 treatment (Fig. 5.15B).
In recipients of proinsulin transgenic HSPC, the proportion of host-type CD25-expressing CD4+ T
cells increased after IL-2/αIL-2 treatment, yet this remained unchanged in recipients of NOD HSPC
(Fig. 5.15D). In the CD8+ T cell compartment, the total number of donor- and host-type
CD25+CD8+ T cells remained unchanged after IL-2/αIL-2 treatment in both recipients of proinsulin
transgenic and NOD HSPC (Fig. 5.17E, F). Similarly, the proportion of donor- and host-type
CD25+CD8+ T cells did not differ between recipients of proinsulin transgenic and NOD HSPC
treated with IL-2/αIL-2 or PBS (Fig. 5.17G, H). Some points require consideration for interpretation
of these data. Initial timecourse studies analysed Treg expansion in secondary lymphoid organs, it is
possible that responses in peripheral blood do not reflect those observed in spleen and LN. In
addition, CD25 expression alone does accurately identify Treg as conventional T cells up-regulate
CD25 upon activation (Williams et al., 2006). However, no difference in the frequency of CD25expressing CD8+ T cells was observed after IL-2/αIL-2 treatment, which may perhaps indicate the
increase in the proportion of host-type CD25+CD4+ T cells in recipients of proinsulin transgenic
HSPC represents a specific expansion of Treg.
159
Donor-type
-3
0 .0 2 0
0 .0 1 5
0 .0 0 0
PBS
IL - 2 /  IL - 2
+
(% o f C D 4 )
H o st C D 4 C D 25
+
+
5
0
0
PBS
IL - 2 /  IL - 2
-3
C D 8 c e lls ( x 1 0
n .s .
0 .0 0 3
0 .0 0 2
IL - 2 /  IL - 2
PBS
p ro In s T g
n .s .
n .s .
0 .0 0 8
***
0 .0 0 6
0 .0 0 4
+
+
Donor CD25
n .s .
0 .0 0 4
0 .0 0 1
0 .0 0 0
PBS
IL - 2 /  IL - 2
PBS
0 .0 0 2
0 .0 0 0
PBS
p ro In s T g
n .s .
n .s .
10
IL - 2 /  IL - 2
8
6
IL - 2 /  IL - 2
NOD
H
PBS
IL - 2 /  IL - 2
p ro In s T g
CD8+ T cells
n .s .
3
n .s .
2
+
+
C D 8 c e lls ( % )
*
5
NOD
F
H ost C D 25
-3
)
p ro In s T g
CD4+ T cells
*
IL - 2 /  IL - 2
PBS
p ro In s T g
10
)
IL - 2 /  IL - 2
IL - 2 /  IL - 2
PBS
15
+
n .s .
PBS
IL - 2 /  IL - 2
NOD
D
0 .0 0 5
4
H ost C D 25
Donor CD25
0 .0 0
PBS
n .s .
10
0 .0 2
p ro In s T g
+
D o n o r C D 2 5 C D 4 + (% o f C D 4 )
IL - 2 /  IL - 2
PBS
15
NOD
G
n .s .
0 .0 4
H ost C D 25
0 .0 0 5
NOD
C D 8 c e lls ( x 1 0
E
0 .0 6
+
0 .0 1 0
NOD
C
**
0 .0 8
C D 4 c e lls ( x 1 0
n .s .
C D 8 c e lls ( % )
-3
C D 4 c e lls ( x 1 0
+
Donor CD25
n .s .
0 .0 2 5
)
B
)
A
Host-type
n .s .
2
0
PBS
IL - 2 /  IL - 2
NOD
PBS
IL - 2 /  IL - 2
p ro In s T g
1
0
PBS
IL - 2 /  IL - 2
NOD
PBS
IL - 2 /  IL - 2
p ro In s T g
Figure 5.15– Expansion of host-type CD25+CD4+ T cells in recipients of proinsulin transgenic HSPC
following IL-2/αIL-2 treatment.
Three-week old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 1x105 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HPC transferred. HPC recipients
were injected with IL-2/αIL-2 complex or PBS i.p. on day 0, 1 and 2 of transfer. Six weeks later, the total
number of donor-type (CD45.1+CD45.2+) CD25+CD4+ T cells (A), total number of host-type
(CD45.1+CD45.2-) CD25+CD4+ T cells (B), donor-type CD25+CD4+ T cells (% of host CD4+) (C), host-type
CD25+CD4+ T cells (% of host CD4+) (D), total number of donor-type CD25+CD8+ T cells (E), total number
of host-type CD25+CD8+ T cells (F), donor-type CD25+CD8+ T cells (% of donor CD8+) (G), host-type
CD25+CD8+ T cells (% of host CD8+) (H) were determined in peripheral blood using flow cytometry. Data
are pooled from two independent experiments. Each point represents a single mouse. Error bars shows S.D.
One-way ANOVA followed by Tukey post-test. (*p<0.05, **p<0.01, ***p<0.001).
160
5.2.10 IL-2/αIL-2 treatment reduces diabetes incidence in 3-week old recipients of NOD HSPC
To test the effect on diabetes development of proinsulin transgenic HSPC transfer to NOD
recipients prior to the onset of insulitis (In't Veld, 2014), 3-week old NOD mice were irradiated
(300cGy), NOD or proinsulin transgenic HSPC transferred and diabetes incidence was assessed.
The time of diabetes onset was similar for recipients of proinsulin transgenic and NOD HSPC
compared to untreated NOD mice, with the exception of the somewhat delayed onset observed for
IL-2/αIL-2-treated NOD HSPC recipients (Fig. 5.16A). IL-2/αIL-2 treatment reduced the overall
incidence of diabetes for recipients of NOD HSPC compared to untreated NOD mice (Fig. 5.16A).
There appeared to be a reduction in diabetes incidence for IL-2/αIL-2-treated recipients of
proinsulin transgenic HSPC compared to untreated NOD mice, yet, this was not significant (Fig.
5.16A) There was no correlation between donor-type leukocyte development and the age of
diabetes onset for neither recipients of NOD nor proinsulin transgenic HSPC (Fig. 5.16B). These
data suggest transfer of proinsulin-encoding HSPC to 3-week old NOD recipients did not alter
diabetes development. However, treatment of recipients of NOD HSPC with IL-2/αIL-2 complex
reduced diabetes development. Since IL-2/αIL-2 complex was not administrated in the absence of
low-dose irradiation and HSPC transfer, it is not known whether this protective effect would be
mediated by IL-2/αIL-2 complex alone, although, this could be explored in the future.
161
A
P e r c e n t d ia b e t ic
100
80
p r o In s T g H P C + IL - 2 / IL - 2 ( n = 7 )
p r o In s T g H P C ( n = 1 4 )
60
N O D H P C + IL - 2 / IL - 2 ( n = 6 )
**
40
N O D H P C (n = 1 0 )
u n tr e a te d ( n = 1 7 )
20
0
0
100
200
300
A g e (d a y s )
Irradiation +
HPC transfer
B
T o t a l d o n o r c e lls ( % )
80
p r o In s T g H P C
N O D HPC
60
r = 0 .4 6
40
r = 0 .7 6
20
0
0
100
200
300
A ge of onset
Figure 5.16 – IL-2/αIL-2 complex reduced diabetes incidence in 3-week old recipients of NOD
HSPC.
Three-week old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 1x105 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HSPC transferred. HSPC
recipients were injected with IL-2/αIL-2 complex or PBS i.p. on day 0, 1 and 2 of transfer. One group
was left untreated. Mice were screened weekly for glycosuria commencing at approximately 80 days
of age. Diabetes was confirmed by two consecutive weeks of BGL>12mM. (A) Survival curve
showing diabetes incidence. (B) Correlation of the proportion of donor-type leukocytes in peripheral
blood 6 weeks after HPC transfer and the age of diabetes onset shown with linear regression. Data are
pooled from three independent experiments. Each point represents a single mouse. Log-rank (MantelCox) test (A) and Pearson correlation (B). (**p<0.01: NOD HPC + IL-2/αIL-2 vs untreated).
162
5.2.11 Donor-type leukocytes develop after transfer of NOD and proinsulin transgenic HSPC to 3day old recipients
Since transfer of proinsulin-encoding HSPC to 3-week old NOD recipients did not protect
from diabetes development, it might be possible that transfer of proinsulin transgenic HSPC to
neonatal NOD recipients will enable development of potentially tolerogenic, proinsulin-expressing
APC at a very early stage in diabetes progression. To test this, 3-day old NOD mice were irradiated
(300cGy) and NOD or proinsulin transgenic HSPC transferred intrahepatically. Development of
donor-type leukocytes was assessed in peripheral blood was assessed six weeks after transfer.
Similar to 3- (Fig. 5.13A) and 6-week old (Fig. 5.10A) NOD recipients, 3-day old recipients of
HSPC showed significantly lower total numbers of leukocytes in peripheral blood compared to
unirradiated NOD mice at six weeks after irradiation and HSPC transfer (Fig. 5.17A). The total
number of donor-type DC, B cells, myeloid cells and T cells did not differ significantly between
recipients of NOD and proinsulin transgenic HSPC (Fig. 5.17B-F). Similarly, no significant
difference was observed for the proportion of donor-type leukocytes between recipients of NOD
and proinsulin transgenic HSPC (Fig. 5.18A). Donor-type leukocyte development was lower for 3day old recipients (approximately 20%, Fig. 5.18A) compared to 3- (approximately 40%, Fig.
5.14A) and 6-week old recipients (approximately 40%, Fig. 5.11A). A higher degree of variability
in the proportion of donor-type leukocytes was observed between replicates in the proinsulin
transgenic HSPC group, which may indicate recipients express varying amounts of transgenic
proinsulin (Fig. 5.18A). No significant differences were observed in the proportions of donor-type
DC, B cells, myeloid cells or T cells between recipients of NOD and proinsulin transgenic HSPC
(Fig. 5.18B-F). These data confirm development of donor-type leukocytes in 3-day old recipients of
both NOD and proinsulin transgenic HSPC. Therefore, by 6 weeks of age, recipients of proinsulin
transgenic HSPC may express transgenic proinsulin at level sufficient to prevent diabetes
development. However, due to time constraints, recipients of NOD and proinsulin transgenic HSPC
were approximately 120-days old at completion of this thesis. Therefore, no solid conclusions can
be drawn as yet regarding the effect of low-dose irradiation and transfer of proinsulin-encoding and
NOD HSPC in 3-day old NOD recipients.
163
Overall
DC
A
B
n .s .
*
0 .0 2 5
)
-3
6
D o n o r D C /u l ( x 1 0
T o t a l c e lls /u l ( x 1 0
-3
)
8
4
2
0
0 .0 2 0
0 .0 1 5
0 .0 1 0
0 .0 0 5
0 .0 0 0
u n ir r . N O D
NOD
p r o In s T g
NOD
Myeloid cells
D
-3
n .s .
D o n o r B c e lls /u l ( x 1 0
-3
)
D o n o r m y e lo id c e lls /u l ( x 1 0
0 .2 0
0 .1 5
0 .1 0
0 .0 5
0 .0 0
NOD
)
B cells
C
p r o In s T g
0 .8
n .s .
0 .6
0 .4
0 .2
0 .0
NOD
CD8+ T cells
CD4+ T cells
)
0 .6
-3
c e ll/u l ( x 1 0
0 .1 5
n .s .
0 .4
+
0 .1 0
Donor CD4
Donor CD8
p r o In s T g
F
n .s .
0 .2 0
+
c e lls /u l (x 1 0
-3
)
E
p r o In s T g
0 .0 5
0 .0 0
NOD
p r o In s T g
0 .2
0 .0
NOD
p r o In s T g
Figure 5.17 – Similar donor-type leukocyte development in sublethally irradiated 3-day old
recipients of NOD and proinsulin transgenic HSPC.
Three-day old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 2x104 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HSPC transferred. Six weeks
later, the total number of total leukocytes (A) and the total number of donor-type (CD45.1+CD45.2+)
DC (B), B cells (C), myeloid cells (D), CD8+ T cells (E) and CD4+ T cells (F) in peripheral blood
were determined using flow cytometry. Data are pooled from two independent experiments. Each
point represents a single mouse. Error bars shows S.D. One-way ANOVA followed by Tukey post-test
(A) and student’s t-test (B-F). (*p<0.05).
164
Overall
B
100
80
n .s .
60
40
n .s .
100
D o n o r D C (% )
T o t a l d o n o r c e lls ( % )
A
DC
20
80
60
40
20
0
0
NOD
p r o In s T g
NOD
B cells
C
n .s .
100
D o n o r m y e lo id c e lls ( % )
D o n o r B c e lls ( % )
Myeloid cells
D
40
30
20
10
0
n .s .
80
60
40
20
0
NOD
p r o In s T g
NOD
CD8+ T cells
E
CD4+ T cells
50
c e lls ( % )
n .s .
n .s .
40
30
+
30
20
Donor CD4
c e lls ( % )
+
p r o In s T g
F
40
Donor CD8
p r o In s T g
10
20
10
0
0
NOD
p r o In s T g
NOD
p r o In s T g
Figure 5.18 – Similar donor-type leukocyte development in sublethally irradiated, 3-day old
recipients of NOD and proinsulin transgenic HSPC.
Three day old NOD (CD45.1+CD45.2-) mice were irradiated (300cGy) and 2x104 NOD
(CD45.1+CD45.2+) or proinsulin transgenic (CD45.1+CD45.2+) purified HSPC transferred. Six weeks
later, the proportion of donor-type (CD45.1+CD45.2+) total cells (A), DC (B), B cells (C), myeloid
cells (D), CD8+ T cells (E) and CD4+ T cells (F) in peripheral blood were determined using flow
cytometry. Data are pooled from two independent experiments. Each point represents a single mouse.
Error bars shows S.D. Student’s t-test.
165
Table 5.1 Summary of BM/HSPC transfer experiments in Chapter 5
Age of recipients
Cells
transferred
Bulk BM
lin-
Sorted
c-kit+ HSPC
Strain
Treatment
16 weeks
10 weeks
NOD
+ PBS
+ IL-2/αIL-2
Proinsulin + PBS
transgenic
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
+ IL-2/αIL-2
NOD
+ PBS
+ IL-2/αIL-2
Proinsulin + PBS
transgenic
n.s.
n/a
n/a
n/a
n.s.
n/a
n/a
n/a
+ IL-2/αIL-2
n/a
n/a
6 weeks
n.s.
n.s.
Diabetes
incidence
increased
n.s.
n.s.
n/a
No
change
from
untreated
NOD
mice
n/a
3 weeks
n/a
n/a
n/a
n/a
n.s.
n.s.
n.s.
Diabetes
incidence
decreased
Figure 5.19 – Summary of main findings from BM/HSPC transfer experiments in Chapter 5.
‘Bulk BM’ refers to Figure 5.8, ‘Sorted lin-c-kit+ HSPC’ refers to Figure 5.12 and Figure 5.16. n.s.
denotes no significant change in diabetes incidence, while n/a denotes that the indicated groups were
not included in experiments.
166
5.3 Discussion
The previous two chapters have demonstrated tolerance induction in naïve and memory
insulin-specific CD8+ T cells after transfer to mice transgenically expressing proinsulin,
respectively. Since insulin-reactive CD8+ T cells might be primary contributors to β-cell destruction
in NOD mice, it is possible that inducing tolerance in these diabetogenic T cells could protect
susceptible mice from disease development. It has been shown that proinsulin-encoding HSPC
transfer can protect NOD mice from autoimmune diabetes after myeloblative conditioning (Chan et
al., 2006; Steptoe et al., 2003). The highly intensive conditioning used eliminates most of the
pathogenic immune cell pool and reconstitution of the immune system occurs from transferred
proinsulin-encoding HSPC. It remains to be addressed whether, and at which stage of disease,
introducing potentially protective, proinsulin-encoding APC will be sufficient to inactivate the
existing insulin-specific T-cell population and whether this will protect from diabetes development.
In a clinical setting, only the mildest level of conditioning would be acceptable due to the not
immediately life threatening nature of T1D and the young age of onset. Therefore, work in this
chapter sought to determine the feasibility of such as approach in an experimental setting.
Since immune responses are mounted toward ubiquitously-targeted antigen, leading to
rejection of BM encoding ubiquitously-targeted antigen (Coleman et al., 2013), targeting antigen
using cell-specific promoters might overcome rejection of gene-modified donor BM. Proinsulinencoding HSPC are not rejected in recipients subjected to high levels of γ-radiation (950 or 650cGy,
respectively) (Chan et al., 2006; Steptoe et al., 2003). Given BM from MII.OVA mice is rejected
when transferred to non-transgenic recipients after low dose irradiation (Coleman et al., 2013), it
was possible that HSPC/BM from proinsulin transgenic mice would be rejected in a similar manner
in wild type NOD recipients. In fact, proinsulin-encoding BM/HSPC were not rejected in nontransgenic NOD recipients, implying proinsulin transgenic HSC were not subjected to the same
anti-transgene immune response as MII.OVA HSC. Evidence for immunogenicity of gene-modified
cells in transplantation and gene therapy is provided by studies employing known-immunogenic
foreign antigens such as green fluorescent protein (GFP) (Eixarch et al., 2009) and OVA
(Brockstedt et al., 1999) or virus-derived antigens such as herpes simplex virus tyrosine kinase
(Berger et al., 2006). The inherent immunogenicity of the transgene may determine the likelihood
of graft rejection (Carpentier et al., 2015). (Pro)insulin is a self-antigen which might explain the
acceptance of proinsulin-encoding BM/HSPC in the present study. Furthermore, the CD8+ T cell
immunodominant epitope of OVA binds to MHC class I with high affinity, in contrast to the lowaffinity MHC class I binding of (pro)insulin-derived peptides, in particular InsB15-23. The difference
in affinity of OVA and (pro)insulin may affect the ability of the transgene to prime a CTL response
capable of rejecting antigen-encoding HSPC, which may account for the acceptance of proinsulin
167
transgenic BM and HSPC in the present study. Despite the ‘self’ nature of insulin, insulin is
immunogenic in NOD mice (Lamont et al., 2014; Wong et al., 2009), BALB/c mice (Abiru et al.,
2001) and T1D-susceptible humans (Kent et al., 2005; Yang et al., 2014), as such, immunogenicity
might not be the single determining factor in immune-mediated rejection of gene-modified HSPC,
particularly in the context of self-antigens and autoimmunity.
While IL-2/αIL-2 treatment expanded CD4+Foxp3+ Treg in the present study, similar to
reports with C57BL/6 mice (Webster et al., 2009), there also appeared to be expansion of
potentially pathologic CD8+ TEM and NK-like cells, which has been observed in studies using
another anti-IL-2 mAb, S4B6, in C57BL/6 mice (Boyman et al., 2006). It is possible that IL-2
signalling defects in NOD mice (Yamanouchi et al., 2007) may account for the more promiscuous
selectivity of IL-2/JES6-1 in NOD compared to C57BL/6 mice. There is the caveat that CD49b was
used to identify NK- and NKT-like cells in NOD mice in the present study. While CD49b has been
used to define NK and NKT cells in NK1.1-negative strains, such as BALB/c (Arase et al., 2001)
and NOD (Gonzalez et al., 2001a), others have demonstrated CD49b-defined NKT cells are not
equivalent to CD1d-tetramer-defined NKT cells (Pellicci et al., 2005). Therefore, a more robust
method for identifying NK and NKT cells would be to use α-galactosylceramide-loaded, CD1dtetramer, or NK1.1-congenic NOD mice. NOD mice have a deficiency of NKT cells (Baxter et al.,
1997) and it is thought that this deficiency is crucial for autoimmune diabetes to progress (Yang et
al., 2001). Activated NKT cells have been shown to inhibit both spontaneous autoimmune diabetes
(Naumov et al., 2001) and adoptively transferred diabetes (Chen et al., 2005). The apparent increase
in NKT cells observed in the present study following IL-2/αIL-2 may be beneficial in inhibiting
diabetes development. The involvement of NK cells is less clear, with evidence to suggest NK cells
are protective (Lee et al., 2008), disease-promoting (Poirot et al., 2004) or unnecessary for diabetes
development (Beilke et al., 2012), depending on the model used. With regard to the effect of IL2/αIL-2 treatment on CD8+ T cells, expansion of TEM may be expected following IL-2/αIL-2
treatment, as effector CD8+ T cells are dependent on IL-2 [reviewed in (Boyman & Sprent, 2012)]
and there was no way of differentiating between effector T cells and true TEM by expression of
CD44 and CD62L as used in the present study. In addition, Treg and non-Treg cell expansion may
differ when IL-2/αIL-2 is administered in conjunction with low-dose irradiation and HSPC/BM
transfer. It is plausible that low-dose irradiation leads to an inflammatory environment in which
different immune cell subsets become more responsive to IL-2 signalling. Indeed, increased IL-2
signalling from treatment with a stimulatory IL-2/αIL-2 complex drives OVA-specific CD8+ T cells
toward autoimmunity (Waithman et al., 2008). Expansion of CD8+ TEM and NK-like cells following
IL-2/αIL-2 treatment might be exacerbated by lymphopenia following low-dose irradiation and this
may further contribute to diabetogenicity. In contrast, expansion of NKT-like cells may be
168
beneficial in aiding the regulation of diabetogenic cells. Since the primary outcome of BM/HSPC
transfer experiments in the present study was overall diabetes incidence, the effect of expanding
individual IL-2/αIL-2-responsive cell subsets was not investigated. It is possible that the outcomes
of IL-2/αIL-2 treatment observed in the present study were due to changes in the balance of IL-2
responsive regulatory and pathogenic cell subsets at different disease stages. Further dissection of
the cellular response to IL-2/αIL-2 following low-dose irradiation is required to confirm population
dynamics and how this effects diabetes incidence in recipients.
In a therapeutic setting, low-dose IL-2 treatment in recent-onset diabetic NOD mice has
been shown to reverse hyperglycemia, but not insulitis, in approximately 60% of treated mice
(Grinberg-Bleyer et al., 2010). Treatment in pre-diabetic NOD mice was associated with an
increased frequency of islet-infiltrating CD4+Foxp3+ Treg with enhanced expression of CD25,
Foxp3, CTLA-4 and GITR. Interestingly, low-dose IL-2 did not further increase Treg frequency in
islets of diabetic NOD mice but enhanced CD25 and Foxp3 expression. These findings suggest lowdose IL-2 is beneficial in new-onset diabetic NOD mice by increasing suppressive function of isletinfiltrating Treg rather than Treg expansion. Grinberg-Bleyer et al. demonstrated benefit of IL-2
treatment in inflammatory sites, in that reduced IFN-γ production by CD8+ T cells was observed in
pancreas, which may explain the inhibition of progression from insulitis to overt diabetes. In the
present study, diabetes incidence appeared to be slightly reduced when IL-2/αIL-2 complex was
coupled with transfer of NOD BM to 16-week old recipients. It is not known which cell subsets are
expanded by IL-2/αIL-2 complex in the context of irradiation-induced lymphopenia, which persists
for at least 6 weeks. Compounding the potential effect of lymphopenia on IL-2/αIL-2 treatment is
the introduction of antigen, that is, proinsulin, by way of proinsulin-encoding BM/HSPC transfer.
This could prime insulin-specific Treg and/or effector T cells that could either regulate or exert
effector function, respectively. Utilisation of peptide-loaded tetramers in combination with
intracellular transcription factor and cytokine analysis could help dissect this further. Interestingly,
IL-2/αIL-2 treatment was most beneficial in 3-week old recipients of NOD and proinsulin
transgenic HSPC compared to other disease stages. Defects in Treg survival have been
demonstrated in 6-week old NOD mice (Tang et al., 2008) and IL-2/αIL-2 treatment might
overcome this survival defect at this early disease stage, rendering the Treg suppressive, leading to
reduced diabetes incidence in 3-week old recipients. Also, at this early disease stage, there are
possibly fewer activated, and thus IL-2 responsive, effector T cells that may be permissive to
expansion following IL-2/αIL-2 treatment, and conversely, T cells are less permissible to regulation
later in diabetes development (You et al., 2005). Both factors may account for increased IL-2/αIL-2
efficacy in early diabetes development and for reduced efficacy at later stages of disease in the
present study. While timecourse experiments (Fig 5.4) showed the dosing regimen used in the
169
present study was sufficient to expand Treg in NOD mice, it is not known whether greater
expansion of Treg would be required to inhibit diabetes development. Greater and/or more
prolonged expansion of Treg using IL-2/αIL2 complex might be achieved through staggering the
dose over the first few weeks following BM/HSPC transfer and/or by administering IL-2/αIL2
complex over more than three days. In human T1D, low-dose IL-2 preferentially increases the
frequency of and CD25, CTLA-4 and GITR expression by Treg (Rosenzwajg et al., 2015).
