International Immunology, Vol. 20, No. 5, pp. 633–644
doi:10.1093/intimm/dxn023
ª The Japanese Society for Immunology. 2008. All rights reserved.
For permissions, please e-mail: [email protected]
Modulation of T cell homeostasis and alloreactivity
under continuous FTY720 exposure
Barbara Metzler1, Patrick Gfeller1, Grazyna Wieczorek1, Jianping Li1, Barbara
Nuesslein-Hildesheim1, Andreas Katopodis1, Matthias Mueller2 and Volker Brinkmann1
1
Department of Autoimmunity and Transplantation and 2Models of Disease Center, Novartis Institutes for Biomedical Research,
4002 Basel, Switzerland
Keywords: homeostatic proliferation, sphingolipid mediators, T cell memory, transplantation
Abstract
The immunomodulator FTY720 inhibits lymph node (LN) and thymic egress, thereby constraining
T cell circulation and reducing peripheral T cell numbers. Here, we analyzed in mouse models the as
yet scarcely characterized impact of long-term (up to 6 months) FTY720 exposure on T cell
homeostasis and possible consequences for alloreactivity. In green fluorescent protein (GFP)
hemopoietic chimeras, the turnover of (initially GFP2) peripheral T cell pools was markedly delayed
under FTY720, while normal homeostatic differences between CD4 and CD8 T cell sub-populations
were retained or amplified further. Homeostatic proliferation was enhanced, and within shrinking
T cell pools, the proportions of effector memory phenotype CD4 T cells (CD4TPEM) increased in
spleens and LNs and of central memory phenotype CD8 T cells (CD8TPCM) in LNs. By contrast, the
fractions of CD8TPEM and CD4TPCM remained stably small under FTY720. The enrichment for CD4TPEM
and CD8TPCM correlated with larger proportions of IFNg-producing T cells upon nonspecific but not
allospecific stimulation. Splenic CD4 T cells from FTY720-treated mice proliferated more strongly
upon transfer to semi-allogeneic hosts. However, heart allograft survival was not compromised in
FTY720 pre-treated recipients. It correlated with reduced intra-graft CD8 T cells, and the longest
surviving transplants contained the highest numbers of CD4 T cells. Thus, continuous FTY720
exposure reveals differential homeostatic responses by memory phenotype CD4 and CD8 T cell
sub-populations, and it may enhance alloreactive CD4 T cell proliferation and tissue infiltration
without accelerating allograft rejection.
Introduction
The sphingolipid mimetic FTY720 inhibits transplant rejection
and autoimmunity in animal models and was shown to be effective in phase II clinical trials of transplantation (Tx) and
multiple sclerosis (1–3). FTY720 is rapidly converted to
FTY720 phosphate in vivo and modulates lymphocyte trafficking via sphingosine-1 phosphate receptor-1 (S1P1),
expressed to similar degrees by CD4 and CD8 T cells
(4, 5). Mature single-positive (SP) thymocytes need S1P1 to
leave the thymus (4, 6). Peripheral T cells transiently downregulate S1P1 upon activation and are thus rendered refractory to sphingosine-1 phosphate (S1P)-mediated lymph node
(LN) exit signals. FTY720 induces S1P1 downregulation,
thereby impairing S1P-dependent thymus and LN egress
(4). Changes in lymphocyte trafficking by S1P1 modulators
have also been attributed to increased endothelial barrier
functions, and the relative significance of T cell autonomous
versus endothelial effects has been debated extensively
Correspondence to: B. Metzler; E-mail: [email protected]
Transmitting editor: A. Cooke
without being fully resolved (7–9). Either mechanism may
contribute to the decline in T cell numbers observed in the
blood and to a lesser extent also in the spleens of FTY720treated rodents (10, 11). The effects of FTY720 are reversible,
however, thereby requiring continuous therapeutic administration, with as yet scarcely characterized consequences for
T cell homeostasis.
In the presence of a functioning thymus, naive peripheral
T cells are continuously replaced by new thymic emigrants
(12, 13). FTY720 was shown to inhibit rather than completely
block thymic egress during a 3-week treatment course (14),
thus raising the question as to what extent a turnover of peripheral T cell pools is still possible under long-term continuous FTY720 exposure. While thymic egress inhibition would
impair peripheral naive T cells (TN) in general, the sequestration of T cells expressing the LN homing receptor CD62L, i.e.
mostly TN, might lead to an additional compartmentalized
Received 22 October 2007, accepted 8 February 2008
Advance Access publication 14 March 2008
634 T cell homeostasis and alloreactivity under FTY720
decline in TN, as shown for peripheral non-lymphoid tissues in
mice and in the blood of FTY720-treated patients (15, 16).
Moreover, shrinking peripheral T cell pools might provoke
homeostatic proliferation, thereby further promoting a concentration of phenotypic and functional memory T cells, possibly including those with auto- and allo-cross-reactive
potential (17, 18).
In this report, we analyzed in mouse models changes in
the turnover and in the naive/memory phenotype composition of peripheral CD4 and CD8 T cell pools under continuous FTY720 exposure for up to 6 months. We further
investigated the functional consequences of these changes
for memory and alloreactivity.
Materials and methods
Mice
Female 6–8-week old C57BL/6 (B6) mice were obtained from
Charles River, L’Arbresle, France; congenic B6CD45.1
(B6.SJL-Ptprca/BoAiTac) from Taconic, Ejby, Denmark and
B6C3F1 (C57BL/6 3 C3H) F1 from Charles River, Sulzfeld,
Germany. Transgenic mice on a B6 background expressing
enhanced green fluorescent protein (GFP) under the b-actin
promoter were generated and bred at Novartis, Basel, Switzerland (19). Mice were housed in conventional facilities in filtertop-protected cages with free access to water and a standard
laboratory diet, in accordance with Swiss Federal law.
In vivo reagents
FTY720 (2-amino-2-[2-(4-octylphenyl) ethylpropane-1,3-diol)
was synthesized at Novartis. FTY720 was supplied in the
drinking water at 3.3 lg ml 1 for an estimated daily dose
of 0.5 mg kg 1. FTY720 is highly stable in water at room
temperature, and FTY720-containing drinking water was
changed weekly. 5-Bromo-2-deoxyuridine (BrdU) (Sigma,
Buchs, Switzerland) was added to the drinking water at 0.8
mg ml 1 for 7 days and exchanged daily.
Hemopoietic chimerism
One day before bone marrow (BM) Tx, B6 recipient mice
were injected intraperitoneally with 30 mg kg 1 busulfan, as
described (20). BM was prepared from tibiae, femurs and
humeri of B6 GFP transgenic donors; 2.5 3 107 viable white
BM cells were injected intravenously (i.v.) per recipient. To
exclude variability in BM engraftment as a reason for differences in GFP T cell chimerism, the blood of BM recipients
was screened, by flow cytometry, for GFP monocyte/granulocyte chimerism (CD11b+) that is not affected by FTY720 and
the relative levels of which parallel the levels of lymphocyte
chimerism among BM recipients (20). Only BM recipients
with similar blood GFP monocyte/granulocyte chimerism
were compared for GFP T cell chimerism.
Flow cytometry for GFP T cell chimerism and BrdU
T cells were analyzed by four-color flow cytometry. RBCs
were lysed with CAL-Lyse (CALTAG Laboratories, Burlingame, CA, USA) or RBC lysing buffer (Sigma). Antibodies
were purchased from BD PharMingen, Basel, Switzerland.
