RAGE Ligation Affects T Cell Activation and Controls T Cell
Yali Chen,§Eitan M. Akirav,* Wei Chen,* Octavian Henegariu,* Bernhard Moser,¶
Dharmesh Desai,§Jane M. Shen,‡Jeffery C. Webster,‡Robert C. Andrews,‡Adnan M. Mjalli,‡
Robert Rothlein,‡Ann Marie Schmidt,¶Raphael Clynes,§and Kevan C. Herold2*†
The pattern recognition receptor, RAGE, has been shown to be involved in adaptive immune responses but its role on the
components of these responses is not well understood. We have studied the effects of a small molecule inhibitor of RAGE and
the deletion of the receptor (RAGE?/? mice) on T cell responses involved in autoimmunity and allograft rejection. Syn-
geneic islet graft and islet allograft rejection was reduced in NOD and B6 mice treated with TTP488, a small molecule RAGE
inhibitor (p < 0.001). RAGE?/? mice with streptozotocin-induced diabetes showed delayed rejection of islet allografts
compared with wild type (WT) mice (p < 0.02). This response in vivo correlated with reduced proliferative responses of
RAGE?/? T cells in MLRs and in WT T cells cultured with TTP488. Overall T cell proliferation following activation with
anti-CD3 and anti-CD28 mAbs were similar in RAGE?/? and WT cells, but RAGE?/? T cells did not respond to co-
stimulation with anti-CD28 mAb. Furthermore, culture supernatants from cultures with anti-CD3 and anti-CD28 mAbs
showed higher levels of IL-10, IL-5, and TNF-? with RAGE?/? compared with WT T cells, and WT T cells showed reduced
production of IFN-? in the presence of TTP488, suggesting that RAGE may be important in the differentiation of T cell
subjects. Indeed, by real-time PCR, we found higher levels of RAGE mRNA expression on clonal T cells activated under Th1
differentiating conditions. We conclude that activation of RAGE on T cells is involved in early events that lead to differ-
entiation of Th1?T cells.
The Journal of Immunology, 2008, 181: 4272–4278.
cells, may stimulate pathogen-associated molecular pattern recep-
tors, such as TLR, that lead to expression of molecules that acti-
vate the adaptive immune responses (1). Activation of innate re-
sponses may also alter the way in which Ags are presented to T
and other cells of the adaptive immune response. The link between
innate and adaptive immune responses has led investigators to pos-
tulate a role of innate responses in regulating autoimmune diseases
such as Type 1 diabetes (2, 3). In support of this hypothesis, de-
ficiency of IL-1 receptor and treatment with IL-1 receptor antag-
onist attenuates the rate of diabetes in NOD mice (4, 5).
Innate immune responses have also been shown to be involved
in the control of adaptive responses. Activation of NK T cells with
?-gal-cer was shown to prevent diabetes in the NOD mouse, and
a biased response of invariant V?24J?Q T cells was identified in
first degree relatives of patients who progressed to diabetes when
compared with nonprogressors (6, 7). In contrast, blockade of
NKG2D prevented diabetes in NOD mice (8). Clearly the progres-
mmune responses to foreign and autoantigens involve acti-
vation of both innate and adaptive immune responses. “Dan-
ger signals,” for example, those initiated by stressed or dying
sion to diabetes in animal models and in humans is primarily de-
pendent on the adaptive responses of T cells, but the way in which
the early innate responses lead to and shape adaptive immune re-
sponses is not clear.
The pattern recognition receptor, receptor for advanced glyca-
tion endproducts (RAGE),3is a potential link between adaptive
and innate responses (9). RAGE was originally identified as the
receptor for ligands for molecules whose concentrations are in-
creased in patients with diabetes, and a role for this molecule in the
development of secondary end-organ complications has been pos-
tulated. However, there are other ligands for RAGE, including
S100/calgranulins and HMGB1, products of cellular destruction
that may be generated during inflammatory responses and released
by inflammatory cells (10). A role for RAGE in innate responses
has been identified. Liliensiek et al. (11) showed that mice defi-
cient in RAGE (RAGE?/?) were protected from the lethal effects
of septic shock following cecal ligation, and suggested that RAGE
plays a role in propagating inflammation. Hofmann et al. (12)
found that interaction of RAGE ligands such as members of the
S100/calgranulin family with cellular RAGE on endothelium,
mononuclear phagocytes, and lymphocytes triggered cellular acti-
vation, with generation of key proinflammatory mediators.
Kokkola et al. (13) found that HMGB1 has the potential to induce
a proinflammatory phenotype in macrophages following interac-
tions with RAGE. More recently, Tian et al. (14) found that in
systemic lupus erythematosus, DNA-containing immune com-
plexes that bind HMGB1-bound RAGE, activating TLR-9, and
caused plasma dendritic cells (DC) to secrete IFN-?.
*Department of Immunobiology and
School of Medicine, New Haven, CT 06520;‡TransTech Pharma, High Point, NC
27265; and§Department of Medicine and¶Department of Surgery, Columbia Uni-
versity, New York, NY 10032
†Department of Medicine, Yale University
Received for publication August 30, 2007. Accepted for publication July 1, 2008.
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.
1This work was supported by Grant 2004-808 from the Juvenile Diabetes Research
2Address correspondence and reprint requests to Dr. Kevan Herold, Department of
Immunobiology, Yale University, 300 Cedar Street, S155B, New Haven, CT 06520.
E-mail address: email@example.com
3Abbreviations used in this paper: RAGE, receptor for advanced glycation endprod-
uct; sRAGE, soluble RAGE; WT, wild type; DC, dendritic cell.
Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00
The Journal of Immunology
It has also been suggested that RAGE ligation may affect adap-
tive immune responses, but the role of this pattern recognition
receptor on T cells has not been well studied, and some investi-
gations have failed to show a role of RAGE in adaptive responses
(11). These studies and previous studies indicate a role of RAGE
in adaptive responses: soluble RAGE (sRAGE) blocked delayed-
type hypersensitivity and inflammatory colitis and a dominant-
negative RAGE rendered an encephalogenic CD4?T cell clone
nonpathogenic (12, 15). We previously showed that transfer of
diabetes by diabetogenic spleen cells could be inhibited with
sRAGE (16). The previous studies generally used sRAGE to block
ligation of RAGE by its ligands, but this approach did not allow us
to directly identify RAGE as the cellular molecule involved in
immune responses, and the specific role of RAGE on T cellular
responses had not been addressed. Moreover, in our previous stud-
ies we found RAGE expressed on T cells, but the relative expres-
sion of RAGE was far greater on other immune cells, which raised
a question, therefore, about the significance of the RAGE expres-
sion on T cells.
Therefore, to determine whether RAGE is involved in adaptive
T cell responses, we studied islet allograft rejection in RAGE?/?
mice and recurrent autoimmune diabetes in NOD mice following
treatment with a specific small molecule RAGE inhibitor. Treat-
ment with a small molecule RAGE inhibitor attenuated syngeneic
islet destruction in an auto- and alloimmune setting, and
RAGE?/? mice showed reduced rates of islet allograft rejection.
Our studies also suggest a previously unrecognized role of RAGE
in T cell activation that may account for the reduced responses.
RAGE?/? T cells show reduced allo-reactive responses in vitro
and a reduced response to CD28 costimulatory signals. Further-
more, RAGE activation appears to be involved in differentiation
of T cell phenotypes since RAGE?/? T cells produce rela-
tively greater amounts of IL-10 in response to TCR activation
and there is increased expression of RAGE on clonal Th1 cells.
These studies show a role of RAGE in modifying activation
pathways during adaptive T cell responses.
Materials and Methods
Mice and reagents
NOD/LtJ, C57BL/6, B10.A.5R, B10.BR, and BALB/c were purchased
from The Jackson Laboratory. RAGE?/? mice, described previously,
were crossed into C57BL/6 for ?10 generations (11, 17). B10.BR-TCR
mice (heterozygous animals carrying the transgenic AND TCR) were bred
in our facility. All the mice were housed under pathogen-free conditions at
our facility. All experiments were approved by the Institutional Animal
Care and Use Committees of Columbia and Yale Universities.
Tech Pharma. The TTP488 or control peptide was administered i.p. daily
at a dose of 100 mcg/d.
Syngeneic and allogeneic islet transplantation
Diabetes was diagnosed when two random glucose levels, measured by a
hand-held meter (Glucometer Elite XL; Bayer), were ?250 mg/dl. Dia-
betic recipient mice received islet grafts within 2 days of the diagnosis.
Islets from nondiabetic donor mice were isolated and transplanted into the
subcapsular space of the right kidneys of recipient mice as previously de-
scribed (16). In each experiment, control mice (that received placebo) or
wild-type (WT) mice received islets from the same pool as TTP488-treated
or RAGE?/? mice. Islet graft function was monitored by serial blood
glucose measurements daily for the first 2 wks after islet transplantation,
followed by every other day thereafter. The immediate graft function after
transplantation was confirmed by finding glucose levels ?200 mg/dl. Graft
loss or recurrent diabetes was determined when blood glucose exceeded
250 mg/dl on two measurements.
In syngeneic islet transplantation experiments, islets were isolated from
young prediabetic NOD/LtJ (6–7 wk old) mice and transplanted into NOD
mice with spontaneous diabetes. The mice were treated with TTP488 or
control peptide as described above. For allogeneic islet transplantation ex-
periments, islets were isolated from WT BALB/c mice (8–10 wk old) and
transplanted into B6 mice that were rendered diabetic by treatment with a
single i.v. injection of streptozotocin (STZ) (Sigma-Aldrich) (200 mg/kg).
T cell activation in vitro
Mixed lymphocyte reactions. To study the responses of T cells to alloan-
tigens, MLRs were performed with purified T cells from RAGE?/? or
WT B6 mice as responders and irradiated (3000 rad) T-depleted BALB/c
splenocytes as stimulators. T cells were isolated using Pan T Cell Iso-
lation Kit (Miltenyi Biotec) following the manufacturer’s methods.
Stimulator cells (T-depleted BALB/c splenocytes) were cultured with
responder cells at a 1:1 ratio (at 1 ? 106/ml) in triplicate cultures in
RPMI 1640 (Invitrogen) medium supplemented with 10% heat-inacti-
vated FCS (Sigma-Aldrich), 25 mM HEPES (Invitrogen), 10 mM L-
glutamine (Invitrogen), 100 ?g/ml penicillin/streptomycin, and 2-ME
(Invitrogen) in 96-well round-bottom plates at 37°C with 5% CO2for 4
days. Supernatants from the cultures were harvested for measurement of
cytokines. One ?Ci of [3H]thymidine was then added to the cultures
which were harvested 18 h later.
T cell activation with anti-CD3 and anti-CD28 mAbs. To study the role of
RAGE in differentiation and activation of T cells, splenic T lymphocytes
were isolated to 99% purity. Splenocytes from WT or RAGE?/? B6 mice
were cultured in complete media on plastic tissue culture plates for 4 h and
the non-adherent cells were then placed on T Cell Isolation columns
(Miltenyi Biotec) twice. The cells purified in this manner were found to be
99% CD3?by flow cytometry. The purified T cells were cultured (1 ?
