| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Immunogenetics and Transplantation Laboratory, Department of Surgery, University of California, San Francisco, CA 94143;
Schepens Eye Research Institute, Harvard Medical School, Boston, MA 02114; and
Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, GA 30322
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Some recent studies suggest that p53-based immunotherapies could be designed to treat different types of cancers. Indeed, different approaches using p53 immunization including vaccination with p53-transfected virus and in vivo transfer of p53-specific CTL have been successfully used to confer some tumor protection in mice (6, 7). In addition, p53-specific CTL generated to a particular tumor were shown to lyse a variety of cancer cells, an observation indicating the existence of p53 antigenic determinants shared by tumors of different origins (8, 9). Collectively, these studies demonstrated that p53 protein does contain antigenic peptides that can activate T cells and enhance antitumor immunity when administered under appropriate conditions (10, 11, 12). However, the actual mechanisms underlying the antigenicity and immunogenicity of p53 in untreated individuals with developing tumors remains unclear. As we gain insights into this question, we may design new methods of therapeutic intervention to maneuver the anti-p53 T cell response toward effective tumor rejection.
It is possible that altered structure and expression of mut p53 in tumors could result in the presentation of some p53 determinants to T cells, thereby rendering cancer cells visible to the immune system (13, 14, 15). Although recent evidence indicates that p53 is immunogenic during tumorigenesis, the precise nature of this immune response needs to be investigated. Most importantly, the reasons why this response is apparently poorly effective in eliminating tumors in nonvaccinated individuals is unknown, an issue that is essential to the design of future p53-based immunotherapy for cancer. One important question is whether wild-type (wt) and mut p53 protein follow the same rules of processing and presentation in normal and cancer cells. It is clear that wt p53 represents a self-protein constitutively expressed in all cells of our body. In normal adult cells, newly synthesized wt p53 is present at low concentration, and its expression is confined to the nucleus. At first glance, these features of wt p53 expression suggest that p53 self-peptides are not presented in MHC class II context at the surface of hematopoeitic cells. Conversely, in neonates, p53 is expressed in large quantities in the developing thymus, a property which may account for presentation of p53 peptides by MHC and for their involvment in positive and negative selection of self-p53-reactive T cells (16). In support of this view, recent studies by Sherman and colleagues (17) comparing anti-p53 CTL responses in normal and p53 knock-out mice have demonstrated that some wt p53 peptides are presented in MHC class I context during ontogeny and mediate deletion/inactivation of corresponding p53-reactive CD8+ T cells. Alternatively, other wt p53 determinants, despite their ability to bind MHC class I molecules, failed to mediate negative selection. Interestingly, vaccination with these p53 self-peptides could promote antitumor immunity without inducing detectable signs of autoimmunity. Therefore, it appears that abundant opportunity exists for endogenous processing and presentation of some self-p53 peptides by MHC class I molecules to CD8+ T cells. However, whether self-p53 peptides also influence shaping of CD4+ T cell repertoire and whether certain undeleted anti-self-p53-reactive T helper cells contribute to antitumor immunity have not been investigated previously.
Many studies have concluded that syngeneic tumors placed in normal mice are nonimmunogenic because they promote T cell anergy, thereby ensuring tumor immune escape. The majority of these studies have examined CD8+ CTL responses directed to a variety of tumor-specific Ags (18, 19). It has been reasoned that most tumors express exclusively MHC class I-presenting molecules necessary for CD8+ T cell recognition. The focus of tumor immunologists on antitumor CTL responses has been further strengthened by the success of vaccination procedures involving adoptive transfer of activated antitumor CTL (18). However, it has become increasingly evident that efficient and long-lasting vaccination against tumors requires activation of tumor-specific CD4+ Th cells. First, effective sensitization of antitumor CTL responses is achieved only when Ag is delivered with adjuvant or in the form of Ag-pulsed dendritic cells (20, 21, 22). Even more compelling is the observation that antitumor vaccines fail to confer protection in CD4-knock-out mice or after Ab-mediated depletion of CD4+ T cells in normal mice. Moreover, it has been reported that eradication of murine leukemias in tumor-bearing hosts can be accomplished by adoptive transfer of activated CD4+ T cell clones (22, 23). Finally, vaccination with peptides containing tumor-derived CD4+ T helper determinants has been shown to protect against subsequent tumor challenge (24). Taken together, these studies demonstrate that, in addition to activation of CTL, induction of CD4+ T cell response to tumor Ags must be achieved to confer long-term and effective antitumor immunity.