Cells were initially incubated with an Fc receptor-blocking
antibody (anti-CD16/CD32, 2.4G2). GFP T cell chimerism
was determined by co-staining with anti-CD3–PE (145-2C11),
anti-CD4–allophycocyanin (RM4-5) and anti-CD8–Peridinin
chlorophyll protein (PerCP) (E13-161.7). T cells in the GFP
transfer experiments were analyzed with allophycocyaninconjugated anti-CD4 or anti-CD8 together with anti-CD62L–
PE (Mel-14) and biotinylated anti-CD44 (IM7) followed by
streptavidin–PerCP. BrdU detection was modified from previously published methods to fully retain GFP fluorescence intensity and surface staining (21, 22). Cells were surface
stained with allophycocyanin-conjugated anti-CD4 or antiCD8 and anti-CD44–biotin and incubated in 1% PFA/PBS,
0.05% Tween 20 overnight at 4C. Samples were incubated
in 125 Kunitz units DNase I (from bovine pancreas, Sigma)
per 250 ll DNase buffer (4.2 mM MgCl2, 0.15 M NaCl; pH 5)
for 1 h at 37C. BrdU staining was performed with antiBrdU–PE (3D4) for 30 min at room temperature in 10% FCS
and 0.5% Tween 20. Cells were finally stained with streptavidin–PerCP for 30 min at 4C. In studies without GFP chimerism, cells were also surface labeled with anti-CD62L–PE
following BrdU staining with anti-BrdU–FITC (3D4). BrdU
staining efficiency was confirmed with thymocytes; negative
reference cells were prepared from mice without BrdU treatment and stained in parallel with the same reagents. Data
were acquired on a FACS Calibur using Cellquest software
(BD PharMingen). Due to the strong T cell depletion, blood
samples from FTY720-treated mice were limited to an acquisition of ;400 gated CD4 or CD8 T cells. All data were analyzed by Flowjo software (Tree Star, Ashland, OR, USA).
Intracellular cytokine assay
Spleen and LN (pooled inguinal, axillary and brachial) suspensions were cultured in RPMI containing 10% FCS at 5 3
105 cells per well in 96-well round-bottom plates in the presence of brefeldin A and activated with 50 ng ml 1 Phorbol
12,13-dibutyrate (Sigma) and 300 ng ml 1 ionomycin
(Calbiochem, Merck Chemicals, Nottingham, UK). For allospecific stimulation, 5 3 105 cells were co-cultured with 106
irradiated (3000 rad) B6C3F1 splenocytes, following a 1-min
spin at 1200 r.p.m. to improve cell–cell contacts. Positive
controls were spleen cells from B6 mice that had rejected
a B6C3F1 skin graft 2 months before. Cells were harvested
after 6 h and surface stained with anti-CD4–allophycocyanin
and anti-CD8–PerCP or with anti-H-2K–biotin–streptavidin–
PerCP and either anti-CD4–allophycocyanin or anti-CD8–
allophycocyanin. Cells were fixed, permeabilized and stained
with anti-IFNc–PE (XMG1.2) or an isotype control using the
Cytofix/Cytoperm Kit according to the manufacturer’s
instructions (BD PharMingen).
Systemic graft versus host-driven T cell proliferation
Spleens were taken from B6 45.1 donors treated with
FTY720 for 3–4 months or from age-matched controls. RBCdepleted cell suspensions were stained with 10 lM carboxyfluorescein succinimidyl ester (CFSE) for 10 min at 37C. To
compensate for lower T cell proportions by a factor of ;0.75
in spleens from FTY720-treated mice, B6 or B6C3F1 recipients were injected i.v. with either 3 3 107 or 4 3 107 spleen
cells from untreated or FTY720-treated donor mice,
T cell homeostasis and alloreactivity under FTY720 635
respectively. Three days later, recipients were sacrificed and
single spleen cell suspensions from individual recipients
stained with anti-CD45.1–PE (A20), anti-CD3–PerCP and either anti-CD4–allophycocyanin or anti-CD8–allophycocyanin.
Heart Tx and histology
Fully vascularized heterotopic hearts from B6C3F1 donors
were transplanted into the abdominal cavity of FTY720-treated or age-matched untreated B6 recipients. Graft survival
was monitored by daily palpation (23). After rejection, heart
tissue was fixed in formalin or frozen according to standard
procedures. Three-micron-thick paraffin sections were
stained with hematoxylin and eosin and trichrome stain. All
sections were examined in a blinded manner under a light
microscope. Immunohistochemistry was performed on
3-lm-thick paraffin sections or 5-lm-thick frozen sections.
Standard procedures were applied using mAbs: rat antihuman CD3 that is cross-reactive with mouse CD3 (CD3-12,
Serotec Ltd, Oxford, UK), rat anti-mouse CD4 (H129.19, BD
PharMingen), rat anti-mouse CD8a (53-6.7, BD PharMingen),
rat anti-mouse B220/CD45R (RA3-6B2, Serotec Ltd) and the
streptavidin–biotin–peroxidase complex technique. C3d indirect immunofluorescence staining was performed on frozen
sections using polyclonal antibody rabbit anti-human C3d
(DakoCytomation Denmark A/S, Glostrup, Denmark). Negative immunohistochemical staining controls were obtained
by omitting the primary antibodies. CD4- and CD8-positive
cells were counted on five representative fields (magnification 3400) per section. Differences in the duration of heart
graft survival, the numbers of infiltrating CD4 and CD8
T cells and CD8:CD4 ratios were evaluated by Wilcoxon–
Mann–Whitney rank-sum statistics.
Detection of anti-donor alloantibodies
Allospecific IgM and IgG were detected by flow cytometry
as described before using purified T cells from C3H or
B6C3F1 mice as target cells (20).
Data modeling
Chimerism data were analyzed in Excel with XLfit software
that uses the Levenberg–Marquardt method for non-linear
regression. The Boltzmann three-parameter logistic equation
(model 600, with the minimum y-axis value locked at zero)
provided the best fit, with an F-test value close to unity, a correlation coefficient >0.99, a low chi-square value and a standard error of parameters of <10% (24).
Results
The turnover of peripheral T cell pools is markedly delayed
under continuous FTY720 and differs for CD4 and CD8 T cells
We initially compared the turnover of peripheral T cell pools
in FTY720-treated and untreated mice. To identify the progeny of newly developed T cells, syngeneic hemopoietic chimerism was induced with BM from enhanced GFP
transgenic B6 donors (19). B6 BM recipients were preconditioned with the alkylating agent busulfan that preferentially
depletes early stem cells and supports hemopoietic chimerism while avoiding peripheral lymphopenia and immune
modulation arising from irradiation-induced tissue injury
(25–27). FTY720 treatment was started 6 days after BM
transfer within a time frame of no detectable thymic and peripheral GFP+ cells. GFP T cell chimerism was monitored
during 5–6 months of continuous FTY720 treatment, with efficacy reflected by ;95–97% blood T cell depletion and a 2to 4-fold increase in SP thymocytes (Fig. 1B and data not
shown). While SP thymocytes in FTY720-treated mice
reached similar steady state GFP chimerism of ;80–90%,
with marginally delayed kinetics, peripheral GFP T cell chimerism differed markedly under FTY720 (Fig. 1Aa and b versus c–h). GFP+ T cells in the blood of FTY720-treated mice
appeared with similarly fast kinetics, yet reached lower
steady state levels (Fig. 1Ac and d). In the spleens and LNs
of FTY720-treated mice, by contrast, GFP T cell chimerism
growth was profoundly delayed but appeared to approach
steady state levels more comparable to those in control mice
(Fig. 1Ae–h). For more instructive quantitative comparisons,
the data were fitted to mathematical models (24, 28). They
were modeled accurately by the Boltzmann logistic equation
for sigmoidal growth. FTY720 treatment approximately
doubled the time required to reach inflection points (half
maximal GFP chimerism in this model), and maximal growth
rates (at the inflection point) were reduced to ;one-half to
one-third of controls (Table 1). FTY720 treatment retained or
further amplified differences in CD4 and CD8 GFP T cell chimerism that were already apparent in control mice. CD4GFP
T cell chimerism increased faster in LNs than in spleens
around the inflection point. By contrast, early CD8GFP T cell
chimerism growth (at 3 weeks) was higher in spleens than
in LNs, and this difference diminished thereafter (Table 1
and Supplementary Table 1, available at International Immunology Online, for data from a second independent experiment). Generally, peripheral T cell numbers declined under
FTY720, so that GFP chimerism developed within T cell
pools of different sizes in controls and FTY720-treated mice.
T cell numbers were most strongly reduced in the blood
by ;95–97%, followed by an ;60–70% decline in spleens
and a more moderate decrease in LN T cells (Fig. 1B, shown
for 12 weeks after GFP+ BM transfer). Similar differences in T
cell numbers were measured during 1–6 months of FTY720
treatment in GFP chimerism experiments as well as for
FTY720-treated mice that had not undergone any other experimental procedures (data not shown).