105/well, in triplicate) for 72 h in U-bottom tissue culture wells coated with
graded concentrations of anti-CD3 mAb with our without anti-CD28 mAb
(mAbs 145–2C11 and Pv-1, both from BD Pharmingen). The culture su-
pernatants were harvested for measurement of cytokines and [3H]thymi-
dine was added for 18 h. The cells were harvested after 72 h in culture, and
the uptake of [3H]thymidine by the cells was measured as described above.
diabetes. A, NOD mice with hyperglycemia received a transplant of syn-
geneic islet grafts and were treated with TTP488 (n ? 8) or control peptide
(n ? 11) as indicated in the Materials and Methods. The percentage of
nondiabetic mice is shown. There was a significant prolongation of the
survival of islets in mice treated with TTP488 (p ? 0.001). B, C57BL/6
mice with STZ-induced diabetes were transplanted with BALB/c islets and
treated with TTP488 (n ? 8) or control protein (n ? 8). There was a
significant prolongation of graft survival with the RAGE antagonist (p ?
Survival of syngeneic or allogeneic islet grafts in mice with
4273The Journal of Immunology
Cytokines IL-2, IL-4, IL-5, IL-10, IFN-?, and TNF-? were measured in the
supernatants of these cultures with Cytokine Multiplex Kit (BioSource
International). The T cell proliferative data are presented as both the
cpm from the cultures as well as the stimulation index of the cells in
response to anti-CD3 mAb with anti-CD28 mAb to determine the ef-
fects of T cell costimulation with anti-CD28 mAb. For this analysis, the
cpm in the presence of anti-CD3 and the indicated doses of anti-CD28
mAb were divided by the cpm in cultures of the same cells in the
absence of anti-CD28 mAb.
In certain MLR’s, TTP488 4 nm was added to cultures of WT or
RAGE?/? mice. The percent inhibition was calculated as: 1 ? (cpm in
cultured with stimulators and TTP488 ? cpm without stimulators/cpm
with stimulators ? cpm without stimulators). In addition, responders from
the MLR’s of WT cells that were cultured with or without TTP488 were
activated with PMA/ionomycin (50 ng/ml and 500 ng/ml). The culture
supernatants were harvested after 16 h and the concentrations of IL-2,
IFN-?, IL-4, and IL-10 were measured with a Luminex cytometer with
cytokine specific beads.
The following mAbs were used for T cell analysis: PE-conjugated rat anti-
mouse CD4, PE-conjugated rat anti-mouse CD8a, PerCP-conjugated rat
anti-mouse CD4, allophycocyanin-conjugated rat anti-mouse CD3e, allo-
phycocyanin-conjugated rat anti-mouse CD25, and PE-conjugated anti-
CD28 (all from BD Pharmingen).
Identification of RAGE transcripts in Th1 and Th2 cells
T cell-depleted APC were prepared by Ab-mediated complement lysis of
B10.A (5R) splenocytes and treated with treated with 50 ?g/ml mitomycin
C (Sigma-Aldrich) before use. CD4 T cells from lymph nodes and spleens
of TCR-transgenic (AND) mice were isolated by immunomagnetic nega-
tive selection using Abs against CD8, CD32/CD16, B220, MHC class II,
and NK cells, followed by incubation with anti-mouse and anti-rat Ig-
coated magnetic beads (Poly-sciences). Purity of the recovered V?11CD4
T cells was 85–95% as determined by staining with anti-CD4 and anti-
V?11 mAbs (18).
Induction of naive T cell differentiation was performed by incubating
mitomycin C-treated APC (1 ? 106/ml) with CD4 T cells (5 ? 105/ml)
from AND TCR transgenic mice in the presence of peptide (moth cyto-
chrome C at 5 ?g/ml l), rIL-2 (25 U/ml), and anti-IFN-? Ab (XMG1.2) for
Th2 or anti-IL4 Ab (11B11) for Th1 (19, 20).
Total RNA was prepared using TRIzol (Invitrogen) and 1–10 ?g RNA
aliquots were further cleaned using the Rneasy Mini kit (Qiagen). RNA
samples were subjected to reverse transcriptase reactions (Superscript II;
Invitrogen) in the presence of oligo(dT) and hexamers. Approximately, 2
ng cDNA template were used for one PCR. All experiments were per-
formed using gene-specific primer pairs and SYBR green I (Molecular
Probes) fluorescence detection in an MX3000P instrument (Stratagene).
The sequence of the gene-specific and reference gene (actin) primers se-
lected is Rage-F: GTCATTCTCACAGGAAGGCAGAAAG; Rage-R:
TTCTTGTTCACAGCAGCACACTTC (228 bp product); Actin-F: GCT
GGTGAAGC (206 bp product). All reactions were done using 40 cycles of
20 s/94°C denaturing, 10 s/61°C annealing and 30 s/74°C extension (17).
For every primer pair, the quality of the PCR amplification was assessed by
visualization of the specific amplification product on a 2.5% agarose gel.
Quantitative PCR of reference gene (?-actin) was use to normalize the
amount of cDNA in each sample tested. For quantitative evaluation of
amplification efficiency, a standard curve of fluorescence vs template
amount was recorded for each primer pair, using decreasing concentrations
of T cell cDNA as template (40; 4; 1; 0.25; 0.0625; 0.015 ng/reaction). At
the end of each quantitative PCR, a melting curve analysis was performed
to analyze product identity after the run and avoid recording false positives.
