This information is current as
of June 15, 2017.
CD4 Help-Independent Induction of
Cytotoxic CD8 Cells to Allogeneic P815
Tumor Cells Is Absolutely Dependent on
Costimulation
Yifan Zhan, Alexandra J. Corbett, Jamie L. Brady, Robyn M.
Sutherland and Andrew M. Lew
J Immunol 2000; 165:3612-3619; ;
doi: 10.4049/jimmunol.165.7.3612
http://www.jimmunol.org/content/165/7/3612
Subscription
Permissions
Email Alerts
This article cites 55 articles, 33 of which you can access for free at:
http://www.jimmunol.org/content/165/7/3612.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2000 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
References
CD4 Help-Independent Induction of Cytotoxic CD8 Cells to
Allogeneic P815 Tumor Cells Is Absolutely Dependent on
Costimulation1
Yifan Zhan, Alexandra J. Corbett, Jamie L. Brady, Robyn M. Sutherland, and
Andrew M. Lew2
C
ytotoxic CD8 cells are a major type of effector cells
against most allografts, and therefore inhibiting the generation of CTL should improve transplant survival. Better understanding of the activation requirements for allogeneic
CTL will benefit the development of effective immune intervention. CTL to alloantigens can be either induced through direct interaction between CD8 T cells and allogeneic MHC class I molecules from donors, or cross-priming of donor Ags by the host (1).
Cross-priming of CD8 cells to a number of Ags has been shown
recently to require help from CD4 cells (2– 4). However, there are
conflicting reports on the requirement of CD4 help in the generation of alloreactive CTL. Some have found that the generation of
alloreactive CTL depends on help from CD4 cells (5, 6), whereas
others found that activation of naive TCR transgenic (Tg)3
CD8 cells on a RAG⫺/⫺ background against alloantigen (H-2Ld)
was intact, indicating the independence of CD4 help for CTL
induction (7).
CD4 help for induction of cross-primed CTL can be potentiated
by costimulatory molecules (such as CD40) on APCs (2– 4). The
B7 and CD40 costimulatory pathways can act independently of
Walter and Eliza Hall Institute of Medical Research, Royal Melbourne Hospital,
Parkville, Australia
Received for publication May 1, 2000. Accepted for publication July 10, 2000.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by Juvenile Diabetes Foundation-National Institutes of
Health and Juvenile Diabetes Foundation Center for Excellence Grants, Angelo & Sue
Alberti Charitable Foundation, Rebecca Cooper Foundation, Diabetes Australia, and
National Health and Medical Research Council, Australia.
2
Address correspondence and reprint requests to Dr. A. Lew, Walter and Eliza Hall
Institute of Medical Research, Post Office, Royal Melbourne Hospital, Parkville 3050,
Australia. E-mail address: [email protected]
3
Abbreviations used in this paper: Tg, transgenic; KLH, keyhole limpet hemocyanin;
NP, (4-hydroxy-3-nitrophenol) acetyl.
Copyright © 2000 by The American Association of Immunologists
each other (8) or influence each other. For example, modulation of
CD40 can up-regulate B7 on APC (9 –11); and CTLA4Ig, a negative regulator of B7/CD28 interaction, can block the ability of
anti-CD40 Ab to stimulate CTL (3). Conversely, ligation of CD28
can induce CD40L expression (12). These findings not only illustrate a pathway whereby CD4 help for CD8 cell activation is
educed via APCs, but also reinforce the importance of costimulation for activation of naive CD8 T cells. Nevertheless, costimulation blockade does not always prevent CD8 activation and allograft rejection. For example, in a TCR Tg/RAG⫺/⫺ mouse,
induction of CTL was not affected by CTLA4Ig treatment (7).
Independence of CD4 help and costimulation for CTL induction
in TCR Tg mice could be simply due to the sheer abundance of
CTL precursors with high intrinsic affinity to a single Ag. In such
mice, costimulation-dependent population expansion that is critical for a normal immune response may not be necessary. Thus, we
reassessed in this study the requirement of CD4 help and costimulation for CTL induction to allogeneic tumor P815 in wild-type
mice (in which the T cell frequency is more representative of an
allograft recipient) compared with a novel CD4-deficient mouse.
To investigate the role of costimulation and CD4 help, we have
generated a novel CD4-deficient mouse that is Tg for a rat antimouse CD4 Ab (GK1.5). These GK mice have a permanent absence of peripheral CD4 cells without the residual Th cells present
in commonly used CD4⫺/⫺ mice (13) and MHC class II⫺/⫺ mice
(14, 15). Requirement of B7/CD28 costimulation for CTL induction in the presence or absence of CD4 cells was evaluated in
CTLA4Ig Tg mice and in double Tg GK/CTLA4Ig mice. The role
of CD40/CD40L interaction in the induction of CTL in the presence and the absence of CD4 cells was evaluated with blocking
anti-CD40L Ab.
We show in this study that although CTL responses to allogeneic P815 tumor cells were optimal and more rapid with the presence of CD4 cells, they can be induced in their absence. This CD4
0022-1767/00/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
Mice made transgenic (Tg) for a rat anti-mouse CD4 Ab (GK mice) represent a novel CD4-deficient model. They not only lack
canonical CD4 cells in the periphery, but also lack the residual aberrant Th cells that are found in CD4ⴚ/ⴚ mice and MHC class
IIⴚ/ⴚ mice. To analyze the role of CD4 help and costimulation for CTL induction against alloantigens, we have assessed the surface
and functional phenotype of CD8 cells in vivo (e.g., clearance of allogeneic P815 cells) and in vitro. In our CD4-deficient GK mice,
CTL responses to allogeneic P815 cells were induced, albeit delayed, and were sufficient to eliminate P815 cells. Induction of CTL
and elimination of allogeneic P815 cells were inhibited both in the presence and absence of CD4 cells by temporary CD40 ligand
blockade. This indicated that direct interaction of CD40/CD40L between APCs and CD8 cells may be an accessory signal in CTL
induction (as well as the indirect pathway via APC/CD4 interaction). Furthermore, whereas in CTLA4Ig single Tg mice P815 cells
were rejected promptly, in the double Tg GK/CTLA4Ig mice CTL were not induced and allogeneic P815 cells were not rejected.