Furthermore, IFN-γ production by proinsulin- and GAD65-specific T cells was generally reduced in
IL-2-responsive patients. It should be noted that the increase in Treg frequency was transient was
returned to baseline levels at a time relative to IL-2 dosage. The transience of Treg expansion may
underpin the lack of metabolic efficacy in IL-2 trials in human T1D (Hartemann et al., 2013). It has
been noted that a longer dosing regimen may increase efficacy and as such a trial testing this in
recent-onset children is underway (ClinicalTrials.gov Identifier: NCT01862120). Similarly, in the
present study Treg expansion following IL-2/αIL-2 treatment was not sustained and may account
for inefficacy when administered to older NOD recipients.
A potential benefit of combining a Treg-boosting therapy with transfer of proinsulinencoding BM/HSPC, as employed in the present study, is expansion of insulin-specific Treg.
Hyperglycemia was reversed in recent-onset NOD mice after treatment with an anti-CD3 mAb and
intranasal administration of human proinsulin II B24-C36 peptide (Bresson et al., 2006). While
InsB9-23-stimulated CD4+CD25+ T cells derived from InsB9-23-immunised NOD mice prevented
diabetes development in an adoptive transfer setting (Mukherjee et al., 2003). Indeed, Treg
displaying a phenotype indicative of suppressive function were expanded by IL-2/αIL-2 complex in
NOD mice at early stages of diabetes pathogenesis in the present study. To test the suppressive
function of IL-2/anti-IL2 complex expanded Tregs, sorted CD4+CD25hi T cells could be sorted from
treated and untreated NOD mice. The ability of these cells to suppress autologous CD25 - T cells
stimulated with anti-CD3, for example, could be examined in vitro. Or, the proliferation of CFSElabelled G9 (or 8.3, IGRP-specific CD8+) T cells in response to peptide immunisation could be
analysed in recipients that have been treated, or not, with IL-2/αIL2 complex as an in vivo approach.
Generation of suppressive Treg in an autoimmune, inflammatory environment is challenging. To
overcome this immune dysregulation, Kasagi et al. employed low-dose irradiation (200cGy) in
combination with transfer of phagocytic cells and GAD65 peptide to induce GAD65-specific Treg
which prevented further hyperglycemia in recently diabetic NOD mice (Kasagi et al., 2014). The
fact that hyperglycemic NOD mice did not return to normoglycemia suggests the GAD65-specific
Treg were regulating the established response without impeding the pathogenic process. There is
evidence for β-cell regeneration succeeding Treg-mediated regulation (Diaz-de-Durana et al., 2013;
Luo et al., 2005). If β-cell regeneration was occurring in the study by Kasagi et al., it may be
170
expected that BGL in ‘cured’ mice would decrease. By design, the BM/HSPC transfer experiments
in the present study did not track BGL over time. In 16-week old recipients it might have been
possible that BM transfer coupled with IL-2/αIL-2 treatment reduced hyperglycemia more so than
recipients of BM alone, potentially by regulating established autoimmunity.
Targeting antigen to fixed APC induces peripheral tolerance in cognate T cells (Turley &
Miller, 2007). As such, targeting insulin in this manner reversed hyperglycemia in recently diabetic
NOD mice but this effect was not observed with GAD65 peptides or the IGRP peptide mimetope,
NRP-V7 (Fife et al., 2006). Hyperglycemic remission was ablated by administration of an anti-PDL1 mAb, but Treg involvement in this setting was PD-L1-independent, indicating Treg alone did
not mediate diabetes protection in this setting. Although, NOD mice have been shown to harbor
anergic T cells which recognise InsB10-23, suggesting anergy may be broken in certain
circumstances (Pauken et al., 2013). Furthermore, splenocytes coupled with both InsB9-23 and
InsB15-23 significantly reduced diabetes incidence compared to sham-coupled splenocytes when
injected into 4-5 week old NOD mice (Prasad et al., 2012). Reduced diabetes incidence was not
observed in mice when IGRP206-214, GAD65509-524 or GAD65524-543 peptides were used. There was
robust protection from insulitis in NOD mice treated with InsB9-23-coupled splenoctyes and Treg
were again implicated in this protection. These data are consistent with the primacy of anti-insulin
responses in autoimmune diabetes and tolerance induction to insulin is necessary for preventing
diabetes development.
Inducing tolerance to insulin by administration of a plasmid DNA vaccine encoding
preproinsulin II delayed diabetes development in hyperglycemic NOD mice (Solvason et al., 2008).
Delayed diabetes development was preproinsulin II-dependent as no effect was observed when the
DNA vaccine encoded preproinsulin I, consistent with observations in transgenic mice overexpressing proinsulin II (French et al., 1997) and mice lacking proinsulin I (Moriyama et al., 2003).
In human T1D, weekly administration of a DNA plasmid encoding proinsulin reduced the
frequency of proinsulin-specific CD8+ T cells (Roep et al., 2013). This immunological outcome was
also associated with conservation of C-peptide in proinsulin plasmid treated-patients and connects
the reduction of pathogenic mediators with metabolic improvement. These studies suggest deleting
(pro)insulin-specific T cells may be an effective means to impede diabetes development. The
transiency of antigen presentation derived from plasmid DNA is disadvantageous for long-term
tolerance induction. Due to the multipotency of HSPC, the tolerogenic capacity of transgenic
antigen expression can be passed on to progeny APC, thus generating a long-lived, renewing source
of APC able to induce tolerance to cognate T cells, in theory, indefinitely. From these plasmid DNA
studies, reduction of insulin-reactive CD8+ T cells might be an indicator of efficacy in autoimmune
diabetes immunotherapy. On the other hand, studies in which whole insulin or insulin peptides were
171
administered implicate Treg induction as crucial for impediment of diabetes development in NOD
mice (Bergerot et al., 1999; Fousteri et al., 2010). Similar trials in human T1D patients
demonstrated that administration of insulin protein did not lead to a sustained metabolic benefit
(Chaillous et al., 2000; Skyler et al., 2005). Therefore, deletion of antigen-reactive CD8+ T cells
might be required for observable metabolic benefit in human T1D, rather than exclusively inducing
Treg. It is possible that an effective immunotherapy will be capable of β-cell antigen-specific T-cell
deletion in addition to Treg induction, synergistically creating a robustly tolerogenic environment to
impede diabetes development.
In another gene-therapy approach, targeting InsB9-23 to hepatocytes using a lentiviral vector
(LV) protected 90% of NOD mice from diabetes (Akbarpour et al., 2015). This protection was not
observed when the LV encoded OVA or when InsB9-23 expression was ubiquitously targeted.
Interestingly, InsB9-23- and OVA-specific CD8+ T cells were primed following transfer of InsB9-23or OVA-encoding LV constructs, respectively. Despite robust production of IFN-γ by these primed
CD8+ T cells, hepatocytes expressing either antigen were not rejected, as indicated by high vector
copy number. In contrast, there was significantly lower copy number of LV vector in liver when
either transgene was ubiquitously targeted. It is known that targeting delivery of transgeneexpressing cells to the liver may protect from immune-mediated rejection (Follenzi et al., 2004;
Mingozzi et al., 2003), probably due to effective T-cell tolerance following activation in liver
(Benseler et al., 2011; Bowen et al., 2004). The authors suggest targeting antigen expression to APC
in NOD mice is immunostimulatory, due to the CTL response primed after transfer of ubiquitously
targeted InsB9-23, this is contrasted to driving antigen expression from the hepatocyte-specific
enhanced transthyretin (ET) promoter. However, CTL were primed in response to ET-promoted
transgene also. Despite the elicitation of InsB9-23-specific CTL, InsB9-23 expression in hepatocytes
induced Foxp3+ Treg expansion, particularly evident in pLN, which mediated protection from
diabetes. It was shown that a dose of an anti-CD3 mAb alongside the InsB-encoding LV construct
was required to reverse hyperglycemia in recently diabetic NOD mice. This suggests reduction of
pathogenic T cells is still required to mediate diabetes reversal and highlights the benefit of a
therapy which specifically depletes pathogenic T cells. These efficacious studies of Treg-dependent
tolerance induction to insulin emphasise the ability of antigen-specific Treg to negate established
islet autoimmunity, which at late stages of disease may be reliant mainly on bystander suppression
than a direct effect on insulin-specific T cells as other β cell-antigen specificities are prominent at
this stage.
Studies demonstrating efficacy of tolerance induction to insulin and its effect on diabetes
development raise the question as to why a similar effect was not observed following transfer of
proinsulin-encoding HSPC/BM in the present study. Perhaps the tolerogenicity of proinsulin172
encoding BM is altered after transfer to recipients facilitated by low-dose irradiation, in that APC
may present proinsulin-derived determinants in an immunostimulatory, rather than tolerogenic,
capacity. Data from the host laboratory indicates inactivation of transferred G9 T cells occurs in
recipients of proinsulin-encoding BM, suggesting transgenic proinsulin expression remains
tolerogenic, at least for TCR transgenic G9 T cells, in the context of low-dose irradiation and BM
transfer. However, the impact of low-dose irradiation and BM transfer may alter the host immune
repertoire of NOD recipients in a different manner than is indicated by the G9 T-cell adoptive
transfer model. That is, the polyclonal repertoire of the NOD mouse would include insulin-reactive
CD8+ T cells with varying TCR affinities, insulin-reactive CD4+ T cells and T cells potentially
responsive to modified insulin-derived epitopes, and the response of such T cells to transgenic
proinsulin expression is not known. The notion of critical checkpoints in progression to diabetes
might help explain the processes involved here (Andre et al., 1996; Hoglund et al., 1999). An
important observation in the present study is the similarity in time of diabetes onset between
different treatment groups within experiments but also between experiments, when BM/HSPC were
transferred to recipients at different disease stages. It might be expected that if an overt anti-insulin
CTL response was primed, which may be expected more so following transfer of proinsulin
transgenic BM than HSPC due to mature proinsulin-expressing APC residing in BM, diabetes onset
might mimic that seen for diabetogenic T-cell adoptive transfer (Wong et al., 2009) and diabetes
might ensue soon thereafter. Rapid diabetes onset was not observed even in 6-week old recipients of
proinsulin transgenic BM, in which overall diabetes incidence was increased compared to recipients
of NOD BM. Also, T cells may reside in BM from proinsulin transgenic mice that are tolerant to
insulin in proinsulin transgenic mice, but in the context of lymphopenia in BM recipients may
become activated and contribute to the increased diabetes penetrance observed in 6-week old
recipients of proinsulin transgenic BM. Therefore, at least the initial checkpoint was not overcome
earlier as a result of proinsulin-encoding BM transfer in 6-week old NOD recipients. Evidence for
dysfunctional antigen presentation as a checkpoint following proinsulin transgenic BM transfer may
be the higher diabetes incidence in recipients of proinsulin transgenic BM compared to proinsulin
transgenic HSPC. This perhaps suggests mature APC presenting proinsulin-derived determinants
residing in BM, but not purified HSPC, mediate events underlying this increased diabetes incidence.
The fact that proinsulin transgenic BM transfer to 10- and 16- week old recipients did not increase
diabetes incidence might be a result of the prominence of the anti-insulin response in 6-week old
NOD recipients (Trudeau et al., 2003). Another point favouring the failure of regulatory
checkpoints, rather than direct priming of an anti-insulin CTL response, in 6-week old recipients of
proinsulin transgenic BM is the engraftment of proinsulin transgenic BM (Fig. 5.1) and HSPC (Fig.
5.10). If anti-insulin CTL were primed, rejection of proinsulin-encoding BM/HSPC might be
173
expected. It is possible that a diabetogenic anti-insulin T-cell response may not recognise and
destroy proinsulin-encoding APC derived from donor BM/HSPC. Therefore, insulitis may develop
‘normally’ in recipients of proinsulin-encoding BM but once insulitis has become destructive, the
checkpoint preventing the descent to hyperglycemia is less effective than in recipients of NOD BM.
Interestingly, IL-2/αIL-2 treatment in 6-week old recipients of NOD BM also leads to increased
diabetes incidence similar to recipients of proinsulin transgenic BM. This suggests a similar
breakdown of checkpoint regulation, plausibly due to an IL-2/αIL-2-mediated effect on insulinreactive T cells, which are prominent in 6-week old NOD mice, or other potentially diabetogenic
immune cells. Some evidence suggests lymphopenia, observed in the present study, is detrimental
for tolerance induction due to the reduced recovery of Treg (Kaminitz et al., 2010), although, other
studies have suggested Treg are less susceptible to irradiation-induced apoptosis (Qu et al., 2010a;
Qu et al., 2010b). The effect of irradiation-induced lymphopenia may contribute to the different
effects of proinsulin-encoding BM/HSPC transfer on diabetes development at different disease
stages.
Due to epitope spreading (Ott et al., 2004; Tian et al., 2006), it might be expected that
inducing tolerance to insulin at a late stage of diabetes progression would not completely resolve
established islet autoimmunity. Inactivating insulin-specific T cells may remain beneficial even at
late time points in diabetes pathogenesis as it may ease the autoimmune pressure on β cells, perhaps
maintaining metabolic activity, without complete reversal of hyperglycemia. As proinsulinencoding BM transfer did not alter established diabetes pathogenesis in the present study, targeting
the range of known β-cell antigens, might prove capable of reversing diabetes development. Once
the islet autoimmune response is reversed, remaining β cells may regenerate (Dor et al., 2004; Teta
et al., 2007), and islet transplantation (Al-Adra et al., 2012; Okitsu et al., 2001) or β-cell
replacement (Lawandi et al., 2015) strategies could be employed more readily for further metabolic
improvement.
In conclusion, this chapter demonstrated the ability of HSPC from mice expressing
proinsulin under control of an MHC class II promoter to engraft in wild type NOD mice facilitated
by low-dose irradiation. Expansion of CD4+Foxp3+ Treg was observed in NOD mice following
treatment with an IL-2/αIL-2 complex. These IL-2/αIL-2-expanded Treg expressed higher levels of
markers associated with Treg suppression including CD25, Foxp3, GITR and TNFRII. Expansion
of CD8+ TEM and NK-like cells was also observed following IL-2/αIL-2 treatment. Transfer of
proinsulin-encoding BM to 10- and 16-week old non-transgenic NOD recipients did not alter
diabetes development. When proinsulin-encoding BM was transferred to 6-week old NOD
recipients, diabetes incidence was increased compared to recipients of NOD BM. Transfer of
proinsulin-encoding HSPC resulted in substantial donor-type leukocyte development in peripheral
174
blood of 6-week old NOD recipients, yet, this had no effect on diabetes development. Interestingly,
IL-2/αIL-2 treatment reduced diabetes incidence in 3-week old recipients of NOD HSPC. Finally,
proinsulin-encoding HSPC were transferred to 3-day old NOD mice, but due to time constraints, the
outcome of diabetes incidence studies was yet to be determined. Data within this chapter suggests
proinsulin-encoding BM transfer facilitated by low-dose irradiation and treatment with an IL-2/αIL2 complex alter diabetes development in an age-dependent manner.
175
5.4 Future directions
For further progress with regard to Treg expansion, the immune cell population dynamics
following IL-2/αIL-2 treatment in the context of low-dose irradiation would need to be further
dissected to determine the potentially beneficial and deleterious effects of IL-2/αIL-2 on diabetes
progression. Since diabetes incidence in 6-week old recipients of proinsulin-encoding BM was
higher than in recipients of proinsulin-encoding HSPC, transfer of proinsulin-encoding HSPC to 10and 16-week old NOD recipients would address whether this different effect of BM or HSPC
transfer is disease stage-dependent. Since IL-2/αIL-2 treatment was most beneficial in 3-week old
NOD recipients of HSPC, it might be possible that IL-2/αIL-2 would be beneficial also in 3-day old
NOD recipients. To test this, 3-day old NOD recipients of proinsulin-encoding HSPC could be
treated with IL-2/αIL-2. To determine the tolerogenic capacity of proinsulin-encoding BM/HSPC
transfer, β cell-specific T cells might be tracked in recipients using β-cell antigen peptide-loaded
tetramers. The phenotype of tetramer-binding cells, for example, expression of TCR, CD44 and PD1, could be determined as indicators of activation status and tolerance induction. Finally, in a
broader approach, the effect of targeting multiple β-cell antigens, such as IGRP, GAD65 and ChgA
together with insulin, to prevent diabetes development could be explored. It may be possible that
inducing tolerance to multiple β-cell antigens increases the therapeutic window of intervention to
potentially inactivate the predominant diabetogenic T-cell specificities which may prevent further
β-cell damage and arrest general islet inflammation, which would be advantageous, particularly in
human T1D.
176
Chapter 6
Concluding remarks
177
6.1 Concluding remarks
T1D results from autoimmune destruction of insulin-producing β cells. CD8+ T cells
recognizing β-cell antigens are important contributors to this autoimmune process through direct
lysis of β cells and proinflammatory cytokine production. The current treatments for T1D are daily
injections of exogenous insulin and in some cases, islet transplantation. However, neither
exogenous insulin nor islet transplantation mediate the underlying cause of T1D, that is, pathogenic
islet autoimmunity, and as a result complications can arise from imperfect glycemic control and
recurrent autoimmunity can destroy newly-transplanted islets. Therefore, there is a need for an
immunotherapy to antigen-specifically inactivate β-cell antigen-specific CD8+ T cells, without
compromising protective immunity. Targeting antigen expression to professional APC rapidly
inactivates cognate CD8+ T cells (Kenna et al., 2010). Since insulin is the primary β-cell antigen
responsible for autoimmune diabetes in NOD mice (French et al., 1997; Nakayama et al., 2005), the
present study sought to determine whether targeting proinsulin expression to MHC class IIexpressing APC induces peripheral tolerance in insulin-specific CD8+ T cells. Also, I examined the
effect of targeted transgenic proinsulin expression on diabetes development in NOD mice after
BM/HSPC transfer facilitated by low-dose irradiation.
Data presented in Chapter 3 and 4 demonstrated that the robustly tolerogenic nature of
antigen expression under control of an MHC class II promoter, previously described in a model
employing a foreign model antigen, OVA (Kenna et al., 2010), translated to a diabetes-relevant, βcell antigen, insulin, in an autoimmune-prone setting. Both naïve and memory G9 T cells
proliferated extensively prior to undergoing rapid deletion after transfer to mice expressing
transgenic proinsulin. Remaining G9 T cells in proinsulin transgenic recipients were functionally
inactive. Interestingly, the frequency of G9 T cells waned over time after transfer to NOD recipients
in that, by twenty one days after transfer, the number of G9 T cells was significantly lower in NOD
recipients compared to B16A recipients. It is possible that transferred G9 T cells circulate to pLN,
encounter insulin presentation and undergo deletion in NOD recipients. It has been demonstrated
that cross-presentation of OVA in pLN was required for deletion of OT-I T cells (Kurts et al.,
1999). Using models with other antigens, it has been demonstrated that defects in CD8+ T-cell
tolerance pathways exist in NOD mice. For example, clone 4 T cells on the NOD background
proliferated and accumulated in pLN of HA transgenic mice (Martinez et al., 2005), while IGRPspecific CD8+ T-cell accumulation was reduced in NOD mice congenic for protective genetic loci
(Hamilton-Williams et al., 2012; Hamilton-Williams et al., 2010). Since NOD mice share several
risk genes with human T1D, it is plausible that such tolerance defects exist in human T1D and
demonstrating tolerance induction in autoimmune-prone NOD mice might indicate that targeting
antigen expression using an MHC class II promoter would overcome tolerance defects observed in
178
human T1D also. The response of G9 T cells, characterised by slow deletion and, possibly effector
function, after transfer to NOD recipients in the present study provided direct comparison to the
rapid inactivation of transferred G9 T cells in proinsulin transgenic recipients. The mechanism
underlying rapid inactivation of cognate CD8+ T cells in response to antigen expression under
control of an MHC class II promoter remains to be precisely elucidated. It is possible that the dose
of antigen is important, or the MHC class II promoter targets antigen expression to a robustly
tolerogenic APC subset. Rapid inactivation of β-cell antigen-specific T cells without transient
effector function would be desirable in terms of a T1D immunotherapy as β-cell damage by T cells
undergoing tolerance induction would be detrimental for metabolic function in T1D patients
(Lachin et al., 2014).
As both naïve and memory G9 T cells were rapidly inactivated after transfer to mice
expressing proinsulin under control of an MHC class II promoter, the effect on diabetes
development of introducing potentially tolerogenic proinsulin expression to non-transgenic NOD
mice was investigated in an immunotherapeutic approach. NOD mice at defined ages, that represent
distinct stages of diabetes pathogenesis, were used as recipients of BM/HSPC to investigate the
window in which targeting insulin might impede diabetes development. BM from mice expressing
transgenic proinsulin engrafted to high levels in non-transgenic NOD recipients when transferred
after conditioning with low-dose irradiation. This was in contrast to the immune rejection of BM
from MII.OVA mice in non-transgenic C57BL/6 recipients (Coleman et al., 2012). Therefore, in
NOD mice, BM/HSPC transfer represents a means to instate tolerogenic proinsulin expression in
non-transgenic mice. Transfer of proinsulin transgenic BM to 6-week old NOD recipients resulted
in increased diabetes penetrance compared to transfer of NOD BM. If this increase in diabetes
penetrance was the result of priming an overt anti-proinsulin transgene CTL response, it might be
expected that diabetes onset would be earlier in recipients of proinsulin transgenic compared to
NOD BM, but this was not the case. Also, an increase in diabetes penetrance was not observed
when proinsulin-encoding HSPC were transferred in the present study, suggesting the breakdown of
critical checkpoints in diabetes progression following transfer of proinsulin transgenic BM was
dependent on proinsulin-encoding cells present in the transferred BM. Transfer of proinsulinencoding HSPC facilitated by high dose irradiation (650-950cGy) protects NOD mice from diabetes
development (Chan et al., 2006; Steptoe et al., 2003), indicating irradiation de-bulks the immune
system of potentially diabetogenic cells. The important contribution of conditioning is highlighted
in the case of transplantation of autologous HSPC in T1D patients (Li et al., 2012; Voltarelli et al.,
2007), where conditioning acted as an immune re-set, resulting in enhanced β-cell function. Yet,
this response was transient, as reconstitution of the immune compartment was derived from
autologous HSPC which are inherently defective and prone to autoimmunity. Targeting antigen
179
expression using an MHC class II promoter to HSPC might overcome this transiency by generating
a self-renewing population of antigen-expressing APC to inactivate diabetogenic T cells in an
antigen-specific manner, thus removing the drivers of β-cell destruction. Furthermore, general
immunosuppression, which might be an outcome from high-intensity conditioning for
transplantation of HSPC, is an undesired consequence which led to the cessation of T1D clinical
trials using non-specific immune modulators including cyclosporine (Bougneres et al., 1988; Stiller
et al., 1984) and anti-CD3 mAb (Keymeulen et al., 2010a). Therefore, it would be highly desirable
for transfer of antigen-encoding HSPC to be facilitated by low-dose conditioning, preserving
protective immunity. This will be challenging, since immune responses can be mounted toward
transgenes (Carpentier et al., 2015; Eixarch et al., 2009), although data in the present study indicate
proinsulin-encoding HSPC were not rejected in sublethally irradiated NOD mice. However, lower
dose conditioning may be achievable as efficacy of antigen-encoding HSPC transfer would not rely
solely on the re-set of the immune system from high-dose conditioning but active T-cell tolerance
induction imparted by antigen-expressing APC.