For relative quantitation, the threshold cycle values of the gene tested and
of the reference gene (?-actin) were exported into the Q-gene software.
Group data are expressed as mean ? SEM. An unpaired Student’s t test
was used to compare average values between groups. Graft survival was
compared between recipient groups by Kaplan-Meier analysis. Cytokine
data was log transformed before performing statistical tests. Values of p ?
0.05 were considered significant.
Diabetes was induced with a single dose of STZ in WT (n ? 14) or
RAGE?/? (n ? 12) B6 mice, which then received a transplant of BALB/c
islets. There was a significant delay in rejection of the allogeneic (BALB/c)
islets in RAGE?/? mice (p ? 0.02).
Survival of allogeneic islets in WT and RAGE?/? mice.
studied in primary mixed lymphocyte reactions after 5 days. The data represent the mean values of four separate experiments. There was a significant
reduction (??, p ? 0.01) in the proliferative response by RAGE?/? T cells when the actual cpm or the stimulation index (cpm with stimulators/cpm without
stimulators) were compared (31.0 ? 1.6 vs 9.0 ? 0.67, p ? 0.001). B, The kinetics of the responses of a single experiment (representative of two) comparing
WT and RAGE?/? cells are shown.
Reduced MLR responses in RAGE?/? mice. A, The proliferative responses of WT or RAGE?/? T cells to BALB/c stimulators were
4274RAGE MODULATION OF T CELL ACTIVATION
Blockade of RAGE by treatment with the small molecule RAGE
antagonist (TTP488) prolongs islet auto and allograft survival
We used two model systems to study the role of RAGE in auto- or
alloimmune adaptive immune responses. We first tested the effects
of a small molecule RAGE antagonist (TTP488) on the survival of
syngeneic islet grafts transplanted into NOD mice with autoim-
mune diabetes (Fig. 1A) or BALB/c allografts transplanted into B6
mice with chemically induced diabetes (Fig. 1B). When diabetic
NOD mice were transplanted with syngeneic islets and treated
daily with TTP488, the median survival time increased from
22.7 ? 0.56 days (n ? 11) to 41.1 ? 1.6 days (n ? 8; p ? 0.001)
(Fig. 1A). In a similar manner, treatment with TTP488 increased
the median survival of BALB/c allografts from 8.8 ? 0.3 days
(n ? 8) to 14.1 ? 0.5 days (n ? 8; p ? 0.001) (Fig. 1B).
We next studied the survival of allogeneic (BALB/c) islets that
were transplanted into WT or RAGE?/? B6 mice with diabetes
induced with a single dose of STZ (Fig. 2). There was a significant
prolongation of allograft survival of the allogeneic islets in
RAGE?/? mice (median survival time ? 15 days, n ? 12) com-
pared with RAGE?/? recipients (median survival time ? 10.5
days, n ? 14, p ? 0.02). The reduced responses were also not the
result of deficiency of T cells in the RAGE?/? mice because the
percentage of CD4?and CD8?T cells was similar in RAGE?/?
(19.3 ? 0.68 and 13.3 ? 1.1%, n ? 5) and WT (18.8 ? 0.9 and
10.9 ? 1.2%, n ? 4) splenocytes respectively.
Proliferative responses in vitro are reduced in the absence of
To determine the basis for the reduced responses in vivo, we com-
pared the responses of WT and RAGE?/? B6 splenocytes to ir-
radiated BALB/c stimulator cells in primary MLRs (WT or
RAGE?/? B6 vs BALB/c) (Fig. 3A). The proliferative responses
were significantly reduced in the RAGE?/? B6 mice compared
with WT B6 mice in response to T-depleted BALB/c stimulator
cells (n ? 4; p ? 0.01 vs WT). The reduced responses that we had
seen on the 5thday of the primary MLR were not simply a differ-
ence in the kinetics of the responses of the RAGE?/? T cells. The
RAGE?/? T cells showed a lower proliferative response after 3
days in culture, and the difference with the WT cells was persistent
even after 10 days when the responses of both the WT and
RAGE?/? T cells were waning (Fig. 3B).
We also studied the effects of the small molecule inhibitor of
RAGE, TTP488, on the proliferative responses of purified WT T
cells in a MLR. The responses to BALB/c stimulators were re-
duced when the TTP488 was added to cultures of WT but not
RAGE?/? T cells (Fig. 4).
cells with TTP488 at the indicated concentrations or a similar dilution of DMSO. The incorporation of [3H]thymidine was measured after 5 days. The data
are expressed as the stimulation index (cpm at the indicated concentration of TTP488 or DMSO/cpm of the responder cells only). The cpm of the responder
cells were WT: 11884, RAGE?/?: 12895). B, The percent inhibition (1?cpm with TTP/cpm with equivalent dilution of DMSO *100) of the WT and
RAGE?/? T cells was calculated. The data shown are from a single experiment representative of three independent experiments.
Inhibition of WT but not RAGE?/? T cells by TTP488. A, Purified WT or RAGE?/? T cells were added to irradiated BALB/c stimulator
(open symbols) mice and added to tissue culture plates coated with the indicated concentrations of anti-CD3 mAb without or with anti-CD28 mAb at the
indicated concentrations. B, The effects of CD28 costimulation were analyzed by dividing the cpm from cultures of T cells from WT or RAGE?/? T cells
in the presence of the indicated concentrations of anti-CD28 mAb by the cpm from cultures without CD28. There was markedly reduced response to CD28
costimulation by the RAGE?/? T cells. The data shown is from a single experiment that is representative of three independent studies. C, To exclude
differences in CD28 expression as the basis for the findings in B, the expression of CD28 was analyzed on WT (dashed line) and RAGE?/? (solid line)
CD4?T cells by flow cytometry (the isotype control is shaded).