These findings suggest that CD40/CD40L is involved in both CD4-dependent and CD4-independent pathways, and that B7/CD28
is pivotal in the CD4-independent pathway of CTL induction against allogeneic P815 cells. The Journal of Immunology, 2000,
165: 3612–3619.
The Journal of Immunology
cell-independent CTL induction, however, is absolutely dependent
on direct costimulation of B7/CD28, whereas CD40/CD40L is involved in both CD4-dependent and CD4-independent pathways.
Materials and Methods
Mice
C57BL/6 mice with mutations in MHC class I:C.H-2bm1 (termed bm1)
were used as wild-type mice. All Tg mice were produced on this bm1
background. The light chain was derived from the rat mAb GK1.5, and the
heavy chain was a fusion of the VH of GK1.5 to C region of mouse IgG2c
(16) and shown to be functional by transfection (17). The genes were
cloned behind the human CMV promoter to generate GK Tg mice.
CTLA4Ig Tg mice have previously been described (18). These mice respectively produce anti-CD4 Ab and CTLA4Ig abundantly into the circulation (see below). Doubly Tg mice were produced by crossing between
homozygous GK and homozygous CTLA4Ig mice. BALB/c mice (H-2d)
were used as donor and source of stimulator for MLR. These animals were
bred under specific pathogen-free conditions at our Institute.
Cell lines and Abs
Assessment of CTL function in vivo by clearance of target cells
Mice were injected i.p. with 5 ⫻ 106 P815 cells and killed at various times.
Peritoneal cells were harvested by lavage and counted; numbers of surviving P815 cells were calculated from the percentage of H-2Dd-positive cells,
as determined by FACS. Spleens were removed for cell culture and FACS
analysis of CD8 cells. In some experiments, mice were treated with 0.5 mg
of anti-CD40L Ab each on days 0 and 3 relative to P815 injection.
Cell culture
Splenocytes from P815-primed and unprimed animals were cultured at 5 ⫻
106 cells/ml in 2 ml vol with 106 irradiated (200 Gy) P815 cells or 5 ␮g/ml
anti-CD3 (145-2C11). Culture supernatants were harvested at various times
after culture, and cytokine levels in the supernatants were evaluated. In
some cases, living cells were harvested from the bulk culture and their lytic
activity was assayed.
Assessment of CTL activity in vitro
For testing direct killing of targets, spleen cells from in vivo primed animals were prepared and used as effectors in the standard 4-h 51Cr release
assay. As targets, P815 and control EL-4 cells were labeled with 51Cr for
90 min at 37°C and 104 cells were incubated with various numbers of
effector cells in 96-well round-bottom plates for 4 h. Supernatants after
plate centrifugation were harvested and assayed for specific lysis. For measuring CTL precursors, splenocytes from in vivo primed animals were
cultured at 5 ⫻ 106 cells/ml in 2 ml vol with 106 irradiated (200 Gy) P815
cells for 6 days. After this in vitro stimulation, cells were assayed for lytic
activity to P815 and EL-4 cells, as above.
FACS analysis of lymphoid cells
Leukocyte suspensions of spleen, lymph nodes, and peritoneal lavage were
stained with mAb to CD4, CD8, CD62L, CD44, and CD25. Accumulation
of P815 cells in peritoneal lavages was evaluated by the percentage of
H-2d⫹ cells. Stained cells were analyzed on a FACScan (Becton Dickson,
Mountain View, CA) with CellQuest software.
Cytokine ELISA
IL-2, IL-5, and IFN-␥ were assayed by ELISA, according to manufacturer’s instruction. Recombinant cytokines as standard, coating Ab, and biotinylated Ab were obtained from PharMingen.
Ab response to NP
To test helper actvity of GK mice, mice were immunized with 50 ␮g of
alum-precipitated (4-hydroxy-3-nitrophenol) acetyl (NP) conjugated to
keyhole limpet hemocyanin (KLH) (NP:KLH ratio 15:1). Serum Abs with
high affinity for NP were detected with NP2-BSA-coated plates, as described previously (20).
Results
Mice Tg for anti-CD4 Ab are a novel model of CD4 cell
deficiency
We generated mice that were Tg for a chimeric rat anti-CD4 Ab
(derived from mAb GK1.5; hence the name GK mice) as a novel
model of CD4 deficiency. Under the human CMV promoter, the
Ab was produced predominantly in the pancreas (by Northern
analysis) and reached levels of 0.1 mg/ml in the circulation. These
mice lack peripheral CD4 cells (Fig. 1A) and lack the residual
helper cell populations mentioned above, viz CD4 cells in MHC
class II⫺/⫺ mice and the double negative population in CD4⫺/⫺
mice (Fig. 1, B and C). The profiles with three anti-CD4 mAbs
(YTA3.1.2, H129, and GK1.5) were similar; critically, YTA3.1.2
sees a different epitope to GK1.5 (19). The CD8 population was
not significantly changed in peripheral lymphoid organs compared
with wild-type mice. Mice doubly Tg for anti-CD4 Ab and
CTLA4Ig (GK/CTLA4Ig) were also produced and their CD4 phenotype was the same as GK mice (Fig. 1A).
To test Th activity in GK mice, they were immunized with NPKLH in alum adjuvant, and helper-dependent IgG response was
monitored. Although GK mice were capable of mounting a reduced IgM response, as expected they failed to mount a IgG response, which was very strong in wild-type mice (Fig. 1D).
CTL induction and allogeneic tumor clearance were delayed in
CD4-deficient mice
To examine whether CD8 CTL to P815 cells could be induced in
the absence of CD4 cells, both wild-type bm1 mice and CD4 celldeficient GK mice were injected i.p. with 5 ⫻ 106 P815 cells.
Allogeneic tumor growth was quantitatively monitored by peritoneal lavage. In bm1 mice, tumor cells were cleared 2 wk after
injection. In contrast, tumor clearance was slower in GK mice
compared with wild-type bm1 mice (Fig. 2). At 1 wk, P815 cells
in GK mice were 7-fold more abundant than in wild-type bm1
mice (Fig. 2A). P815 cells had been totally cleared in wild-type
mice by 2 wk (Fig. 2A), whereas it required an additional 1–2 wk
in GK mice. The rejection of P815 cells in GK mice was indeed
mediated by CD8 cells, as depletion of CD8 cells by mAb treatment prevented the elimination of allogeneic tumor cells (Fig. 2B).