Translating antigen-encoding HSPC transfer from transgenic mouse models to human T1D
patients would require viral, posssibly lentivial (Bigger & Wynn, 2014), transduction of autologous
HSPC that could generate a self-renewing population of potentially tolerogenic antigen-expressing
APC after engraftment. Gene therapy approaches to (re)introduce absent or defective gene products
have been trialed in patients with genetic disorders including Wiskott-Aldrich syndrome and Xlinked adrenoleukodystrophy (Aiuti et al., 2013; Cartier et al., 2009). Recent advances with LV
which target receptors expressed by LT-HSC in humans, such as CD105 and CD133, which
resulted in a high level of transduction in granulocyte-colony stimulating factor (G-CSF)-mobilised
CD34+ cells (Brendel et al., 2015; Kays et al., 2015), might benefit such an immunotherapeutic
approach by potentially facilitating high levels of transgene expression in recipients. There are
obvious considerations for application of such an approach including the fidelity of tolerogenic
antigen expression, that is, ensuring antigen expression remains tolerogenic even in the presence of
inflammatory stimuli, defining an optimal conditioning regimen allowing for high levels of
transduced cell engraftment within safe parameters, and general safety considerations of using LV.
Most β-cell antigen-specific immunotherapeutic approaches to date have involved nasal,
oral, subcutaneous or intramuscular administration of peptide, including those from insulin, IGRP
and GAD65. Peptide immunotherapies generally reply on induction of regulation including Treg as
the means of tolerance induction (Hjorth et al., 2011). While protection from diabetes development
has been observed in NOD mice treated with β-cell peptides (Bresson et al., 2006; Daniel &
Wegmann, 1996; Tian et al., 1996), generally, these findings did not translate to human T1D and if
efficacy was observed, this was transient (Ludvigsson et al., 2011; Pozzilli et al., 2000; Skyler et al.,
180
2005). Since defects have been observed in human T1D with regard to Treg (Brusko et al., 2005;
Jailwala et al., 2009; Lindley et al., 2005), it might be possible that relying solely on induction of
regulation is ineffective for adequate control of diabetogenic responses. Actively deleting β cellspecific T cells might be a more robust means to impede islet autoimmunity by elimination of
contributors to β-cell destruction and proinflammatory cytokine production. Inducing regulation
concurrently with T-cell deletion might provide the opportunity to halt further β-cell damage due to
Treg by-stander suppression. Administration of a proinsulin-encoding plasmid DNA vaccine led to
deletion of proinsulin-specific CD8+ T cells and an increase in C-peptide compared to placebo
(Roep et al., 2013), while treatment alefacept (Rigby et al., 2013) showed deletion of CD8+ and
CD4+ Tmem and increased C-peptide levels. Therefore, deleting CD8+ T cells in human T1D is
feasible and appears to benefit residual β-cell function.
Another important consideration in the development of an antigen-specific immunotherapy
in T1D is the occurrence of epitope spreading. T-cell responses toward insulin are required for
diabetes development in NOD mice (French et al., 1997; Moriyama et al., 2003; Nakayama et al.,
2005) and priming of T cells recognising other β-cell antigens, including IGRP and GAD65, are
primed downstream of the anti-insulin response (Krishnamurthy et al., 2006; Prasad et al., 2012).
Due to the primacy of anti-insulin responses in autoimmune diabetes in the NOD mouse, work in
this thesis sought to determine whether inducing tolerance to insulin could impede diabetes
development. Since transfer of proinsulin-encoding HSPC transfer might only impede diabetes
development when performed in very young recipients, it might be possible that inducing tolerance
to multiple β-cell antigens would widen the window for intervention to alter diabetes development.
An approach in which several myelin-associated peptides coupled to peripheral blood mononuclear
cells (PBMC) has been trialed in human MS patients indicating such an approach is safe and
reduces antigen-specific T-cell responses (Lutterotti et al., 2013). However, a multiple-peptide
therapy showed reduced efficacy in preventing syngeneic islet graft rejection in NOD mice
compared to InsB9-23 therapy alone (Fousteri et al., 2015). Therefore, more preclinical studies in
NOD mice using multiple peptides are required to determine the feasibility of such a therapy in
autoimmune diabetes.
It is becoming increasingly evident that a ‘cure’ for T1D will incorporate therapies to
modulate the immune system along with strategies to maintain or enhance β-cell function and this
would provide a multi-pronged approach to address both the immunological and metabolic
components of T1D. Constant, tolerogenic antigen presentation would be an attractive approach to
purge the diabetogenic T-cell repertoire, and might also maintain T-cell tolerance by continuously
inactivating autoreactive T cells recently emerged from the thymus. There is evidence to suggest the
diagnosis of T1D encapsulates multiple disease types such as absence of autoantibodies at diagnosis
181
and heterogeneity in C-peptide preservation (Gianani et al., 2010). Furthermore, there is evidence to
suggest some T1D patients may retain some form of β-cell functionality long after T1D diagnosis
(Keenan et al., 2010). Inactivating β cell-specific T-cell responses might be beneficial even after
T1D diagnosis in that remaining β-cell mass might be preserved and β-cell function may be restored
following inactivation of diabetogenic T-cell responses (Krogvold et al., 2015), thereby reducing
the risk of complication development and islet graft survival might be increased in the absence of
recurrent autoimmunity. While prevention of T1D with an immunotherapy to induce diabetogenic
T-cell tolerance would be ideal, treatment of patients with established disease might be beneficial
also.
Despite extensive studies using preclinical animal models and clinical trials in human T1D,
an immunotherapy to halt diabetes progression, within the parameters of tolerable side-effects, in
human patients has remained elusive. However, recent studies have yielded promising results by
means of depleting diabetogenic T cells in recent-onset T1D patients. Therefore, the demonstration
of rapid deletion of insulin-specific CD8+ Tmem by transgenic expression of proinsulin in this
thesis has potential implications in developing immunotherapeutic approaches toward T1D and may
provide insight to help overcome issues associated with current therapeutic approaches including
transient efficacy and impaired protective immunity.
182
Chapter 7
References
183
Abiru, N., Maniatis, A. K., Yu, L., Miao, D., Moriyama, H., Wegmann, D., & Eisenbarth, G. S. (2001).
Peptide and major histocompatibility complex-specific breaking of humoral tolerance to native
insulin with the B9-23 peptide in diabetes-prone and normal mice. Diabetes, 50(6), 1274-1281.
Abreu, J. R., Martina, S., Verrijn Stuart, A. A., Fillie, Y. E., Franken, K. L., Drijfhout, J. W., & Roep, B. O.
(2012). CD8 T cell autoreactivity to preproinsulin epitopes with very low human leucocyte antigen
class I binding affinity. Clin Exp Immunol, 170(1), 57-65. doi: 10.1111/j.1365-2249.2012.04635.x
Acha-Orbea, H., & McDevitt, H. O. (1987). The first external domain of the nonobese diabetic mouse class
II I-A beta chain is unique. Proc Natl Acad Sci U S A, 84(8), 2435-2439.
Afzali, B., Mitchell, P. J., Scotta, C., Canavan, J., Edozie, F. C., Fazekasova, H., Lord, G. M., John, S.,
Barber, L. D., Hernandez-Fuentes, M. P., Lechler, R. I., & Lombardi, G. (2011). Relative resistance
of human CD4(+) memory T cells to suppression by CD4(+) CD25(+) regulatory T cells. Am J
Transplant, 11(8), 1734-1742. doi: 10.1111/j.1600-6143.2011.03635.x
Aiuti, A., Biasco, L., Scaramuzza, S., Ferrua, F., Cicalese, M. P., Baricordi, C., Dionisio, F., Calabria, A.,
Giannelli, S., Castiello, M. C., Bosticardo, M., Evangelio, C., Assanelli, A., Casiraghi, M., Di
Nunzio, S., Callegaro, L., Benati, C., Rizzardi, P., Pellin, D., Di Serio, C., Schmidt, M., Von Kalle,
C., Gardner, J., Mehta, N., Neduva, V., Dow, D. J., Galy, A., Miniero, R., Finocchi, A., Metin, A.,
Banerjee, P. P., Orange, J. S., Galimberti, S., Valsecchi, M. G., Biffi, A., Montini, E., Villa, A.,
Ciceri, F., Roncarolo, M. G., & Naldini, L. (2013). Lentiviral hematopoietic stem cell gene therapy
in patients with Wiskott-Aldrich syndrome. Science, 341(6148), 1233151. doi:
10.1126/science.1233151
Akbarpour, M., Goudy, K. S., Cantore, A., Russo, F., Sanvito, F., Naldini, L., Annoni, A., & Roncarolo, M.
G. (2015). Insulin B chain 9-23 gene transfer to hepatocytes protects from type 1 diabetes by
inducing Ag-specific FoxP3+ Tregs. Sci Transl Med, 7(289), 289ra281. doi:
10.1126/scitranslmed.aaa3032
Al-Adra, D. P., Pawlick, R., Shapiro, A. M., & Anderson, C. C. (2012). Targeting cells causing split
tolerance allows fully allogeneic islet survival with minimal conditioning in NOD mixed chimeras.
Am J Transplant, 12(12), 3235-3245. doi: 10.1111/j.1600-6143.2012.04260.x
al-Kassab, A. S., & Raziuddin, S. (1990). Immune activation and T cell subset abnormalities in circulation of
patients with recently diagnosed type I diabetes mellitus. Clin Exp Immunol, 81(2), 267-271.
Alard, P., Manirarora, J. N., Parnell, S. A., Hudkins, J. L., Clark, S. L., & Kosiewicz, M. M. (2006).
Deficiency in NOD antigen-presenting cell function may be responsible for suboptimal CD4+CD25+
T-cell-mediated regulation and type 1 diabetes development in NOD mice. Diabetes, 55(7), 20982105. doi: 10.2337/db05-0810
Alkemade, G. M., Clemente-Casares, X., Yu, Z., Xu, B. Y., Wang, J., Tsai, S., Wright, J. R., Jr., Roep, B.
O., & Santamaria, P. (2013). Local autoantigen expression as essential gatekeeper of memory T-cell
recruitment to islet grafts in diabetic hosts. Diabetes, 62(3), 905-911. doi: 10.2337/db12-0600
Anderson, A. C., Chandwaskar, R., Lee, D. H., & Kuchroo, V. K. (2008). Cutting edge: the Idd3 genetic
interval determines regulatory T cell function through CD11b+CD11c- APC. J Immunol, 181(11),
7449-7452.
Anderson, M. S., & Bluestone, J. A. (2005). The NOD mouse: a model of immune dysregulation. Annu Rev
Immunol, 23, 447-485. doi: 10.1146/annurev.immunol.23.021704.115643
Anderson, M. S., Venanzi, E. S., Klein, L., Chen, Z., Berzins, S. P., Turley, S. J., von Boehmer, H., Bronson,
R., Dierich, A., Benoist, C., & Mathis, D. (2002). Projection of an immunological self shadow
within the thymus by the aire protein. Science, 298(5597), 1395-1401. doi: 10.1126/science.1075958
Andre, I., Gonzalez, A., Wang, B., Katz, J., Benoist, C., & Mathis, D. (1996). Checkpoints in the progression
of autoimmune disease: lessons from diabetes models. Proc Natl Acad Sci U S A, 93(6), 2260-2263.
Araki, K., Turner, A. P., Shaffer, V. O., Gangappa, S., Keller, S. A., Bachmann, M. F., Larsen, C. P., &
Ahmed, R. (2009a). mTOR regulates memory CD8 T-cell differentiation. Nature, 460(7251), 108112. doi: 10.1038/nature08155
Araki, M., Chung, D., Liu, S., Rainbow, D. B., Chamberlain, G., Garner, V., Hunter, K. M., Vijayakrishnan,
L., Peterson, L. B., Oukka, M., Sharpe, A. H., Sobel, R., Kuchroo, V. K., & Wicker, L. S. (2009b).
Genetic evidence that the differential expression of the ligand-independent isoform of CTLA-4 is the
molecular basis of the Idd5.1 type 1 diabetes region in nonobese diabetic mice. J Immunol, 183(8),
5146-5157. doi: 10.4049/jimmunol.0802610
Arase, H., Saito, T., Phillips, J. H., & Lanier, L. L. (2001). Cutting edge: the mouse NK cell-associated
antigen recognized by DX5 monoclonal antibody is CD49b (alpha 2 integrin, very late antigen-2). J
Immunol, 167(3), 1141-1144.
184
Asseman, C., Mauze, S., Leach, M. W., Coffman, R. L., & Powrie, F. (1999). An essential role for
interleukin 10 in the function of regulatory T cells that inhibit intestinal inflammation. J Exp Med,
190(7), 995-1004.
Bach, J. F. (1994). Insulin-dependent diabetes mellitus as an autoimmune disease. Endocr Rev, 15(4), 516542.
Baeyens, A., Perol, L., Fourcade, G., Cagnard, N., Carpentier, W., Woytschak, J., Boyman, O., Hartemann,
A., & Piaggio, E. (2013). Limitations of IL-2 and rapamycin in immunotherapy of type 1 diabetes.
Diabetes, 62(9), 3120-3131. doi: 10.2337/db13-0214
Barron, L., Dooms, H., Hoyer, K. K., Kuswanto, W., Hofmann, J., O'Gorman, W. E., & Abbas, A. K. (2010).
Cutting edge: mechanisms of IL-2-dependent maintenance of functional regulatory T cells. J
Immunol, 185(11), 6426-6430. doi: 10.4049/jimmunol.0903940
Battaglia, M., Stabilini, A., Draghici, E., Migliavacca, B., Gregori, S., Bonifacio, E., & Roncarolo, M. G.
(2006). Induction of tolerance in type 1 diabetes via both CD4+CD25+ T regulatory cells and T
regulatory type 1 cells. Diabetes, 55(6), 1571-1580. doi: 10.2337/db05-1576
Baxter, A. G., Kinder, S. J., Hammond, K. J., Scollay, R., & Godfrey, D. I. (1997). Association between
alphabetaTCR+CD4-CD8- T-cell deficiency and IDDM in NOD/Lt mice. Diabetes, 46(4), 572-582.
Becker, T. C., Coley, S. M., Wherry, E. J., & Ahmed, R. (2005). Bone marrow is a preferred site for
homeostatic proliferation of memory CD8 T cells. J Immunol, 174(3), 1269-1273.
Beilhack, G. F., Scheffold, Y. C., Weissman, I. L., Taylor, C., Jerabek, L., Burge, M. J., Masek, M. A., &
Shizuru, J. A. (2003). Purified allogeneic hematopoietic stem cell transplantation blocks diabetes
pathogenesis in NOD mice. Diabetes, 52(1), 59-68.
Beilke, J. N., Meagher, C. T., Hosiawa, K., Champsaur, M., Bluestone, J. A., & Lanier, L. L. (2012). NK
cells are not required for spontaneous autoimmune diabetes in NOD mice. PLoS One, 7(4), e36011.
doi: 10.1371/journal.pone.0036011
Belghith, M., Bluestone, J. A., Barriot, S., Megret, J., Bach, J. F., & Chatenoud, L. (2003). TGF-betadependent mechanisms mediate restoration of self-tolerance induced by antibodies to CD3 in overt
autoimmune diabetes. Nat Med, 9(9), 1202-1208. doi: 10.1038/nm924
Benseler, V., Warren, A., Vo, M., Holz, L. E., Tay, S. S., Le Couteur, D. G., Breen, E., Allison, A. C., van
Rooijen, N., McGuffog, C., Schlitt, H. J., Bowen, D. G., McCaughan, G. W., & Bertolino, P. (2011).
Hepatocyte entry leads to degradation of autoreactive CD8 T cells. Proc Natl Acad Sci U S A,
108(40), 16735-16740. doi: 10.1073/pnas.1112251108
Berger, C., Flowers, M. E., Warren, E. H., & Riddell, S. R. (2006). Analysis of transgene-specific immune
responses that limit the in vivo persistence of adoptively transferred HSV-TK-modified donor T cells
after allogeneic hematopoietic cell transplantation. Blood, 107(6), 2294-2302. doi: 10.1182/blood2005-08-3503
Bergerot, I., Arreaza, G. A., Cameron, M. J., Burdick, M. D., Strieter, R. M., Chensue, S. W., Chakrabarti,
S., & Delovitch, T. L. (1999). Insulin B-chain reactive CD4+ regulatory T-cells induced by oral
insulin treatment protect from type 1 diabetes by blocking the cytokine secretion and pancreatic
infiltration of diabetogenic effector T-cells. Diabetes, 48(9), 1720-1729.
Bertolino, P., Bowen, D. G., McCaughan, G. W., & Fazekas de St Groth, B. (2001). Antigen-specific
primary activation of CD8+ T cells within the liver. J Immunol, 166(9), 5430-5438.
Bigger, B. W., & Wynn, R. F. (2014). Novel approaches and mechanisms in hematopoietic stem cell gene
therapy. Discov Med, 17(94), 207-215.
Bjornstad, P., Snell-Bergeon, J. K., Nadeau, K. J., & Maahs, D. M. (2015). Insulin sensitivity and
complications in type 1 diabetes: New insights. World J Diabetes, 6(1), 8-16. doi:
10.4239/wjd.v6.i1.8
Bluestone, J. A., Herold, K., & Eisenbarth, G. (2010). Genetics, pathogenesis and clinical interventions in
type 1 diabetes. Nature, 464(7293), 1293-1300.
Bonifaz, L., Bonnyay, D., Mahnke, K., Rivera, M., Nussenzweig, M. C., & Steinman, R. M. (2002).
Efficient targeting of protein antigen to the dendritic cell receptor DEC-205 in the steady state leads
to antigen presentation on major histocompatibility complex class I products and peripheral CD8+ T
cell tolerance. J Exp Med, 196(12), 1627-1638.
Bottazzo, G. F., Florin-Christensen, A., & Doniach, D. (1974). Islet-cell antibodies in diabetes mellitus with
autoimmune polyendocrine deficiencies. Lancet, 2(7892), 1279-1283.
Bougneres, P. F., Carel, J. C., Castano, L., Boitard, C., Gardin, J. P., Landais, P., Hors, J., Mihatsch, M. J.,
Paillard, M., Chaussain, J. L., & et al. (1988). Factors associated with early remission of type I
185
diabetes in children treated with cyclosporine. N Engl J Med, 318(11), 663-670. doi:
10.1056/nejm198803173181103
Bowen, D. G., Zen, M., Holz, L., Davis, T., McCaughan, G. W., & Bertolino, P. (2004). The site of primary
T cell activation is a determinant of the balance between intrahepatic tolerance and immunity. J Clin
Invest, 114(5), 701-712. doi: 10.1172/jci21593
Boyman, O., Kovar, M., Rubinstein, M. P., Surh, C. D., & Sprent, J. (2006). Selective stimulation of T cell
subsets with antibody-cytokine immune complexes. Science, 311(5769), 1924-1927. doi:
10.1126/science.1122927
Boyman, O., & Sprent, J. (2012). The role of interleukin-2 during homeostasis and activation of the immune
system. Nat Rev Immunol, 12(3), 180-190. doi: 10.1038/nri3156
Brendel, C., Goebel, B., Daniela, A., Brugman, M., Kneissl, S., Schwable, J., Kaufmann, K. B., MullerKuller, U., Kunkel, H., Chen-Wichmann, L., Abel, T., Serve, H., Bystrykh, L., Buchholz, C. J., &
Grez, M. (2015). CD133-targeted gene transfer into long-term repopulating hematopoietic stem
cells. Mol Ther, 23(1), 63-70. doi: 10.1038/mt.2014.173
Bresson, D., Fousteri, G., Manenkova, Y., Croft, M., & von Herrath, M. (2011). Antigen-specific prevention
of type 1 diabetes in NOD mice is ameliorated by OX40 agonist treatment. J Autoimmun, 37(4), 342351. doi: 10.1016/j.jaut.2011.10.001
Bresson, D., Togher, L., Rodrigo, E., Chen, Y., Bluestone, J. A., Herold, K. C., & von Herrath, M. (2006).
Anti-CD3 and nasal proinsulin combination therapy enhances remission from recent-onset
autoimmune diabetes by inducing Tregs. J Clin Invest, 116(5), 1371-1381. doi: 10.1172/jci27191
Brinkman, C. C., Rouhani, S. J., Srinivasan, N., & Engelhard, V. H. (2013). Peripheral tissue homing
receptors enable T cell entry into lymph nodes and affect the anatomical distribution of memory
cells. J Immunol, 191(5), 2412-2425. doi: 10.4049/jimmunol.1300651
Brockstedt, D. G., Podsakoff, G. M., Fong, L., Kurtzman, G., Mueller-Ruchholtz, W., & Engleman, E. G.
(1999). Induction of immunity to antigens expressed by recombinant adeno-associated virus depends
on the route of administration. Clin Immunol, 92(1), 67-75. doi: 10.1006/clim.1999.4724
Brusko, T. M., Wasserfall, C. H., Clare-Salzler, M. J., Schatz, D. A., & Atkinson, M. A. (2005). Functional
defects and the influence of age on the frequency of CD4+ CD25+ T-cells in type 1 diabetes.
Diabetes, 54(5), 1407-1414.
Bulek, A. M., Cole, D. K., Skowera, A., Dolton, G., Gras, S., Madura, F., Fuller, A., Miles, J. J., Gostick, E.,
Price, D. A., Drijfhout, J. W., Knight, R. R., Huang, G. C., Lissin, N., Molloy, P. E., Wooldridge, L.,
Jakobsen, B. K., Rossjohn, J., Peakman, M., Rizkallah, P. J., & Sewell, A. K. (2012). Structural
basis for the killing of human beta cells by CD8(+) T cells in type 1 diabetes. Nat Immunol, 13(3),
283-289. doi: 10.1038/ni.2206
Cai, Z., Kishimoto, H., Brunmark, A., Jackson, M. R., Peterson, P. A., & Sprent, J. (1997). Requirements for
peptide-induced T cell receptor downregulation on naive CD8+ T cells. J Exp Med, 185(4), 641-651.
Calzascia, T., Pellegrini, M., Lin, A., Garza, K. M., Elford, A. R., Shahinian, A., Ohashi, P. S., & Mak, T.
W. (2008). CD4 T cells, lymphopenia, and IL-7 in a multistep pathway to autoimmunity. Proc Natl
Acad Sci U S A, 105(8), 2999-3004. doi: 10.1073/pnas.0712135105
Cao, X., Cai, S. F., Fehniger, T. A., Song, J., Collins, L. I., Piwnica-Worms, D. R., & Ley, T. J. (2007).
Granzyme B and perforin are important for regulatory T cell-mediated suppression of tumor
clearance. Immunity, 27(4), 635-646. doi: 10.1016/j.immuni.2007.08.014
Carbone, F. R. (2015). Tissue-Resident Memory T Cells and Fixed Immune Surveillance in Nonlymphoid
Organs. J Immunol, 195(1), 17-22. doi: 10.4049/jimmunol.1500515
Carpentier, M., Lorain, S., Chappert, P., Lalfer, M., Hardet, R., Urbain, D., Peccate, C., Adriouch, S., Garcia,
L., Davoust, J., & Gross, D. A. (2015). Intrinsic transgene immunogenicity gears CD8(+) T-cell
priming after rAAV-mediated muscle gene transfer. Mol Ther, 23(4), 697-706. doi:
10.1038/mt.2014.235
Carrio, R., Bathe, O. F., & Malek, T. R. (2004). Initial antigen encounter programs CD8+ T cells competent
to develop into memory cells that are activated in an antigen-free, IL-7- and IL-15-rich environment.
J Immunol, 172(12), 7315-7323.
Cartier, N., Hacein-Bey-Abina, S., Bartholomae, C. C., Veres, G., Schmidt, M., Kutschera, I., Vidaud, M.,
Abel, U., Dal-Cortivo, L., Caccavelli, L., Mahlaoui, N., Kiermer, V., Mittelstaedt, D., Bellesme, C.,
Lahlou, N., Lefrere, F., Blanche, S., Audit, M., Payen, E., Leboulch, P., l'Homme, B., Bougneres, P.,
Von Kalle, C., Fischer, A., Cavazzana-Calvo, M., & Aubourg, P. (2009). Hematopoietic stem cell
gene therapy with a lentiviral vector in X-linked adrenoleukodystrophy. Science, 326(5954), 818823. doi: 10.1126/science.1171242
186
Casey, K. A., Fraser, K. A., Schenkel, J. M., Moran, A., Abt, M. C., Beura, L. K., Lucas, P. J., Artis, D.,
Wherry, E. J., Hogquist, K., Vezys, V., & Masopust, D. (2012). Antigen-independent differentiation
and maintenance of effector-like resident memory T cells in tissues. J Immunol, 188(10), 4866-4875.
doi: 10.4049/jimmunol.1200402
Chaillous, L., Lefevre, H., Thivolet, C., Boitard, C., Lahlou, N., Atlan-Gepner, C., Bouhanick, B., Mogenet,
A., Nicolino, M., Carel, J. C., Lecomte, P., Marechaud, R., Bougneres, P., Charbonnel, B., & Sai, P.