Diminished responsiveness to CD28 costimulation by RAGE?/? T cells. A, T cells were isolated from B6 (filled symbols) or RAGE?/?
4275The Journal of Immunology
The reduced proliferative responses we observed in the MLR’s
could be due to differences in RAGE?/? T or non-T cells. There-
fore, we purified splenic T cells and tested their proliferative and
cytokine responses to anti-CD3 and anti-CD28 mAbs (Fig. 5).
Overall, the responses of the RAGE?/? T cells in response to
activation with maximal doses of anti-CD3 (3 ?g/ml) and anti-
CD28 (1 ?g/ml) mAbs were only marginally different in
RAGE?/? and WT mice (Fig. 5A). However, the proliferative
responses of RAGE?/? T cells to anti-CD3 mAb in the absence
of anti-CD28 mAb were generally higher (31%) when compared
with WT T cells. We found a striking difference in the T cell
responses to costimulation with anti-CD28 mAb (Fig. 5B). There
was little enhancement of the proliferative responses of
RAGE?/? T cells by costimulation with anti-CD28 mAb above
proliferation that was induced with anti-CD3 mAb alone. At a
concentration of 0.3 ?g/ml of immobilized anti-CD3 mAb, anti-
CD28 mAb (@1 ?g/ml) increased the proliferative responses by
5.71 ? 0.52-fold over the responses to anti-CD3 mAb alone in
WT, but only 1.50 ? 0.08-fold in RAGE?/? mice (n ? 3 inde-
pendent experiments, p ? 0.001). Likewise, the stimulation index
was 1.25 ? 0.08 in WT cells stimulated with anti-CD3 @ 0.03
?g/ml with anti-CD28 @ 1 ?g/ml but the RAGE?/? T cells the
responses were not higher than those without anti-CD28 mAb
(p ? 0.03). These findings indicate that RAGE?/? T cells are
unresponsive to costimulation with anti-CD28 mAb. The unre-
sponsiveness to anti-CD28 mAb was not due to reduced expres-
sion of CD28 on RAGE?/? T cells on T cells. The expression of
CD28 on CD4?cells was similar in WT and RAGE?/? spleno-
cytes (Fig. 5C).
RAGE is associated with differentiation of T cells
To determine the mechanisms by which RAGE deficiency affects
the development of T cell effector responses, we measured the
levels of cytokines in the supernatants of purified T cells from WT
or RAGE?/? mice that were activated with anti-CD3 and anti-
CD28 mAbs (Fig. 6A). The levels of IL-10 and IL-5 (p ? 0.05)
and TNF-? (p ? 0.01) were significantly increased in the super-
natants from T cells from RAGE?/? compared with WT mice.
The levels of IL-4 were also higher on average in the cultures with
RAGE ?/? cells but the difference with the WT cells was not
statistically significant (p ? 0.24).
We also studied the cytokine responses of responders of a MLR
in the presence or absence of TTP488. WT C57BL/6 T cells were
stimulated with T-depleted BALB/c splenocytes in the presence or
absence of TTP488 and the production of cytokines by the re-
sponders after activation with PMA/ionomycin was measured. The
levels of IFN-? were significantly lower in the presence of the
TTP488 (n ? 3 mice/group, p ? 0.01) (Fig. 6B). There was also
a reduction in the concentration of IL-2 and an increase in IL-4 in
the supernatants from the cells cultured in the presence of TTP488
compared with those cultured in its absence but the differences did
not reach statistical significance.
These findings suggested that RAGE may be associated with
differentiation of Th1 cells. We therefore studied RAGE expres-
sion in clonal T cells differentiated under conditions leading to Th1
or Th2 phenotypes. We studied the expression of RAGE by real-
time PCR in AND transgenic T cells that were activated with Ag
(mouse cytochrome C) and mitomycin C-treated APC’s under con-
ditions leading to Th1 or Th2 phenotypes (Fig. 7). The expression
of RAGE mRNA was increased on cells that were cultured with
peptide, IL-2, IL-12, and anti-IL-4 mAb compared with the same
activated with anti-CD3 and anti-CD28 mAbs and the culture supernatants were harvested for measurement of cytokine concentrations (mean ? SEM, n ?
4/group). There was a significant increase in the concentration of TNF-?, IL-5, and IL-10 in the cultures from RAGE?/? mice compared with WT mice
(?, p ? 0.05; ??, p ? 0.01). B, Primary MLR’s with purified WT T cells were cultured in the presence of TTP488 or the excipient DMSO as described.
The responders were harvested after 5 days and activated with PMA/ionomycin overnight. The concentrations in the supernatants were measured (n ?
3/group). There was a significant reduction in the production of IFN-? in responders that had been cultured with TTP488.
Cytokine concentrations in the supernatants of activated T cells. A, Purified RAGE?/? (solid bars) and WT (stippled bars) T cells were
cultured under Th1 and Th2 differentiating conditions. RAGE transcripts
were measured by real-time PCR and compared with actin in naive AND
transgenic T cells cultured with mitomycin C-treated APC’s and Ag under
Th1 (fine hatched bar) and Th2 (coarse hatched bar) skewing conditions as
described in Materials and Methods. The data are from a single experiment
that is representative of two independent experiments.