As a measurement of in vivo effector function, splenocytes were
tested for direct killing immediately ex vivo in a chromium release
assay (no in vitro culture). CTL effectors to P815 cells in bm1 mice
were not detected at day 6 after P815 injection, but appeared at day
10. CTL effectors appeared later in GK mice: they were not detected at day 10, but appeared at day 14 (Fig. 3A). A standard CTL
assay (6 days in vitro stimulation followed by chromium release
assay) was used to detect CTL precursors. CTL precursors to P815
cells could be detected much earlier (6 days after injection) for
both bm1 mice and GK mice. Whereas splenocytes harvested at
that time showed no direct killing of P815 cells, CTL effectors did
develop after in vitro stimulation (Fig. 3B).
Evidence of CD8 cell activation in CD4 cell-deficient GK mice
was also found by FACS analysis. Activated CD8 cells downregulate CD62L expression (21). The proportion of CD62Llow
CD8⫹ cell population in spleen was increased after exposure to
P815 cells in both wild-type mice and GK mice (Fig. 4B). It was
noted that CD8 expression was also down-regulated on the
CD62Llow cells (Fig. 4B). This is not surprising, as activated CD8
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
P815 (H-2d) is a mastocytoma-derived tumor line; EL-4 (H-2b) is a thymoma-derived cell line. Cell lines were maintained in DMEM supplemented with 10% FCS and 5 mM HEPES. The mAb 53-6.7 and YT169
(anti CD8), and MR1 (anti-CD40L) for in vivo experiments were produced
in our laboratory. YTA3.1.2 is an anti-CD4 Ab that recognizes a different
epitope to GK1.5 (19) and was used to confirm depletion of CD4 cells. The
conjugated Abs 53-6.7 (anti-CD8) and H129 (anti-CD4) were purchased
from Sigma (St. Louis, MO); the conjugated Abs Mel-14 (anti-CD62L),
IM7 (anti-CD44), PC61 (anti-CD25), H57-595 (anti-TCR␤), and 34-2-12
(anti-H-2Dd) were purchased from PharMingen (San Diego, CA).
3613
3614
COSTIMULATION REQUIREMENT FOR CTL INDUCTION
cells have been found previously to down-regulate CD8 (22). Activated CD8 cells were high in CD44 and CD25 expression as well
(data not shown).
The above results confirmed that CD8 can be primed into CTL
effectors in the absence of CD4 cells, but that CD4 cells were
important for optimal CTL induction. One way that CD4 cells can
help CTL induction is via IL-2 production. Bulk culture (24 h) of
splenocytes from P815-injected mice showed that unlike cells
from wild-type bm1 mice that produced substantial amounts of
IL-2 (80 pg/ml), cells from GK mice did not produce detectable
amounts of IL-2 in vitro (⬍0.2 pg/ml). Likewise, Ag-specific
IFN-␥ production by splenocytes after 72-h culture was also only
detected in bm1 mice (2.2 ng/ml), but not in GK mice (⬍0.1 ng/
ml). In support of this role of IL-2, addition of exogenous IL-2
improved the response of CD8 cells from our CD4-deficient mice
in vitro (see below).
Treatment with blocking anti-CD40L Ab suppressed CTL
induction and clearance of allogeneic cells in wild-type and
CD4-deficient mice
During cross-priming of CD8 cells, CD40L and CD40 interaction
between CD4 cells and APC results in APC activation and then
indirectly helps CD8 activation (2– 4). In some cases, direct
CD40L/CD40 interaction between CD8 cells and APC can occur
(23, 24). To examine how CD40L/CD40 interaction contributes to
CTL induction to allogeneic MHC Ag, both bm1 mice and GK
mice were treated with anti-CD40L Ab, MR1. Two injections of
0.5 mg MR1 resulted in delayed tumor rejection in both bm1 mice
and GK mice, although numbers of tumor cells in GK mice were
significantly higher than in bm1 mice. At 2 wk after P815 injection, the numbers of P815 cells that accumulated in the peritoneal
cavity of MR1-treated wild-type mice and GK mice were 22 and
50 million, respectively (Fig. 5A). As a measurement of in vivo
effector function, splenocytes were tested for direct killing immediately ex vivo in a chromium release assay (no in vitro culture).
CTL induction in MR1-treated mice was greatly inhibited for both
wild-type and GK mice (Fig. 5B). The proportion of CD62Llow
CD8 cells was also reduced (Fig. 4B). Furthermore, presumably
because of reduced expansion after priming, the percentage of
CD8 cells in MR1-treated P815-primed mice (whether wild-type
or GK) was one-half that in corresponding untreated P815-primed
mice (Fig. 4A). With no significant change in total numbers of
splenocytes (all about 1.2 ⫻ 108), the decreased percentage reflects
a decrease in absolute numbers of CD8 cells. We conclude that
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 1. A, Lack of CD4 T cells in the periphery of GK mice. Splenocytes and lymph node cells
were prepared from bm1 mice, GK mice, and GK/
CTLA4Ig doubly Tg mice. Cells were stained for
CD4 and CD8. YTA3.1.2 was used for CD4 staining, as it sees a different epitope to GK1.5 (19); thus,
the lack of CD4 staining cannot be due to masking.
Both GK and GK/CTLA4Ig mice lack peripheral
CD4 cells and have a normal CD8 cell population
compared with bm1 mice. Numbers represent the
percentages of CD4 cells of total lymphoid cells. B,
Increased double negative population in CD4⫺/⫺
mice. Lymph node cells were prepared from bm1,
GK, MHC II⫺/⫺, and CD4⫺/⫺ mice and stained for
CD4/CD8 vs TCR␣␤. Numbers represent the percentages of CD4⫺CD8⫺TCR␣␤⫹ cells (framed) of
total lymphoid cells. C, Residual CD4 population
existed in MHC II⫺/⫺ mice. Lymph node cells from
MHC II⫺/⫺ mice and CD4⫺/⫺ mice were stained for
CD4 and CD8. Numbers represent the percentages
of CD4 cells of total lymphoid cells. D, Ab responses to NP by ELISA. bm1 and GK mice were
immunized i.p. with alum-precipitated NP-KLH (50
␮g). Production of IgG and IgM to NP-KLH was
evaluated 2 wk after immunization. Data represent
the OD means and SD of four mice in each group.