Chronic FTY720 increases homeostatic T cell proliferation
Subsequent analyses addressed possible consequences of
reduced sizes and slower turnover of peripheral T cell pools
under FTY720, in particular altered compositions of T cell
compartments and homeostatic proliferation. Indications of
homeostatic T cell proliferation under FTY720 initially came
from GFP chimerism studies showing that a fraction of
splenic CD8 T cells in FTY720-treated mice displayed lower
co-receptor expression levels (Fig. 2A). Co-receptor downregulation was more pronounced among GFP+CD8 T cells
than GFP CD8 T cells and to a lesser extent also noted for
CD4 T cells from FTY720-treated mice (Fig. 2Aa–d and data
not shown). CD8lo T cells also showed lower surface CD3
expression, they expressed high levels of the activation/
636 T cell homeostasis and alloreactivity under FTY720
A
CD8 SP thymocytes
CD4 SP thymocytes
Thymus
% GFP chimerism
100
100
a
b
80
80
60
60
40
40
20
20
0
0
0
4
8
12
16
20
24
0
4
8
CD4 T cell
16
20
24
CD8 T cell
c
80
12
80
60
60
40
40
20
20
0
0
d
Blood
e
Spleen
% GFP chimerism
80
60
60
40
40
20
20
0
0
f
80
g
80
LN
80
60
60
40
40
20
20
h
0
0
0
4
8
12
16
20
0
24
4
8
12
16
20
24
Weeks after GFP+ BM transfer
B
untreated
CD4 T cells
CD8 T cells
1.0
0.6
12
0.8
CD8 T cells x 10-6
25
CD4 T cells x 10-6
FTY720
20
0.6
15
0.4
10
0.2
5
blood
8
6
0.2
4
2
0
0
0
10
0.4
spleen
LN
0
blood
spleen
LN
Fig. 1. Delayed turnover of peripheral T cell pools under continuous FTY720 exposure. Busulfan-conditioned C57BL/6 mice received syngeneic
BM from GFP transgenic donors. They were left untreated (open circles) or given FTY720 continuously in the drinking water from day 6 after BM
transfer (closed circles). At the indicated times, thymi (a and b), blood (c and d), spleens (e and f) and LNs (g and h) from two mice per group
were analyzed, by flow cytometry, for GFP+ cells among gated CD3+CD4+ and CD3+CD8+ cells. (A) Each symbol represents an individual data
point. The smoothed lines connect the arithmetic means of the two data points for untreated (solid line) and FTY720-treated (dashed line) mice.
(B) Total (GFP+ and GFP ) numbers of CD4 or CD8 T cells in the blood (separate scale), spleens and LNs (standardized for both inguinal, axillary
and brachial per mouse) are shown for untreated (gray circles) and FTY720-treated (black circles) GFP BM recipients at the 12-week time point.
Similar data were obtained in another independent GFP BM chimerism experiment.
T cell homeostasis and alloreactivity under FTY720 637
Table 1. Differences in GFP T cell chimerism between untreated and FTY720-treated mice and between CD4 and CD8 T cell
pools
Group
Untreated
T cell type
CD4
CD8
FTY720
CD4
CD8
Location
Inflection point (% GFP+;
week after BM Tx)a
GFP+ chimerism growthb at
3 weeks
Inflection point
8 weeks
12 weeks
Spleen
LN
Spleen
LN
35;
40;
35;
35;
6.7
6.7
7.5
7.6
3.0
3.0
2.5
2.1
13.6
16.7
11.3
11.9
11.0
13.4
11.1
11.8
0.6
0.6
1.9
1.7
Spleen
LN
Spleen
LN
27;
43;
39;
37;
16.1
17.5
15.1
16.7
0.5
0.4
0.7
0.3
3.2
6.1
5.5
5.8
1.4
1.4
2.3
1.3
2.6
3.7
4.7
3.7
Data were fitted to the Boltzmann form of the logistic equation using XLfit software: f(t) = A/(1 + Exp((c t)/d)), where A is the plateau (steady
state) of GFP+ chimerism, c is the time t after GFP+ BM transfer at A/2 and d is a slope factor. Inflection points are f(ti) = A/2; ti = c.
a
Inflection points were derived from the Boltzmann equation. b Chimerism growth was calculated using its first derivative f#(t) = 1/d 3 1/A 3 f(t) 3
(A f(t)).
memory marker CD44 and they were enriched for recently
proliferated cells, as indicated by the incorporation of the
nucleotide analog BrdU during 1 week of BrdU exposure
(Fig. 2Ae and f and data not shown). Larger proportions of
recently proliferated cells were detected among both CD8
and CD4 T cells from FTY720-treated mice (Fig. 2B). We
confirmed in another model that lower T cell co-receptor expression indicated a history of homeostatic proliferation
rather than an FTY720-related artifact. CD8 and CD3 downregulations were also observed during homeostatic proliferation of donor BALB/c T cells within SCID recipient mice on
a BALB/c background. By contrast, proliferating alloreactive
B6 T cells in BALB/c SCID hosts kept high CD8 expression
levels (Supplementary Figure 1, available at International Immunology Online).
Differential changes in the composition of CD4 and CD8 naive
and memory phenotype T cell pools
The combined effects of thymic egress inhibition and increased homeostatic proliferation would be conducive to
smaller T cell pools that are more concentrated in memory
phenotype T cells (TPM). In the GFP T cell chimerism studies,
CD4 T cells became highly concentrated in effector TPM
(TPEM, CD62LloCD44hi) in the spleens and LN, whereas CD8
T cells contained enlarged proportions of central TPM (TPCM,
(CD62LhiCD44hi) in LN but not in spleens. These differences
were already apparent among GFP-negative (i.e. mostly
‘older’) T cells from control animals, suggesting that FTY720
exposure further enhanced physiological—rather than implementing new—differences in the homeostasis of CD4 and
CD8 T cell sub-populations (Supplementary Figure 2, available at International Immunology Online). It was important,
however, to exclude the possibility of artifacts related to busulfan treatment and GFP BM transfer. Therefore, the composition of T cell pools was analyzed for age-matched
untreated and FTY720-treated but otherwise unmanipulated
mice based on CD62L and CD44 expression (Fig. 3, histogram insets). The decline in total T cell numbers after 3
months on FTY720 was comparable to previous experiments
(e.g. Fig. 1B), reducing T cell numbers to ;3, 30 and 60% of
normal values in the blood, spleen and LN, respectively.
FTY720 exposure preferentially depleted CD4 and CD8 TN
from the blood, resulting in an over-representation of CD4
and CD8 TPEM and of CD8TPCM among the remaining blood
T cells (Fig. 3a and b). Due to the strong overall blood T cell
depletion, however, the numbers of TPM also decreased,
e.g. CD4 T cells CD4TPEM from 7% of 100% numbers = 7 to
65% of 3% numbers = 2. In spleens, the composition of the
CD8 T cell pool did not change with respect to CD62L/
CD44 phenotyping (Fig. 3d). By contrast, TN were preferentially lost from the splenic CD4 T cell pool, thereby further
enriching CD4TPEM from ;20% in controls to ;50% in
FTY720-treated mice (Fig. 3c). Thus, within the degree of
data variability, splenic CD4TPEM numbers remained relatively stable or declined moderately (20% of 100% numbers
versus 50% of 30% numbers = 20 versus 15). In LN, TPM
displayed a strong dichotomy between an exclusive proportional rise in TPEM among CD4 T cells and of TPCM among
CD8 T cells (Fig. 3e and f). Given a more variable and
smaller drop in LN T cell numbers, these proportional
increases, on average, resulted in stable or moderately elevated numbers of LN CD4TPEM and a reproducible increase
in LN CD8TPCM (6% of 100% numbers versus 15% of 60%
numbers = 6 versus 9; and 6% of 100% numbers versus
20% of 60% numbers = 6 versus 12). It seemed noteworthy
that CD8TPEM were not enriched in either spleen or LN and
that the small proportions of CD4TPCM did not increase in
any compartment tested (including the BM, not depicted).
An analysis of BrdU uptake by individual T cell subsets
showed that physiological and FTY720-enhanced homeostatic proliferation was largely confined to TPM, and it was
particularly prominent for CD4TPCM from the blood and BM.
Homeostatic proliferation was not increased for splenic
CD4TPEM and LN CD8TPCM, indicating that their strong local
enrichment did not result from enhanced proliferation at
those sites (data not shown).