(2000). Oral insulin administration and residual beta-cell function in recent-onset type 1 diabetes: a
multicentre randomised controlled trial. Diabete Insuline Orale group. Lancet, 356(9229), 545-549.
Chan, J., Clements, W., Field, J., Nasa, Z., Lock, P., Yap, F., Toh, B. H., & Alderuccio, F. (2006).
Transplantation of bone marrow genetically engineered to express proinsulin II protects against
autoimmune insulitis in NOD mice. J Gene Med, 8(11), 1281-1290. doi: 10.1002/jgm.968
Charlton, J. J., Chatzidakis, I., Tsoukatou, D., Boumpas, D. T., Garinis, G. A., & Mamalaki, C. (2013).
Programmed death-1 shapes memory phenotype CD8 T cell subsets in a cell-intrinsic manner. J
Immunol, 190(12), 6104-6114. doi: 10.4049/jimmunol.1201617
Charlton, J. J., Tsoukatou, D., Mamalaki, C., & Chatzidakis, I. (2015). Programmed death 1 regulates
memory phenotype CD4 T cell accumulation, inhibits expansion of the effector memory phenotype
subset and modulates production of effector cytokines. PLoS One, 10(3), e0119200. doi:
10.1371/journal.pone.0119200
Chatenoud, L., Thervet, E., Primo, J., & Bach, J. F. (1994). Anti-CD3 antibody induces long-term remission
of overt autoimmunity in nonobese diabetic mice. Proc Natl Acad Sci U S A, 91(1), 123-127.
Chatila, T. A., Blaeser, F., Ho, N., Lederman, H. M., Voulgaropoulos, C., Helms, C., & Bowcock, A. M.
(2000). JM2, encoding a fork head-related protein, is mutated in X-linked autoimmunity-allergic
disregulation syndrome. J Clin Invest, 106(12), R75-81. doi: 10.1172/jci11679
Chee, J., Ko, H. J., Skowera, A., Jhala, G., Catterall, T., Graham, K. L., Sutherland, R. M., Thomas, H. E.,
Lew, A. M., Peakman, M., Kay, T. W., & Krishnamurthy, B. (2014). Effector-memory T cells
develop in islets and report islet pathology in type 1 diabetes. J Immunol, 192(2), 572-580. doi:
10.4049/jimmunol.1302100
Chen, W., Jin, W., Hardegen, N., Lei, K. J., Li, L., Marinos, N., McGrady, G., & Wahl, S. M. (2003).
Conversion of peripheral CD4+CD25- naive T cells to CD4+CD25+ regulatory T cells by TGF-beta
induction of transcription factor Foxp3. J Exp Med, 198(12), 1875-1886. doi: 10.1084/jem.20030152
Chen, X., Makala, L. H., Jin, Y., Hopkins, D., Muir, A., Garge, N., Podolsky, R. H., & She, J. X. (2008a).
Type 1 diabetes patients have significantly lower frequency of plasmacytoid dendritic cells in the
peripheral blood. Clin Immunol, 129(3), 413-418. doi: 10.1016/j.clim.2008.08.013
Chen, X., Subleski, J. J., Kopf, H., Howard, O. M., Mannel, D. N., & Oppenheim, J. J. (2008b). Cutting
edge: expression of TNFR2 defines a maximally suppressive subset of mouse CD4+CD25+FoxP3+
T regulatory cells: applicability to tumor-infiltrating T regulatory cells. J Immunol, 180(10), 64676471.
Chen, Y., Heeger, P. S., & Valujskikh, A. (2004). In vivo helper functions of alloreactive memory CD4+ T
cells remain intact despite donor-specific transfusion and anti-CD40 ligand therapy. J Immunol,
172(9), 5456-5466.
Chen, Y. G., Choisy-Rossi, C. M., Holl, T. M., Chapman, H. D., Besra, G. S., Porcelli, S. A., Shaffer, D. J.,
Roopenian, D., Wilson, S. B., & Serreze, D. V. (2005). Activated NKT cells inhibit autoimmune
diabetes through tolerogenic recruitment of dendritic cells to pancreatic lymph nodes. J Immunol,
174(3), 1196-1204.
Christianson, S. W., Shultz, L. D., & Leiter, E. H. (1993). Adoptive transfer of diabetes into
immunodeficient NOD-scid/scid mice. Relative contributions of CD4+ and CD8+ T-cells from
diabetic versus prediabetic NOD.NON-Thy-1a donors. Diabetes, 42(1), 44-55.
Churlaud, G., Pitoiset, F., Jebbawi, F., Lorenzon, R., Bellier, B., Rosenzwajg, M., & Klatzmann, D. (2015).
Human and Mouse CD8(+)CD25(+)FOXP3(+) Regulatory T Cells at Steady State and during
Interleukin-2 Therapy. Front Immunol, 6, 171. doi: 10.3389/fimmu.2015.00171
Coleman, M. A., Bridge, J. A., Lane, S. W., Dixon, C. M., Hill, G. R., Wells, J. W., Thomas, R., & Steptoe,
R. J. (2012). Tolerance induction with gene-modified stem cells and immune-preserving
conditioning in primed mice: restricting antigen to differentiated antigen-presenting cells permits
efficacy. Blood. doi: 10.1182/blood-2012-06-434100
Coleman, M. A., Bridge, J. A., Lane, S. W., Dixon, C. M., Hill, G. R., Wells, J. W., Thomas, R., & Steptoe,
R. J. (2013). Tolerance induction with gene-modified stem cells and immune-preserving
187
conditioning in primed mice: restricting antigen to differentiated antigen-presenting cells permits
efficacy. Blood, 121(6), 1049-1058. doi: 10.1182/blood-2012-06-434100
Collison, L. W., Workman, C. J., Kuo, T. T., Boyd, K., Wang, Y., Vignali, K. M., Cross, R., Sehy, D.,
Blumberg, R. S., & Vignali, D. A. (2007). The inhibitory cytokine IL-35 contributes to regulatory Tcell function. Nature, 450(7169), 566-569. doi: 10.1038/nature06306
Concannon, P., Rich, S. S., & Nepom, G. T. (2009). Genetics of type 1A diabetes. N Engl J Med, 360(16),
1646-1654. doi: 10.1056/NEJMra0808284
Coppieters, K. T., Dotta, F., Amirian, N., Campbell, P. D., Kay, T. W., Atkinson, M. A., Roep, B. O., & von
Herrath, M. G. (2012). Demonstration of islet-autoreactive CD8 T cells in insulitic lesions from
recent onset and long-term type 1 diabetes patients. J Exp Med, 209(1), 51-60. doi:
10.1084/jem.20111187
Croft, M., Bradley, L. M., & Swain, S. L. (1994). Naive versus memory CD4 T cell response to antigen.
Memory cells are less dependent on accessory cell costimulation and can respond to many antigenpresenting cell types including resting B cells. J Immunol, 152(6), 2675-2685.
Curtsinger, J. M., Lins, D. C., & Mescher, M. F. (1998). CD8+ memory T cells (CD44high, Ly-6C+) are
more sensitive than naive cells to (CD44low, Ly-6C-) to TCR/CD8 signaling in response to antigen.
J Immunol, 160(7), 3236-3243.
Czabotar, P. E., Colman, P. M., & Huang, D. C. (2009). Bax activation by Bim? Cell Death Differ, 16(9),
1187-1191. doi: 10.1038/cdd.2009.83
D'Angeli, M. A., Merzon, E., Valbuena, L. F., Tirschwell, D., Paris, C. A., & Mueller, B. A. (2010).
Environmental factors associated with childhood-onset type 1 diabetes mellitus: an exploration of
the hygiene and overload hypotheses. Arch Pediatr Adolesc Med, 164(8), 732-738. doi:
10.1001/archpediatrics.2010.115
Daniel, D., Gill, R. G., Schloot, N., & Wegmann, D. (1995). Epitope specificity, cytokine production profile
and diabetogenic activity of insulin-specific T cell clones isolated from NOD mice. Eur J Immunol,
25(4), 1056-1062. doi: 10.1002/eji.1830250430
Daniel, D., & Wegmann, D. R. (1996). Protection of nonobese diabetic mice from diabetes by intranasal or
subcutaneous administration of insulin peptide B-(9-23). Proc Natl Acad Sci U S A, 93(2), 956-960.
Danke, N. A., Koelle, D. M., Yee, C., Beheray, S., & Kwok, W. W. (2004). Autoreactive T cells in healthy
individuals. J Immunol, 172(10), 5967-5972.
Davey, G. M., Kurts, C., Miller, J. F., Bouillet, P., Strasser, A., Brooks, A. G., Carbone, F. R., & Heath, W.
R. (2002). Peripheral deletion of autoreactive CD8 T cells by cross presentation of self-antigen
occurs by a Bcl-2-inhibitable pathway mediated by Bim. J Exp Med, 196(7), 947-955.
Day, C. L., Kaufmann, D. E., Kiepiela, P., Brown, J. A., Moodley, E. S., Reddy, S., Mackey, E. W., Miller,
J. D., Leslie, A. J., DePierres, C., Mncube, Z., Duraiswamy, J., Zhu, B., Eichbaum, Q., Altfeld, M.,
Wherry, E. J., Coovadia, H. M., Goulder, P. J., Klenerman, P., Ahmed, R., Freeman, G. J., &
Walker, B. D. (2006). PD-1 expression on HIV-specific T cells is associated with T-cell exhaustion
and disease progression. Nature, 443(7109), 350-354. doi: 10.1038/nature05115
de Oliveira, G. L., Malmegrim, K. C., Ferreira, A. F., Tognon, R., Kashima, S., Couri, C. E., Covas, D. T.,
Voltarelli, J. C., & de Castro, F. A. (2012). Up-regulation of fas and fasL pro-apoptotic genes
expression in type 1 diabetes patients after autologous haematopoietic stem cell transplantation. Clin
Exp Immunol, 168(3), 291-302. doi: 10.1111/j.1365-2249.2012.04583.x
De Riva, A., Varley, M. C., Bluck, L. J., Cooke, A., Deery, M. J., & Busch, R. (2013). Accelerated turnover
of MHC class II molecules in nonobese diabetic mice is developmentally and environmentally
regulated in vivo and dispensable for autoimmunity. J Immunol, 190(12), 5961-5971. doi:
10.4049/jimmunol.1300551
Delong, T., Baker, R. L., He, J., Barbour, G., Bradley, B., & Haskins, K. (2012). Diabetogenic T-cell clones
recognize an altered peptide of chromogranin A. Diabetes, 61(12), 3239-3246. doi: 10.2337/db120112
Dengler, T. J., & Pober, J. S. (2000). Human vascular endothelial cells stimulate memory but not naive
CD8+ T cells to differentiate into CTL retaining an early activation phenotype. J Immunol, 164(10),
5146-5155.
Deotare, U., Al-Dawsari, G., Couban, S., & Lipton, J. H. (2015). G-CSF-primed bone marrow as a source of
stem cells for allografting: revisiting the concept. Bone Marrow Transplant. doi:
10.1038/bmt.2015.80
188
Derbinski, J., Schulte, A., Kyewski, B., & Klein, L. (2001). Promiscuous gene expression in medullary
thymic epithelial cells mirrors the peripheral self. Nat Immunol, 2(11), 1032-1039. doi:
10.1038/ni723
Diabetes Prevention Trial--Type 1 Diabetes Study, G. (2002). Effects of insulin in relatives of patients with
type 1 diabetes mellitus. N Engl J Med, 346(22), 1685-1691. doi: 10.1056/NEJMoa012350
Diana, J., Simoni, Y., Furio, L., Beaudoin, L., Agerberth, B., Barrat, F., & Lehuen, A. (2013). Crosstalk
between neutrophils, B-1a cells and plasmacytoid dendritic cells initiates autoimmune diabetes. Nat
Med, 19(1), 65-73. doi: 10.1038/nm.3042
Diaz-de-Durana, Y., Lau, J., Knee, D., Filippi, C., Londei, M., McNamara, P., Nasoff, M., DiDonato, M.,
Glynne, R., & Herman, A. E. (2013). IL-2 immunotherapy reveals potential for innate beta cell
regeneration in the non-obese diabetic mouse model of autoimmune diabetes. PLoS One, 8(10),
e78483. doi: 10.1371/journal.pone.0078483
DiLorenzo, T. P. (2011). Multiple antigens versus single major antigen in type 1 diabetes: arguing for
multiple antigens. Diabetes Metab Res Rev, 27(8), 778-783. doi: 10.1002/dmrr.1251
Diz, R., Garland, A., Vincent, B. G., Johnson, M. C., Spidale, N., Wang, B., & Tisch, R. (2012).
Autoreactive effector/memory CD4+ and CD8+ T cells infiltrating grafted and endogenous islets in
diabetic NOD mice exhibit similar T cell receptor usage. PLoS One, 7(12), e52054. doi:
10.1371/journal.pone.0052054
Dor, Y., Brown, J., Martinez, O. I., & Melton, D. A. (2004). Adult pancreatic beta-cells are formed by selfduplication rather than stem-cell differentiation. Nature, 429(6987), 41-46. doi:
10.1038/nature02520
Doyle, H. A., & Mamula, M. J. (2001). Post-translational protein modifications in antigen recognition and
autoimmunity. Trends Immunol, 22(8), 443-449.
Drake, D. R., 3rd, Ream, R. M., Lawrence, C. W., & Braciale, T. J. (2005). Transient loss of MHC class I
tetramer binding after CD8+ T cell activation reflects altered T cell effector function. J Immunol,
175(3), 1507-1515.
Durinovic-Bello, I., Jelinek, E., Schlosser, M., Eiermann, T., Boehm, B. O., Karges, W., Marchand, L., &
Polychronakos, C. (2005). Class III alleles at the insulin VNTR polymorphism are associated with
regulatory T-cell responses to proinsulin epitopes in HLA-DR4, DQ8 individuals. Diabetes, 54
Suppl 2, S18-24.
Eixarch, H., Gomez, A., Kadar, E., George, M., Martinez, N., Espejo, C., Petriz, J., Gimeno, R., &
Barquinero, J. (2009). Transgene expression levels determine the immunogenicity of transduced
hematopoietic grafts in partially myeloablated mice. Mol Ther, 17(11), 1904-1909. doi:
10.1038/mt.2009.198
Elyaman, W., Kivisakk, P., Reddy, J., Chitnis, T., Raddassi, K., Imitola, J., Bradshaw, E., Kuchroo, V. K.,
Yagita, H., Sayegh, M. H., & Khoury, S. J. (2008). Distinct functions of autoreactive memory and
effector CD4+ T cells in experimental autoimmune encephalomyelitis. Am J Pathol, 173(2), 411422. doi: 10.2353/ajpath.2008.080142
Erlich, H., Valdes, A. M., Noble, J., Carlson, J. A., Varney, M., Concannon, P., Mychaleckyj, J. C., Todd, J.
A., Bonella, P., Fear, A. L., Lavant, E., Louey, A., & Moonsamy, P. (2008). HLA DR-DQ
haplotypes and genotypes and type 1 diabetes risk: analysis of the type 1 diabetes genetics
consortium families. Diabetes, 57(4), 1084-1092. doi: 10.2337/db07-1331
Every, A. L., Kramer, D. R., Mannering, S. I., Lew, A. M., & Harrison, L. C. (2006). Intranasal vaccination
with proinsulin DNA induces regulatory CD4+ T cells that prevent experimental autoimmune
diabetes. J Immunol, 176(8), 4608-4615.
Fallarino, F., Grohmann, U., Hwang, K. W., Orabona, C., Vacca, C., Bianchi, R., Belladonna, M. L., Fioretti,
M. C., Alegre, M. L., & Puccetti, P. (2003). Modulation of tryptophan catabolism by regulatory T
cells. Nat Immunol, 4(12), 1206-1212. doi: 10.1038/ni1003
Feng, Y., Wang, D., Yuan, R., Parker, C. M., Farber, D. L., & Hadley, G. A. (2002). CD103 expression is
required for destruction of pancreatic islet allografts by CD8(+) T cells. J Exp Med, 196(7), 877-886.
Ferraro, A., Socci, C., Stabilini, A., Valle, A., Monti, P., Piemonti, L., Nano, R., Olek, S., Maffi, P., Scavini,
M., Secchi, A., Staudacher, C., Bonifacio, E., & Battaglia, M. (2011). Expansion of Th17 cells and
functional defects in T regulatory cells are key features of the pancreatic lymph nodes in patients
with type 1 diabetes. Diabetes, 60(11), 2903-2913. doi: 10.2337/db11-0090
Feuerstein, N., Huang, D., & Prystowsky, M. B. (1995). Rapamycin selectively blocks interleukin-2-induced
proliferating cell nuclear antigen gene expression in T lymphocyte. Evidence for inhibition of
CREB/ATF binding activities. J Biol Chem, 270(16), 9454-9458.
189
Fife, B. T., Guleria, I., Gubbels Bupp, M., Eagar, T. N., Tang, Q., Bour-Jordan, H., Yagita, H., Azuma, M.,
Sayegh, M. H., & Bluestone, J. A. (2006). Insulin-induced remission in new-onset NOD mice is
maintained by the PD-1-PD-L1 pathway. J Exp Med, 203(12), 2737-2747. doi:
10.1084/jem.20061577
Follenzi, A., Battaglia, M., Lombardo, A., Annoni, A., Roncarolo, M. G., & Naldini, L. (2004). Targeting
lentiviral vector expression to hepatocytes limits transgene-specific immune response and establishes
long-term expression of human antihemophilic factor IX in mice. Blood, 103(10), 3700-3709. doi:
10.1182/blood-2003-09-3217
Fotino, N., Fotino, C., & Pileggi, A. (2015). Re-engineering islet cell transplantation. Pharmacol Res, 98, 7685. doi: 10.1016/j.phrs.2015.02.010
Fourcade, J., Sun, Z., Pagliano, O., Guillaume, P., Luescher, I. F., Sander, C., Kirkwood, J. M., Olive, D.,
Kuchroo, V., & Zarour, H. M. (2012). CD8(+) T cells specific for tumor antigens can be rendered
dysfunctional by the tumor microenvironment through upregulation of the inhibitory receptors
BTLA and PD-1. Cancer Res, 72(4), 887-896. doi: 10.1158/0008-5472.can-11-2637
Fousteri, G., Dave, A., Bot, A., Juntti, T., Omid, S., & von Herrath, M. (2010). Subcutaneous insulin B:923/IFA immunisation induces Tregs that control late-stage prediabetes in NOD mice through IL-10
and IFNgamma. Diabetologia, 53(9), 1958-1970. doi: 10.1007/s00125-010-1777-x
Fousteri, G., Jofra, T., Di Fonte, R., & Battaglia, M. (2015). Combination of an Antigen-Specific Therapy
and an Immunomodulatory Treatment to Simultaneous Block Recurrent Autoimmunity and
Alloreactivity in Non-Obese Diabetic Mice. PLoS One, 10(6), e0127631. doi:
10.1371/journal.pone.0127631
French, M. B., Allison, J., Cram, D. S., Thomas, H. E., Dempsey-Collier, M., Silva, A., Georgiou, H. M.,
Kay, T. W., Harrison, L. C., & Lew, A. M. (1997). Transgenic expression of mouse proinsulin II
prevents diabetes in nonobese diabetic mice. Diabetes, 46(1), 34-39.
Gardner, J. M., Devoss, J. J., Friedman, R. S., Wong, D. J., Tan, Y. X., Zhou, X., Johannes, K. P., Su, M. A.,
Chang, H. Y., Krummel, M. F., & Anderson, M. S. (2008). Deletional tolerance mediated by
extrathymic Aire-expressing cells. Science, 321(5890), 843-847. doi: 10.1126/science.1159407
Gavathiotis, E., Suzuki, M., Davis, M. L., Pitter, K., Bird, G. H., Katz, S. G., Tu, H. C., Kim, H., Cheng, E.
H., Tjandra, N., & Walensky, L. D. (2008). BAX activation is initiated at a novel interaction site.
Nature, 455(7216), 1076-1081. doi: 10.1038/nature07396
Gianani, R., Campbell-Thompson, M., Sarkar, S. A., Wasserfall, C., Pugliese, A., Solis, J. M., Kent, S. C.,
Hering, B. J., West, E., Steck, A., Bonner-Weir, S., Atkinson, M. A., Coppieters, K., von Herrath,
M., & Eisenbarth, G. S. (2010). Dimorphic histopathology of long-standing childhood-onset
diabetes. Diabetologia, 53(4), 690-698. doi: 10.1007/s00125-009-1642-y
Gonzalez, A., Andre-Schmutz, I., Carnaud, C., Mathis, D., & Benoist, C. (2001a). Damage control, rather
than unresponsiveness, effected by protective DX5+ T cells in autoimmune diabetes. Nat Immunol,
2(12), 1117-1125. doi: 10.1038/ni738
Gonzalez, J., Harris, T., Childs, G., & Prystowsky, M. B. (2001b). Rapamycin blocks IL-2-driven T cell
cycle progression while preserving T cell survival. Blood Cells Mol Dis, 27(3), 572-585. doi:
10.1006/bcmd.2001.0420
Gottlieb, P. A., Delong, T., Baker, R. L., Fitzgerald-Miller, L., Wagner, R., Cook, G., Rewers, M. R.,
Michels, A., & Haskins, K. (2014). Chromogranin A is a T cell antigen in human type 1 diabetes. J
Autoimmun, 50, 38-41. doi: 10.1016/j.jaut.2013.10.003
Gottschalk, R. A., Corse, E., & Allison, J. P. (2012). Expression of Helios in peripherally induced Foxp3+
regulatory T cells. J Immunol, 188(3), 976-980. doi: 10.4049/jimmunol.1102964
Goudy, K. S., Johnson, M. C., Garland, A., Li, C., Samulski, R. J., Wang, B., & Tisch, R. (2011). Reduced
IL-2 expression in NOD mice leads to a temporal increase in CD62Llo FoxP3+ CD4+ T cells with
limited suppressor activity. Eur J Immunol, 41(5), 1480-1490. doi: 10.1002/eji.201040890
Graser, R. T., DiLorenzo, T. P., Wang, F., Christianson, G. J., Chapman, H. D., Roopenian, D. C.,
Nathenson, S. G., & Serreze, D. V. (2000). Identification of a CD8 T cell that can independently
mediate autoimmune diabetes development in the complete absence of CD4 T cell helper functions.
J Immunol, 164(7), 3913-3918.
Gregori, S., Giarratana, N., Smiroldo, S., & Adorini, L. (2003). Dynamics of pathogenic and suppressor T
cells in autoimmune diabetes development. J Immunol, 171(8), 4040-4047.
Grinberg-Bleyer, Y., Baeyens, A., You, S., Elhage, R., Fourcade, G., Gregoire, S., Cagnard, N., Carpentier,
W., Tang, Q., Bluestone, J., Chatenoud, L., Klatzmann, D., Salomon, B. L., & Piaggio, E. (2010).
190
IL-2 reverses established type 1 diabetes in NOD mice by a local effect on pancreatic regulatory T
cells. J Exp Med, 207(9), 1871-1878. doi: 10.1084/jem.20100209
Grossman, W. J., Verbsky, J. W., Tollefsen, B. L., Kemper, C., Atkinson, J. P., & Ley, T. J. (2004).
Differential expression of granzymes A and B in human cytotoxic lymphocyte subsets and T
regulatory cells. Blood, 104(9), 2840-2848. doi: 10.1182/blood-2004-03-0859
Hamilton-Williams, E. E., Cheung, J., Rainbow, D. B., Hunter, K. M., Wicker, L. S., & Sherman, L. A.