Expression of RAGE transcripts in AND transgenic T cells
4276RAGE MODULATION OF T CELL ACTIVATION
cells that were cultured with peptide, IL-4, and anti-IFN-? mAb.
The RAGE mRNA expression was greatest on T cells during pri-
mary cell activation. RAGE mRNA expression was negligible in
cells during secondary activation with Ag (data not shown).
We have shown a role for RAGE in adaptive immune responses to
allo- and autoantigens. A small molecule inhibitor of RAGE re-
duced the destruction of syngeneic islets that were transplanted
into NOD mice with spontaneous diabetes, and RAGE-deficient
B6 mice rejected allogeneic BALB/c islet grafts at a reduced rate
compared with WT B6 mice. RAGE?/? T cells were not anergic.
Their response in a MLR was reduced, but they showed the same
proliferative responses to activation with full doses of anti-CD3
and anti-CD28 mAbs as WT T cells. However, there were mark-
edly reduced responses of the RAGE?/? T cells to CD28 co-
stimulatory signals with anti-CD3 mAb that was not explained by
differences in the expression of CD28 on the surfaces of
RAGE?/? T cells. In response to anti-CD3 and anti-CD28 mAbs,
RAGE?/? T cells produce relatively greater amounts of IL-10,
IL-5, and TNF-? in response to TCR ligation, which is similar to
our findings previously in which there was higher levels of IL-10
in the pancreata of NOD/SCID mice that were treated with sRAGE
and showed reduced development of diabetes after receiving dia-
betogenic splenocytes (16). The small molecule RAGE inhibitor
TTP488 also inhibited responses of T cells in a MLR, and inhibi-
tion was not seen with RAGE?/? cells. Moreover, the responders
from MLR’s that had been cultured with TTP488 secreted reduced
levels of IFN-? in response to activation with PMA/ionomycin
consistent with a role of RAGE in differentiation of Th1 cells. The
discrepancies between the cytokine pattern in the RAGE?/? T
cells activated with anti-CD3 and anti-CD28 and responders in a
MLR in the presence of the RAGE inhibitor are likely due to the
different culture systems used to activate the cells – one involving
the direct activation of naive cells and the other the secondary
activation after differentiation in the presence of the RAGE inhib-
itor. Indeed, in a previous report, we found that OTII cells on a
RAGE?/? background produced reduced levels of IFN-? and
IL-2 when activated with PMA/ionomycin after priming with pep-
tide pulsed DC’s (21). Finally, our finding of increased RAGE
expression on activated clonal populations of cells under condi-
tions that promote development of Th1 compared with Th2 cells is
consistent with the role of this receptor in early events that lead to
development of T cell phenotypes, specifically polarization toward
a Th1 phenotype.
A number of recent studies have highlighted the role of RAGE
on the activation and maturation of DCs and other APCs, but few
have addressed the role of RAGE on T cells. Exposure of neutro-
phils, monocytes, or macrophages to the RAGE ligand HMGB1
enhances expression of proinflammatory cytokines by these cells,
but RAGE was shown to play only a minor role in activation of
these cells by HMGB1 (22). Likewise, these same investigators
found that HMGB1 interacts with TLRs 2 and 4 but they were
unable to show binding of HMGB1 to RAGE by fluorescence res-
onance energy transfer and immunoprecipitation. However, Tian
et al. (14) recently reported that DNA-containing immune com-
plexes in systemic lupus erythematosus bind HMGB1 and RAGE
activating TLR9 on plasma DCs causing them to secrete IFN?. A
role for RAGE in T cell activation was suggested by previous
studies by Yan et al. (15) in which a dominant-negative RAGE,
expressed on CD4?T cells, blocked induction of experimental
allergic encephalomyelitis. We previously showed that sRAGE at-
tenuated the adoptive transfer of diabetes by diabetogenic T cells
into NOD/SCID recipients (16). However, sRAGE did not prevent
diabetes induced with a clonal population of CD4?T cells, which
raised the possibility that the role of RAGE was indirect, rather
than directly on T cells, or alternatively that RAGE played a role
in earlier stages of T cell differentiation. These new studies, how-
ever, establish a role of RAGE signaling on the early stages of T
cell differentiation. Moreover, in other studies, we have also found
that RAGE?/? OVA-reactive OTII cells show diminished prolif-
erative responses to Ag when transferred into WT recipients (21).
Studies in RAGE null mice have indicated that RAGE is in-
volved in perpetuation of cellular responses (9). This conclusion is
consistent with our previous studies in autoimmune diabetes and
the present studies of recurrent diabetes in diabetic NOD recipients
of syngeneic islet grafts in which the responses of previously ac-
tivated effector cells were inhibited. However, in the current stud-
ies, involving responses to allografts, alloantigens, and TCR sig-
naling, the differences in cytokine production and RAGE
expression on T cell phenotypes were in primary responses. There
was negligible expression of RAGE on T cells studied after sec-
ondary activation (not shown).
CD28 expression is required for the formation of the mature
immunologic synapse – central supramolecular activation cluster
localization of PKC ? that is found in CD28?T cells is absent in
CD28?T cells (23). CD28 engagement leads to the redistribution
and clustering of membrane and intracellular kinase-rich raft mi-
crodomains at the site of TCR engagements. This results in higher
rates and more stable tyrosine phosphorylation of several sub-
strates and higher consumption of Lck (24). These important
events in TCR signaling may be absent in the RAGE?/? mice,
but further studies of TCR signaling will help to define the basis
for the loss of CD28 responsiveness in the absence of RAGE.