Chronic FTY720 exposure increases the proportions of
IFNc-producing T cells
We next tested whether the enrichment of CD4TPEM in
spleens and LN and of CD8TPCM in LNs of FTY720-treated
mice correlated with proportional increases in functional
638 T cell homeostasis and alloreactivity under FTY720
A
FTY720
Untreated
a
No. cells
No. cells
GFP+
b
CD8lo
100
101
103
102
104
c
101
102
103
104
101
102
103
104
No. cells
d
No. cells
GFP-
100
CD8lo
100
101
103
102
104
100
e
f
% of Max
CD8
% of Max
CD8
100
101
103
102
104
33
10
100
101
CD44
gated CD8
gated CD8hi
lo
gated CD8lo
isotype control
no BrdU
CD4 T cells
CD8 T cells
a
b
% of Max
GFP+
Enhanced proliferation by splenic CD4 T cells from
FTY720-treated donors within semi-allogeneic recipients
34
19
4
1
BrdU
BrdU
c
d
% of Max
GFP-
104
BrdU
gated CD8hi
B
103
102
20
8
7
0
10
1
10
2
10
2
103
memory T cells. Splenic and LN cells from untreated and
3-month FTY720-treated mice were analyzed, by flow cytometry, for intracellular IFNc after stimulation with phorbol ester
and ionomycin for 6 h (Fig. 4). During this time, only preexisting memory T cells are capable of producing IFNc.
CD4 T cells in LNs and spleens of FTY720-treated mice contained 2–3 times larger proportions of IFNc+ T cells than the
same lymphoid organs from age-matched untreated mice.
CD8 T cells generally contained larger fractions of IFNc+
cells than CD4 T cells, and these were increased further in
LNs (; doubled) but not consistently in spleens of FTY720treated mice. In addition to overall functional T cell memory
as revealed by nonspecific activation, we also screened for
allo-cross-reactive memory T cell to facilitate the interpretation of subsequent Tx experiments. Allo-cross-reactive memory T cells may form upon activation with infectious agents.
Although our animals were free from infections, they were
housed in conventional facilities, i.e. not germ free or specific pathogen free and rather old at the time of many experiments. A small possibility remained, therefore, that under
long-term FTY720 exposure, allo-cross-reactive memory
T cells might have arisen through homeostatic proliferation
and/or had been enriched under continuous FTY720 treatment. However, no IFNc+ T cells above background were
detected upon stimulation with irradiated B6C3F1 spleen
cells (Fig. 4a–d). High IFNc expression in response to
B6C3F1 stimulators was only detected among T cells from
B6 mice that had rejected a B6C3F1 skin graft 2 months
before and that served as positive controls for this assay
(Fig. 4, histogram inset).
4
10
0
10
1
10
102
103
104
BrdU
BrdU
untreated + BrdU
FTY720 + BrdU
untreated, no BrdU
Fig. 2. Continuous FTY720 treatment induces homeostatic proliferation. Three months after the transfer of GFP transgenic syngeneic
donor BM, as described in Fig. 1, mice received BrdU in their drinking
water for 1 week. (A) CD8 expression by splenic T cells from FTY720treated (a and c) or untreated mice (b and d) gated on GFP+ (a and b)
and GFP cells (c and d). CD8hi and CD8lo T cells were analyzed for
co-expression of CD44 (e) and BrdU (f); plots (e) and (f) show splenic
CD8 T cells from a FTY720-treated mouse. (B) BrdU staining of splenic
CD4 and CD8 T cells among GFP+ (a and b) and GFP (c and d)
gated cells. Numbers within histograms indicate percentage of
BrdU+ cells.
Because FTY720 does not inhibit T cell egress from the
spleen, splenic memory T cells might still respond rapidly to
allografts other than through immediate effector functions,
such as through enhanced alloreactive proliferation. On the
other hand, within FTY720-treated hosts themselves, lower
T cell numbers might outweigh qualitative changes of T cell
sub-populations that could jeopardize allograft integrity. To
test for in vivo alloreactive proliferation independently of differences in T cell numbers, equivalent numbers of CFSElabeled splenic T cells from 3- to 4-month FTY720-treated or
untreated B6CD45.1 donors were transferred to (untreated)
congenic B6 or B6C3F1 recipients. Three days later, donorderived CD45.1+ CD4 and CD8 T cells from recipient
spleens were analyzed for their proliferation patterns, as indicated by the 50% loss of CFSE intensity with every cell division (Fig. 5). In congenic B6 hosts, both donor-derived
CD4 and CD8 T cells retained maximal CFSE fluorescence
intensity, regardless as to whether they originated from untreated or FTY720-treated donors. Upon transfer to B6C3F1
recipients, about 40–45% of CD4 and CD8 T cells from untreated donors as well as CD8 T cells from FTY720-treated
donor spleens displayed similar CFSE staining patterns consistent with 1–7 cell divisions. In marked contrast, CD4
T cells from FTY720-treated donors proliferated more vigorously, with >70% cycling cells and with the majority having
undergone high numbers (5–7) of divisions.
T cell homeostasis and alloreactivity under FTY720 639
% of CD4T cells
a
TPCM
TPEM
c
TPCM
TN
TPEM
e
80
80
60
60
60
40
40
80
*
*
TPCM
TN
TPEM
Isotype
*
low
40
high
*
*
20
20
20
0
0
0
*
100
b
% of CD8T cells
LN
Spleen
Blood
TN
d
80
*
40
20
0
*
*
102
103
104
CD62L
f
80
80
60
101
60
60
40
40
20
20
0
0
Low
*
high
untreated
FTY720
*
100
101
102
103
104
CD44
isotype control
gated CD62Llo
gatedCD62Lhi
Fig. 3. Differential increases in the proportions of TPM subsets. Blood, spleens and LN cells were collected from untreated (gray bars) and
3-month FTY720-treated (black bars) mice. CD4 and CD8 T cells were phenotyped according to their combined CD62L/CD44 expression pattern
(histogram insets). The marker settings delineate three populations, CD62LhiCD44lo, TN; CD62LhiCD44hi, TPCM and CD62LloCD44hi, TPEM. The
proportions of the indicated T cell sub-populations are expressed with respect to 100% CD4 T cells (upper row) and 100% CD8 T cells (bottom
row). Bar graphs and error bars are means and standard deviations of data collected from three mice per group. Asterisks indicate statistically
significant differences (by Wilcoxon–Mann–Whitney rank-sum statistics) between data from untreated and FTY720-treated mice. The data are
representative of three independent experiments.
Continuous FTY720 treatment prolongs heart allograft survival
The enhanced capacity of splenic CD4 T cells from FTY720
pre-treated mice for alloreactive proliferation raised questions regarding their impact on the survival of a solid organ
allograft within FTY720 pre-treated recipients themselves.
This was tested with FTY720-treated B6 recipients of fully
vascularized B6C3F1 heart grafts. When 3- to 4-month pretreatment with FTY720 was stopped 1 week before heart Tx,
allograft rejection was moderately delayed by 4 days compared with recipients that had never received FTY720 (days
of rejection 14–17 versus 8–12 after Tx), suggesting that
modifications in the composition and function of T cell pools
per se did not accelerate allograft rejection, and underpinning the need for continuous FTY720 treatment for effective
transplant protection. For recipients on continuous FTY720
after Tx, allograft survival was compared for two groups;
one that had been pre-treated with FTY720 for 3–4 months
and a second group put on FTY720 only 1 week prior to
heart Tx. T cells in the former recipient group had therefore
undergone an extended period of enhanced homeostatic
proliferation and had reached maximal TPM enrichment in all
locations. One week of FTY720 treatment, by contrast, only
achieved minimal T cell numbers and maximal TPM skewing
in the blood but not yet in spleens and LNs (data not
shown). Recipients from both pre-treatment groups showed
significantly delayed allograft rejection, with a tendency for
longer graft survival with the 1-week pre-treatment protocol
(days 17–87 and 35–114 versus days 8–12, Table 2). The
histopathology of rejected heart grafts was similar for both
groups under continuous FTY720 and differed markedly
from untreated mice. Untreated recipients lost their heart
grafts through severe acute interstitial cellular rejection with
endothelialitis obliterating the arterial lumen. By contrast,
rejected heart allografts from recipients on continuous
FTY720 showed various grades of mild to moderate–severe
acute interstitial cellular rejection in combination with fibrinoid necrosis of arteries and mild intimal hyperplasia
(Fig. 6). Mild chronic rejection was only observed in the two
longest surviving heart allografts. The main overall mechanism of rejection under FTY720 in this model, therefore, was
delayed cellular rather than humoral or chronic rejection. Although allospecific IgM and IgG antibodies were not reduced in the blood of FTY720-treated recipients, there were
no differences in the intensity or distribution of C3d deposits
in untreated or FTY720-treated mice, and infiltrating B cells
(B220+) were extremely rare in all groups (data not shown).