Three similar experiments were performed.
The Journal of Immunology
3615
CTLA4Ig prevented CTL induction and clearance of allogeneic
cells in CD4-deficient mice, but not wild-type mice
direct CD40L/CD40 interaction between CD8 cells and APC is
critical for CTL induction in the absence or presence of CD4 help.
Transient MR1 treatment did not induce tolerance in this study.
Even in the absence of CD4 cells, in MR1-treated GK mice the
tumor cells were cleared by 4 wk (data not shown). It may be that
continuous blockade of CD40 and CD40L interaction is required
for long-term suppression of CTL induction.
CD8 cells from GK/CTLA4Ig doubly Tg mice can respond in
vitro in the presence of IL-2
As shown above, splenocytes isolated from wild-type and GK
mice early on (6 days post-P815 injection) contained very few
mature CTL effectors, but they did contain CTL precursors that
FIGURE 3. CTL activity of splenocytes from P815-primed mice. Mice were injected i.p. with 5 ⫻ 106 P815 cells (H-2d). Spleens were harvested at
indicated times and single cell suspensions were prepared. A, Direct killing. CTL effectors in freshly isolated splenocytes were tested for specific killing
of P815 cells in 4-h 51Cr release assay. Each data point represents the mean of triplicate wells. Direct killing to P815 cells by splenocytes from the wild-type
and GK mice that were injected with P815 cells for 6 days was ⬍5% (data not shown), and killing of control EL-4 cells by splenocytes from all groups
was ⬍5%. B, Killing after in vitro stimulation. Splenocytes from the wild-type and GK mice that were injected with P815 cells for 6 days were cultured
at 5 ⫻ 106 cells/ml in 2 ml in the presence of 106 irradiated (200 Gy) P815 cells in vitro for 6 days. Recovered viable cells were tested for cytotoxicity
to P815. Killing of EL-4 cells by cells from all groups was ⬍5% (data not shown).
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 2. A, Allogeneic tumor clearance in wild-type bm1 and GK
mice. Mice were injected i.p. with 5 ⫻ 106 P815 cells (H-2d). At indicated
times, peritoneal cells were harvested by lavage. Absolute numbers of
P815 cells were calculated from total number of peritoneal lavage cells and
the percentage of H-2d-positive cells (as determined by FACS). Data represent the means and SD of three mice in each group. Three similar experiments were performed. B, Tumor cells from GK mice treated with
anti-CD8 Ab. GK mice were given three doses of anti-CD8 Ab (total 3 mg)
on days 0, 3, and 7 relative to P815 injection. Treated and untreated mice
were injected i.p. with 5 ⫻ 106 P815 cells (H-2d). Peritoneal lavage cells
were harvested 14 days after P815 injection.
The B7/CD28 costimulation pathway has been studied extensively
and represents one of the major pathways for T cell costimulation
(25). B7 blockade with fusion protein CTLA4Ig has been shown to
inhibit or prevent the rejection of allo- and xenografts (26, 27).
However, graft rejection can still occur in the absence of B7/CD28
costimulation (7, 28). We examined in this study in an allogeneic
tumor model whether CD8 cell activation in the absence of CD4
help is sensitive to costimulation blockade by CTLA4Ig.
We have previously produced mice Tg for CTLA4Ig under the
rat insulin promoter (18). Islets from these mice inhibited MLR in
vitro (18). In these mice, CTLA4Ig secreted by the pancreatic islets led to circulating CTLA4Ig that could be detected by binding
splenocytes, previously stimulated by LPS to up-regulate CD86
(B7-2) (Fig. 6A). Because the detection method could potentially
underestimate levels of CTLA4Ig due to saturation of B7 epitopes,
serum samples were serial diluted for testing their CTLA4Ig levels. Based on mean fluorescence intensity at sample dilution (1 in
10) that B7 binding was unsaturated, Fig. 6B showed that
CTLA4Ig Tg mice and GK/CTLA4Ig doubly Tg mice had a comparably high level of circulating CTLA4Ig before and 2 wk after
injection of tumor cells. The levels of CTLA4Ig in Tg mice correspond to 0.1 mg/ml serum.
When mice were challenged i.p. with 5 ⫻ 106 P815 cells, both
GK and CTLA4Ig single Tg mice rejected their tumor cells at 2 wk
and CD8 cells appeared to be activated (high percentage of CD8
cells were CD62Llow) (Fig. 7). In contrast, GK/CTLA4Ig doubly
Tg mice failed to reject the P815 cells; indeed, the P815 cells
multiplied to be 2 ⫻ 108 in the peritoneum (Fig. 7A). By about 4
wk, these mice succumbed to the tumor burden. This was in accord
with the low numbers of activated CD8 cells (Fig. 7B) and the lack
of CTL induction in GK/CTLA4Ig Tg mice (Fig. 8B).
3616
matured into effectors after in vitro culture. It is possible that
splenocytes from doubly Tg mice were also primed to become
CTL precursors, but did not mature into effectors in vivo. To investigate this possibility, splenocytes from primed doubly Tg mice
were cultured with irradiated P815 cells and/or IL-2. Addition of
exogenous IL-2 (1 ng/ml) was required for CD8 cells to expand, to
become killers, and to secrete IFN- ␥; there was no response when
P815 cells alone were used (Fig. 8).
Discussion
Anti-CD4 Ab Tg mice were used in this study as the model of
CD4 cell deficiency. The advantage of the model is that mice
FIGURE 5. Tumor clearance and CTL induction in MR1-treated bm1 and GK mice. Experimental groups were the same as in Fig. 4. Peritoneal cells and spleens were harvested 2 wk
after P815 injection. Tumor growth and CTL activity were assayed as in Figs. 2 and 3A. A, Tumor growth; B, direct killing of P815 cells (without in vitro stimulation of splenocytes).