Two observations support the view that CD4+ T cell responses directed to p53 are elicited during the process of tumorigenesis in vivo. First, anti-p53 Abs displaying IgG isotypes have been detected in the blood of cancer patients (25, 26, 27, 28, 29). Because B cell activation and differentiation are known to require help from activated Ag-specific CD4+ T cells, it is likely that some anti-p53 CD4+ Th cells had been stimulated during tumor development. Second, in patients with breast tumors expressing mut p53, PBMC have been shown to proliferate in vitro in the presence of p53 protein (27). These studies provide indirect evidence suggesting that some anti-p53 CD4+ T cells become activated during cancer. However, the identity of the p53 determinants recognized by these CD4+ T cells as well as the functional properties of these T cells are still unknown.
In this article, we analyzed MHC class II-restricted CD4+ T cell response to wt and mut mouse p53 in healthy and tumor-bearing mice. We demonstrated that anti-p53 T cells specific to certain wt p53 determinants were present in the periphery of adult immune system and could be specifically activated after p53 peptide immunization. Most importantly, we showed that mice inoculated with syngeneic J774 metastatic sarcomas mounted potent CD4+ T cell responses to p53. This response was mediated by T cells recognizing the mutated portion of p53 and by T cells directed to formerly cryptic self-p53 determinants. Interestingly, we observed that anti-p53 Th response was directed toward distinct p53 peptides depending upon the stage of tumorigenesis. The implications of these findings in understanding the basis for immunogenicity of tumor Ags in vivo during cancer are discussed.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
BALB/c (H-2d) and BALB/c J-Trp 53tm/tyj (p53 KO) mice were purchased from The Jackson Laboratory (Bar Harbor, ME) and were housed at University of California (San Francisco, CA) animal facilities. Mice of either sex were used at 6–8 wk of age. The mice were immunized in their hind footpads with 50 µg of the p53 peptide emulsified in CFA (Difco, Detroit, MI). The care of all animals involved in this study was in accordance with institutional guidelines.
Peptides
Peptides were synthesized utilizing Fmoc chemistry by Research
Genetics (Huntsville, AL) and purified by HPLC (purity, >95%). The
amino acid sequences of the peptides were as follows: p53.167–181,
TEVVRRCPHHERCSD; p53.277–291, RDRREEEENFRKKEV; p53.316–333,
KKKPLDGEYFTLKIRGRK; p53.365–384, YLKTKKGQSTSRHKKTMVKKV; wt
p53.225–239, EYTTIHYKYMCNSSC; mut p53.225–239, EYTTIHHKYMCNSSC; and
repressor peptide, P12–26, LEDARRLKAIYEKKK.
Peptide binding to MHC class II molecules
p53 peptides were tested for their ability to competitively inhibit the binding of known MHC class II binding peptides to their specific MHC class II molecules, Ad and Ed. Briefly, each p53 peptide was incubated at serial concentrations (0.01–500 µM) in the presence of a given purified MHC class II molecule (100 nM) and a fixed concentration (0.5 µM) of a biotinylated version of a known MHC class II-binding peptide. For Ad binding, biotinylated sperm whale myoglobine-derived peptide Myo106–118 was used as a reporter peptide. In Ed-binding assays, hen eggwhite lysozyme HEL104–120 peptide was used. Both reporter and competitor peptides were incubated with MHC class II for 18 h at pH 5.0. In each experiment, nonbiotinylated reporter peptides were used as positive control competitors. In all experiments, microplates were coated with anti-Ad (MKD6) or anti-Ed (14-4-4S) mAbs and incubated with the solution containing MHC class II and peptides. The amount of MHC-bound biotinylated peptides was then determined by europium fluorescence as decribed elsewhere (30).
Inoculation of tumors
The J774 BALB/c-derived monocyte-like sarcoma cell line used in this study was obtained from the American Type Culture Collection (Manassas, VA) and was maintained in modified DMEM supplemented with 2 mM glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin (Life Technologies, Grand Island, NY), and 10% FCS (Gemini Bioproducts, Calabasas, CA).
BALB/c mice were inoculated i.p. with 106 live tumor cells at days 0, 7, and 14. The development of solid tumors in the abdominal area was observed daily by palpation. Ascites became visible within 21–25 days after initial inoculation and grew progressively. Tumor-bearing mice were sacrificed according to institutional animal care committee guidelines.