T cell infiltrates in rejected allografts contained more CD8+
than CD4+ cells, albeit with clear differences between
groups (Table 2). With 3- to 4-month FTY720 pre-treatment,
both CD8 and CD4 T cell infiltrates tended to be smaller. In
marked contrast, 1-week FTY720 pre-treatment resulted in
lower numbers of infiltrating CD8 T cells, whereas CD4 T cell
numbers were higher, thereby resulting in the lowest
CD8:CD4 T cell ratios within the longest surviving allografts.
Discussion
In this report, we analyzed in mouse models the impact of
long-term continuous FTY720 exposure on T cell homeostasis and properties pertaining to T cell memory and alloreactivity. In initial studies, syngeneic GFP T cell chimerism
indicated the replacement of peripheral T cells by thymic
640 T cell homeostasis and alloreactivity under FTY720
a
b
LN, CD4 T cells
16
CD4
spleen, CD4 T cells
20
CD8
untreated
FTY720
Untreated donor
B6 (CD45.1)
to B6
16
12
12
8
8
4
0
0
PdBu+
Iono
c
none
allo
PdBu+
Iono
allo
spleen, CD8 T cells
d
LN, CD8 T cells
none
FTY720-treated
donor B6 (CD45.1)
to B6
20
16
16
12
12
untreated donor
B6 (CD45.1)
to B6C3F1
8
8
4
Relative cell number
4
45 %
42 %
4
0
0
PdBu+
Iono
none
allo
PdBu+
Iono
none
allo
Stimulation
CD8 T cells
CD4 T cells
Isotype ctrl
3%
FTY720-treated
donor B6 (CD45.1)
to B6C3F1
72 %
41 %
100 101 102 103 104 100 101 102 103 104
CFSE
10
0
10
1
10
2
10
3
4
10
Fig. 4. Continuous FTY720 treatment increases the proportions of
IFNc-producing T cells. LN (a and c) and spleen cells (b and d) from
3-month untreated (gray circles) or FTY720-treated (black circles) B6
mice were activated for 6 h in vitro with PdBu and ionomycin (Iono) or
with irradiated B6C3F1 ‘allo’ spleen cells in the presence of brefeldin
A. Cells were surface stained for CD4 or CD8 and from B6C3F1 cocultures also for H-2Kk. Fixed and permeabilized cells were stained
for intracellular IFNc and analyzed by flow cytometry. Controls for alloantigen-driven IFNc production were spleen cells from a B6 mouse
that had rejected a B6C3F1 skin graft 2 months previously (histogram
inset). Symbols represent data from two individual mice per group.
The data are representative of three independent experiments.
emigrants and their progeny. It followed sigmoidal growth,
characteristic of competitive population dynamics, the rules
of which are readily applicable to the competition among
lymphocytes for limited space and resources (28, 29). TN, although long lived in the absence of a thymus, were shown to
be replaced rapidly in euthymic mice and with kinetics similar to normal GFP T cell chimerism in our model (30, 31).
Chronic FTY720 treatment placed challenges on T cell homeostasis, as TN were strongly diminished in the blood, with
little replacement by mature thymocytes. This resulted in
lower blood steady state levels of GFP+ T cells and their
delayed kinetics in spleens and LNs. The kinetics of CD4
and CD8 T cell GFP chimerism differed in spleens and LNs
of both untreated and FTY720-treated mice. Because TN
Fig. 5. Proliferation of T cells from FTY720-treated donor spleens
upon transfer to semi-allogeneic hosts. CFSE-stained splenocytes
prepared from 3- to 4-month FTY720-treated or untreated B6CD45.1
donors were injected i.v. into B6 or B6C3F1 recipients. For the transfer
of equivalent numbers of T cells, recipients were injected with either
3 3 107 spleen cells from untreated or 4 3 107 spleen cells from
FTY720-treated donors. Three days later, recipient spleens were
analyzed for proliferation by CFSE fluorescence intensity of donorderived CD45.1+ CD4 and CD8 T cells. Histograms are representative
of data from two syngeneic and three semi-allogeneic recipients per
group. Markers indicate the percentage of T cells with loss of CFSE
staining intensity. Similar data were obtained in another independent
experiment.
and memory T cells are under separate homeostatic control
(32, 33), the contribution of thymic emigrants to the turnover
of peripheral T cell pools is determined by the proportions
of peripheral TN. Thus, the faster replacement of CD4 T cells
in LNs by (initially naive) thymic descendants was presumably favored by larger proportions of CD4TN in LNs over
spleens. In addition, it may also reflect that CD4TN but not
CD8TN absolutely depend on LNs for their survival (34).
CD8TN were more similarly distributed in spleens and LNs,
although slightly higher proportions of splenic CD8TN might
have contributed to initially faster CD8GFP chimerism in
spleens. CD8 T cell chimerism within LNs presumably
‘caught up’ through growing contributions from CD8TPCM,
i.e. a memory phenotype including the LN homing receptor
CD62L. In the light of differential CD4 and CD8 T cell subset
compartmentalization and declining total T cell numbers
T cell homeostasis and alloreactivity under FTY720 641
Table 2. Prolonged heart allograft survival and altered T cell
infiltrates in FTY720-treated mice
Treatment of
heart
allograft
recipients
CD8+
CD4+
CD8:CD4
Day of
cell
ratiod
rejectiona cell
numbersb numbersc
Untreated, n = 4
8
11
11
12
1285
458
614
704
157
89
78
82
8.2
5.1
7.9
8.6
FTY720, from 14-week
pre-Tx, n = 4
17
45
56
87
357
289
266
839
41
56
87
80
8.7
5.2
3.1
10.5
FTY720, from 1-week
pre-Tx, n = 4
35
91
92
114
318
598
450
280
87
224
215
160
3.7
2.7
2.1
1.8
B6 recipients of B6C3F1 heart allografts were untreated or kept on
FTY720 following Tx and a 1- or 14-week pre-treatment period with
FTY720 before Tx. Data within each row refer to the same recipient
and graft tissue. P-values are derived from Wilcoxon-Mann-Whitney
rank-sum statistics.
a
P = 0.014, untreated versus FTY720 14-week pre-Tx and untreated
versus FTY720 1-week pre-Tx. bCell numbers are cumulative of
five representative fields per section stained for CD8 by
immunohistochemistry. P = 0.029, untreated versus FTY720 1-week
pre-Tx. cCell numbers are cumulative of five representative fields per
section stained for CD4 by immunohistochemistry. P = 0.057, FTY720
1-week pre-Tx versus untreated. dP = 0.014, untreated versus FTY720
1-week pre-Tx; P = 0.029, FTY720 1-week versus 14-week pre-Tx.
Fig. 6. Histopathology of rejected heart allografts from untreated and
FTY720-treated recipients. B6 recipients of B6C3F1 heart allografts
were untreated or on FTY720 for 3–4 months before heart Tx, with
FTY720 treatment continued thereafter. On the day of rejection,
grafted hearts were fixed in formalin and 3lm sections stained with
hematoxylin and eosin (H&E, original magnification 3200) or
trichrome (original magnification 3400).
under FTY720, it seemed possible that altered survival rates
as a consequence of changing competition for survival factors also influenced chimerism growth rates. We measured
similarly and moderately increased cell sizes (by forward
scatter, not depicted) for CD4 and CD8 GFP+ and GFP TN
from FTY720-treated mice. In a previous report, an increase
in cell size by non-proliferating CD8TN cultured in the presence of IL15 was taken to reflect the pro-survival effect of
this cytokine (35). Although CD4TN do not utilize IL15 in addition to IL7 for their survival, increased sizes of non-proliferating TN might generally reflect an improved availability of
appropriate survival factors. Furthermore, T cells from
FTY720-treated mice had a slightly higher expression of the
anti-apoptotic factor Bcl-2, which is up-regulated by IL7 and
IL15 (36, 37). Thus, these observations might indicate a modest generally improved T cell survival capacity under
FTY720, yet without favoring any of the CD4 or CD8 GFP+
or GFP T cell subsets, as would be required for a considerable impact on chimerism growth rates.