(2012). Cellular mechanisms of restored beta-cell tolerance mediated by protective alleles of Idd3
and Idd5. Diabetes, 61(1), 166-174. doi: 10.2337/db11-0790
Hamilton-Williams, E. E., Palmer, S. E., Charlton, B., & Slattery, R. M. (2003). Beta cell MHC class I is a
late requirement for diabetes. Proc Natl Acad Sci U S A, 100(11), 6688-6693. doi:
10.1073/pnas.1131954100
Hamilton-Williams, E. E., Wong, S. B., Martinez, X., Rainbow, D. B., Hunter, K. M., Wicker, L. S., &
Sherman, L. A. (2010). Idd9.2 and Idd9.3 protective alleles function in CD4+ T-cells and
nonlymphoid cells to prevent expansion of pathogenic islet-specific CD8+ T-cells. Diabetes, 59(6),
1478-1486. doi: 10.2337/db09-1801
Hanninen, A., Jalkanen, S., Salmi, M., Toikkanen, S., Nikolakaros, G., & Simell, O. (1992). Macrophages, T
cell receptor usage, and endothelial cell activation in the pancreas at the onset of insulin-dependent
diabetes mellitus. J Clin Invest, 90(5), 1901-1910. doi: 10.1172/jci116067
Harding, F. A., McArthur, J. G., Gross, J. A., Raulet, D. H., & Allison, J. P. (1992). CD28-mediated
signalling co-stimulates murine T cells and prevents induction of anergy in T-cell clones. Nature,
356(6370), 607-609. doi: 10.1038/356607a0
Harjutsalo, V., Sjoberg, L., & Tuomilehto, J. (2008). Time trends in the incidence of type 1 diabetes in
Finnish children: a cohort study. Lancet, 371(9626), 1777-1782. doi: 10.1016/s0140-6736(08)607655
Harrison, L. C., Honeyman, M. C., Steele, C. E., Stone, N. L., Sarugeri, E., Bonifacio, E., Couper, J. J., &
Colman, P. G. (2004). Pancreatic beta-cell function and immune responses to insulin after
administration of intranasal insulin to humans at risk for type 1 diabetes. Diabetes Care, 27(10),
2348-2355.
Hartemann, A., Bensimon, G., Payan, C. A., Jacqueminet, S., Bourron, O., Nicolas, N., Fonfrede, M.,
Rosenzwajg, M., Bernard, C., & Klatzmann, D. (2013). Low-dose interleukin 2 in patients with type
1 diabetes: a phase 1/2 randomised, double-blind, placebo-controlled trial. Lancet Diabetes
Endocrinol, 1(4), 295-305. doi: 10.1016/s2213-8587(13)70113-x
Haskins, K., Portas, M., Bradley, B., Wegmann, D., & Lafferty, K. (1988). T-lymphocyte clone specific for
pancreatic islet antigen. Diabetes, 37(10), 1444-1448.
Hawiger, D., Inaba, K., Dorsett, Y., Guo, M., Mahnke, K., Rivera, M., Ravetch, J. V., Steinman, R. M., &
Nussenzweig, M. C. (2001). Dendritic cells induce peripheral T cell unresponsiveness under steady
state conditions in vivo. J Exp Med, 194(6), 769-779.
Hayward, A. R., & Shriber, M. (1992). Reduced incidence of insulitis in NOD mice following anti-CD3
injection: requirement for neonatal injection. J Autoimmun, 5(1), 59-67.
Heath, V. L., Moore, N. C., Parnell, S. M., & Mason, D. W. (1998). Intrathymic expression of genes
involved in organ specific autoimmune disease. J Autoimmun, 11(4), 309-318. doi:
10.1006/jaut.1998.0210
Hehmke, B., Michaelis, D., Gens, E., Laube, F., & Kohnert, K. D. (1995). Aberrant activation of CD8+ Tcell and CD8+ T-cell subsets in patients with newly diagnosed IDDM. Diabetes, 44(12), 1414-1419.
Heissmeyer, V., Macian, F., Im, S. H., Varma, R., Feske, S., Venuprasad, K., Gu, H., Liu, Y. C., Dustin, M.
L., & Rao, A. (2004). Calcineurin imposes T cell unresponsiveness through targeted proteolysis of
signaling proteins. Nat Immunol, 5(3), 255-265. doi: 10.1038/ni1047
Hernandez, J., Aung, S., Redmond, W. L., & Sherman, L. A. (2001). Phenotypic and functional analysis of
CD8(+) T cells undergoing peripheral deletion in response to cross-presentation of self-antigen. J
Exp Med, 194(6), 707-717.
Herold, K. C., Bluestone, J. A., Montag, A. G., Parihar, A., Wiegner, A., Gress, R. E., & Hirsch, R. (1992).
Prevention of autoimmune diabetes with nonactivating anti-CD3 monoclonal antibody. Diabetes,
41(3), 385-391.
Herold, K. C., Hagopian, W., Auger, J. A., Poumian-Ruiz, E., Taylor, L., Donaldson, D., Gitelman, S. E.,
Harlan, D. M., Xu, D., Zivin, R. A., & Bluestone, J. A. (2002). Anti-CD3 monoclonal antibody in
new-onset type 1 diabetes mellitus. N Engl J Med, 346(22), 1692-1698. doi:
10.1056/NEJMoa012864
191
Herold, K. C., Pescovitz, M. D., McGee, P., Krause-Steinrauf, H., Spain, L. M., Bourcier, K., Asare, A., Liu,
Z., Lachin, J. M., Dosch, H. M., & Type 1 Diabetes TrialNet Anti, C. D. S. G. (2011). Increased T
cell proliferative responses to islet antigens identify clinical responders to anti-CD20 monoclonal
antibody (rituximab) therapy in type 1 diabetes. J Immunol, 187(4), 1998-2005. doi:
10.4049/jimmunol.1100539
Hjorth, M., Axelsson, S., Ryden, A., Faresjo, M., Ludvigsson, J., & Casas, R. (2011). GAD-alum treatment
induces GAD65-specific CD4+CD25highFOXP3+ cells in type 1 diabetic patients. Clin Immunol,
138(1), 117-126. doi: 10.1016/j.clim.2010.10.004
Hoglund, P., Mintern, J., Waltzinger, C., Heath, W., Benoist, C., & Mathis, D. (1999). Initiation of
autoimmune diabetes by developmentally regulated presentation of islet cell antigens in the
pancreatic lymph nodes. J Exp Med, 189(2), 331-339.
Hogquist, K. A., Jameson, S. C., Heath, W. R., Howard, J. L., Bevan, M. J., & Carbone, F. R. (1994). T cell
receptor antagonist peptides induce positive selection. Cell, 76(1), 17-27.
Hori, S., Nomura, T., & Sakaguchi, S. (2003). Control of regulatory T cell development by the transcription
factor Foxp3. Science, 299(5609), 1057-1061. doi: 10.1126/science.1079490
Hu, C. Y., Rodriguez-Pinto, D., Du, W., Ahuja, A., Henegariu, O., Wong, F. S., Shlomchik, M. J., & Wen,
L. (2007). Treatment with CD20-specific antibody prevents and reverses autoimmune diabetes in
mice. J Clin Invest, 117(12), 3857-3867. doi: 10.1172/jci32405
Huang, C. T., Workman, C. J., Flies, D., Pan, X., Marson, A. L., Zhou, G., Hipkiss, E. L., Ravi, S.,
Kowalski, J., Levitsky, H. I., Powell, J. D., Pardoll, D. M., Drake, C. G., & Vignali, D. A. (2004).
Role of LAG-3 in regulatory T cells. Immunity, 21(4), 503-513. doi: 10.1016/j.immuni.2004.08.010
Huang, L., Soldevila, G., Leeker, M., Flavell, R., & Crispe, I. N. (1994). The liver eliminates T cells
undergoing antigen-triggered apoptosis in vivo. Immunity, 1(9), 741-749.
Ikehara, S., Ohtsuki, H., Good, R. A., Asamoto, H., Nakamura, T., Sekita, K., Muso, E., Tochino, Y., Ida, T.,
Kuzuya, H., & et al. (1985). Prevention of type I diabetes in nonobese diabetic mice by allogenic
bone marrow transplantation. Proc Natl Acad Sci U S A, 82(22), 7743-7747.
Ilonen, J., Surcel, H. M., & Kaar, M. L. (1991). Abnormalities within CD4 and CD8 T lymphocytes subsets
in type 1 (insulin-dependent) diabetes. Clin Exp Immunol, 85(2), 278-281.
In't Veld, P. (2014). Insulitis in human type 1 diabetes: a comparison between patients and animal models.
Semin Immunopathol, 36(5), 569-579. doi: 10.1007/s00281-014-0438-4
Irvine, K. M., Gallego, P., An, X., Best, S. E., Thomas, G., Wells, C., Harris, M., Cotterill, A., & Thomas, R.
(2012). Peripheral blood monocyte gene expression profile clinically stratifies patients with recentonset type 1 diabetes. Diabetes, 61(5), 1281-1290. doi: 10.2337/db11-1549
Itoh, N., Hanafusa, T., Miyazaki, A., Miyagawa, J., Yamagata, K., Yamamoto, K., Waguri, M., Imagawa, A.,
Tamura, S., Inada, M., & et al. (1993). Mononuclear cell infiltration and its relation to the expression
of major histocompatibility complex antigens and adhesion molecules in pancreas biopsy specimens
from newly diagnosed insulin-dependent diabetes mellitus patients. J Clin Invest, 92(5), 2313-2322.
doi: 10.1172/jci116835
Jaeckel, E., Klein, L., Martin-Orozco, N., & von Boehmer, H. (2003). Normal incidence of diabetes in NOD
mice tolerant to glutamic acid decarboxylase. J Exp Med, 197(12), 1635-1644. doi:
10.1084/jem.20030215
Jailwala, P., Waukau, J., Glisic, S., Jana, S., Ehlenbach, S., Hessner, M., Alemzadeh, R., Matsuyama, S.,
Laud, P., Wang, X., & Ghosh, S. (2009). Apoptosis of CD4+ CD25(high) T cells in type 1 diabetes
may be partially mediated by IL-2 deprivation. PLoS One, 4(8), e6527. doi:
10.1371/journal.pone.0006527
Jellison, E. R., Turner, M. J., Blair, D. A., Lingenheld, E. G., Zu, L., Puddington, L., & Lefrancois, L.
(2012). Distinct mechanisms mediate naive and memory CD8 T-cell tolerance. Proc Natl Acad Sci U
S A, 109(52), 21438-21443. doi: 10.1073/pnas.1217409110
Jenkins, M. K., & Schwartz, R. H. (1987). Antigen presentation by chemically modified splenocytes induces
antigen-specific T cell unresponsiveness in vitro and in vivo. J Exp Med, 165(2), 302-319.
Jin, Y., Chen, X., Podolsky, R., Hopkins, D., Makala, L. H., Muir, A., & She, J. X. (2009). APC dysfunction
is correlated with defective suppression of T cell proliferation in human type 1 diabetes. Clin
Immunol, 130(3), 272-279. doi: 10.1016/j.clim.2008.10.005
Joetham, A., Takeda, K., Taube, C., Miyahara, N., Matsubara, S., Koya, T., Rha, Y. H., Dakhama, A., &
Gelfand, E. W. (2007). Naturally occurring lung CD4(+)CD25(+) T cell regulation of airway allergic
responses depends on IL-10 induction of TGF-beta. J Immunol, 178(3), 1433-1442.
192
Jordan, M. S., Boesteanu, A., Reed, A. J., Petrone, A. L., Holenbeck, A. E., Lerman, M. A., Naji, A., &
Caton, A. J. (2001). Thymic selection of CD4+CD25+ regulatory T cells induced by an agonist selfpeptide. Nat Immunol, 2(4), 301-306. doi: 10.1038/86302
Kaech, S. M., Wherry, E. J., & Ahmed, R. (2002). Effector and memory T-cell differentiation: implications
for vaccine development. Nat Rev Immunol, 2(4), 251-262. doi: 10.1038/nri778
Kaminitz, A., Mizrahi, K., Yaniv, I., Stein, J., & Askenasy, N. (2010). Immunosuppressive therapy
exacerbates autoimmunity in NOD mice and diminishes the protective activity of regulatory T cells.
J Autoimmun, 35(2), 145-152. doi: 10.1016/j.jaut.2010.06.002
Kanagawa, O., Martin, S. M., Vaupel, B. A., Carrasco-Marin, E., & Unanue, E. R. (1998). Autoreactivity of
T cells from nonobese diabetic mice: an I-Ag7-dependent reaction. Proc Natl Acad Sci U S A, 95(4),
1721-1724.
Kared, H., Adle-Biassette, H., Fois, E., Masson, A., Bach, J. F., Chatenoud, L., Schneider, E., & Zavala, F.
(2006). Jagged2-expressing hematopoietic progenitors promote regulatory T cell expansion in the
periphery through notch signaling. Immunity, 25(5), 823-834. doi: 10.1016/j.immuni.2006.09.008
Kasagi, S., Zhang, P., Che, L., Abbatiello, B., Maruyama, T., Nakatsukasa, H., Zanvit, P., Jin, W., Konkel, J.
E., & Chen, W. (2014). In vivo-generated antigen-specific regulatory T cells treat autoimmunity
without compromising antibacterial immune response. Sci Transl Med, 6(241), 241ra278. doi:
10.1126/scitranslmed.3008895
Katoh, H., Zheng, P., & Liu, Y. (2013). FOXP3: Genetic and epigenetic implications for autoimmunity. J
Autoimmun. doi: 10.1016/j.jaut.2012.12.004
Katz, J., Benoist, C., & Mathis, D. (1993a). Major histocompatibility complex class I molecules are required
for the development of insulitis in non-obese diabetic mice. Eur J Immunol, 23(12), 3358-3360. doi:
10.1002/eji.1830231244
Katz, J. D., Wang, B., Haskins, K., Benoist, C., & Mathis, D. (1993b). Following a diabetogenic T cell from
genesis through pathogenesis. Cell, 74(6), 1089-1100.
Kaufman, D. L., Clare-Salzler, M., Tian, J., Forsthuber, T., Ting, G. S., Robinson, P., Atkinson, M. A.,
Sercarz, E. E., Tobin, A. J., & Lehmann, P. V. (1993). Spontaneous loss of T-cell tolerance to
glutamic acid decarboxylase in murine insulin-dependent diabetes. Nature, 366(6450), 69-72. doi:
10.1038/366069a0
Kawamura, K., McLaughlin, K. A., Weissert, R., & Forsthuber, T. G. (2008). Myelin-reactive type B T cells
and T cells specific for low-affinity MHC-binding myelin peptides escape tolerance in HLA-DR
transgenic mice. J Immunol, 181(5), 3202-3211.
Kawasaki, E., Nakamura, K., Kuriya, G., Satoh, T., Kobayashi, M., Kuwahara, H., Abiru, N., Yamasaki, H.,
Matsuura, N., Miura, J., Uchigata, Y., & Eguchi, K. (2011). Differences in the humoral
autoreactivity to zinc transporter 8 between childhood- and adult-onset type 1 diabetes in Japanese
patients. Clin Immunol, 138(2), 146-153. doi: 10.1016/j.clim.2010.10.007
Kays, S. K., Kaufmann, K. B., Abel, T., Brendel, C., Bonig, H., Grez, M., Buchholz, C. J., & Kneissl, S.
(2015). CD105 is a surface marker for receptor-targeted gene transfer into human long-term
repopulating hematopoietic stem cells. Stem Cells Dev, 24(6), 714-723. doi: 10.1089/scd.2014.0455
Keenan, H. A., Sun, J. K., Levine, J., Doria, A., Aiello, L. P., Eisenbarth, G., Bonner-Weir, S., & King, G. L.
(2010). Residual insulin production and pancreatic ss-cell turnover after 50 years of diabetes: Joslin
Medalist Study. Diabetes, 59(11), 2846-2853. doi: 10.2337/db10-0676
Keir, M. E., Freeman, G. J., & Sharpe, A. H. (2007). PD-1 regulates self-reactive CD8+ T cell responses to
antigen in lymph nodes and tissues. J Immunol, 179(8), 5064-5070.
Kenna, T. J., Thomas, R., & Steptoe, R. J. (2008). Steady-state dendritic cells expressing cognate antigen
terminate memory CD8+ T-cell responses. Blood, 111(4), 2091-2100. doi: 10.1182/blood-2007-07103200
Kenna, T. J., Waldie, T., McNally, A., Thomson, M., Yagita, H., Thomas, R., & Steptoe, R. J. (2010).
Targeting antigen to diverse APCs inactivates memory CD8+ T cells without eliciting tissuedestructive effector function. J Immunol, 184(2), 598-606. doi: 10.4049/jimmunol.0900032
Kent, S. C., Chen, Y., Bregoli, L., Clemmings, S. M., Kenyon, N. S., Ricordi, C., Hering, B. J., & Hafler, D.
A. (2005). Expanded T cells from pancreatic lymph nodes of type 1 diabetic subjects recognize an
insulin epitope. Nature, 435(7039), 224-228. doi: 10.1038/nature03625
Keymeulen, B., Candon, S., Fafi-Kremer, S., Ziegler, A., Leruez-Ville, M., Mathieu, C.,
Vandemeulebroucke, E., Walter, M., Crenier, L., Thervet, E., Legendre, C., Pierard, D., Hale, G.,
Waldmann, H., Bach, J. F., Seigneurin, J. M., Pipeleers, D., & Chatenoud, L. (2010a). Transient
193
Epstein-Barr virus reactivation in CD3 monoclonal antibody-treated patients. Blood, 115(6), 11451155. doi: 10.1182/blood-2009-02-204875
Keymeulen, B., Walter, M., Mathieu, C., Kaufman, L., Gorus, F., Hilbrands, R., Vandemeulebroucke, E.,
Van de Velde, U., Crenier, L., De Block, C., Candon, S., Waldmann, H., Ziegler, A. G., Chatenoud,
L., & Pipeleers, D. (2010b). Four-year metabolic outcome of a randomised controlled CD3-antibody
trial in recent-onset type 1 diabetic patients depends on their age and baseline residual beta cell mass.
Diabetologia, 53(4), 614-623.
Kikutani, H., & Makino, S. (1992). The murine autoimmune diabetes model: NOD and related strains. Adv
Immunol, 51, 285-322.
Klebanoff, C. A., Gattinoni, L., Torabi-Parizi, P., Kerstann, K., Cardones, A. R., Finkelstein, S. E., Palmer,
D. C., Antony, P. A., Hwang, S. T., Rosenberg, S. A., Waldmann, T. A., & Restifo, N. P. (2005).
Central memory self/tumor-reactive CD8+ T cells confer superior antitumor immunity compared
with effector memory T cells. Proc Natl Acad Sci U S A, 102(27), 9571-9576. doi:
10.1073/pnas.0503726102
Kleijwegt, F. S., Jansen, D. T., Teeler, J., Joosten, A. M., Laban, S., Nikolic, T., & Roep, B. O. (2013).
Tolerogenic dendritic cells impede priming of naive CD8(+) T cells and deplete memory CD8(+) T
cells. Eur J Immunol, 43(1), 85-92. doi: 10.1002/eji.201242879
Klein, L., Kyewski, B., Allen, P. M., & Hogquist, K. A. (2014). Positive and negative selection of the T cell
repertoire: what thymocytes see (and don't see). Nat Rev Immunol, 14(6), 377-391. doi:
10.1038/nri3667
Knip, M., Virtanen, S. M., Seppa, K., Ilonen, J., Savilahti, E., Vaarala, O., Reunanen, A., Teramo, K.,
Hamalainen, A. M., Paronen, J., Dosch, H. M., Hakulinen, T., Akerblom, H. K., & Finnish, T. S. G.
(2010). Dietary intervention in infancy and later signs of beta-cell autoimmunity. N Engl J Med,
363(20), 1900-1908. doi: 10.1056/NEJMoa1004809
Kockum, I., Lernmark, A., Dahlquist, G., Falorni, A., Hagopian, W. A., Landin-Olsson, M., Li, L. C.,
Luthman, H., Palmer, J. P., Sanjeevi, C. B., Sundkvist, G., & Ostman, I. J. (1996). Genetic and
immunological findings in patients with newly diagnosed insulin-dependent diabetes mellitus. The
Swedish Childhood Diabetes Study Group and The Diabetes Incidence in Sweden Study (DISS)
Group. Horm Metab Res, 28(7), 344-347. doi: 10.1055/s-2007-979811
Kontos, S., Kourtis, I. C., Dane, K. Y., & Hubbell, J. A. (2013). Engineering antigens for in situ erythrocyte
binding induces T-cell deletion. Proc Natl Acad Sci U S A, 110(1), E60-68. doi:
10.1073/pnas.1216353110
Kreuwel, H. T., Aung, S., Silao, C., & Sherman, L. A. (2002). Memory CD8(+) T cells undergo peripheral
tolerance. Immunity, 17(1), 73-81.
Kreuwel, H. T., Biggs, J. A., Pilip, I. M., Pamer, E. G., Lo, D., & Sherman, L. A. (2001). Defective CD8+ T
cell peripheral tolerance in nonobese diabetic mice. J Immunol, 167(2), 1112-1117.
Krishnamurthy, B., Chee, J., Jhala, G., Fynch, S., Graham, K. L., Santamaria, P., Morahan, G., Allison, J.,
Izon, D., Thomas, H. E., & Kay, T. W. (2012). Complete diabetes protection despite delayed thymic
tolerance in NOD8.3 TCR transgenic mice due to antigen-induced extrathymic deletion of T cells.
Diabetes, 61(2), 425-435. doi: 10.2337/db11-0948
Krishnamurthy, B., Dudek, N. L., McKenzie, M. D., Purcell, A. W., Brooks, A. G., Gellert, S., Colman, P.
G., Harrison, L. C., Lew, A. M., Thomas, H. E., & Kay, T. W. (2006). Responses against islet
antigens in NOD mice are prevented by tolerance to proinsulin but not IGRP. J Clin Invest, 116(12),
3258-3265. doi: 10.1172/jci29602
Krishnamurthy, B., Mariana, L., Gellert, S. A., Colman, P. G., Harrison, L. C., Lew, A. M., Santamaria, P.,
Thomas, H. E., & Kay, T. W. (2008). Autoimmunity to both proinsulin and IGRP is required for
diabetes in nonobese diabetic 8.3 TCR transgenic mice. J Immunol, 180(7), 4458-4464.
Krogvold, L., Skog, O., Sundstrom, G., Edwin, B., Buanes, T., Hanssen, K. F., Ludvigsson, J., Grabherr, M.,
Korsgren, O., & Dahl-Jorgensen, K. (2015). Function of Isolated Pancreatic Islets From Patients at
Onset of Type 1 Diabetes: Insulin Secretion Can Be Restored After Some Days in a
Nondiabetogenic Environment In Vitro: Results From the DiViD Study. Diabetes, 64(7), 2506-2512.
doi: 10.2337/db14-1911
Kurts, C., Heath, W. R., Kosaka, H., Miller, J. F., & Carbone, F. R. (1998). The peripheral deletion of
autoreactive CD8+ T cells induced by cross-presentation of self-antigens involves signaling through
CD95 (Fas, Apo-1). J Exp Med, 188(2), 415-420.
194
Kurts, C., Kosaka, H., Carbone, F. R., Miller, J. F., & Heath, W. R. (1997). Class I-restricted crosspresentation of exogenous self-antigens leads to deletion of autoreactive CD8(+) T cells. J Exp Med,
186(2), 239-245.
Kurts, C., Sutherland, R. M., Davey, G., Li, M., Lew, A. M., Blanas, E., Carbone, F. R., Miller, J. F., &
Heath, W. R. (1999). CD8 T cell ignorance or tolerance to islet antigens depends on antigen dose.
Proc Natl Acad Sci U S A, 96(22), 12703-12707.
Lachin, J. M., McGee, P., & Palmer, J. P. (2014). Impact of C-peptide preservation on metabolic and clinical
outcomes in the Diabetes Control and Complications Trial. Diabetes, 63(2), 739-748. doi:
10.2337/db13-0881
Lamb, J. R., Skidmore, B. J., Green, N., Chiller, J. M., & Feldmann, M. (1983). Induction of tolerance in
influenza virus-immune T lymphocyte clones with synthetic peptides of influenza hemagglutinin. J
Exp Med, 157(5), 1434-1447.
Lamont, D., Mukherjee, G., Kumar, P. R., Samanta, D., McPhee, C. G., Kay, T. W., Almo, S. C., DiLorenzo,
T. P., & Serreze, D. V. (2014). Compensatory mechanisms allow undersized anchor-deficient class I
MHC ligands to mediate pathogenic autoreactive T cell responses. J Immunol, 193(5), 2135-2146.
doi: 10.4049/jimmunol.1400997
Lawandi, J., Tao, C., Ren, B., Williams, P., Ling, D., Swan, M. A., Nassif, N. T., Torpy, F. R., O'Brien, B.