Lymph node and spleen T cell proliferation assays
Popliteal lymph node and spleen cells were obtained 10–21 days after tumor or peptide immunization and were used in Ag-induced proliferation assays. Suspensions of 5 x 105 lymph node and 106 spleen cells were prepared and washed in serum-free AIM-V medium (Life Technologies). The cells were then cultured in 0.2 ml of medium alone, in the presence of the serial dilutions of p53 peptides, or with a control peptide (P12–26) in 96-well culture dishes for 4 days. Ag-induced proliferation was assessed by determining the incorporation of 1 µCi [3H]thymidine during the last 18 h of culture.
Cytokine measurement
Lymph node and spleen cells were harvested from either
peptide-immunized or tumor-bearing mice. Suspensions of 5 x
105 lymph node and 106
spleen cells were plated in 96-well dishes either in AIM-V medium alone
or in the presence of p53 peptides. Forty-eight hours later, the
concentration of IFN-
and IL-5 in culture supernatants was
determined using an ELISA assay. Briefly, ELISA plates (Corning Glass,
Corning, NY) were coated overnight with either rat anti-mouse
IFN- (R4-6A2)-capturing or rat anti-mouse IL-5
(TFRK-4)-capturing mAbs. Supernatants from cell cultures were added to
the wells and incubated overnight at 4°C. For detection, biotinylated
rat anti-mouse IFN- mAb (XMG 1.2) or rat anti-mouse IL-5 mAb
(TFRK-5) was used before incubation with streptavidin D HRP (Vector,
Burlingame, CA). All mAbs were purchased from PharMingen (San Diego,
CA). Peroxidase activity was revealed with
2,2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) substrate (Sigma,
St. Louis, MO) containing
H2O2. Mouse recombinant
IFN- and IL-5 (PharMingen) were used as standards. Absorbance was
measured at 405 nm.
ELISA spot analyses were performed with ELISA spot plates (CTL,
Cleveland, OH) coated with the capture Ab in sterile PBS overnight.
R46A2 and TRFK4 mAb were used to capture IFN- and IL-5, respectively
(PharMingen). The plates were then blocked for 1 h with sterile
PBS containing 1% BSA and were washed three times with sterile PBS. A
total of 5 x 105 purified T cells in 200
µl of AIM-V medium were then placed in each well with or without p53
peptides in the presence of irradiated (2000 rad) BALB/c splenocytes
(as APCs) and were cultured for 48 h at 37°C in 5%
CO2. After washing, biotinylated
anti-lymphokine detection Abs were added overnight. XMG1.2 and
TRFK5 biotinylated mAbs were used for IFN- and IL-5, respectively (5
µg/ml, PharMingen). Streptavidin-HRP (Vector; 1:2000 in PBS 0.025%
Tween 20 for 2 h at room temperature) was then added to each well.
Finally, the plates were developed using 800 µl AEC (10 mg dissolved
in 1 ml dimethyl formamide; Pierce, Rockford, IL) mixed in 24 ml 0.1 M
sodium acetate (pH 5.0) containing 12 µl
H202. The resulting spots
were counted on a computer-assisted ELISA spot image analyzer (T Spot
Image Analyzer; CTL).
The phenotype of responding T cells and their MHC restriction were determined using Ab-mediated blocking experiments. The following mAbs were added to the T cell cultures at serial dilutions: anti-CD4 (GK1.5), anti-CD8 (53.6.7), anti-Ad (MKD6), and anti-Ed (14-4-4s) (PharMingen). No blocking of cytokine production was observed with anti-MHC class I Dd (34-5-8S) Ab.
Indirect immunofluorescence and flow cytometry
J774 tumor cells were monitored for intracellular expression of p53 protein by indirect immunofluorescence and were analyzed on a Becton Dickinson (San Jose, CA) flow cytometer. Before staining with anti-p53 mAb, cells were permeabilized using 0.5% saponin solution (Sigma). All subsequent steps were performed in the presence of 0.5% of saponin. Cells were washed in PBS containing 2% FCS and then were incubated for 30 min (5 x 105 cells/tube) in the presence of pAb 246 and pAb 240 mouse anti-p53 mAb (1–5 µg/ml; Oncogen Research Products, Cambridge, MA) and then in the presence of corresponding FITC-conjugated goat anti-mouse IgG Abs.
p53-specific CD4+ T cell line preparation
The CD4+ T cell line specific for mut p53.225–239 peptide was obtained from BALB/c mice. Mice were immunized in their hind footpads with 50 µg of mut p53.225–239 peptide emulsified in CFA. Nine days later, popliteal lymph node cells were harvested and cultured at 5 x 106 cells/ml in complete DMEM supplemented with 2 x 10-5 M 2-ME and 10% FCS. The T cell line was stimulated every other week with either Ag (peptide at final concentration, 20 µM), syngeneic irradiated splenocytes (2 x 106 cells/ml) as feeders and 20 U/ml of human rIL-2 (Genzyme, Cambridge, MA), or IL-2 alone (25 U/ml). The CD4+ phenotype of the T cell line was shown by two-color fluorescence analysis using FITC-conjugated rat anti-mouse CD4 mAb (RM4-4) and PE-conjugated rat anti-CD8 mAb (53.6.7) (PharMingen).