In addition to delaying the turnover of peripheral T cell
pools, continuous FTY720 exposure led to a decline in peripheral T cell numbers, thereby provoking an increase in homeostatic proliferation. This was associated with co-receptor
and CD3 down-regulation, which was most pronounced
among GFP+CD8 T cells. The reasons for stronger proliferation among GFP+ T cells remain speculative. A previous
study excluded different intrinsic proliferative capacities between wild-type and similar (actin promoter-driven) GFP
transgenic T cells (38). Possibly, ‘younger’ T cells, as represented by GFP+ T cells, proliferate more readily than ‘old’
peripheral T cells in response to partial lymphopenia. Coreceptor down-regulation was generally observed for a fraction of T cells from FTY720-treated mice in all experiments
and was not, therefore, peculiar to GFP chimerism studies.
Nor was it directly induced by FTY720, as it was also demonstrated for T cells proliferating in syngeneic SCID hosts,
whereas CD8 appeared stably expressed during the proliferation of allo-mismatched T cells in the same SCID transfer
model. Regular health reports ruled out infections as a reason for increased proliferation. TCR down-regulation during
homeostatic proliferation was described previously (39), and
down-regulation of CD8 and the TCR by T cells encountering
their cognate antigen in the periphery was identified as
a mechanism of peripheral tolerance (40). Given the (necessary and sufficient) role of self-peptide–MHC in driving
homeostatic proliferation (41, 42), co-receptor/CD3 downregulation may reflect one strategy to preserve peripheral
tolerance to self-antigens.
Although homeostatic proliferation was shown to promote
transplant rejection and autoimmunity under some conditions (17, 18), continuous FTY720 application resulted in prolonged rather than accelerated allograft survival, even when
discontinued before Tx. Furthermore, tissues and organs
from mice continuously treated with FTY720 for up to 2 years
bore no signs of autoimmunity (our unpublished observations). Several reasons might account for these apparent
discrepancies, such as different degrees of T cell depletion
and homeostatic proliferation as well as the clonality/specificity of the proliferating T cells. A limited T cell repertoire
with strong skewing toward allo-cross-reactive or self-reactive pathogenic T cells might arise through the expansion of
small T cell numbers within an initially T-deplete or near-deplete space. Unlike TCR transgenic allospecific T cells that
caused enhanced allograft rejection following homeostatic
proliferation within T-deplete recipients (18), the degrees of
642 T cell homeostasis and alloreactivity under FTY720
T cell depletion and homeostatic proliferation under FTY720
were modest, by comparison, and the T cell repertoire of
FTY720-treated mice was not intentionally skewed. Nor is
the wild-type B6 strain used in our experiments susceptible
to autoimmunity, unlike the non-obese diabetic mouse model
where diabetes was promoted by homeostatic proliferation
(17). In other studies, strong T cell depletion of 80–90% triggered high degrees of homeostatic proliferation that interfered with the induction of Tx tolerance by co-stimulation
blockade (43). It remains to be shown weather tolerance induction would be impaired by a more moderate reduction of
T cell numbers and less vigorous homeostatic proliferation.
The highest degrees of homeostatic proliferation were
found among CD4TPCM, resulting in ;80 and 95% BrdU+
CD4TPCM in the blood and BM of FTY720-treated mice. Nevertheless, their contribution to the whole CD4 T cell pool
never exceeded 4–5% at any site in controls or under longterm FTY720 exposure. The high death rate that must be
inferred for CD4TPCM may be supported by their very low
levels of Bcl-2 that were lowest among all T cell subsets
(our unpublished data). CD8TPEM also retained similarly
small proportions in the spleens and LNs of controls and
FTY720-treated mice. As total T cell numbers declined under
FTY720, the absolute numbers of CD8TPEM and CD4TPCM also
decreased accordingly, possibly indicating that maintaining
stable small proportions of CD8TPEM, and in particular of
CD4TPCM may be critical to the integrity of the immune system.
The strongest FTY720-induced modulation of local T cell
pool compositions (other than in the blood) were a further
concentration of CD4TPEM in spleens and LNs and of
CD8TPCM within LNs. Homeostatic proliferation of CD4TPEM
and CD8TPCM was also increased under FTY720, albeit not
at those sites of their predominant enrichment. Factors that
contributed substantially to the concentration of CD4TPEM
presumably included the preferential loss of CD4TN from
spleens and LNs. The large and selective proportional increase of CD8TPCM within LNs and not spleens translated into reduced total numbers of splenic CD8TPCM and higher
numbers of CD8TPCM within LNs, suggesting that their migration to and sequestration within LNs of FTY720-treated
mice played a role. Although the proportions of CD4TPEM
and CD8TPCM increased, their overall numbers did not, as
implemented by the separate homeostatic control of TN and
memory T cells that prevents memory T cells from occupying
the void left by contracting TN pools (32, 33). In addition to
FTY720-induced T cell redistribution, it remains formally possible that phenotype conversion might have contributed to
changes in the compositions of TPCM and TPEM. Insight into
phenotype conversion in mice largely comes from studies in
TCR transgenic models, and prominent proportional
changes of effector memory T cells (TEM) and central memory T cells (TCM) through phenotype conversion appeared to
be favored by non-physiologically high T responder frequencies in vivo and in vitro (44, 45). Other reports demonstrated
various degrees of phenotype flexibility to accommodate
T cell migration and location patterns (46, 47). The tendencies
towards more prominent CD4TPEM and CD8TPCM among
‘naturally occurring’ memory T cells in this and previous
studies might imply, however, that endogenous TPEM/TPCM
patterns are formed during initial rounds of homeostatic pro-
liferation, i.e. during the conversion from a naive to an either
central or effector memory phenotype rather than interconversion between memory phenotypes (47–49). Furthermore, as indicated by the GFP-negative T cell pool in the
chimerism studies, these CD4TPEM/CD8TPCM biases become
more prominent as the proportions of TN decline with the
age of the T cell pool. Thus, given the preferential loss of TN
and increased homeostatic proliferation under FTY720, the
resulting proportional increases in CD4TPEM and CD8TPCM
presumably further accentuate physiological tendencies
rather than reflect memory T cell conversion. Enlarged proportions of CD4TPEM and CD8TPCM in our model correlated
with higher percentages of rapidly IFNc-producing CD4 T
cell from spleens and LNs and CD8 T cells from LNs of
FTY720-treated mice. Although ‘central’ and ‘effector’ memory phenotypes define functionally distinct human memory T
cells, with the former lacking immediate effector functions
(50), this does not apply to murine CD8TCM and CD8TPCM
that were shown capable of mediating effector functions
(51, 52). IFNc-producing T cells were only obtained upon
nonspecific activation and not in the presence of semiallogeneic (B6C3F1) cells. This was expected since allocross-reactive IFNc-producing memory T cells induced by
infectious agents, often viruses, are rare in ‘clean’ mice but
may be a serious obstacle in clinical Tx (53). Surprisingly,
however, CD4 T cells from FTY720-treated donor spleens
proliferated more vigorously within B6C3F1 hosts than CD4
T cells from untreated controls. This might be accounted for
by the larger content of CD4TPEM among donor spleen cells
from FTY720-treated mice and/or the strong proliferative capacity of CD4TPCM. In either case, these data might suggest
that all or certain subsets of CD4TPM, i.e. cells with a history
of self-MHC-driven homeostatic proliferation, can gain functional memory for allo-cross-reactive proliferation. Nevertheless, despite an enhanced alloreactive proliferation potential
of splenic T cells, i.e. those not ‘trappable’ by FTY720, heart
allograft rejection was strongly delayed. Cellular graft infiltrates were dominated by CD8 T cells, and prolonged allograft survival correlated with lower numbers of intra-graft
CD8 T cells rather than CD4 T cells. Presumably, FTY720-induced LN sequestration of CD8TPCM contributed to allograft
protection, consistent with previous studies demonstrating
that murine allospecific CD8TCM were far more potent mediators of allograft rejection than CD8TEM (54). The tendency
of the longest surviving allografts to contain higher numbers
of CD4 T cells might reflect an intrinsically weaker productive
alloreactive potential at the level of individual intra-graft CD4
T cells, thereby requiring larger numbers of them to accumulate for effective allograft rejection. Because FTY720 does
not inhibit T cell effector functions directly (55), this would
imply that the short FTY720 pre-treatment protocol promoted
CD4 T cell infiltrates with inherently weaker alloreactivity at
the single cell (rather than the population) level. There is no
precedence, however, to support this possibility. Alternatively, the enlarged CD4 T cell infiltrates, even though not capable of mediating indefinite allograft survival, included an
active protective component, as might be achieved with
contributions from regulator T cells (Tregs). Because alloantibodies were not reduced in FTY720-treated mice, putative
Tregs more likely acted within the allograft rather than
T cell homeostasis and alloreactivity under FTY720 643
lymphoid tissue. Transplant protection in the presence of
Tregs and alloantibodies was reported previously, as was,
within allografts, the accumulation of Tregs that modulate the
effector rather than the induction phase of cellular immunity
(56–58). The properties of graft-infiltrating T cells in this
model are currently under investigation. It also remains to be
clarified why short and not long pre-treatment with FTY720
favored larger allograft CD4 T cell infiltrates. This could result
from reduced total numbers of a potential Treg type or Treg
‘precursor’, as would apply to CD4TN, CD4TPCM or a subpopulation thereof. The properties of CD4TPCM overlap with
those of CD62LhiCD25+CD4 Tregs, shown to contain cells
that cycle rapidly in response to self-antigen (59).