A., & Simpson, A. M. (2015). Reversal of diabetes following transplantation of an insulin-secreting
human liver cell line: Melligen cells. Mol Ther Methods Clin Dev, 2, 15011. doi:
10.1038/mtm.2015.11
Lawson, J. M., Tremble, J., Dayan, C., Beyan, H., Leslie, R. D. G., Peakman, M., & Tree, T. I. M. (2008).
Increased resistance to CD4+CD25hi regulatory T cell-mediated suppression in patients with type 1
diabetes. Clin Exp Immunol, 154(3), 353-359. doi: 10.1111/j.1365-2249.2008.03810.x
Lee, I. F., Qin, H., Priatel, J. J., & Tan, R. (2008). Critical role for IFN-gamma in natural killer cell-mediated
protection from diabetes. Eur J Immunol, 38(1), 82-89. doi: 10.1002/eji.200737189
Lee, L. F., Logronio, K., Tu, G. H., Zhai, W., Ni, I., Mei, L., Dilley, J., Yu, J., Rajpal, A., Brown, C., Appah,
C., Chin, S. M., Han, B., Affolter, T., & Lin, J. C. (2012a). Anti-IL-7 receptor-alpha reverses
established type 1 diabetes in nonobese diabetic mice by modulating effector T-cell function. Proc
Natl Acad Sci U S A, 109(31), 12674-12679. doi: 10.1073/pnas.1203795109
Lee, M. S., Kwon, H. J., & Kim, H. S. (2012b). Macrophages from nonobese diabetic mouse have a selective
defect in IFN-gamma but not IFN-alpha/beta receptor pathway. J Clin Immunol, 32(4), 753-761. doi:
10.1007/s10875-012-9682-3
Leventhal, J., Miller, J., Abecassis, M., Tollerud, D. J., & Ildstad, S. T. (2013). Evolving approaches of
hematopoietic stem cell-based therapies to induce tolerance to organ transplants: the long road to
tolerance. Clin Pharmacol Ther, 93(1), 36-45. doi: 10.1038/clpt.2012.201
Li, H., Kaufman, C. L., Boggs, S. S., Johnson, P. C., Patrene, K. D., & Ildstad, S. T. (1996). Mixed
allogeneic chimerism induced by a sublethal approach prevents autoimmune diabetes and reverses
insulitis in nonobese diabetic (NOD) mice. J Immunol, 156(1), 380-388.
Li, J., Figueira, S. K., Vrazo, A. C., Binkowski, B. F., Butler, B. L., Tabata, Y., Filipovich, A., Jordan, M. B.,
& Risma, K. A. (2014). Real-time detection of CTL function reveals distinct patterns of caspase
activation mediated by Fas versus granzyme B. J Immunol, 193(2), 519-528. doi:
10.4049/jimmunol.1301668
Li, L., Shen, S., Ouyang, J., Hu, Y., Hu, L., Cui, W., Zhang, N., Zhuge, Y. Z., Chen, B., Xu, J., & Zhu, D.
(2012). Autologous hematopoietic stem cell transplantation modulates immunocompetent cells and
improves beta-cell function in Chinese patients with new onset of type 1 diabetes. J Clin Endocrinol
Metab, 97(5), 1729-1736. doi: 10.1210/jc.2011-2188
Li, Y., Zhou, L., Li, Y., Zhang, J., Guo, B., Meng, G., Chen, X., Zheng, Q., Zhang, L., Zhang, M., & Wang,
L. (2015). Identification of autoreactive CD8(+) T cell responses targeting chromogranin A in
humanized NOD mice and type 1 diabetes patients. Clin Immunol, 159(1), 63-71. doi:
10.1016/j.clim.2015.04.017
Lieberman, S. M., Evans, A. M., Han, B., Takaki, T., Vinnitskaya, Y., Caldwell, J. A., Serreze, D. V.,
Shabanowitz, J., Hunt, D. F., Nathenson, S. G., Santamaria, P., & DiLorenzo, T. P. (2003).
Identification of the β cell antigen targeted by a prevalent population of pathogenic CD8+ T cells in
autoimmune diabetes. Proc Natl Acad Sci U S A, 100(14), 8384-8388. doi:
10.1073/pnas.0932778100
Lindley, S., Dayan, C. M., Bishop, A., Roep, B. O., Peakman, M., & Tree, T. I. (2005). Defective suppressor
function in CD4(+)CD25(+) T-cells from patients with type 1 diabetes. Diabetes, 54(1), 92-99.
195
Liu, K., Catalfamo, M., Li, Y., Henkart, P. A., & Weng, N. P. (2002). IL-15 mimics T cell receptor
crosslinking in the induction of cellular proliferation, gene expression, and cytotoxicity in CD8+
memory T cells. Proc Natl Acad Sci U S A, 99(9), 6192-6197. doi: 10.1073/pnas.092675799
Liu, W., Xiao, X., Demirci, G., Madsen, J., & Li, X. C. (2012). Innate NK cells and macrophages recognize
and reject allogeneic nonself in vivo via different mechanisms. J Immunol, 188(6), 2703-2711. doi:
10.4049/jimmunol.1102997
Lo, D., Freedman, J., Hesse, S., Palmiter, R. D., Brinster, R. L., & Sherman, L. A. (1992). Peripheral
tolerance to an islet cell-specific hemagglutinin transgene affects both CD4+ and CD8+ T cells. Eur
J Immunol, 22(4), 1013-1022. doi: 10.1002/eji.1830220421
London, C. A., Lodge, M. P., & Abbas, A. K. (2000). Functional responses and costimulator dependence of
memory CD4+ T cells. J Immunol, 164(1), 265-272.
Long, S. A., Buckner, J. H., & Greenbaum, C. J. (2013). IL-2 therapy in type 1 diabetes: "Trials" and
tribulations. Clin Immunol, 149(3), 324-331. doi: 10.1016/j.clim.2013.02.005
Long, S. A., Cerosaletti, K., Bollyky, P. L., Tatum, M., Shilling, H., Zhang, S., Zhang, Z. Y., Pihoker, C.,
Sanda, S., Greenbaum, C., & Buckner, J. H. (2010). Defects in IL-2R signaling contribute to
diminished maintenance of FOXP3 expression in CD4(+)CD25(+) regulatory T-cells of type 1
diabetic subjects. Diabetes, 59(2), 407-415. doi: 10.2337/db09-0694
Long, S. A., Cerosaletti, K., Wan, J. Y., Ho, J. C., Tatum, M., Wei, S., Shilling, H. G., & Buckner, J. H.
(2011). An autoimmune-associated variant in PTPN2 reveals an impairment of IL-2R signaling in
CD4(+) T cells. Genes Immun, 12(2), 116-125. doi: 10.1038/gene.2010.54
Long, S. A., Rieck, M., Sanda, S., Bollyky, J. B., Samuels, P. L., Goland, R., Ahmann, A., Rabinovitch, A.,
Aggarwal, S., Phippard, D., Turka, L. A., Ehlers, M. R., Bianchine, P. J., Boyle, K. D., Adah, S. A.,
Bluestone, J. A., Buckner, J. H., & Greenbaum, C. J. (2012). Rapamycin/IL-2 combination therapy
in patients with type 1 diabetes augments Tregs yet transiently impairs beta-cell function. Diabetes,
61(9), 2340-2348. doi: 10.2337/db12-0049
Lowe, C. E., Cooper, J. D., Brusko, T., Walker, N. M., Smyth, D. J., Bailey, R., Bourget, K., Plagnol, V.,
Field, S., Atkinson, M., Clayton, D. G., Wicker, L. S., & Todd, J. A. (2007). Large-scale genetic fine
mapping and genotype-phenotype associations implicate polymorphism in the IL2RA region in type
1 diabetes. Nat Genet, 39(9), 1074-1082. doi: 10.1038/ng2102
Lowin, B., Hahne, M., Mattmann, C., & Tschopp, J. (1994). Cytolytic T-cell cytotoxicity is mediated
through perforin and Fas lytic pathways. Nature, 370(6491), 650-652. doi: 10.1038/370650a0
Luce, S., Lemonnier, F., Briand, J. P., Coste, J., Lahlou, N., Muller, S., Larger, E., Rocha, B., Mallone, R., &
Boitard, C. (2011). Single insulin-specific CD8+ T cells show characteristic gene expression profiles
in human type 1 diabetes. Diabetes, 60(12), 3289-3299. doi: 10.2337/db11-0270
Ludvigsson, J., Hjorth, M., Cheramy, M., Axelsson, S., Pihl, M., Forsander, G., Nilsson, N. O., Samuelsson,
B. O., Wood, T., Aman, J., Ortqvist, E., & Casas, R. (2011). Extended evaluation of the safety and
efficacy of GAD treatment of children and adolescents with recent-onset type 1 diabetes: a
randomised controlled trial. Diabetologia, 54(3), 634-640. doi: 10.1007/s00125-010-1988-1
Luo, X., Yang, H., Kim, I. S., Saint-Hilaire, F., Thomas, D. A., De, B. P., Ozkaynak, E., Muthukumar, T.,
Hancock, W. W., Crystal, R. G., & Suthanthiran, M. (2005). Systemic transforming growth factorbeta1 gene therapy induces Foxp3+ regulatory cells, restores self-tolerance, and facilitates
regeneration of beta cell function in overtly diabetic nonobese diabetic mice. Transplantation, 79(9),
1091-1096.
Luton, F., Buferne, M., Davoust, J., Schmitt-Verhulst, A. M., & Boyer, C. (1994). Evidence for protein
tyrosine kinase involvement in ligand-induced TCR/CD3 internalization and surface redistribution. J
Immunol, 153(1), 63-72.
Lutterotti, A., Yousef, S., Sputtek, A., Sturner, K. H., Stellmann, J. P., Breiden, P., Reinhardt, S., Schulze,
C., Bester, M., Heesen, C., Schippling, S., Miller, S. D., Sospedra, M., & Martin, R. (2013).
Antigen-specific tolerance by autologous myelin peptide-coupled cells: a phase 1 trial in multiple
sclerosis. Sci Transl Med, 5(188), 188ra175. doi: 10.1126/scitranslmed.3006168
Lyman, M. A., Nugent, C. T., Marquardt, K. L., Biggs, J. A., Pamer, E. G., & Sherman, L. A. (2005). The
fate of low affinity tumor-specific CD8+ T cells in tumor-bearing mice. J Immunol, 174(5), 25632572.
Ma, S., Zhao, D. L., Yin, Z. Q., Mukherjee, R., Singh, B., Qin, H. Y., Stiller, C. R., & Jevnikar, A. M.
(1997). Transgenic plants expressing autoantigens fed to mice to induce oral immune tolerance. Nat
Med, 3(7), 793-796.
196
MacCuish, A. C., Irvine, W. J., Barnes, E. W., & Duncan, L. J. (1974). Antibodies to pancreatic islet cells in
insulin-dependent diabetics with coexistent autoimmune disease. Lancet, 2(7896), 1529-1531.
Mai, H. L., Boeffard, F., Longis, J., Danger, R., Martinet, B., Haspot, F., Vanhove, B., Brouard, S., &
Soulillou, J. P. (2014). IL-7 receptor blockade following T cell depletion promotes long-term
allograft survival. J Clin Invest, 124(4), 1723-1733. doi: 10.1172/jci66287
Makino, S., Kunimoto, K., Muraoka, Y., Mizushima, Y., Katagiri, K., & Tochino, Y. (1980). Breeding of a
non-obese, diabetic strain of mice. Jikken Dobutsu, 29(1), 1-13.
Manjunath, N., Shankar, P., Wan, J., Weninger, W., Crowley, M. A., Hieshima, K., Springer, T. A., Fan, X.,
Shen, H., Lieberman, J., & von Andrian, U. H. (2001). Effector differentiation is not prerequisite for
generation of memory cytotoxic T lymphocytes. J Clin Invest, 108(6), 871-878. doi:
10.1172/jci13296
Mannering, S. I., Harrison, L. C., Williamson, N. A., Morris, J. S., Thearle, D. J., Jensen, K. P., Kay, T. W.,
Rossjohn, J., Falk, B. A., Nepom, G. T., & Purcell, A. W. (2005). The insulin A-chain epitope
recognized by human T cells is posttranslationally modified. J Exp Med, 202(9), 1191-1197. doi:
10.1084/jem.20051251
Mannering, S. I., Morris, J. S., Stone, N. L., Jensen, K. P., PM, V. A. N. E., & Harrison, L. C. (2004). CD4+
T cell proliferation in response to GAD and proinsulin in healthy, pre-diabetic, and diabetic donors.
Ann N Y Acad Sci, 1037, 16-21. doi: 10.1196/annals.1337.003
Martinez, G. J., Pereira, R. M., Aijo, T., Kim, E. Y., Marangoni, F., Pipkin, M. E., Togher, S., Heissmeyer,
V., Zhang, Y. C., Crotty, S., Lamperti, E. D., Ansel, K. M., Mempel, T. R., Lahdesmaki, H., Hogan,
P. G., & Rao, A. (2015). The transcription factor NFAT promotes exhaustion of activated CD8(+) T
cells. Immunity, 42(2), 265-278. doi: 10.1016/j.immuni.2015.01.006
Martinez, X., Kreuwel, H. T., Redmond, W. L., Trenney, R., Hunter, K., Rosen, H., Sarvetnick, N., Wicker,
L. S., & Sherman, L. A. (2005). CD8+ T cell tolerance in nonobese diabetic mice is restored by
insulin-dependent diabetes resistance alleles. J Immunol, 175(3), 1677-1685.
Mazo, I. B., Honczarenko, M., Leung, H., Cavanagh, L. L., Bonasio, R., Weninger, W., Engelke, K., Xia, L.,
McEver, R. P., Koni, P. A., Silberstein, L. E., & von Andrian, U. H. (2005). Bone marrow is a major
reservoir and site of recruitment for central memory CD8+ T cells. Immunity, 22(2), 259-270. doi:
10.1016/j.immuni.2005.01.008
McGinty, J. W., Chow, I. T., Greenbaum, C., Odegard, J., Kwok, W. W., & James, E. A. (2014).
Recognition of posttranslationally modified GAD65 epitopes in subjects with type 1 diabetes.
Diabetes, 63(9), 3033-3040. doi: 10.2337/db13-1952
McHugh, R. S., Whitters, M. J., Piccirillo, C. A., Young, D. A., Shevach, E. M., Collins, M., & Byrne, M. C.
(2002). CD4(+)CD25(+) immunoregulatory T cells: gene expression analysis reveals a functional
role for the glucocorticoid-induced TNF receptor. Immunity, 16(2), 311-323.
McInerney, M. F., Rath, S., & Janeway, C. A., Jr. (1991). Exclusive expression of MHC class II proteins on
CD45+ cells in pancreatic islets of NOD mice. Diabetes, 40(5), 648-651.
McNally, A., Hill, G. R., Sparwasser, T., Thomas, R., & Steptoe, R. J. (2011). CD4+CD25+ regulatory T
cells control CD8+ T-cell effector differentiation by modulating IL-2 homeostasis. Proc Natl Acad
Sci U S A, 108(18), 7529-7534. doi: 10.1073/pnas.1103782108
Mehlhop-Williams, E. R., & Bevan, M. J. (2014). Memory CD8+ T cells exhibit increased antigen threshold
requirements for recall proliferation. J Exp Med, 211(2), 345-356. doi: 10.1084/jem.20131271
Meng, Q., Wang, W., Shi, X., Jin, Y., & Zhang, Y. (2011). Protection against autoimmune diabetes by
silkworm-produced GFP-tagged CTB-insulin fusion protein. Clin Dev Immunol, 2011, 831704. doi:
10.1155/2011/831704
Metzger, T. C., & Anderson, M. S. (2011). Control of central and peripheral tolerance by Aire. Immunol Rev,
241(1), 89-103. doi: 10.1111/j.1600-065X.2011.01008.x
Minami, Y., Samelson, L. E., & Klausner, R. D. (1987). Internalization and cycling of the T cell antigen
receptor. Role of protein kinase C. J Biol Chem, 262(27), 13342-13347.
Mingozzi, F., Liu, Y. L., Dobrzynski, E., Kaufhold, A., Liu, J. H., Wang, Y., Arruda, V. R., High, K. A., &
Herzog, R. W. (2003). Induction of immune tolerance to coagulation factor IX antigen by in vivo
hepatic gene transfer. J Clin Invest, 111(9), 1347-1356. doi: 10.1172/jci16887
Montane, J., Bischoff, L., Soukhatcheva, G., Dai, D. L., Hardenberg, G., Levings, M. K., Orban, P. C.,
Kieffer, T. J., Tan, R., & Verchere, C. B. (2011). Prevention of murine autoimmune diabetes by
CCL22-mediated Treg recruitment to the pancreatic islets. J Clin Invest, 121(8), 3024-3028. doi:
10.1172/jci43048
197
Monti, P., Scirpoli, M., Maffi, P., Ghidoli, N., De Taddeo, F., Bertuzzi, F., Piemonti, L., Falcone, M., Secchi,
A., & Bonifacio, E. (2008). Islet transplantation in patients with autoimmune diabetes induces
homeostatic cytokines that expand autoreactive memory T cells. J Clin Invest, 118(5), 1806-1814.
doi: 10.1172/jci35197
Monti, P., Scirpoli, M., Rigamonti, A., Mayr, A., Jaeger, A., Bonfanti, R., Chiumello, G., Ziegler, A. G., &
Bonifacio, E. (2007). Evidence for in vivo primed and expanded autoreactive T cells as a specific
feature of patients with type 1 diabetes. J Immunol, 179(9), 5785-5792.
Moriyama, H., Abiru, N., Paronen, J., Sikora, K., Liu, E., Miao, D., Devendra, D., Beilke, J., Gianani, R.,
Gill, R. G., & Eisenbarth, G. S. (2003). Evidence for a primary islet autoantigen (preproinsulin 1) for
insulitis and diabetes in the nonobese diabetic mouse. Proc Natl Acad Sci U S A, 100(18), 1037610381. doi: 10.1073/pnas.1834450100
Mottram, P. L., Murray-Segal, L. J., Han, W., Maguire, J., & Stein-Oakley, A. N. (2002). Remission and
pancreas isograft survival in recent onset diabetic NOD mice after treatment with low-dose anti-CD3
monoclonal antibodies. Transpl Immunol, 10(1), 63-72.
Mukherjee, G., Geliebter, A., Babad, J., Santamaria, P., Serreze, D. V., Freeman, G. J., Tarbell, K. V.,
Sharpe, A., & DiLorenzo, T. P. (2013). DEC-205-mediated antigen targeting to steady-state
dendritic cells induces deletion of diabetogenic CD8(+) T cells independently of PD-1 and PD-L1.
Int Immunol, 25(11), 651-660. doi: 10.1093/intimm/dxt031
Mukherjee, R., Chaturvedi, P., Qin, H. Y., & Singh, B. (2003). CD4+CD25+ regulatory T cells generated in
response to insulin B:9-23 peptide prevent adoptive transfer of diabetes by diabetogenic T cells. J
Autoimmun, 21(3), 221-237.
Mukhopadhaya, A., Hanafusa, T., Jarchum, I., Chen, Y. G., Iwai, Y., Serreze, D. V., Steinman, R. M.,
Tarbell, K. V., & DiLorenzo, T. P. (2008). Selective delivery of beta cell antigen to dendritic cells in
vivo leads to deletion and tolerance of autoreactive CD8+ T cells in NOD mice. Proc Natl Acad Sci
U S A, 105(17), 6374-6379. doi: 10.1073/pnas.0802644105
Nagata, M., & Yoon, J. W. (1992). Studies on autoimmunity for T-cell-mediated beta-cell destruction.
Distinct difference in beta-cell destruction between CD4+ and CD8+ T-cell clones derived from
lymphocytes infiltrating the islets of NOD mice. Diabetes, 41(8), 998-1008.
Naik, R. G., Brooks-Worrell, B. M., & Palmer, J. P. (2009). Latent autoimmune diabetes in adults. J Clin
Endocrinol Metab, 94(12), 4635-4644. doi: 10.1210/jc.2009-1120
Nakayama, M., Abiru, N., Moriyama, H., Babaya, N., Liu, E., Miao, D., Yu, L., Wegmann, D. R., Hutton, J.
C., Elliott, J. F., & Eisenbarth, G. S. (2005). Prime role for an insulin epitope in the development of
type 1 diabetes in NOD mice. Nature, 435(7039), 220-223. doi: 10.1038/nature03523
Nakayama, M., Beilke, J. N., Jasinski, J. M., Kobayashi, M., Miao, D., Li, M., Coulombe, M. G., Liu, E.,
Elliott, J. F., Gill, R. G., & Eisenbarth, G. S. (2007). Priming and effector dependence on insulin
B:9–23 peptide in NOD islet autoimmunity. J Clin Invest, 117(7), 1835-1843. doi: 10.1172/jci31368
Nanto-Salonen, K., Kupila, A., Simell, S., Siljander, H., Salonsaari, T., Hekkala, A., Korhonen, S., Erkkola,
R., Sipila, J. I., Haavisto, L., Siltala, M., Tuominen, J., Hakalax, J., Hyoty, H., Ilonen, J., Veijola, R.,
Simell, T., Knip, M., & Simell, O. (2008). Nasal insulin to prevent type 1 diabetes in children with
HLA genotypes and autoantibodies conferring increased risk of disease: a double-blind, randomised
controlled trial. Lancet, 372(9651), 1746-1755. doi: 10.1016/s0140-6736(08)61309-4
Naramura, M., Jang, I. K., Kole, H., Huang, F., Haines, D., & Gu, H. (2002). c-Cbl and Cbl-b regulate T cell
responsiveness by promoting ligand-induced TCR down-modulation. Nat Immunol, 3(12), 11921199. doi: 10.1038/ni855
Nasreen, M., Waldie, T. M., Dixon, C. M., & Steptoe, R. J. (2010). Steady-state antigen-expressing dendritic
cells terminate CD4+ memory T-cell responses. Eur J Immunol, 40(7), 2016-2025. doi:
10.1002/eji.200940085
Naumov, Y. N., Bahjat, K. S., Gausling, R., Abraham, R., Exley, M. A., Koezuka, Y., Balk, S. B.,
Strominger, J. L., Clare-Salzer, M., & Wilson, S. B. (2001). Activation of CD1d-restricted T cells
protects NOD mice from developing diabetes by regulating dendritic cell subsets. Proc Natl Acad
Sci U S A, 98(24), 13838-13843. doi: 10.1073/pnas.251531798
Nishimoto, H., Kikutani, H., Yamamura, K., & Kishimoto, T. (1987). Prevention of autoimmune insulitis by
expression of I-E molecules in NOD mice. Nature, 328(6129), 432-434. doi: 10.1038/328432a0
Oderup, C., Cederbom, L., Makowska, A., Cilio, C. M., & Ivars, F. (2006). Cytotoxic T lymphocyte antigen4-dependent down-modulation of costimulatory molecules on dendritic cells in CD4+ CD25+
regulatory T-cell-mediated suppression. Immunology, 118(2), 240-249. doi: 10.1111/j.13652567.2006.02362.x
198
Oeser, J. K., Parekh, V. V., Wang, Y., Jegadeesh, N. K., Sarkar, S. A., Wong, R., Lee, C. E., Pound, L. D.,
Hutton, J. C., Van Kaer, L., & O'Brien, R. M. (2011). Deletion of the G6pc2 gene encoding the isletspecific glucose-6-phosphatase catalytic subunit-related protein does not affect the progression or
incidence of type 1 diabetes in NOD/ShiLtJ mice. Diabetes, 60(11), 2922-2927. doi: 10.2337/db110220
Ohashi, P. S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C. T., Odermatt, B., Malissen, B., Zinkernagel, R.
M., & Hengartner, H. (1991). Ablation of "tolerance" and induction of diabetes by virus infection in
viral antigen transgenic mice. Cell, 65(2), 305-317.
Okitsu, T., Bartlett, S. T., Hadley, G. A., Drachenberg, C. B., & Farney, A. C. (2001). Recurrent
autoimmunity accelerates destruction of minor and major histoincompatible islet grafts in nonobese
diabetic (NOD) mice. Am J Transplant, 1(2), 138-145.
Oldstone, M. B., Nerenberg, M., Southern, P., Price, J., & Lewicki, H. (1991). Virus infection triggers
insulin-dependent diabetes mellitus in a transgenic model: role of anti-self (virus) immune response.