RT-PCR and sequencing
The J774 tumor cell line was maintained in culture as described above, and total RNA was extracted using the RNA STAT-60 kit (Tel-Test, Friendswood, TX). Approximately 3 µg of total RNA was then retrotranscribed using the Superscript II polymerase (Life Technologies) into c-DNA via oligo dT priming. A total of 10% of the reaction mix was then used in PCR to amplify the full-length p53 open reading frame using primers designed from the wt mouse p53 sequence. A 1.2-kb cDNA fragment encoding the murine p53 protein from J774 was ligated into the PUC57 vector (MBI Fermentas), and several clones were sequenced by dideoxy sequencing. GeneWorks software (IntelliGenetics, Mountain View, CA) was used to compare wt and mut p53 cDNA sequences and to translate them. The oligos used as primers for sequencing were as follows (5' to 3', only one strand shown): M13F, GTAAAACGACGGCCAGT; M13R, CAGGAAACAGCTATGAC; P53–919, GGAAGAGGCGCTTGTGCAGGT; and P53–414, CCTGTGCAGTTGTGGGTCAG.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We first investigated the existence of potential antigenic
CD4+ T cell determinants on p53 protein in BALB/c
(H-2d) mice. To address this question, the wt
mouse p53 amino acid sequence was screened for the presence of peptides
containing MHC class II (Ad and
Ed) peptide-binding motifs. An
Ed motif has been previously determined using
sequences of synthetic peptides that bind with high affinity to the
Ed molecule and natural peptide ligands eluted
from the Ed binding groove (31).
This motif contained multiple basic residues, and the following
preferential anchor residues at positions P1, P4, P6, and P9: a bulky
aromatic amino acid (P1), a positively charged amino acid (P4), an
aliphatic amino acid or glycine (P6), and a basic amino acid (P9) (Fig. 1
). Four regions along the p53 amino acid
sequence were found to contain the Ed-binding
motif: p53.167–181, 277–291, 316–334, and 365–384. The first two
peptides belong to "hot-spot" mutational portions of p53, whereas
the two others are derived from conserved portions of the p53 protein
(32). Additionally, 167–181 and 277–291 p53 peptides
were also shown to contain a consensus motif for binding to
Ad MHC class II molecule (alternating of
hydrophobic and polar amino acids) (Ref. 33 and Fig. 1).
The four p53 peptides containing H-2d MHC class
II binding motifs were synthesized and tested for their immunogenicity
in BALB/c mice.
|
A) and IFN- production
(Fig. 2B) were observed in response to p53.316–334 and
p53.365–384 peptides. This response was inhibited by anti-CD4 but
not anti-CD8 Abs and by anti-MHC class II
(Ed) but not anti-MHC class I
(Dd) Abs (data not shown). This demonstrates that
anti-p53 response to these peptides was mediated by MHC class II
Ed-restricted CD4+ T cells.
Immunization with p53.167–181 peptide resulted in strong IFN-
production (Fig. 2B) but modest T cell proliferation (Fig. 2A).
|
, A and B), was
nonimmunogenic in BALB/c mice (Fig. 2
). This suggested that this p53
peptide had been efficiently presented by MHC class II during ontogeny
and had negatively selected corresponding CD4+ T
cells. Confirming this hypothesis, when seen as a foreign Ag,
p53.277–291 induced CD4+ T cell response in p53
knock-out mice after immunization with p53 protein-containing J774 cell
lysates (Fig. 3C). Therefore, this peptide could be
efficiently presented by the Ed molecule and
could induce an in vivo T cell response in mice whose T cells, specific
for this p53 peptide/Ed determinant, had not been
rendered tolerant. Based upon data in p53 KO mice, it is likely that
p53.277–291 self-determinant had been presented in MHC class II
context during thymic development and had ensured
CD4+ T cell tolerance (dominant self-determinant)
(35, 36). In contrast, other p53 self-peptides
(p53.167–181, 316–334, and 365–384) induce potent
CD4+ MHC class II-restricted T cell responses in
syngeneic BALB/c mice. Apparently, these peptides bind to self-MHC
class II molecules, but they did not reach the threshold of
presentation required for deletion/inactivation of corresponding T
cells (cryptic self-determinants) (35, 36). Therefore,
self-p53, in spite of its low expression and nuclear localization,
contributes to the CD4+ T cell repertoire
selection during ontogeny. These results also suggest that under
appropriate conditions, anti-self-p53 Th cell response can be
elicited in adult BALB/c mice.