In conclusion, enhanced homeostatic proliferation and
skewing toward phenotypic and functional memory T cells
under continuous FTY720 exposed homeostatic differences
between CD4 and CD8 T cell subsets but did not promote
allograft rejection. Ongoing studies investigate whether homeostatic tuning under FTY720 may be harnessed rationally
to promote Treg tissue infiltrates for therapeutic purposes.
3
4
5
6
7
8
9
10
Supplementary data
Supplementary Figures 1 and 2 and Table 1 are available at
International Immunology Online.
11
Disclosures
All authors are employees of Novartis, Basel, Switzerland.
12
Acknowledgements
We thank Ed Palmer and Anna Huttenlocher for their critical reading
and discussions of the manuscript and Martin Beibel for reviewing the
data modeling sections. We also thank Sonia Deveze, Peter Gugger,
Antje Marcantonio and Marinette Erard for expert technical support.
Abbreviations
BM
BrdU
CD8TPCM
CD4TPEM
CFSE
GFP
i.v.
LN
PdBu
PerCP
S1P
SP
S1P1
TCM
TEM
TN
TPM
Treg
Tx
bone marrow
5-bromo-2-deoxyuridine
central memory phenotype CD8 T cells
effector memory phenotype CD4 T cells
carboxyfluorescein succinimidyl ester
green fluorescent protein
intravenously
lymph node
phorbol 12,13-dibutyrate
peridinin chlorophyll protein
sphingosine-1 phosphate
single positive
sphingosine-1 phosphate receptor-1
central memory T cells
effector memory T cells
naive T cells
memory phenotype T cells
regulator T cell
transplantation
References
1 Brinkmann, V. and Lynch, K. R. 2002. FTY720: targeting G-proteincoupled receptors for sphingosine 1-phosphate in transplantation
and autoimmunity. Curr. Opin. Immunol. 14:569.
2 Tedesco-Silva, H., Mourad, G., Kahan, B. D. et al. 2005. FTY720,
a novel immunomodulator: efficacy and safety results from the first
13
14
15
16
17
18
19
20
21
22
phase 2A study in de novo renal transplantation. Transplantation
79:1553.
Kappos, L., Antel, J., Comi, G. et al. 2006. Oral fingolimod (FTY720)
for relapsing multiple sclerosis. N. Engl. J. Med. 355:1124.
Matloubian, M., Lo, C. G., Cinamon, G. et al. 2004. Lymphocyte
egress from thymus and peripheral lymphoid organs is dependent
on S1P receptor 1. Nature 427:355.
Graeler, M. and Goetzl, E. J. 2002. Activation-regulated expression and chemotactic function of sphingosine 1-phosphate
receptors in mouse splenic T cells. FASEB J. 16:1874.
Allende, M. L., Dreier, J. L., Mandala, S. and Proia, R. L. 2004.
Expression of the sphingosine-1-phosphate receptor, S1P1, on
T-cells controls thymic emigration. J. Biol. Chem. 279:15396.
Sanchez, T., Estrada-Hernandez, T., Paik, J. H. et al. 2003.
Phosphorylation and action of the immunomodulator FTY720
inhibits vascular endothelial cell growth factor-induced vascular
permeability. J. Biol. Chem. 278:47281.
Wei, S. H., Rosen, H., Matheu, M. P. et al. 2005. Sphingosine
1-phosphate type 1 receptor agonism inhibits transendothelial
migration of medullary T cells to lymphatic sinuses. Nat. Immunol.
6:1228.
Rosen, H. and Goetzl, E. J. 2005. Sphingosine 1-phosphate and
its receptors: an autocrine and paracrine network. Nat. Rev.
Immunol. 5:560.
Chiba, K., Yanagawa, Y., Masubuchi, Y. et al. 1998. FTY720,
a novel immunosuppressant, induces sequestration of circulating
mature lymphocytes by acceleration of lymphocyte homing in
rats. I. FTY720 selectively decreases the number of circulating
mature lymphocytes by acceleration of lymphocyte homing.
J. Immunol. 160:5037.
Morris, M. A., Gibb, D. R., Picard, F., Brinkmann, V., Straume, M.
and Ley, K. 2005. Transient T cell accumulation in lymph nodes
and sustained lymphopenia in mice treated with FTY720. Eur. J.
Immunol. 35:3570.
Tanchot, C., Le Campion, A., Martin, B., Leaument, S., Dautigny,
N. and Lucas, B. 2002. Conversion of naive T cells to a memorylike phenotype in lymphopenic hosts is not related to a homeostatic mechanism that fills the peripheral naive T cell pool.
J. Immunol. 168:5042.
Ge, Q., Hu, H., Eisen, H. N. and Chen, J. 2002. Different
contributions of thymopoiesis and homeostasis-driven proliferation to the reconstitution of naive and memory T cell compartments. Proc. Natl Acad. Sci. USA. 99:2989.
Yagi, H., Kamba, R., Chiba, K. et al. 2000. Immunosuppressant
FTY720 inhibits thymocyte emigration. Eur. J. Immunol 30:1435.
Hofmann, M., Brinkmann, V. and Zerwes, H. G. 2006. FTY720
preferentially depletes naive T cells from peripheral and lymphoid
organs. Int. Immunopharmacol. 6:1902.
Bohler, T., Waiser, J., Schutz, M., Dragun, D., Neumayer, H. H. and
Budde, K. 2004. FTY720 mediates apoptosis-independent lymphopenia in human renal allograft recipients: different effects on
CD62L+ and CCR5+ T lymphocytes. Transplantation 77:1424.
King, C., Ilic, A., Koelsch, K. and Sarvetnick, N. 2004. Homeostatic expansion of T cells during immune insufficiency generates
autoimmunity. Cell. 117:265.
Maile, R., Barnes, C. M., Nielsen, A. I., Meyer, A. A., Frelinger, J. A.
and Cairns, B. A. 2006. Lymphopenia-induced homeostatic
proliferation of CD8+ T cells is a mechanism for effective allogeneic
skin graft rejection following burn injury. J. Immunol. 176:6717.
Shmerling, D., Danzer, C. P., Mao, X. et al. 2005. Strong and
ubiquitous expression of transgenes targeted into the beta-actin
locus by Cre/lox cassette replacement. Genesis. 42:229.
Metzler, B., Gfeller, P., Bigaud, M. et al. 2004. Combinations of
anti-LFA-1, everolimus, anti-CD40 ligand, and allogeneic bone
marrow induce central transplantation tolerance through hemopoietic chimerism, including protection from chronic heart
allograft rejection. J. Immunol. 173:7025.
Tough, D. F. and Sprent, J. 1994. Turnover of naive and memoryphenotype T cells. J. Exp. Med. 179:1127.