Cell, 65(2), 319-331.
Oling, V., Marttila, J., Ilonen, J., Kwok, W. W., Nepom, G., Knip, M., Simell, O., & Reijonen, H. (2005).
GAD65- and proinsulin-specific CD4+ T-cells detected by MHC class II tetramers in peripheral
blood of type 1 diabetes patients and at-risk subjects. J Autoimmun, 25(3), 235-243. doi:
10.1016/j.jaut.2005.09.018
Olson, J. A., McDonald-Hyman, C., Jameson, S. C., & Hamilton, S. E. (2013). Effector-like CD8(+) T cells
in the memory population mediate potent protective immunity. Immunity, 38(6), 1250-1260. doi:
10.1016/j.immuni.2013.05.009
Orban, T., Farkas, K., Jalahej, H., Kis, J., Treszl, A., Falk, B., Reijonen, H., Wolfsdorf, J., Ricker, A.,
Matthews, J. B., Tchao, N., Sayre, P., & Bianchine, P. (2010). Autoantigen-specific regulatory T
cells induced in patients with type 1 diabetes mellitus by insulin B-chain immunotherapy. J
Autoimmun, 34(4), 408-415. doi: 10.1016/j.jaut.2009.10.005
Ott, P. A., Dittrich, M. T., Herzog, B. A., Guerkov, R., Gottlieb, P. A., Putnam, A. L., Durinovic-Bello, I.,
Boehm, B. O., Tary-Lehmann, M., & Lehmann, P. V. (2004). T cells recognize multiple GAD65 and
proinsulin epitopes in human type 1 diabetes, suggesting determinant spreading. J Clin Immunol,
24(4), 327-339. doi: 10.1023/b:joci.0000029120.77824.41
Padgett, L. E., Burg, A. R., Lei, W., & Tse, H. M. (2015). Loss of NADPH oxidase-derived superoxide
skews macrophage phenotypes to delay type 1 diabetes. Diabetes, 64(3), 937-946. doi:
10.2337/db14-0929
Palmer, J. P., Asplin, C. M., Clemons, P., Lyen, K., Tatpati, O., Raghu, P. K., & Paquette, T. L. (1983).
Insulin antibodies in insulin-dependent diabetics before insulin treatment. Science, 222(4630), 13371339.
Parish, I. A., Rao, S., Smyth, G. K., Juelich, T., Denyer, G. S., Davey, G. M., Strasser, A., & Heath, W. R.
(2009). The molecular signature of CD8+ T cells undergoing deletional tolerance. Blood, 113(19),
4575-4585. doi: 10.1182/blood-2008-10-185223
Pauken, K. E., Linehan, J. L., Spanier, J. A., Sahli, N. L., Kalekar, L. A., Binstadt, B. A., Moon, J. J.,
Mueller, D. L., Jenkins, M. K., & Fife, B. T. (2013). Cutting edge: type 1 diabetes occurs despite
robust anergy among endogenous insulin-specific CD4 T cells in NOD mice. J Immunol, 191(10),
4913-4917. doi: 10.4049/jimmunol.1301927
Pearson, T., Greiner, D. L., & Shultz, L. D. (2008). Creation of "humanized" mice to study human immunity.
Curr Protoc Immunol, Chapter 15, Unit 15.21. doi: 10.1002/0471142735.im1521s81
Pellicci, D. G., Hammond, K. J., Coquet, J., Kyparissoudis, K., Brooks, A. G., Kedzierska, K., Keating, R.,
Turner, S., Berzins, S., Smyth, M. J., & Godfrey, D. I. (2005). DX5/CD49b-positive T cells are not
synonymous with CD1d-dependent NKT cells. J Immunol, 175(7), 4416-4425.
Penaranda, C., Kuswanto, W., Hofmann, J., Kenefeck, R., Narendran, P., Walker, L. S., Bluestone, J. A.,
Abbas, A. K., & Dooms, H. (2012). IL-7 receptor blockade reverses autoimmune diabetes by
promoting inhibition of effector/memory T cells. Proc Natl Acad Sci U S A, 109(31), 12668-12673.
doi: 10.1073/pnas.1203692109
Peng, Y., Laouar, Y., Li, M. O., Green, E. A., & Flavell, R. A. (2004). TGF-beta regulates in vivo expansion
of Foxp3-expressing CD4+CD25+ regulatory T cells responsible for protection against diabetes.
Proc Natl Acad Sci U S A, 101(13), 4572-4577. doi: 10.1073/pnas.0400810101
Pescovitz, M. D., Greenbaum, C. J., Krause-Steinrauf, H., Becker, D. J., Gitelman, S. E., Goland, R.,
Gottlieb, P. A., Marks, J. B., McGee, P. F., Moran, A. M., Raskin, P., Rodriguez, H., Schatz, D. A.,
Wherrett, D., Wilson, D. M., Lachin, J. M., Skyler, J. S., & Type 1 Diabetes TrialNet Anti, C. D. S.
199
G. (2009). Rituximab, B-lymphocyte depletion, and preservation of beta-cell function. N Engl J
Med, 361(22), 2143-2152. doi: 10.1056/NEJMoa0904452
Petersen, J. S., Bregenholt, S., Apostolopolous, V., Homann, D., Wolfe, T., Hughes, A., De Jongh, K.,
Wang, M., Dyrberg, T., & Von Herrath, M. G. (2003). Coupling of oral human or porcine insulin to
the B subunit of cholera toxin (CTB) overcomes critical antigenic differences for prevention of type
I diabetes. Clin Exp Immunol, 134(1), 38-45.
Petersen, L. D., Duinkerken, G., Bruining, G. J., van Lier, R. A., de Vries, R. R., & Roep, B. O. (1996).
Increased numbers of in vivo activated T cells in patients with recent onset insulin-dependent
diabetes mellitus. J Autoimmun, 9(6), 731-737.
Peterson, P., & Peltonen, L. (2005). Autoimmune polyendocrinopathy syndrome type 1 (APS1) and AIRE
gene: new views on molecular basis of autoimmunity. J Autoimmun, 25 Suppl, 49-55. doi:
10.1016/j.jaut.2005.09.022
Pihoker, C., Gilliam, L. K., Hampe, C. S., & Lernmark, A. (2005). Autoantibodies in diabetes. Diabetes, 54
Suppl 2, S52-61.
Pinkse, G. G., Tysma, O. H., Bergen, C. A., Kester, M. G., Ossendorp, F., van Veelen, P. A., Keymeulen, B.,
Pipeleers, D., Drijfhout, J. W., & Roep, B. O. (2005). Autoreactive CD8 T cells associated with beta
cell destruction in type 1 diabetes. Proc Natl Acad Sci U S A, 102(51), 18425-18430. doi:
10.1073/pnas.0508621102
Poirot, L., Benoist, C., & Mathis, D. (2004). Natural killer cells distinguish innocuous and destructive forms
of pancreatic islet autoimmunity. Proc Natl Acad Sci U S A, 101(21), 8102-8107. doi:
10.1073/pnas.0402065101
Polanski, M., Melican, N. S., Zhang, J., & Weiner, H. L. (1997). Oral administration of the immunodominant
B-chain of insulin reduces diabetes in a co-transfer model of diabetes in the NOD mouse and is
associated with a switch from Th1 to Th2 cytokines. J Autoimmun, 10(4), 339-346. doi:
10.1006/jaut.1997.0148
Poulton, L. D., Smyth, M. J., Hawke, C. G., Silveira, P., Shepherd, D., Naidenko, O. V., Godfrey, D. I., &
Baxter, A. G. (2001). Cytometric and functional analyses of NK and NKT cell deficiencies in NOD
mice. Int Immunol, 13(7), 887-896.
Powell, B. R., Buist, N. R., & Stenzel, P. (1982). An X-linked syndrome of diarrhea, polyendocrinopathy,
and fatal infection in infancy. J Pediatr, 100(5), 731-737.
Pozzilli, P., Pitocco, D., Visalli, N., Cavallo, M. G., Buzzetti, R., Crino, A., Spera, S., Suraci, C., Multari, G.,
Cervoni, M., Manca Bitti, M. L., Matteoli, M. C., Marietti, G., Ferrazzoli, F., Cassone Faldetta, M.
R., Giordano, C., Sbriglia, M., Sarugeri, E., & Ghirlanda, G. (2000). No effect of oral insulin on
residual beta-cell function in recent-onset type I diabetes (the IMDIAB VII). IMDIAB Group.
Diabetologia, 43(8), 1000-1004.
Prasad, S., Kohm, A. P., McMahon, J. S., Luo, X., & Miller, S. D. (2012). Pathogenesis of NOD diabetes is
initiated by reactivity to the insulin B chain 9-23 epitope and involves functional epitope spreading.
J Autoimmun, 39(4), 347-353. doi: 10.1016/j.jaut.2012.04.005
Probst, H. C., Lagnel, J., Kollias, G., & van den Broek, M. (2003). Inducible transgenic mice reveal resting
dendritic cells as potent inducers of CD8+ T cell tolerance. Immunity, 18(5), 713-720.
Qu, Y., Jin, S., Zhang, A., Zhang, B., Shi, X., Wang, J., & Zhao, Y. (2010a). Gamma-ray resistance of
regulatory CD4+CD25+Foxp3+ T cells in mice. Radiat Res, 173(2), 148-157. doi: 10.1667/rr0978.1
Qu, Y., Zhang, B., Liu, S., Zhang, A., Wu, T., & Zhao, Y. (2010b). 2-Gy whole-body irradiation
significantly alters the balance of CD4+ CD25- T effector cells and CD4+ CD25+ Foxp3+ T
regulatory cells in mice. Cell Mol Immunol, 7(6), 419-427. doi: 10.1038/cmi.2010.45
Rabinovitch, A., Suarez-Pinzon, W. L., Shapiro, A. M., Rajotte, R. V., & Power, R. (2002). Combination
therapy with sirolimus and interleukin-2 prevents spontaneous and recurrent autoimmune diabetes in
NOD mice. Diabetes, 51(3), 638-645.
Ramadan, A., & Paczesny, S. (2015). Various forms of tissue damage and danger signals following
hematopoietic stem-cell transplantation. Front Immunol, 6, 14. doi: 10.3389/fimmu.2015.00014
Ramiya, V. K., Shang, X. Z., Wasserfall, C. H., & Maclaren, N. K. (1997). Effect of oral and intravenous
insulin and glutamic acid decarboxylase in NOD mice. Autoimmunity, 26(3), 139-151.
Redmond, W. L., Hernandez, J., & Sherman, L. A. (2003). Deletion of naive CD8 T cells requires persistent
antigen and is not programmed by an initial signal from the tolerogenic APC. J Immunol, 171(12),
6349-6354.
Redmond, W. L., Marincek, B. C., & Sherman, L. A. (2005). Distinct requirements for deletion versus
anergy during CD8 T cell peripheral tolerance in vivo. J Immunol, 174(4), 2046-2053.
200
Redmond, W. L., Wei, C. H., Kreuwel, H. T., & Sherman, L. A. (2008). The apoptotic pathway contributing
to the deletion of naive CD8 T cells during the induction of peripheral tolerance to a cross-presented
self-antigen. J Immunol, 180(8), 5275-5282.
Richer, M. J., Lavallee, D. J., Shanina, I., & Horwitz, M. S. (2012). Immunomodulation of antigen
presenting cells promotes natural regulatory T cells that prevent autoimmune diabetes in NOD mice.
PLoS One, 7(2), e31153. doi: 10.1371/journal.pone.0031153
Rigby, M. R., DiMeglio, L. A., Rendell, M. S., Felner, E. I., Dostou, J. M., Gitelman, S. E., Patel, C. M.,
Griffin, K. J., Tsalikian, E., Gottlieb, P. A., Greenbaum, C. J., Sherry, N. A., Moore, W. V.,
Monzavi, R., Willi, S. M., Raskin, P., Moran, A., Russell, W. E., Pinckney, A., Keyes-Elstein, L.,
Howell, M., Aggarwal, S., Lim, N., Phippard, D., Nepom, G. T., McNamara, J., & Ehlers, M. R.
(2013). Targeting of memory T cells with alefacept in new-onset type 1 diabetes (T1DAL study): 12
month results of a randomised, double-blind, placebo-controlled phase 2 trial. Lancet Diabetes
Endocrinol, 1(4), 284-294. doi: 10.1016/s2213-8587(13)70111-6
Robertson, R. P., Davis, C., Larsen, J., Stratta, R., & Sutherland, D. E. (2000). Pancreas and islet
transplantation for patients with diabetes. Diabetes Care, 23(1), 112-116.
Robertson, R. P., Davis, C., Larsen, J., Stratta, R., Sutherland, D. E., & American Diabetes, A. (2006).
Pancreas and islet transplantation in type 1 diabetes. Diabetes Care, 29(4), 935.
Roep, B. O., Solvason, N., Gottlieb, P. A., Abreu, J. R., Harrison, L. C., Eisenbarth, G. S., Yu, L., Leviten,
M., Hagopian, W. A., Buse, J. B., von Herrath, M., Quan, J., King, R. S., Robinson, W. H., Utz, P.
J., Garren, H., & Steinman, L. (2013). Plasmid-encoded proinsulin preserves C-peptide while
specifically reducing proinsulin-specific CD8(+) T cells in type 1 diabetes. Sci Transl Med, 5(191),
191ra182. doi: 10.1126/scitranslmed.3006103
Rosenzwajg, M., Churlaud, G., Mallone, R., Six, A., Derian, N., Chaara, W., Lorenzon, R., Long, S. A.,
Buckner, J. H., Afonso, G., Pham, H. P., Hartemann, A., Yu, A., Pugliese, A., Malek, T. R., &
Klatzmann, D. (2015). Low-dose interleukin-2 fosters a dose-dependent regulatory T cell tuned
milieu in T1D patients. J Autoimmun, 58, 48-58. doi: 10.1016/j.jaut.2015.01.001
Rosenzweig, M., Connole, M., Glickman, R., Yue, S. P., Noren, B., DeMaria, M., & Johnson, R. P. (2001).
Induction of cytotoxic T lymphocyte and antibody responses to enhanced green fluorescent protein
following transplantation of transduced CD34(+) hematopoietic cells. Blood, 97(7), 1951-1959.
Sachdeva, N., Paul, M., Badal, D., Kumar, R., Jacob, N., Dayal, D., Bhansali, A., Arora, S. K., & Bhadada,
S. K. (2015). Preproinsulin specific CD8+ T cells in subjects with latent autoimmune diabetes show
lower frequency and different pathophysiological characteristics than those with type 1 diabetes.
Clin Immunol, 157(1), 78-90. doi: 10.1016/j.clim.2015.01.005
Safford, M., Collins, S., Lutz, M. A., Allen, A., Huang, C. T., Kowalski, J., Blackford, A., Horton, M. R.,
Drake, C., Schwartz, R. H., & Powell, J. D. (2005). Egr-2 and Egr-3 are negative regulators of T cell
activation. Nat Immunol, 6(5), 472-480. doi: 10.1038/ni1193
Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M., & Toda, M. (1995). Immunologic self-tolerance
maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25). Breakdown of a
single mechanism of self-tolerance causes various autoimmune diseases. J Immunol, 155(3), 11511164.
Sallusto, F., Lenig, D., Forster, R., Lipp, M., & Lanzavecchia, A. (1999). Two subsets of memory T
lymphocytes with distinct homing potentials and effector functions. Nature, 401(6754), 708-712.
doi: 10.1038/44385
Salmond, R. J., Brownlie, R. J., Morrison, V. L., & Zamoyska, R. (2014). The tyrosine phosphatase PTPN22
discriminates weak self peptides from strong agonist TCR signals. Nat Immunol, 15(9), 875-883.
doi: 10.1038/ni.2958
Sancho, D., Gomez, M., & Sanchez-Madrid, F. (2005). CD69 is an immunoregulatory molecule induced
following activation. Trends Immunol, 26(3), 136-140. doi: 10.1016/j.it.2004.12.006
Schneider, A., Rieck, M., Sanda, S., Pihoker, C., Greenbaum, C., & Buckner, J. H. (2008). The effector T
cells of diabetic subjects are resistant to regulation via CD4+ FOXP3+ regulatory T cells. J Immunol,
181(10), 7350-7355.
Schuster, C., Brosi, H., Stifter, K., Boehm, B. O., & Schirmbeck, R. (2013). A missing PD-L1/PD-1
coinhibition regulates diabetes induction by preproinsulin-specific CD8 T-cells in an epitopespecific manner. PLoS One, 8(8), e71746. doi: 10.1371/journal.pone.0071746
Schwartz, R. H. (2003). T cell anergy. Annu Rev Immunol, 21, 305-334. doi:
10.1146/annurev.immunol.21.120601.141110
201
Serreze, D. V., Chapman, H. D., Varnum, D. S., Hanson, M. S., Reifsnyder, P. C., Richard, S. D., Fleming,
S. A., Leiter, E. H., & Shultz, L. D. (1996). B lymphocytes are essential for the initiation of T cellmediated autoimmune diabetes: analysis of a new "speed congenic" stock of NOD.Ig mu null mice. J
Exp Med, 184(5), 2049-2053.
Serreze, D. V., Gaedeke, J. W., & Leiter, E. H. (1993). Hematopoietic stem-cell defects underlying abnormal
macrophage development and maturation in NOD/Lt mice: defective regulation of cytokine
receptors and protein kinase C. Proc Natl Acad Sci U S A, 90(20), 9625-9629.
Serreze, D. V., Leiter, E. H., Christianson, G. J., Greiner, D., & Roopenian, D. C. (1994). Major
histocompatibility complex class I-deficient NOD-B2mnull mice are diabetes and insulitis resistant.
Diabetes, 43(3), 505-509.
Shalev, I., Schmelzle, M., Robson, S. C., & Levy, G. (2011). Making sense of regulatory T cell suppressive
function. Semin Immunol, 23(4), 282-292. doi: 10.1016/j.smim.2011.04.003
Shamim, M., Nanjappa, S. G., Singh, A., Plisch, E. H., LeBlanc, S. E., Walent, J., Svaren, J., Seroogy, C., &
Suresh, M. (2007). Cbl-b regulates antigen-induced TCR down-regulation and IFN-gamma
production by effector CD8 T cells without affecting functional avidity. J Immunol, 179(11), 72337243.
Sherry, N., Hagopian, W., Ludvigsson, J., Jain, S. M., Wahlen, J., Ferry, R. J., Jr., Bode, B., Aronoff, S.,
Holland, C., Carlin, D., King, K. L., Wilder, R. L., Pillemer, S., Bonvini, E., Johnson, S., Stein, K.
E., Koenig, S., Herold, K. C., Daifotis, A. G., & Protege Trial, I. (2011). Teplizumab for treatment of
type 1 diabetes (Protege study): 1-year results from a randomised, placebo-controlled trial. Lancet,
378(9790), 487-497. doi: 10.1016/s0140-6736(11)60931-8
Shiina, T., Inoko, H., & Kulski, J. K. (2004). An update of the HLA genomic region, locus information and
disease associations: 2004. Tissue Antigens, 64(6), 631-649. doi: 10.1111/j.1399-0039.2004.00327.x
Singh, K., Hjort, M., Thorvaldson, L., & Sandler, S. (2015). Concomitant analysis of Helios and Neuropilin1 as a marker to detect thymic derived regulatory T cells in naive mice. Sci Rep, 5, 7767. doi:
10.1038/srep07767
Skyler, J. S., Krischer, J. P., Wolfsdorf, J., Cowie, C., Palmer, J. P., Greenbaum, C., Cuthbertson, D., RafkinMervis, L. E., Chase, H. P., & Leschek, E. (2005). Effects of oral insulin in relatives of patients with
type 1 diabetes: The Diabetes Prevention Trial--Type 1. Diabetes Care, 28(5), 1068-1076.
Smith, T. R., Verdeil, G., Marquardt, K., & Sherman, L. A. (2014). Contribution of TCR signaling strength
to CD8+ T cell peripheral tolerance mechanisms. J Immunol, 193(7), 3409-3416. doi:
10.4049/jimmunol.1401194
Sojka, D. K., Hughson, A., & Fowell, D. J. (2009). CTLA-4 is required by CD4+CD25+ Treg to control
CD4+ T-cell lymphopenia-induced proliferation. Eur J Immunol, 39(6), 1544-1551. doi:
10.1002/eji.200838603
Solvason, N., Lou, Y. P., Peters, W., Evans, E., Martinez, J., Ramirez, U., Ocampo, A., Yun, R., Ahmad, S.,
Liu, E., Yu, L., Eisenbarth, G., Leviten, M., Steinman, L., & Garren, H. (2008). Improved efficacy
of a tolerizing DNA vaccine for reversal of hyperglycemia through enhancement of gene expression
and localization to intracellular sites. J Immunol, 181(12), 8298-8307.
Spatz, M., Eibl, N., Hink, S., Wolf, H. M., Fischer, G. F., Mayr, W. R., Schernthaner, G., & Eibl, M. M.
(2003). Impaired primary immune response in type-1 diabetes. Functional impairment at the level of
APCs and T-cells. Cell Immunol, 221(1), 15-26.
Stadinski, B. D., Delong, T., Reisdorph, N., Reisdorph, R., Powell, R. L., Armstrong, M., Piganelli, J. D.,
Barbour, G., Bradley, B., Crawford, F., Marrack, P., Mahata, S. K., Kappler, J. W., & Haskins, K.
(2010). Chromogranin A is an autoantigen in type 1 diabetes. Nat Immunol, 11(3), 225-231. doi:
10.1038/ni.1844
Starr, T. K., Jameson, S. C., & Hogquist, K. A. (2003). Positive and negative selection of T cells. Annu Rev
Immunol, 21, 139-176. doi: 10.1146/annurev.immunol.21.120601.141107
Steffes, M. W., Sibley, S., Jackson, M., & Thomas, W. (2003). Beta-cell function and the development of
diabetes-related complications in the diabetes control and complications trial. Diabetes Care, 26(3),
832-836.
Steptoe, R. J., Ritchie, J. M., & Harrison, L. C. (2003). Transfer of hematopoietic stem cells encoding
autoantigen prevents autoimmune diabetes. J Clin Invest, 111(9), 1357-1363. doi: 10.1172/jci15995
Steptoe, R. J., Ritchie, J. M., Wilson, N. S., Villadangos, J. A., Lew, A. M., & Harrison, L. C. (2007).
Cognate CD4+ help elicited by resting dendritic cells does not impair the induction of peripheral
tolerance in CD8+ T cells. J Immunol, 178(4), 2094-2103.
202
Steptoe, R. J., Stankovic, S., Lopaticki, S., Jones, L. K., Harrison, L. C., & Morahan, G. (2004). Persistence
of recipient lymphocytes in NOD mice after irradiation and bone marrow transplantation. J
Autoimmun, 22(2), 131-138. doi: 10.1016/j.jaut.2003.12.003
Stiller, C. R., Dupre, J., Gent, M., Jenner, M. R., Keown, P. A., Laupacis, A., Martell, R., Rodger, N. W.,
von Graffenried, B., & Wolfe, B. M. (1984). Effects of cyclosporine immunosuppression in insulindependent diabetes mellitus of recent onset. Science, 223(4643), 1362-1367.
Stritesky, G. L., Jameson, S. C., & Hogquist, K. A. (2012). Selection of self-reactive T cells in the thymus.
Annu Rev Immunol, 30, 95-114. doi: 10.1146/annurev-immunol-020711-075035
Stritesky, G. L., Xing, Y., Erickson, J. R., Kalekar, L. A., Wang, X., Mueller, D. L., Jameson, S. C., &
Hogquist, K. A. (2013). Murine thymic selection quantified using a unique method to capture
deleted T cells. Proc Natl Acad Sci U S A, 110(12), 4679-4684. doi: 10.1073/pnas.1217532110
Takiishi, T., Gysemans, C., Bouillon, R., & Mathieu, C. (2010). Vitamin D and diabetes. Endocrinol Metab
Clin North Am, 39(2), 419-446, table of contents. doi: 10.1016/j.ecl.2010.02.013
Takiishi, T., Korf, H., Van Belle, T. L., Robert, S., Grieco, F. A., Caluwaerts, S., Galleri, L., Spagnuolo, I.,
Steidler, L., Van Huynegem, K., Demetter, P., Wasserfall, C., Atkinson, M. A., Dotta, F., Rottiers,
P., Gysemans, C., & Mathieu, C. (2012). Reversal of autoimmune diabetes by restoration of antigenspecific tolerance using genetically modified Lactococcus lactis in mice. J Clin Invest, 122(5), 17171725. doi: 10.1172/jci60530
Tan, J. T., Ernst, B., Kieper, W. C., LeRoy, E., Sprent, J., & Surh, C. D. (2002). Interleukin (IL)-15 and IL-7
jointly regulate homeostatic proliferation of memory phenotype CD8+ cells but are not required for
memory phenotype CD4+ cells. J Exp Med, 195(12), 1523-1532.