|
The results obtained in normal mice prompted us to examine the involvement of p53-specific MHC class II-restricted CD4+ T cell responses in antitumor response. BALB/c mice were inoculated with the highly tumorigenic and metastatic BALB/c-derived monocyte-macrophage cell line J774. Two weeks after i.p. injection of 106 live J774 cells into syngeneic hosts, multiple solid tumors were found in the mouse abdominal area. By the third week after cell inoculation, J774 tumor progressed to ascites.
First, we used a series of anti-p53 mAbs and flow cytometry to
determine whether p53 was expressed in mutated form in J774 cells. pAb
246 mAb recognizes a conformational epitope formed by residues at
position 108–109 on wt mouse p53. Alternatively, pAb 240 mAb does not
bind to wt p53 protein but interacts with an epitope (213–217 aa)
exposed exclusively on mut p53. Here, J774 tumor cells were
permeabilized and incubated with mAb 240 and mAb 246. FITC-labeled
anti-Ig Abs were used to reveal the intracellular binding of
anti-p53 Abs, and the fluorescence was determined using FACS
analysis. We observed that J774 cells did not express the epitope
recognized by pAb 246, a result suggesting that native p53 protein
conformation was altered in this cell line (Fig. 4A). In contrast, the vast
majority of J774 cells were labeled with pAb 240 mAb, a result
demonstrating that p53 was expressed in mutated form (Fig. 4B). Using immunofluorescent microscopy, mutant p53 protein
could be visualized both in the nucleus and in the cytoplasm of J774
cells (data not shown).
|
Next, we investigated whether mutant p53 protein expressed
in J774 cells elicits CD4+ T cell-mediated
response during tumorigenesis. BALB/c mice were injected i.p. with live
nonirradiated J774 tumor cells. Two weeks later, multiple solid tumors
were detected in the abdominal areas of these mice. At this time, mouse
spleen T cells were tested for their reactivity to p53 peptides.
Production of IFN- and IL-5 cytokines was monitored using ELISA
assay. We observed that among four p53 peptides tested, only
p53.167–181 peptide restimulated T cells. As shown in Fig. 5, both IFN- (Fig. 5A) and
IL-5 cytokines (Fig. 5B) were detected, with higher levels
of IL-5. This response was elicited by MHC class II-restricted T cells
because it was blocked by anti-CD4 (not anti-CD8) and
anti-MHC class II Ad Abs (not anti-MHC
class I or anti-MHC class II Ed mAbs) (data
not shown). We conclude that p53.167–181 peptide contains a
well-processed determinant presented in association with
Ad MHC class II molecule to
CD4+ T cells during tumorigenesis in vivo.
Therefore, although p53.167–181 was not processed and presented to T
cells in normal mice, it became a dominant determinant on mut p53
during J774 tumor development. It is important to note that although
s.c. p53.167–181 peptide immunization in CFA induced a Th1 response
(Fig. 2B), spontaneous presentation of this determinant
during cancer biased the cytokine response toward Th2 (Fig. 5B). This indicates that induction of polarized Th2 response
was not an intrinsic property of the p53 peptide but was a feature of
antitumor response.