FIGURE 6. A, Binding by Tg CTLA4Ig. Sera were prepared from wildtype bm1 mice (dotted line), CTLA4Ig Tg, or GK/CTLA4Ig doubly Tg
mice. B7-binding activity in sera was evaluated by FACS with LPS-stimulated splenic B220⫹ cells. B, Estimation of serum Tg CTLA4Ig. Sera
from four mice in each group before or 14 days after injection of P815 cells
were tested by FACS. Data represent the means and SD of mean fluorescence intensity at 1/10 dilution.
lack the aberrant T cell population that is found in MHC class
II⫺/⫺ mice and CD4⫺/⫺ mice. MHC class II⫺/⫺ mice have
residual CD1-restricted CD4 population (15, 29), and CD4⫺/⫺
mice have increased numbers of an abnormal population of
CD4⫺CD8⫺TCR␣␤⫹ cells (13, 30, 31). CD8 cells in CD4⫺/⫺
mice can also be MHC class II restricted (32). These minor populations can function as helpers, and therefore may represent a
problem in interpretation of results, especially in vivo, in which
such minor populations can be readily expanded. This may also
explain why allograft rejection is found to be delayed in mice
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 4. FACS of splenocytes from mice injected i.p. with 5 ⫻ 106
P815 cells. Some groups of mice also had two i.p. injections of MR1 (0.5
mg each at day 0 and day 3 relative to P815 injection). Splenocytes from
P815-primed and unprimed animals were double stained with PE anti-CD4
plus FITC anti-CD8 (A) and PE anti-CD8 plus FITC anti-CD62L (B). The
annotated numbers indicate the percentages of a particular population. Activated cells are CD62Llow.
COSTIMULATION REQUIREMENT FOR CTL INDUCTION
The Journal of Immunology
3617
whose CD4 cells are depleted by Ab, but not in MHC class IIdeficient mice and CD4-deficient mice (33).
Our results indicate that CD4 cells provide help in the induction
of CTL against allogeneic P815 cells. In response to allogeneic
cells, requirement of CD4 help for CTL induction may vary with
precursor frequency. When there are abundant Ag-specific cells,
e.g., in anti-Ld TCR/RAG Tg mice (2C), and thus expansion is less
important, optimal CTL response and P815 tumor clearance are
achieved without the involvement of CD4 cells (7). When there are
fewer cells, e.g., in adoptive transfer experiments into syngeneic
mice with only few (4 – 8) million of TCR Tg cells, alloreactive
stimulated expansion of the transferred cells was critically dependent on cotransfer of CD4 cells (6). In our study in which H-2d⫹
P815 cells are injected into fully mismatched MHC wild-type bm1
mice, the abundance of T cells capable of recognizing alloantigen
in recipient mice probably falls between the above two systems.
An alloreactive CTL response develops in mice lacking peripheral
CD4 cells, and the response is sufficient to clear the tumor cells.
However, development of CTL response in the absence of CD4
cells is suboptimal and delayed when compared with wild-type
mice, reflecting the need of CD4 help for expansion of activated
CD8 cells and optimal CTL response. The delay in CTL induction
may have been the reason that CTL were not detected in CD4
cell-depleted mice in an earlier study (5). The results from CD4
cell-deficient mice are reminiscent of CTL responses to some viral
infections that are believed to be CD4 independent. For example,
although both CD4 cell-deficient and wild-type mice are capable
of mounting CTL to clear influenza infection, the numbers of responding CTL precursors were much lower in CD4 cell-deficient
mice (34). Similarly, a primary CTL response to acute lymphocytic choriomeningitis virus infection was produced in CD4 cell-
FIGURE 8. In vitro stimulation of splenocytes from P815-injected GK/
CTLA4Ig mice. GK/CTLA4Ig mice were i.p. injected with 5 ⫻ 106 P815
cells for 2 wk. Splenocytes were cultured at 5 ⫻ 106/ml in 2 ml vol with
medium alone, 2.5 ⫻ 106 irradiated (200 Gy) P815 cells, 1 ng/ml IL-2
(2.5–7.5 ⫻ 108 U/mg; PharMingen, San Diego, LA), or both. Cells were
cultured for 5 days. A, Numbers of recovered CD8 cells. Viable cells were
harvested, enumerated, and stained for CD8. Data represent the numbers of
CD8 cells recovered, calculated from total cell numbers and percentages of
CD8 cells. The starting number of CD8 cells put into culture is indicated
by an arrow. B, Cytotoxicity of cultured cells. Viable cells recovered from
5-day cultures were assayed for cytotoxicity to P815 cells. Mean and SD
of triplicates for each group are shown. Killing of control EL-4 cells was
less than 5% (data not shown). C, IFN-␥ production. Supernatants from
5-day cultures were assayed for their IFN-␥ levels. The IFN-␥ standards
used were purchased from PharMingen (0.2–1 ⫻ 108 U/mg). Mean and SD
of triplicates for each group are shown. Detection limit ⫽ 0.3 ng/ml.
deficient mice, and this was sufficient to clear infection. However,
in the absence of CD4 cells, the CTL response was not sustained
and infection could persist (35, 36). One conclusion that befits all
of these findings is that the major contribution of CD4 cell help for
optimal CTL induction is facilitating population expansion.
One way that CD4 cells provide help is through the provision of
cytokines such as IL-2 (37, 38). In the above 2C T cell adoptive
transfer study, in vivo expansion of transferred cells was largely
related to the ability of cotransferred CD4 cells to produce IL-2, as
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
FIGURE 7. A, Tumor clearance and CTL induction in GK mice and
GK/CTLA4Ig mice. Mice were i.p. injected with 5 ⫻ 106 P815 cells. Two
weeks after, cells from the peritoneal lavage cells were thoroughly flushed
and harvested. P815 cells were calculated from the total number of peritoneal lavage and the percentage of H-2Dd-positive cells. Data represent
the means and SD of three to four mice in each group. Two similar experiments were performed. B, FACS of CTL effectors. Splenocytes were
double stained for CD8 and CD62L. Percentages of CD8 cells (inside
frame) and CD62LlowCD8⫹ cells (arrow) were calculated and indicated at
the bottom of the figure. There were no differences in CD62L expression on
CD8 cells among three strains of unimmunized mice.
3618
cell help is critically dependent on the B7/CD28 costimulation
pathway.
Interestingly, CD8 activation is blocked by Tg CTLA4Ig, only
when CD4 cells are absent. This is reminiscent of some studies that
show that CTLA4Ig is only effective in blocking transplant rejection when host CD4 cells are depleted (52, 53). We do not believe
that this is simply due to CD4 cells consuming CTLA4Ig, because
we can easily detect free CTLA4Ig in non-CD4-deficient mice.