Carayon, P. and Bord, A. 1992. Identification of DNA-replicating
lymphocyte subsets using a new method to label the bromodeoxyuridine incorporated into the DNA. J. Immunol. Methods
147:225.
644 T cell homeostasis and alloreactivity under FTY720
23 Corry, R. J., Winn, H. J. and Russell, P. S. 1973. Primarily
vascularized allografts of hearts in mice. The role of H-2D, H-2K,
and non-H-2 antigens in rejection. Transplantation 16:343.
24 Reich, J. 1992. Curve Fitting and Modeling for Scientists and
Engineers. McGraw-Hill, USA.
25 Botnick, L. E., Hannon, E. C., Vigneulle, R. and Hellman, S. 1981.
Differential effects of cytotoxic agents on hematopoietic progenitors. Cancer Res. 41:2338.
26 Adams, A. B., Durham, M. M., Kean, L. et al. 2001. Costimulation
blockade, busulfan, and bone marrow promote titratable macrochimerism, induce transplantation tolerance, and correct genetic
hemoglobinopathies with minimal myelosuppression. J. Immunol.
167:1103.
27 Samlowski, W. E., Araneo, B. A., Butler, M. O., Fung, M. C. and
Johnson, H. M. 1989. Peripheral lymph node helper T-cell recovery
after syngeneic bone marrow transplantation in mice prepared
with either gamma-irradiation or busulfan. Blood 74:1436.
28 Berryman, A. 1999. Principles of Population Dynamics and Their
Application. Stanley Thornes Ltd, Cheltenham, UK.
29 Freitas, A. A. and Rocha, B. 2000. Population biology of
lymphocytes: the flight for survival. Annu. Rev. Immunol. 18:83.
30 von Boehmer, H. and Hafen, K. 1993. The life span of naive alpha/
beta T cells in secondary lymphoid organs. J. Exp. Med. 177:891.
31 Tanchot, C. and Rocha, B. 1997. Peripheral selection of T cell
repertoire: the role of continuous thymic output. J. Exp. Med.
186:1099.
32 Tanchot, C. and Rocha, B. 1995. The peripheral T cell repertoire:
independent homeostatic regulation of virgin and activated CD8+
T cell pools. Eur. J. Immunol. 25:2127.
33 Bourgeois, C., Kassiotis, G. and Stockinger, B. 2005. A major role
for memory CD4 T cells in the control of lymphopenia-induced
proliferation of naive CD4 T cells. J. Immunol. 174:5316.
34 Dai, Z. and Lakkis, F. G. 2001. Cutting edge: secondary lymphoid
organs are essential for maintaining the CD4, but not CD8, naive
T cell pool. J. Immunol. 167:6711.
35 Berard, M., Brandt, K., Bulfone-Paus, S. and Tough, D. F. 2003. IL15 promotes the survival of naive and memory phenotype CD8+
T cells. J. Immunol. 170:5018.
36 Schluns, K. S., Kieper, W. C., Jameson, S. C. and Lefrancois, L.
2000. Interleukin-7 mediates the homeostasis of naive and
memory CD8 T cells in vivo. Nat. Immunol. 1:426.
37 Wu, T. S., Lee, J. M., Lai, Y. G. et al. 2002. Reduced expression of
Bcl-2 in CD8+ T cells deficient in the IL-15 receptor alpha-chain.
J. Immunol. 168:705.
38 Boursalian, T. E., Golob, J., Soper, D. M., Cooper, C. J. and Fink, P.
J. 2004. Continued maturation of thymic emigrants in the
periphery. Nat. Immunol. 5:418.
39 Oehen, S. and Brduscha-Riem, K. 1999. Naive cytotoxic
T lymphocytes spontaneously acquire effector function in lymphocytopenic recipients: a pitfall for T cell memory studies? Eur.
J. Immunol. 29:608.
40 Schonrich, G., Kalinke, U., Momburg, F. et al. 1991. Downregulation of T cell receptors on self-reactive T cells as a novel
mechanism for extrathymic tolerance induction. Cell 65:293.
41 Dobber, R., Hertogh-Huijbregts, A., Rozing, J., Bottomly, K. and
Nagelkerken, L. 1992. The involvement of the intestinal microflora
in the expansion of CD4+ T cells with a naive phenotype in the
periphery. Dev. Immunol. 2:141.
42 Ernst, B., Lee, D. S., Chang, J. M., Sprent, J. and Surh, C. D. 1999.
The peptide ligands mediating positive selection in the thymus
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
control T cell survival and homeostatic proliferation in the
periphery. Immunity. 11:173.
Wu, Z., Bensinger, S. J., Zhang, J. et al. 2004. Homeostatic
proliferation is a barrier to transplantation tolerance. Nat. Med.
10:87.
Marzo, A. L., Klonowski, K. D., Le Bon, A., Borrow, P., Tough, D. F.
and Lefrancois, L. 2005. Initial T cell frequency dictates memory
CD8+ T cell lineage commitment. Nat. Immunol. 6:793.
Moulton, V. R., Bushar, N. D., Leeser, D. B., Patke, D. S. and
Farber, D. L. 2006. Divergent generation of heterogeneous
memory CD4 T cells. J. Immunol. 177:869.
Marzo, A. L., Yagita, H. and Lefrancois, L. 2007. Cutting edge:
migration to nonlymphoid tissues results in functional conversion
of central to effector memory CD8 T cells. J. Immunol. 179:36.
Kassiotis, G. and Stockinger, B. 2004. Anatomical heterogeneity of
memory CD4+ T cells due to reversible adaptation to the
microenvironment. J. Immunol. 173:7292.
Seddon, B., Tomlinson, P. and Zamoyska, R. 2003. Interleukin 7
and T cell receptor signals regulate homeostasis of CD4 memory
cells. Nat. Immunol. 4:680.
Hamilton, S. E., Wolkers, M. C., Schoenberger, S. P. and Jameson,
S. C. 2006. The generation of protective memory-like CD8(+) T
cells during homeostatic proliferation requires CD4(+) T cells. Nat.
Immunol. 7:475.
Sallusto, F., Lenig, D., Forster, R., Lipp, M. and Lanzavecchia, A.
1999. Two subsets of memory T lymphocytes with distinct homing
potentials and effector functions. Nature 401:708.
Unsoeld, H., Krautwald, S., Voehringer, D., Kunzendorf, U. and
Pircher, H. 2002. Cutting edge: CCR7+ and CCR7 memory
T cells do not differ in immediate effector cell function. J. Immunol.
169:638.
Boyman, O., Cho, J. H., Tan, J. T., Surh, C. D. and Sprent, J. 2006.
A major histocompatibility complex class I-dependent subset of
memory phenotype CD8+ cells. J. Exp. Med. 203:1817.
Heeger, P. S., Greenspan, N. S., Kuhlenschmidt, S. et al. 1999.
Pretransplant frequency of donor-specific, IFN-gamma-producing
lymphocytes is a manifestation of immunologic memory and
correlates with the risk of posttransplant rejection episodes.
J. Immunol. 163:2267.
Adams, A. B., Williams, M. A., Jones, T. R. et al. 2003.
Heterologous immunity provides a potent barrier to transplantation
tolerance. J. Clin. Invest. 111:1887.
Brinkmann, V., Chen, S., Feng, L., Pinschewer, D., Nikolova, Z.
and Hof, R. 2001. FTY720 alters lymphocyte homing and protects
allografts without inducing general immunosuppression. Transplant. Proc. 33:530.
VanBuskirk, A. M., Wakely, M. E., Sirak, J. H. and Orosz, C. G.
1998. Patterns of allosensitization in allograft recipients: long-term
cardiac allograft acceptance is associated with active alloantibody production in conjunction with active inhibition of alloreactive
delayed-type hypersensitivity. Transplantation 65:1115.
Sawitzki, B., Lehmann, M., Vogt, K. et al. 2002. Bag-1 upregulation in anti-CD4 mAb-treated allo-activated T cell confers
resistance to activation-induced cell death (AICD). Transpl.
Immunol. 9:83.
Graca, L., Cobbold, S. P. and Waldmann, H. 2002. Identification of
regulatory T cells in tolerated allografts. J. Exp. Med. 195:1641.
Fisson, S., Darrasse-Jeze, G., Litvinova, E. et al. 2003. Continuous
activation of autoreactive CD4+ CD25+ regulatory T cells in the
steady state. J. Exp. Med. 198:737.