|
could still be detected after challenge with p53.167–181,
IL-5 production was no longer observed. Alternatively, another peptide,
p53.365–384, elicited potent T cell response characterized by low
IFN- release (Fig. 5C) but a high level of IL-5 secretion
(Fig. 5D). Therefore, during the course of tumorigenesis,
anti-p53 CD4+ T cell response remained
consistently biased toward Th2 type. However, although initial
anti-p53 T cell response was exclusively directed to 167–181
peptide, most of the response at a late phase of tumorigenesis shifted
to the 365–384 p53 determinant. We conclude that during J774 tumor
progression, the determinant hierarchy on p53 had changed. Although
CD4+ T cell response to p53.167–181 was no
longer detectable, another formerly cryptic peptide, p53.365–384, had
become immunodominant. This phenomenon may reflect temporal variation
in p53 Ag processing and presentation and the subsequent recruitment of
new T cells. Cloning and sequencing of p53 cDNA in J774 BALB/c-derived tumor cells
It has been reported for certain tumors that mutations in p53 resulted in the presentation of p53 "neo-determinants" by MHC class I molecules (9, 37). In this study, it was important to determine whether mutation in p53 sequence in J774 tumor cells could account for the creation of a new CD4+ T cell determinant recognized as foreign by the immune system. To address this question, we investigated the nature and the position of the p53 mutation(s) in J774 cells. p53 cDNA was prepared from J774 cell line and cloned. Total RNA was extracted from actively dividing J774 cells and reverse transcribed into cDNA. A 1.2-kb DNA fragment, corresponding to the full-length p53 cDNA, was amplified by PCR and subcloned into PUC57 vectors. To avoid sequencing errors introduced during the PCR reaction, two separate RT-PCR reactions were performed, and their products were independently ligated into PUC57 vectors, sequenced, and compared with wt p53 transcripts.
Thirteen of the clones obtained from the first PCR reaction were sequenced from their 3' region using M13R as primer, and four clones (from the same PCR) were sequenced from their 5' region using M13F. By overlapping sequencing results obtained from each clone on both strands, the full cDNA sequence (1.2 kb) was obtained. Although some of the clones were wt, three clones showed one point mutation at position 691, where a T to C transition was detected. To exclude any sequencing artifact, we also confirmed the sequence of the point-mutated clones by sequencing the region of interest with internal primers (P53–414 and P53–919) and on both strands. Finally, to rule out the possibility of an error introduced during the amplification process, we also sequenced four of the clones that were generated from the second PCR reaction. Three of the clones were wt, but one of them showed the exact same point mutation (T to C) at position 691. Taken together, these data demonstrate that this cell line expresses both wt and mut p53 alleles. Mut p53 gene contains a single nucleotide change at position 691. This point mutation resolves into the substitution of a tyrosine with a histidine at amino acid position 231 of p53 protein.
Immunogenicity of p53 peptides derived from mut p53 region in normal mice
We next examined whether p53 peptides derived from p53 region
containing a mutation were immunogenic in BALB/c mice. Two peptides
were synthesized: wt p53.225–239 derived from wt p53 sequence and mut
p53.225–239 containing a histidine at position 231. Using direct
binding assay, both wt and mutated peptides were shown to bind with
high affinity to MHC class II Ed molecules (Fig. 6B). Additionally, mut
p53.225–239 bound to Ad with intermediate
affinity, whereas wt p53 was a poor Ad binder
(Fig. 6A). Each peptide was injected s.c. in the presence of
CFA to BALB/c mice. Ten days later, mouse spleen cell suspensions were
challenged in vitro with the immunizing peptide and were tested for
proliferation and IFN- production. As shown in Fig. 6, mut
p53.225–239 peptide elicited both T cell proliferative response (Fig. 6C) and IFN- release (Fig. 6D). This response
was mediated by CD4+ Th cells as it was blocked
by anti-CD4 and anti-MHC class II Ed Abs
(data not shown). In contrast, no response was obtained after
immunization and challenge with wt p53.225–239 (Fig. 6, C
and D). These results indicate that wt p53.225–239 contains
a dominant self-determinant whose presentation during development had
tolerized corresponding autoreactive p53-reactive T cells. Apparently,
during cancer, mutation at position 231 in p53 had created a new
determinant that could be recognized as foreign and had induced T cell
responses in vivo. In addition, it was possible that mut p53
peptide/MHC class II complexes, owing to their high affinity to TCR
compared with that of their wt counterparts, activated some undeleted
and normally resting p53.225–239-specific autoreactive T cell
clones.
|
T cell response to mut p53 region-derived peptides in J774
tumor-bearing mice was measured using ELISA spot assay. This technique
allows determination of the actual frequency of cytokine-producing T
cells after Ag recognition. After inoculation of J774 tumor cell line,
BALB/c mice mounted CD4+ T cell responses to both
mut and wt p53.225–239 determinants (Fig. 7). This response was comprised of T
cells secreting both IFN- and IL-5, although the frequency of
IFN--producing cells was much higher. These data showed that
differential expression of p53 in tumor-bearing mice had resulted in
induction of T cell response to the mutated portion of p53 and
concomitant breakdown of tolerance to the wt p53.225–239 peptide. It
is conceivable that presentation of mut p53 peptide had stimulated some
low-avidity T cell clones that had not been deleted during thymic
development. Furthermore, an increased pool of p53 protein, due to its
stabilization upon interaction with mutated chain, may have boosted
processing of this protein and increased the amount of p53 determinants
available for presentation to T cells. Also, the number of MHC class
II/p53.225–239 peptides formed may have reached the threshold
necessary to activate low-avidity T cells specific for this
peptide.
|
).