RAGE ligands, such as HMGB1, have previously been shown to
affect differentiation of DCs and HMGB1 released by DCs is
needed for clonal expansion, survival, and functional polarization
of naive T cells (25). Release of HMGB1 by human DCs was
necessary for up-regulation of CD80 and CD86, ligands for CD28.
RAGE was required for this effect of HMGB1 on DCs, suggesting
that the innate HMGB1/RAGE signaling pathway results in adap-
tive immune responses (25).
Our findings differ from those of Liliensiek et al. (11) who found
that deletion of RAGE provides protection from the lethal effects
of septic shock caused by cecal ligation and puncture, but their
studies did not support a role of RAGE in adaptive immune re-
sponses. However, these previous studies were done in mice on a
mixed genetic background, whereas our studies were done on mice
in which the RAGE mutation was backcrossed to B6 for more than
10 generations so that differences between WT and KO mice could
be identified more clearly (11). Consistent with our findings,
Dumitriu et al. (25) found that HMGB1 signaling through RAGE
was needed for clonal expansion, survival, and functional polar-
ization of naive T cells. These findings also support a direct effect
of RAGE on T cell activation and differentiation rather than just T
cell migration or localization as has been suggested in other model
systems (15, 17). Likewise, we have also found that clonal popu-
lations of RAGE?/? T cells show reduced responses when adop-
tively transferred into WT recipients, indicating that the role of
RAGE on T cell activation is primary (21).
Our findings suggest a novel testable hypothesis for the rela-
tionship between metabolic control of Type 1 diabetes and the
immunologic progression of disease in patients. Studies from the
Diabetes Control and Complications Study showed that individu-
als with new onset Type 1 diabetes, who maintained intensive
glycemic control, had reduced loss of C-peptide responses com-
pared with individuals with conventional metabolic treatment (26).
Reduced availability of RAGE ligands with strict glycemic control
4277 The Journal of Immunology
would be expected to modulate adaptive responses that may be Download full-text
responsible for destruction of ? cells. Indeed, the results in these
experimental settings indicate that RAGE activation can affect
even previously primed responses.
In summary, we have demonstrated a role of RAGE in the ac-
tivation of adaptive immune responses to auto- and alloantigens.
RAGE expression on T cells affects the early events its absence is
manifest by reduced activation of T cells to alloantigens in vivo
and in vitro. RAGE is involved in the differentiation of T cells
along a Th1 phenotype and RAGE mRNA is more abundant in
Th1 compared with Th2 cells. The reduced rate of recurrent dia-
betes and prolonged allograft survival by RAGE blockade and in
RAGE-deficient mice suggests that RAGE may be an important
new target for therapeutic strategies to prevent adaptive immune
responses. The activation of RAGE by ligands, such as HMGB1,
released from dying cells implies a broad role for this pathway in
immune homeostasis. Moreover, the increased availability of
RAGE ligands during hyperglycemia suggests a hypothesis for the
amplification of autoimmune responses in Type 1 diabetes that
may be tested in future clinical settings.
The authors have no financial conflict of interest.
1. Zanin-Zhorov, A., R. Bruck, G. Tal, S. Oren, H. Aeed, R. Hershkoviz,
I. R. Cohen, and O. Lider. 2005. Heat shock protein 60 inhibits Th1-mediated
hepatitis model via innate regulation of Th1/Th2 transcription factors and cyto-
kines. J. Immunol. 174: 3227–3236.
2. Alyanakian, M. A., F. Grela, A. Aumeunier, C. Chiavaroli, C. Gouarin, E. Bardel,
G. Normier, L. Chatenoud, N. Thieblemont, and J. F. Bach. 2006. Transforming
growth factor-? and natural killer T-cells are involved in the protective effect of
a bacterial extract on type 1 diabetes. Diabetes 55: 179–185.
3. Zipris, D., E. Lien, A. Nair, J. X. Xie, D. L. Greiner, J. P. Mordes, and
A. A. Rossini. 2007. TLR9-signaling pathways are involved in Kilham rat virus-
induced autoimmune diabetes in the biobreeding diabetes-resistant rat. J. Immu-
nol. 178: 693–701.
4. Thomas, H. E., W. Irawaty, R. Darwiche, T. C. Brodnicki, P. Santamaria,
J. Allison, and T. W. Kay. 2004. IL-1 receptor deficiency slows progression to
diabetes in the NOD mouse. Diabetes 53: 113–121.
5. Sandberg, J. O., D. L. Eizirik, and S. Sandler. 1997. IL-1 receptor antagonist
inhibits recurrence of disease after syngeneic pancreatic islet transplantation to
spontaneously diabetic non-obese diabetic (NOD) mice. Clin. Exp. Immunol. 108:
6. Wilson, S. B., S. C. Kent, K. T. Patton, T. Orban, R. A. Jackson, M. Exley,
S. Porcelli, D. A. Schatz, M. A. Atkinson, S. P. Balk, et al. 1998. Extreme Th1
bias of invariant V?24J?Q T cells in type 1 diabetes. Nature 391: 177–181.
7. Wang, B., Y. B. Geng, and C. R. Wang. 2001. CD1-restricted NK T cells protect
nonobese diabetic mice from developing diabetes. J. Exp. Med. 194: 313–320.
8. Ogasawara, K., J. A. Hamerman, L. R. Ehrlich, H. Bour-Jordan, P. Santamaria,
J. A. Bluestone, and L. L. Lanier. 2004. NKG2D blockade prevents autoimmune
diabetes in NOD mice. Immunity 20: 757–767.
9. Schmidt, A. M., S. D. Yan, S. F. Yan, and D. M. Stern. 2001. The multiligand
receptor RAGE as a progression factor amplifying immune and inflammatory
responses. J. Clin. Invest. 108: 949–955.