One possible explanation is that in the absence of CD28, costimulation of CD4 cells can be mediated via CD40 (54, 55), and so the
lack of B7/CD28 costimulation in CD8 cells is compensated for by
this CD4 help.
The current study concludes that generation of alloreactive CD8
effectors can be independent of CD4 help, and that CD8 cell activation without CD4 cell help is more sensitive to costimulation
blockade. In our allogeneic system, CD40/CD40L has a role in
CTL priming for both CD4-dependent and CD4-independent responses, but B7/CD28 is critical for the CD4-independent pathway
only. These influences should be considered in alloimmunity under
different transplantation settings in the hope of more successful
intervention.
Acknowledgments
We thank Drs. Bill Heath (Walter and Eliza Hall Institute of Medical Research, Parkville, Australia) and Frank Carbone (University of Melbourne,
Melbourne, Australia) for reading the manuscript.
References
1. Gould, D. S., and H. Auchincloss, Jr. 1999. Direct and indirect recognition: the
role of MHC antigens in graft rejection. Immunol. Today 20:77.
2. Bennett, S. R., F. R. Carbone, F. Karamalis, R. A. Flavell, J. F. Miller, and
W. R. Heath. 1998. Help for cytotoxic-T-cell responses is mediated by CD40
signalling. Nature 393:478.
3. Ridge, J. P., F. Di Rosa, and P. Matzinger. 1998. A conditioned dendritic cell can
be a temporal bridge between a CD4⫹ T-helper and a T-killer cell. Nature 393:
474.
4. Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, and C. J. Melief.
1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L
interactions. Nature 393:480.
5. Kerkvliet, N. I., L. Baecher-Steppan, D. M. Shepherd, J. A. Oughton,
B. A. Vorderstrasse, and G. K. DeKrey. 1996. Inhibition of TC-1 cytokine production, effector cytotoxic T lymphocyte development and alloantibody production by 2,3,7,8-tetrachlorodibenzo-p-dioxin. J. Immunol. 157:2310.
6. Buhlmann, J. E., M. Gonzalez, B. Ginther, A. Panoskaltsis-Mortari, B. R. Blazar,
D. L. Greiner, A. A. Rossini, R. Flavell, and R. J. Noelle. 1999. Cutting edge:
sustained expansion of CD8⫹ T cells requires CD154 expression by Th cells in
acute graft versus host disease. J. Immunol. 162:4373.
7. Manning, T. C., L. A. Rund, M. M. Gruber, F. Fallarino, T. F. Gajewski, and
D. M. Kranz. 1997. Antigen recognition and allogeneic tumor rejection in CD8⫹
TCR transgenic/RAG⫺/⫺ mice. J. Immunol. 159:4665.
8. Saito, K., J. Sakurai, J. Ohata, T. Kohsaka, H. Hashimoto, K. Okumura, R. Abe,
and M. Azuma. 1998. Involvement of CD40 ligand-CD40 and CTLA4 –B7 pathways in murine acute graft-versus-host disease induced by allogeneic T cells
lacking CD28. J. Immunol. 160:4225.
9. Caux, C., C. Massacrier, B. Vanbervliet, B. Dubois, C. Van Kooten, I. Durand,
and J. Banchereau. 1994. Activation of human dendritic cells through CD40
cross-linking. J. Exp. Med. 180:1263.
10. Kiener, P. A., P. Moran-Davis, B. M. Rankin, A. F. Wahl, A. Aruffo, and
D. Hollenbaugh. 1995. Stimulation of CD40 with purified soluble gp39 induces
proinflammatory responses in human monocytes. J. Immunol. 155:4917.
11. McLellan, A. D., R. V. Sorg, L. A. Williams, and D. N. Hart. 1996. Human
dendritic cells activate T lymphocytes via a CD40:CD40 ligand-dependent pathway. Eur. J. Immunol. 26:1204.
12. Yin, D., L. Zhang, R. Wang, L. Radvanyi, C. Haudenschild, Q. Fang,
M. R. Kehry, and Y. Shi. 1999. Ligation of CD28 in vivo induces CD40 ligand
expression and promotes B cell survival. J. Immunol. 163:4328.
13. Locksley, R. M., S. L. Reiner, F. Hatam, D. R. Littman, and N. Killeen. 1993.
Helper T cells without CD4: control of leishmaniasis in CD4-deficient mice.
Science 261:1448.
14. Cardell, S., M. Merkenschlager, H. Bodmer, S. Chan, D. Cosgrove, C. Benoist,
and D. Mathis. 1994. The immune system of mice lacking conventional MHC
class II molecules. Adv. Immunol. 55:423.
15. Cardell, S., S. Tangri, S. Chan, M. Kronenberg, C. Benoist, and D. Mathis. 1995.
CD1-restricted CD4⫹ T cells in major histocompatibility complex class II-deficient mice. J. Exp. Med. 182:993.
16. Martin, R. M., A. Silva, and A. M. Lew. 1997. The Igh-1 sequence of the nonobese diabetic (NOD) mouse assigns it to the IgG2c isotype. Immunogenetics
46:167.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
CD4 cells from IL-2⫺/⫺ mice did not result in the expansion of 2C
cells (6). In our study, P815-primed bm1 splenocytes that contain
both CD4 and CD8 cells produced much more IL-2 than cells from
CD4 cell-deficient mice. It is plausible that IL-2 produced by CD4
cells may enhance CTL responses by expansion of activated precursors. Indeed, we have shown that CD8 cells, either naive or
primed, proliferate and become effectors in vitro in response to
stimulation by P815 alloantigen and IL-2 (Fig. 8), whereas naive
CD8 cells did not respond to alloantigens from P815 without exogenous IL-2 (data not shown). Although IL-2 can also directly
enhance cellular cytotoxicity of CD8 cells as well (39), the amount
of IL-2 required for generation of killer activity is 10-fold lower
than the amounts of IL-2 required for proliferation (40).
Another pathway in which CD4 cells provide help for CD8 cells
is via activation of APCs. CD40/CD40L interaction between CD4
cells and APCs has been demonstrated to be critical for in vivo
CD8 priming (2– 4), although CD40-independent mechanisms
may also exist (41). Ligation of CD40 by an agonist anti-CD40 Ab
can restore CD8 CTL activities in CD4 T cell-depleted mice (2– 4).