However, as shown in Fig. 8, over 10-fold more wt peptide than its
mutated counterpart was required to achieve a similar level of T cell
activation. This result suggests that high-affinity mut p53
peptide-specific T cells can be activated by the wt p53 peptide when it
is presented in sufficient amounts. Taken together, these data strongly
support the view that presentation of mut p53.225–238 during J774
tumorigenesis had broken CD4+ T cell tolerance to
wt p53.225–238 by lowering the threshold necessary to activate some
low-affinity p53 autoreactive T cell clones in BALB/c mice. This shows
that not all T cells reactive to dominant self-peptide p53.225–238 had
been deleted during thymic selection and that a fraction of this
autoreactive repertoire can be activated during tumorigenesis. It also
suggests that "neo-determinants" on tumor-associated proteins can
act as agonists of self-peptides and can activate some undeleted
self-reactive T cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study, we showed that despite a low level of expression and nuclear localization of p53 protein, p53 peptide 277–291 is efficiently processed and presented in MHC class II context to developing CD4+ T cells during ontogeny. Elevated p53 expression associated with massive apoptosis in the neonatal thymus may account for the efficient presentation and subsequent tolerogenicity of this p53 determinant. In contrast, p53 peptides 167–181, 316–334, and 365–384 encompass cryptic self-determinants. These self-p53 peptides, despite their high binding affinity to H-2d MHC class II molecules, did not elicit negative selection of corresponding autoreactive T cells, presumably owing to their poor processing in BALB/c thymic APCs. Consequently, after s.c. peptide immunization, potent CD4+ T cell responses to these three self-p53 determinants were induced in adult BALB/c mice. This shows that CD4+ T cell immunogenicity and tolerogenicity of p53 self-determinants follow the rules of immunodominance previously described for other self-Ags (34, 35). In another study, Theobald et al. (17) have shown that HLA A2.1-restricted CD8+ T cells specific to p53.187–197 self-peptide were negatively selected in HLA-A/2.1-Kb transgenic mice. In contrast, CD8+ T cells recognizing peptide 261–269 of p53 had not been tolerized and displayed CTL activity after peptide immunization. Similarly, other investigators reported that CD8+ and CD4+ T cell lines generated from healthy humans contain self-p53 peptide-reactive T cells (38, 39). Together with our study, these observations show that during thymic selection certain portions of self-p53 are well processed and presented (dominant-self) and induce CD4 and CD8 T cell tolerance. In contrast, other poorly processed p53 determinants remain cryptic and fail to delete corresponding autoreactive anti-p53-specific T cells. It is possible that, during ontogeny, cryptic self-p53 peptides never get processed and presented in MHC class II context, thereby sparing some p53-reactive T cells from negative selection. Alternatively, poor processing of these peptides may result in low-avidity interaction of some p53 peptide/MHC class II complexes with corresponding TCR on T cells, a phenomenon leading to positive selection in the developing thymus. Further analysis of the p53-specific TCR repertoire, which recognizes dominant and cryptic self-p53 determinants in normal and p53 KO mice, may bring some insights to this question.
Under appropriate circumstances, cryptic self-determinants can become efficiently processed and presented in MHC class II context to T cells. Conformational changes and accumulation of large amounts of mut p53 protein in the cytoplasm of J774 cells presumably account for quantitative and/or qualitative alterations in p53 Ag processing and presentation and subsequent activation of p53 autoreactive CD4+ T cells. Three non-mutually exclusive events may contribute to this phenomenon: 1) the number of p53 peptide/MHC complexes formed have become sufficient to activate some low-affinity anti-p53 autoreactive T cell clones, 2) cytoplasmic localization allows access of mut p53 to new processing compartments, and 3) changes in p53 conformation and sequence may reveal or create new sites for degradation by proteolytic enzymes. Supporting these possibilities, in another model, Theobald et al. (40) have reported that proteosomal processing and subsequent presentation of the self-p53.264–272 epitope in MHC class I context is profoundly affected by mutation at the flanking residue 273. In this study, we observed that after inoculation of live J774 tumor in BALB/c mice, in the absence of any peptide immunization, potent CD4+ T cell responses were elicited after 15 and 21 days to the cryptic p53.167–181 and p53.365–384 peptides, respectively. Therefore, during tumorigenesis, these normally cryptic p53 determinants have become dominant. We conclude that tumorigenesis is associated with reversal of determinant hierarchy on p53 protein in that some formerly cryptic p53 determinants have become dominant in tumor-bearing mice. The precise influence of this phenomenon on tumor development remains to be determined.