Tanchot, C., Lemonnier, F. A., Perarnau, B., Freitas, A. A., & Rocha, B. (1997). Differential requirements
for survival and proliferation of CD8 naive or memory T cells. Science, 276(5321), 2057-2062.
Tanemura, M., Ohmura, Y., Deguchi, T., Machida, T., Tsukamoto, R., Wada, H., Kobayashi, S., Marubashi,
S., Eguchi, H., Ito, T., Nagano, H., Mori, M., & Doki, Y. (2012). Rapamycin causes upregulation of
autophagy and impairs islets function both in vitro and in vivo. Am J Transplant, 12(1), 102-114.
doi: 10.1111/j.1600-6143.2011.03771.x
Tang, Q., Adams, J. Y., Penaranda, C., Melli, K., Piaggio, E., Sgouroudis, E., Piccirillo, C. A., Salomon, B.
L., & Bluestone, J. A. (2008). Central role of defective interleukin-2 production in the triggering of
islet autoimmune destruction. Immunity, 28(5), 687-697. doi: 10.1016/j.immuni.2008.03.016
Teta, M., Rankin, M. M., Long, S. Y., Stein, G. M., & Kushner, J. A. (2007). Growth and regeneration of
adult beta cells does not involve specialized progenitors. Dev Cell, 12(5), 817-826. doi:
10.1016/j.devcel.2007.04.011
Thebault-Baumont, K., Dubois-Laforgue, D., Krief, P., Briand, J. P., Halbout, P., Vallon-Geoffroy, K.,
Morin, J., Laloux, V., Lehuen, A., Carel, J. C., Jami, J., Muller, S., & Boitard, C. (2003).
Acceleration of type 1 diabetes mellitus in proinsulin 2-deficient NOD mice. J Clin Invest, 111(6),
851-857. doi: 10.1172/jci16584
Thomas, H. E., Parker, J. L., Schreiber, R. D., & Kay, T. W. (1998). IFN-gamma action on pancreatic beta
cells causes class I MHC upregulation but not diabetes. J Clin Invest, 102(6), 1249-1257. doi:
10.1172/jci2899
Thomas, I. J., Petrich de Marquesini, L. G., Ravanan, R., Smith, R. M., Guerder, S., Flavell, R. A., Wraith,
D. C., Wen, L., & Wong, F. S. (2007). CD86 has sustained costimulatory effects on CD8 T cells. J
Immunol, 179(9), 5936-5946.
Thornton, A. M., Korty, P. E., Tran, D. Q., Wohlfert, E. A., Murray, P. E., Belkaid, Y., & Shevach, E. M.
(2010). Expression of Helios, an Ikaros transcription factor family member, differentiates thymicderived from peripherally induced Foxp3+ T regulatory cells. J Immunol, 184(7), 3433-3441. doi:
10.4049/jimmunol.0904028
Tian, J., Atkinson, M. A., Clare-Salzler, M., Herschenfeld, A., Forsthuber, T., Lehmann, P. V., & Kaufman,
D. L. (1996). Nasal administration of glutamate decarboxylase (GAD65) peptides induces Th2
responses and prevents murine insulin-dependent diabetes. J Exp Med, 183(4), 1561-1567.
Tian, J., Dang, H., & Kaufman, D. L. (2011). Combining antigen-based therapy with GABA treatment
synergistically prolongs survival of transplanted ss-cells in diabetic NOD mice. PLoS One, 6(9),
e25337. doi: 10.1371/journal.pone.0025337
Tian, J., Dang, H., Nguyen, A. V., Chen, Z., & Kaufman, D. L. (2014). Combined therapy with GABA and
proinsulin/alum acts synergistically to restore long-term normoglycemia by modulating T-cell
autoimmunity and promoting beta-cell replication in newly diabetic NOD mice. Diabetes, 63(9),
3128-3134. doi: 10.2337/db13-1385
203
Tian, J., Dang, H., von Boehmer, H., Jaeckel, E., & Kaufman, D. L. (2009). Transgenically induced GAD
tolerance curtails the development of early beta-cell autoreactivities but causes the subsequent
development of supernormal autoreactivities to other beta-cell antigens. Diabetes, 58(12), 28432850. doi: 10.2337/db08-0851
Tian, J., Zekzer, D., Lu, Y., Dang, H., & Kaufman, D. L. (2006). B cells are crucial for determinant
spreading of T cell autoimmunity among beta cell antigens in diabetes-prone nonobese diabetic
mice. J Immunol, 176(4), 2654-2661.
Todd, J. A., Bell, J. I., & McDevitt, H. O. (1987). HLA-DQ beta gene contributes to susceptibility and
resistance to insulin-dependent diabetes mellitus. Nature, 329(6140), 599-604. doi:
10.1038/329599a0
Todd, J. A., & Wicker, L. S. (2001). Genetic protection from the inflammatory disease type 1 diabetes in
humans and animal models. Immunity, 15(3), 387-395.
Torres-Aguilar, H., Sanchez-Torres, C., Jara, L. J., Blank, M., & Shoenfeld, Y. (2010). IL-10/TGF-betatreated dendritic cells, pulsed with insulin, specifically reduce the response to insulin of CD4+
effector/memory T cells from type 1 diabetic individuals. J Clin Immunol, 30(5), 659-668. doi:
10.1007/s10875-010-9430-5
Traitanon, O., Gorbachev, A., Bechtel, J. J., Keslar, K. S., Baldwin, W. M., 3rd, Poggio, E. D., & Fairchild,
R. L. (2014). IL-15 induces alloreactive CD28(-) memory CD8 T cell proliferation and CTLA4-Ig
resistant memory CD8 T cell activation. Am J Transplant, 14(6), 1277-1289. doi: 10.1111/ajt.12719
Trautmann, L., Janbazian, L., Chomont, N., Said, E. A., Gimmig, S., Bessette, B., Boulassel, M. R., Delwart,
E., Sepulveda, H., Balderas, R. S., Routy, J. P., Haddad, E. K., & Sekaly, R. P. (2006). Upregulation
of PD-1 expression on HIV-specific CD8+ T cells leads to reversible immune dysfunction. Nat Med,
12(10), 1198-1202. doi: 10.1038/nm1482
Trudeau, J. D., Kelly-Smith, C., Verchere, C. B., Elliott, J. F., Dutz, J. P., Finegood, D. T., Santamaria, P., &
Tan, R. (2003). Prediction of spontaneous autoimmune diabetes in NOD mice by quantification of
autoreactive T cells in peripheral blood. J Clin Invest, 111(2), 217-223. doi: 10.1172/jci16409
Tsushima, F., Yao, S., Shin, T., Flies, A., Flies, S., Xu, H., Tamada, K., Pardoll, D. M., & Chen, L. (2007).
Interaction between B7-H1 and PD-1 determines initiation and reversal of T-cell anergy. Blood,
110(1), 180-185. doi: 10.1182/blood-2006-11-060087
Turley, D. M., & Miller, S. D. (2007). Peripheral tolerance induction using ethylenecarbodiimide-fixed
APCs uses both direct and indirect mechanisms of antigen presentation for prevention of
experimental autoimmune encephalomyelitis. J Immunol, 178(4), 2212-2220.
Turley, S., Poirot, L., Hattori, M., Benoist, C., & Mathis, D. (2003). Physiological beta cell death triggers
priming of self-reactive T cells by dendritic cells in a type-1 diabetes model. J Exp Med, 198(10),
1527-1537. doi: 10.1084/jem.20030966
Ueda, H., Howson, J. M., Esposito, L., Heward, J., Snook, H., Chamberlain, G., Rainbow, D. B., Hunter, K.
M., Smith, A. N., Di Genova, G., Herr, M. H., Dahlman, I., Payne, F., Smyth, D., Lowe, C., Twells,
R. C., Howlett, S., Healy, B., Nutland, S., Rance, H. E., Everett, V., Smink, L. J., Lam, A. C.,
Cordell, H. J., Walker, N. M., Bordin, C., Hulme, J., Motzo, C., Cucca, F., Hess, J. F., Metzker, M.
L., Rogers, J., Gregory, S., Allahabadia, A., Nithiyananthan, R., Tuomilehto-Wolf, E., Tuomilehto,
J., Bingley, P., Gillespie, K. M., Undlien, D. E., Ronningen, K. S., Guja, C., Ionescu-Tirgoviste, C.,
Savage, D. A., Maxwell, A. P., Carson, D. J., Patterson, C. C., Franklyn, J. A., Clayton, D. G.,
Peterson, L. B., Wicker, L. S., Todd, J. A., & Gough, S. C. (2003). Association of the T-cell
regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature, 423(6939), 506-511. doi:
10.1038/nature01621
Uehira, M., Uno, M., Kurner, T., Kikutani, H., Mori, K., Inomoto, T., Uede, T., Miyazaki, J., Nishimoto, H.,
Kishimoto, T., & et al. (1989). Development of autoimmune insulitis is prevented in E alpha d but
not in A beta k NOD transgenic mice. Int Immunol, 1(2), 209-213.
Unger, W. W., Pearson, T., Abreu, J. R., Laban, S., van der Slik, A. R., der Kracht, S. M., Kester, M. G.,
Serreze, D. V., Shultz, L. D., Griffioen, M., Drijfhout, J. W., Greiner, D. L., & Roep, B. O. (2012).
Islet-specific CTL cloned from a type 1 diabetes patient cause beta-cell destruction after engraftment
into HLA-A2 transgenic NOD/scid/IL2RG null mice. PLoS One, 7(11), e49213. doi:
10.1371/journal.pone.0049213
Vafiadis, P., Bennett, S. T., Todd, J. A., Nadeau, J., Grabs, R., Goodyer, C. G., Wickramasinghe, S., Colle,
E., & Polychronakos, C. (1997). Insulin expression in human thymus is modulated by INS VNTR
alleles at the IDDM2 locus. Nat Genet, 15(3), 289-292. doi: 10.1038/ng0397-289
204
Valitutti, S., Muller, S., Cella, M., Padovan, E., & Lanzavecchia, A. (1995). Serial triggering of many T-cell
receptors by a few peptide-MHC complexes. Nature, 375(6527), 148-151. doi: 10.1038/375148a0
Valujskikh, A., Lantz, O., Celli, S., Matzinger, P., & Heeger, P. S. (2002). Cross-primed CD8(+) T cells
mediate graft rejection via a distinct effector pathway. Nat Immunol, 3(9), 844-851. doi:
10.1038/ni831
van der Merwe, P. A., Bodian, D. L., Daenke, S., Linsley, P., & Davis, S. J. (1997). CD80 (B7-1) binds both
CD28 and CTLA-4 with a low affinity and very fast kinetics. J Exp Med, 185(3), 393-403.
Varanasi, V., Avanesyan, L., Schumann, D. M., & Chervonsky, A. V. (2012). Cytotoxic mechanisms
employed by mouse T cells to destroy pancreatic beta-cells. Diabetes, 61(11), 2862-2870. doi:
10.2337/db11-1784
Verdaguer, J., Schmidt, D., Amrani, A., Anderson, B., Averill, N., & Santamaria, P. (1997). Spontaneous
autoimmune diabetes in monoclonal T cell nonobese diabetic mice. J Exp Med, 186(10), 1663-1676.
Vincent, B. G., Young, E. F., Buntzman, A. S., Stevens, R., Kepler, T. B., Tisch, R. M., Frelinger, J. A., &
Hess, P. R. (2010). Toxin-coupled MHC class I tetramers can specifically ablate autoreactive CD8+
T cells and delay diabetes in nonobese diabetic mice. J Immunol, 184(8), 4196-4204. doi:
10.4049/jimmunol.0903931
Voltarelli, J. C., Couri, C. E., Stracieri, A. B., Oliveira, M. C., Moraes, D. A., Pieroni, F., Coutinho, M.,
Malmegrim, K. C., Foss-Freitas, M. C., Simoes, B. P., Foss, M. C., Squiers, E., & Burt, R. K.
(2007). Autologous nonmyeloablative hematopoietic stem cell transplantation in newly diagnosed
type 1 diabetes mellitus. JAMA, 297(14), 1568-1576. doi: 10.1001/jama.297.14.1568
Waithman, J., Gebhardt, T., Davey, G. M., Heath, W. R., & Carbone, F. R. (2008). Cutting edge: Enhanced
IL-2 signaling can convert self-specific T cell response from tolerance to autoimmunity. J Immunol,
180(9), 5789-5793.
Wan, Y. Y., & Flavell, R. A. (2007). Regulatory T-cell functions are subverted and converted owing to
attenuated Foxp3 expression. Nature, 445(7129), 766-770. doi: 10.1038/nature05479
Wang, J., Yoshida, T., Nakaki, F., Hiai, H., Okazaki, T., & Honjo, T. (2005). Establishment of NOD-Pdcd1/- mice as an efficient animal model of type I diabetes. Proc Natl Acad Sci U S A, 102(33), 1182311828. doi: 10.1073/pnas.0505497102
Watson, C. J. (2015). The current challenges for pancreas transplantation for diabetes mellitus. Pharmacol
Res, 98, 45-51. doi: 10.1016/j.phrs.2015.01.005
Webster, K. E., Walters, S., Kohler, R. E., Mrkvan, T., Boyman, O., Surh, C. D., Grey, S. T., & Sprent, J.
(2009). In vivo expansion of T reg cells with IL-2-mAb complexes: induction of resistance to EAE
and long-term acceptance of islet allografts without immunosuppression. J Exp Med, 206(4), 751760. doi: 10.1084/jem.20082824
Welfare, A. I. o. H. a. (2011). Diabetes prevalence in Australia: detailed estimates for 2007-2008. .
Wells, J. W., Cowled, C. J., Farzaneh, F., & Noble, A. (2008). Combined triggering of dendritic cell
receptors results in synergistic activation and potent cytotoxic immunity. J Immunol, 181(5), 34223431.
Weninger, W., Crowley, M. A., Manjunath, N., & von Andrian, U. H. (2001). Migratory properties of naive,
effector, and memory CD8(+) T cells. J Exp Med, 194(7), 953-966.
Wenzlau, J. M., Juhl, K., Yu, L., Moua, O., Sarkar, S. A., Gottlieb, P., Rewers, M., Eisenbarth, G. S., Jensen,
J., Davidson, H. W., & Hutton, J. C. (2007). The cation efflux transporter ZnT8 (Slc30A8) is a major
autoantigen in human type 1 diabetes. Proc Natl Acad Sci U S A, 104(43), 17040-17045. doi:
10.1073/pnas.0705894104
Wherry, E. J., Teichgraber, V., Becker, T. C., Masopust, D., Kaech, S. M., Antia, R., von Andrian, U. H., &
Ahmed, R. (2003). Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat
Immunol, 4(3), 225-234. doi: 10.1038/ni889
Wicker, L. S., Todd, J. A., & Peterson, L. B. (1995). Genetic control of autoimmune diabetes in the NOD
mouse. Annu Rev Immunol, 13, 179-200. doi: 10.1146/annurev.iy.13.040195.001143
Williams, M. A., & Bevan, M. J. (2007). Effector and memory CTL differentiation. Annu Rev Immunol, 25,
171-192. doi: 10.1146/annurev.immunol.25.022106.141548
Williams, M. A., Tyznik, A. J., & Bevan, M. J. (2006). Interleukin-2 signals during priming are required for
secondary expansion of CD8+ memory T cells. Nature, 441(7095), 890-893. doi:
10.1038/nature04790
Wills, M. R., Okecha, G., Weekes, M. P., Gandhi, M. K., Sissons, P. J., & Carmichael, A. J. (2002).
Identification of naive or antigen-experienced human CD8(+) T cells by expression of costimulation
205
and chemokine receptors: analysis of the human cytomegalovirus-specific CD8(+) T cell response. J
Immunol, 168(11), 5455-5464.
Wong, F. S., Karttunen, J., Dumont, C., Wen, L., Visintin, I., Pilip, I. M., Shastri, N., Pamer, E. G., &
Janeway, C. A., Jr. (1999). Identification of an MHC class I-restricted autoantigen in type 1 diabetes
by screening an organ-specific cDNA library. Nat Med, 5(9), 1026-1031. doi: 10.1038/12465
Wong, F. S., Moustakas, A. K., Wen, L., Papadopoulos, G. K., & Janeway, C. A., Jr. (2002). Analysis of
structure and function relationships of an autoantigenic peptide of insulin bound to H-2K(d) that
stimulates CD8 T cells in insulin-dependent diabetes mellitus. Proc Natl Acad Sci U S A, 99(8),
5551-5556. doi: 10.1073/pnas.072037299
Wong, F. S., Siew, L. K., Scott, G., Thomas, I. J., Chapman, S., Viret, C., & Wen, L. (2009). Activation of
insulin-reactive CD8 T-cells for development of autoimmune diabetes. Diabetes, 58(5), 1156-1164.
doi: 10.2337/db08-0800
Wong, F. S., Visintin, I., Wen, L., Flavell, R. A., & Janeway, C. A., Jr. (1996). CD8 T cell clones from
young nonobese diabetic (NOD) islets can transfer rapid onset of diabetes in NOD mice in the
absence of CD4 cells. J Exp Med, 183(1), 67-76.
Woo, S. R., Turnis, M. E., Goldberg, M. V., Bankoti, J., Selby, M., Nirschl, C. J., Bettini, M. L., Gravano, D.
M., Vogel, P., Liu, C. L., Tangsombatvisit, S., Grosso, J. F., Netto, G., Smeltzer, M. P., Chaux, A.,
Utz, P. J., Workman, C. J., Pardoll, D. M., Korman, A. J., Drake, C. G., & Vignali, D. A. (2012).
Immune inhibitory molecules LAG-3 and PD-1 synergistically regulate T-cell function to promote
tumoral immune escape. Cancer Res, 72(4), 917-927. doi: 10.1158/0008-5472.can-11-1620
Worsaae, A., Hejnaes, K., Moody, A., Ludvigsson, J., Pociot, F., Lorenzen, T., & Dyrberg, T. (1995). T cell
proliferative responses to glutamic acid decarboxylase-65 in IDDM are negatively associated with
HLA DR3/4. Autoimmunity, 22(3), 183-189.
Yamamoto, T., Yamato, E., Tashiro, F., Sato, T., Noso, S., Ikegami, H., Tamura, S., Yanagawa, Y., &
Miyazaki, J. I. (2004). Development of autoimmune diabetes in glutamic acid decarboxylase 65
(GAD65) knockout NOD mice. Diabetologia, 47(2), 221-224. doi: 10.1007/s00125-003-1296-0
Yamanouchi, J., Rainbow, D., Serra, P., Howlett, S., Hunter, K., Garner, V. E., Gonzalez-Munoz, A., Clark,
J., Veijola, R., Cubbon, R., Chen, S. L., Rosa, R., Cumiskey, A. M., Serreze, D. V., Gregory, S.,
Rogers, J., Lyons, P. A., Healy, B., Smink, L. J., Todd, J. A., Peterson, L. B., Wicker, L. S., &
Santamaria, P. (2007). Interleukin-2 gene variation impairs regulatory T cell function and causes
autoimmunity. Nat Genet, 39(3), 329-337. doi: 10.1038/ng1958
Yang, J., Brook, M. O., Carvalho-Gaspar, M., Zhang, J., Ramon, H. E., Sayegh, M. H., Wood, K. J., Turka,
L. A., & Jones, N. D. (2007). Allograft rejection mediated by memory T cells is resistant to
regulation. Proc Natl Acad Sci U S A, 104(50), 19954-19959. doi: 10.1073/pnas.0704397104
Yang, J., Chow, I. T., Sosinowski, T., Torres-Chinn, N., Greenbaum, C. J., James, E. A., Kappler, J. W.,
Davidson, H. W., & Kwok, W. W. (2014). Autoreactive T cells specific for insulin B:11-23
recognize a low-affinity peptide register in human subjects with autoimmune diabetes. Proc Natl
Acad Sci U S A, 111(41), 14840-14845. doi: 10.1073/pnas.1416864111
Yang, J., Danke, N. A., Berger, D., Reichstetter, S., Reijonen, H., Greenbaum, C., Pihoker, C., James, E. A.,
& Kwok, W. W. (2006). Islet-specific glucose-6-phosphatase catalytic subunit-related proteinreactive CD4+ T cells in human subjects. J Immunol, 176(5), 2781-2789.
Yang, J., James, E. A., Sanda, S., Greenbaum, C., & Kwok, W. W. (2012a). CD4+ T cells recognize diverse
epitopes within GAD65: implications for repertoire development and diabetes monitoring.
Immunology. doi: 10.1111/imm.12034
Yang, S. B., Lee, H. Y., Young, D. M., Tien, A. C., Rowson-Baldwin, A., Shu, Y. Y., Jan, Y. N., & Jan, L.
Y. (2012b). Rapamycin induces glucose intolerance in mice by reducing islet mass, insulin content,
and insulin sensitivity. J Mol Med (Berl), 90(5), 575-585. doi: 10.1007/s00109-011-0834-3
Yang, Y., Bao, M., & Yoon, J. W. (2001). Intrinsic defects in the T-cell lineage results in natural killer T-cell
deficiency and the development of diabetes in the nonobese diabetic mouse. Diabetes, 50(12), 26912699.
Yi, Z., Li, L., Garland, A., He, Q., Wang, H., Katz, J. D., Tisch, R., & Wang, B. (2012). IFN-gamma
receptor deficiency prevents diabetes induction by diabetogenic CD4+, but not CD8+, T cells. Eur J
Immunol, 42(8), 2010-2018. doi: 10.1002/eji.201142374
206
You, S., Belghith, M., Cobbold, S., Alyanakian, M. A., Gouarin, C., Barriot, S., Garcia, C., Waldmann, H.,
Bach, J. F., & Chatenoud, L. (2005). Autoimmune diabetes onset results from qualitative rather than
quantitative age-dependent changes in pathogenic T-cells. Diabetes, 54(5), 1415-1422.
Zabransky, D. J., Nirschl, C. J., Durham, N. M., Park, B. V., Ceccato, C. M., Bruno, T. C., Tam, A. J.,
Getnet, D., & Drake, C. G. (2012). Phenotypic and functional properties of Helios+ regulatory T
cells. PLoS One, 7(3), e34547. doi: 10.1371/journal.pone.0034547
Zehn, D., & Bevan, M. J. (2006). T cells with low avidity for a tissue-restricted antigen routinely evade
central and peripheral tolerance and cause autoimmunity. Immunity, 25(2), 261-270. doi:
10.1016/j.immuni.2006.06.009
Zhang, Y., Bandala-Sanchez, E., & Harrison, L. C. (2012). Revisiting regulatory T cells in type 1 diabetes.
Curr Opin Endocrinol Diabetes Obes, 19(4), 271-278. doi: 10.1097/MED.0b013e328355a2d5
Zhang, Z. J., Davidson, L., Eisenbarth, G., & Weiner, H. L. (1991). Suppression of diabetes in nonobese
diabetic mice by oral administration of porcine insulin. Proc Natl Acad Sci U S A, 88(22), 1025210256.
Zhao, Y., Scott, N. A., Quah, H. S., Krishnamurthy, B., Bond, F., Loudovaris, T., Mannering, S. I., Kay, T.
W., & Thomas, H. E. (2015). Mouse pancreatic beta cells express MHC class II and stimulate CD4 T
cells to proliferate. Eur J Immunol. doi: 10.1002/eji.201445378
Zhou, X., Bailey-Bucktrout, S. L., Jeker, L. T., Penaranda, C., Martinez-Llordella, M., Ashby, M.,
Nakayama, M., Rosenthal, W., & Bluestone, J. A. (2009). Instability of the transcription factor
Foxp3 leads to the generation of pathogenic memory T cells in vivo. Nat Immunol, 10(9), 10001007. doi: 10.1038/ni.1774
207