In our study, interruption of CD40/CD40L interaction by antiCD40L Ab resulted in failure to clear tumor cells and to induce
CTL response, notably both in the presence and the absence of
CD4 cells. This suggested that in addition to the known role of
CD40/CD40L in CTL induction via indirect interaction through
APC and CD4 cells (2– 4), there can be direct interaction of CD40/
CD40L between APC and CD8 cells (23, 24), although in normal
circumstances this direct pathway may be masked by the CD4
pathway. As for strong CTL inducers such as lymphocytic choriomeningitis virus, Pichinde virus, or vesicular stomatitis virus,
mice can mount CTL responses in CD40L-knockout mice, but the
memory CTL response was much less efficient than in wild-type
mice (42).
The development of CTL responses to allogeneic Ags in the
absence of CD4 help poses the important question of whether the
CD4 cell-independent CTL response is dependent on costimulation. Optimal T cell activation is believed to require two signals:
ligation of TCR to MHC/peptide complex (signal 1) and costimulation (signal 2). Several reports have described activation of CD8
cells without the need for costimulation, but this requires high dose
or persistent Ag exposure and the use of TCR Tg mice (7, 43– 46).
In contrast, only in the presence of costimulation can low concentrations of allopeptides induce high CD69 expression, proliferation, and high IL-2 production (43, 45). Others have also pointed
to a critical role of costimulation in induction of CD8 CTL (47,
48). A cogent example is in a syngeneic tumor system, in which
without CD4 help, induction of CTL responses was effective, only
when the MHC class I⫹ and class II⫺ tumor cells were transfected
with B7 (47) (although this may still require involvement of B7expressing host APC (49, 50)). As discussed above for CD4 dependence, these results can be largely reconciled by the premise
that lack of costimulation can be compensated for by a very strong
signal 1 at the total population level (abundant specific T cells and
high Ag dose). In our study, CTL induction to allogeneic tumor
cells in the absence of CD4 cells was totally abrogated by B7 being
blocked by circulating CTLA4Ig; thus, the allogeneic tumor cells
persisted for a long time in our study and indeed mice were overcome by the tumor burden. The main contribution of CD28-mediated signaling is to promote the release of IL-2, which in turn
expands activated CTL (51). As already mentioned above, the expansion process that may be less important when the cells are very
abundant (preexpanded) such as in TCR Tg. We believe that without using TCR Tg models, our experimental setting is more representative of an allograft into a transplant recipient and that induction of CTL to allogeneic tumor cells in the absence of CD4
COSTIMULATION REQUIREMENT FOR CTL INDUCTION
The Journal of Immunology
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
toxic T lymphocytes after immunization and show diminished resistance to subsequent virus challenge. J. Virol. 70:1072.
Wagner, H., and M. Rollinghoff. 1978. T-T-cell interactions during the vitro
cytotoxic allograft responses. I. Soluble products from activated Lyl⫹ T cells
trigger autonomously antigen-primed Ly23⫹ T cells to cell proliferation and cytolytic activity. J. Exp. Med. 148:1523.
Baker, P. E., S. Gillis, M. M. Ferm, and K. A. Smith. 1978. The effect of T cell
growth factor on the generation of cytolytic T cells. J. Immunol. 121:2168.
Grimm, E. A., and L. Owen-Schaub. 1991. The IL-2 mediated amplification of
cellular cytotoxicity. J. Cell. Biochem. 45:335.
Van Gool, S. W., A. Kasran, G. Wallays, M. de Boer, and J. L. Ceuppens. 1995.
Accessory signalling by B7-1 for T cell activation induced by anti-CD2: evidence
for IL-2-independent CTL generation and CsA-resistant cytokine production.
Scand. J. Immunol. 41:23.
Lu, Z., L. Yuan, X. Zhou, E. Sotomayor, H. I. Levitsky, and D. M. Pardoll. 2000.
CD40-independent pathways of T cell help for priming of CD8(⫹) cytotoxic T
lymphocytes. J. Exp. Med. 191:541.
Borrow, P., A. Tishon, S. Lee, J. Xu, I. S. Grewal, M. B. Oldstone, and
R. A. Flavell. 1996. CD40L-deficient mice show deficits in antiviral immunity
and have an impaired memory CD8⫹ CTL response. J. Exp. Med. 183:2129.
Cai, Z., A. Brunmark, M. R. Jackson, D. Loh, P. A. Peterson, and J. Sprent. 1996.
Transfected Drosophila cells as a probe for defining the minimal requirements for
stimulating unprimed CD8⫹ T cells. Proc. Natl. Acad. Sci. USA 93:14736.
Kundig, T. M., A. Shahinian, K. Kawai, H. W. Mittrucker, E. Sebzda,
M. F. Bachmann, T. W. Mak, and P. S. Ohashi. 1996. Duration of TCR stimulation determines costimulatory requirement of T cells. Immunity 5:41.
Luxembourg, A. T., A. Brunmark, Y. Kong, M. R. Jackson, P. A. Peterson,
J. Sprent, and Z. Cai. 1998. Requirements for stimulating naive CD8⫹ T cells via
signal 1 alone. J. Immunol. 161:5226.
Wang, B., R. Maile, R. Greenwood, E. J. Collins, and J. A. Frelinger. 2000. Naive
CD8⫹ T cells do not require costimulation for proliferation and differentiation
into cytotoxic effector cells. J. Immunol. 164:1216.
Gajewski, T. F., F. Fallarino, C. Uyttenhove, and T. Boon. 1996. Tumor rejection
requires a CTLA4 ligand provided by the host or expressed on the tumor: superiority of B7-1 over B7-2 for active tumor immunization. J. Immunol. 156:2909.
Krummel, M. F., W. R. Heath, and J. Allison. 1999. Differential coupling of
second signals for cytotoxicity and proliferation in CD8⫹ T cell effectors: amplification of the lytic potential by B7. J. Immunol. 163:2999.
La Motte, R. N., A. H. Sharpe, J. A. Bluestone, and M. B. Mokyr. 1998. Importance of B7-1-expressing host antigen-presenting cells for the eradication of B7-2
transfected P815 tumor cells. J. Immunol. 161:6552.
La Motte, R. N., A. H. Sharpe, J. A. Bluestone, and M. B. Mokyr. 1999. Host
B7-1 and B7-2 costimulatory molecules contribute to the eradication of B7-1transfected P815 tumor cells via a CD8⫹ T cell-dependent mechanism. J. Immunol. 162:4817.