10. Schmidt, A. M., S. D. Yan, S. F. Yan, and D. M. Stern. 2000. The biology of the
receptor for advanced glycation end products and its ligands. Biochim. Biophys.
Acta 1498: 99–111.
11. Liliensiek, B., M. A. Weigand, A. Bierhaus, W. Nicklas, M. Kasper, S. Hofer,
J. Plachky, H. J. Grone, F. C. Kurschus, A. M. Schmidt, et al. 2004. Receptor for
advanced glycation end products (RAGE) regulates sepsis but not the adaptive
immune response. J. Clin. Invest. 113: 1641–1650.
12. Hofmann, M. A., S. Drury, C. Fu, W. Qu, A. Taguchi, Y. Lu, C. Avila,
N. Kambham, A. Bierhaus, P. Nawroth, et al. 1999. RAGE mediates a novel
proinflammatory axis: a central cell surface receptor for S100/calgranulin
polypeptides. Cell 97: 889–901.
13. Kokkola, R., A. Andersson, G. Mullins, T. Ostberg, C. J. Treutiger, B. Arnold,
P. Nawroth, U. Andersson, R. A. Harris, and H. E. Harris. 2005. RAGE is the
major receptor for the proinflammatory activity of HMGB1 in rodent macro-
phages. Scand. J. Immunol. 61: 1–9.
14. Tian, J., A. M. Avalos, S. Y. Mao, B. Chen, K. Senthil, H. Wu, P. Parroche,
S. Drabic, D. Golenbock, C. Sirois, et al. 2007. Toll-like receptor 9-dependent
activation by DNA-containing immune complexes is mediated by HMGB1 and
RAGE. Nat. Immunol. 8: 487–496.
15. Yan, S. S., Z. Y. Wu, H. P. Zhang, G. Furtado, X. Chen, S. F. Yan,
A. M. Schmidt, C. Brown, A. Stern, J. LaFaille, et al. 2003. Suppression of
experimental autoimmune encephalomyelitis by selective blockade of encepha-
litogenic T-cell infiltration of the central nervous system. Nat. Med. 9: 287–293.
16. Chen, Y., S. S. Yan, J. Colgan, H. P. Zhang, J. Luban, A. M. Schmidt, D. Stern,
and K. C. Herold. 2004. Blockade of late stages of autoimmune diabetes by
inhibition of the receptor for advanced glycation end products. J. Immunol. 173:
17. Chavakis, T., A. Bierhaus, N. Al-Fakhri, D. Schneider, S. Witte, T. Linn,
M. Nagashima, J. Morser, B. Arnold, K. T. Preissner, et al. 2003. The pattern
recognition receptor (RAGE) is a counterreceptor for leukocyte integrins: a novel
pathway for inflammatory cell recruitment. J. Exp. Med. 198: 1507–1515.
18. Kaye, J., M. L. Hsu, M. E. Sauron, S. C. Jameson, N. R. Gascoigne, and
S. M. Hedrick. 1989. Selective development of CD4?T cells in transgenic mice
expressing a class II MHC-restricted antigen receptor. Nature 341: 746–749.
19. Levin, D., S. Constant, T. Pasqualini, R. Flavell, and K. Bottomly. 1993. Role of
dendritic cells in the priming of CD4?T lymphocytes to peptide antigen in vivo.
J. Immunol. 151: 6742–6750.
20. Balamuth, F., J. L. Brogdon, and K. Bottomly. 2004. CD4 raft association and
signaling regulate molecular clustering at the immunological synapse site. J. Im-
munol. 172: 5887–5892.
21. Moser, B., D. D. Desai, M. P. Downie, Y. Chen, S. F. Yan, K. Herold,
A. M. Schmidt, and R. Clynes. 2007. Receptor for advanced glycation end prod-
ucts expression on T cells contributes to antigen-specific cellular expansion in
vivo. J. Immunol. 179: 8051–8058.
22. Park, J. S., D. Svetkauskaite, Q. He, J. Y. Kim, D. Strassheim, A. Ishizaka, and
E. Abraham. 2004. Involvement of toll-like receptors 2 and 4 in cellular activa-
tion by high mobility group box 1 protein. J. Biol. Chem. 279: 7370–7377.
23. Huang, J., P. F. Lo, T. Zal, N. R. Gascoigne, B. A. Smith, S. D. Levin, and
H. M. Grey. 2002. CD28 plays a critical role in the segregation of PKC ? within
the immunologic synapse. Proc. Natl. Acad. Sci. USA 99: 9369–9373.
24. Viola, A., S. Schroeder, Y. Sakakibara, and A. Lanzavecchia. 1999. T lympho-
cyte costimulation mediated by reorganization of membrane microdomains. Sci-
ence 283: 680–682.
25. Dumitriu, I. E., P. Baruah, B. Valentinis, R. E. Voll, M. Herrmann, P. P. Nawroth,
B. Arnold, M. E. Bianchi, A. A. Manfredi, and P. Rovere-Querini. 2005. Release
of high mobility group box 1 by dendritic cells controls T cell activation via the
receptor for advanced glycation end products. J. Immunol. 174: 7506–7515.
26. The Diabetes Control and Complications Trial Research Group. 1998. Effect of
intensive therapy on residual ?-cell function in patients with type 1 diabetes in the
diabetes control and complications trial: a randomized, controlled trial. Ann. In-
tern. Med. 128: 517–523.
4278 RAGE MODULATION OF T CELL ACTIVATION