Harding, F. A., and J. P. Allison. 1993. CD28 –B7 interactions allow the induction of CD8⫹ cytotoxic T lymphocytes in the absence of exogenous help. J. Exp.
Med. 177:1791.
Yin, D., and C. G. Fathman. 1995. Induction of tolerance to heart allografts in
high responder rats by combining anti-CD4 with CTLA4Ig. J. Immunol. 155:
1655.
Yin, D. P., H. N. Sankary, J. Williams, N. Krieger, and C. G. Fathman. 1996.
Induction of tolerance to small bowel allografts in high-responder rats by combining anti-CD4 with CTLA4Ig. Transplantation 62:1537.
Hancock, W. W., M. H. Sayegh, X. G. Zheng, R. Peach, P. S. Linsley, and
L. A. Turka. 1996. Costimulatory function and expression of CD40 ligand, CD80,
and CD86 in vascularized murine cardiac allograft rejection. Proc. Natl. Acad.
Sci. USA 93:13967.
Ding, L., and E. M. Shevach. 1996. Activated B cells express CD28/B7-independent costimulatory activity. J. Immunol. 157:1389.
Downloaded from http://www.jimmunol.org/ by guest on June 15, 2017
17. McKenzie, A., J. Brady, R. Martin, H. Georgiou, and A. Lew. 2000. Local secretion of a chimeric anti-CD4 antibody protects against graft rejection in the
NOD mouse. Transplantation 69:1745.
18. Sutherland, R. M., J. L. Brady, H. M. Georgiou, H. Thomas, and A. M. Lew.
2000. Protective effect of CTLA4Ig secreted by transgenic fetal pancreas allografts. Transplantation 69:1806.
19. Darby, C. R., P. J. Morris, and K. J. Wood. 1992. Evidence that long-term cardiac
allograft survival induced by anti-CD4 monoclonal antibody does not require
depletion of CD4⫹ T cells. Transplantation 54:483.
20. Ridderstad, A., G. J. Nossal, and D. M. Tarlinton. 1996. The xid mutation diminishes memory B cell generation but does not affect somatic hypermutation
and selection. J. Immunol. 157:3357.
21. Mobley, J. L., and M. O. Dailey. 1992. Regulation of adhesion molecule expression by CD8 T cells in vivo. I. Differential regulation of gp90MEL-14 (LECAM1), Pgp-1, LFA-1, and VLA- 4␣ during the differentiation of cytotoxic T lymphocytes induced by allografts. J. Immunol. 148:2348.
22. Lucas, B., F. Vasseur, and C. Penit. 1995. Stochastic coreceptor shut-off is restricted to the CD4 lineage maturation pathway. J. Exp. Med. 181:1623.
23. Lefrancois, L., S. Olson, and D. Masopust. 1999. A critical role for CD40-CD40
ligand interactions in amplification of the mucosal CD8 T cell response. J. Exp.
Med. 190:1275.
24. Eck, S. C., and L. A. Turka. 1999. Generation of protective immunity against an
immunogenic carcinoma requires CD40/CD40L and B7/CD28 interactions but
not CD4(⫹) T cells. Cancer Immunol. Immunother. 48:336.
25. Bluestone, J. A. 1995. New perspectives of CD28 –B7-mediated T cell costimulation. Immunity 2:555.
26. Lenschow, D. J., Y. Zeng, J. R. Thistlethwaite, A. Montag, W. Brady,
M. G. Gibson, P. S. Linsley, and J. A. Bluestone. 1992. Long-term survival of
xenogeneic pancreatic islet grafts induced by CTLA4lg. Science 257:789.
27. Pearson, T. C., D. Z. Alexander, K. J. Winn, P. S. Linsley, R. P. Lowry, and
C. P. Larsen. 1994. Transplantation tolerance induced by CTLA4-Ig. Transplantation 57:1701.
28. Kawai, K., A. Shahinian, T. W. Mak, and P. S. Ohashi. 1996. Skin allograft
rejection in CD28-deficient mice. Transplantation 61:352.
29. Schofield, L., M. J. McConville, D. Hansen, A. S. Campbell, B. Fraser-Reid,
M. J. Grusby, and S. D. Tachado. 1999. CD1d-restricted immunoglobulin G
formation to GPI-anchored antigens mediated by NKT cells. Science 283:225.
30. Takamoto, M., Z. X. Wang, N. Watanabe, A. Matsuzawa, H. Nariuchi, and
K. Sugane. 1998. Eosinophilia, IgE production, and cytokine production by lung
T cells in surface CD4-deficient mutant mice infected with Toxocara canis. Immunology 95:97.
31. Schilham, M. W., W. P. Fung-Leung, A. Rahemtulla, T. Kuendig, L. Zhang,
J. Potter, R. G. Miller, H. Hengartner, and T. W. Mak. 1993. Alloreactive cytotoxic T cells can develop and function in mice lacking both CD4 and CD8. Eur.
J. Immunol. 23:1299.
32. Heemskerk, M. H., M. W. Schilham, H. M. Schoemaker, G. Spierenburg,
W. J. Spaan, and C. J. Boog. 1995. Activation of virus-specific major histocompatibility complex class II-restricted CD8⫹ cytotoxic T cells in CD4-deficient
mice. Eur. J. Immunol. 25:1109.
33. He, G., J. Hart, O. S. Kim, G. L. Szot, C. T. Siegel, J. R. Thistlethwaite, and
K. A. Newell. 1999. The role of CD8 and CD4 T cells in intestinal allograft
rejection: a comparison of monoclonal antibody-treated and knockout mice.
Transplantation 67:131.
34. Tripp, R. A., S. R. Sarawar, and P. C. Doherty. 1995. Characteristics of the
influenza virus-specific CD8⫹ T cell response in mice homozygous for disruption
of the H-2lAb gene. J. Immunol. 155:2955.
35. Matloubian, M., R. J. Concepcion, and R. Ahmed. 1994. CD4⫹ T cells are required to sustain CD8⫹ cytotoxic T-cell responses during chronic viral infection.
J. Virol. 68:8056.
36. Von Herrath, M. G., M. Yokoyama, J. Dockter, M. B. Oldstone, and
J. L. Whitton. 1996. CD4-deficient mice have reduced levels of memory cyto-
3619