Interestingly, we observed that the presentation of p53 determinants to CD4+ T cells is a dynamic process in that different p53 determinants were immunogenic depending upon the stage of tumor progression. These differences in the nature of the anti-p53 immune response correlated with the degree of malignancy of J774 cells in mice (development of ascites). Changes in the determinant hierarchy on p53 may reflect further alterations in p53 protein expression during J774 tumor progression. To our knowledge, this is the first demonstration that tumor progression is associated with sequential presentation of new determinants on tumor Ags. This feature of anti-p53 response could be utilized to determine the stage of tumor development in patients. This also suggests that, depending upon the stage of cancer, p53-based tumor vaccines and T cell treatments should target different p53 determinants.
During J774 tumorigenesis, we detected MHC class II-restricted T cell
response to the mutated portion of p53.225–239 (histidine at position
231). Interestingly, although the nonmutated wt p53.225–239 peptide
(tyrosine at position 231) could not elicit any response in healthy
mice, it stimulated vigorous CD4+ T cell response
in tumor-bearing mice. Based upon data obtained with the
anti-p53.225–239 peptide-specific T cell line, it is likely that
mut p53 peptide had activated some undeleted low-affinity T cell
clones, thereby breaking tolerance to this peptide. These data show
that not all CD4+ T cells reactive to dominant
p53 peptides are deleted during thymic development. Apparently, some
low-affinity Th cell clones are spared from deletion and can become
activated during tumorigenesis. Similar observations have been reported
for antitumor CTL responses (41). It is noteworthy that,
unlike the T cell response to cryptic p53 peptides, the response to
p53.225–239 was characterized by the presence of IFN- lymphokine.
This suggests that p53.225–239 may represent a potential candidate for
the design of a cancer vaccine.
It is widely believed that live metastatic tumors do not elicit
immune responses to tumor-specific Ags in vivo (23, 42).
To circumvent this lack of immunogenicity, scientists have vaccinated
rodents with irradiated or apoptotic tumor cells. In some cases,
productive immunization has requested the presence of adjuvants. Other
investigators have transfected cells with genes encoding for
costimulatory molecules or heat shock proteins to render tumors
immunogenic (21, 43, 44, 45, 46, 47). In this article, we show that
live unmodified tumor cells induce potent CD4+ T
cell immune responses to a series of determinants corresponding to
self- and mutated portions of p53 protein. It is noteworthy that Th
cell response to p53 determinants was characterized by partial or
complete absence of proliferation and IL-2 and IFN- production.
Alternatively, in all cases massive release of IL-5 was recorded,
denoting a response that is biased toward Th2-type cytokine profile.
This was not an intrinsic property of these p53 peptides because s.c.
injection of these peptides in adjuvant induced vigorous Th1 responses
in normal mice. Predominance of Th2 cells in antitumor T cell response
may account for the apparent lack of immunogenicity of live tumors
reported in many studies (23, 48). It may also explain
why, despite the presence of an active antitumor immune response,
BALB/c mice failed to reject J774 tumor cells. Moreover, Th2 cells
traditionally thought to ensure self-tolerance may tolerize the host
toward tumor Ags, thereby protecting the tumor from immune rejection.
Supporting this view, studies by Qin et al. (49) suggest
that Th2-type cytokines mediate "nonproductive" humoral B cell
antitumor responses. Together with our work, these studies suggest that
blockade or deletion of undesirable antitumor Th2 cells may represent a
useful strategy for abrogating immune escape of tumors.
In summary, our study demonstrates that different self- and mutated regions of p53 induce MHC class II-restricted CD4+ T cell responses in J774 tumor-bearing BALB/c mice. This response was dominated by T cells displaying a Th2 phenotype, a feature that may account for lack of immune rejection of tumors in these mice. Manipulation of T cell response to these p53 determinants clearly represents an attractive approach in cancer immunotherapy given that mut p53 is expressed in a large variety of tumors of different origins. However, it remains to be investigated whether tumors harboring different mutations on p53 elicit CD4+ T cell responses to the same regions of p53 protein. In addition, we report that the nature of the p53 determinant recognized depends upon the stage of cancer. These features of anti-p53 immune responses should be taken into consideration in the design of future p53-based cancer vaccines.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Cellular and Molecular Immunology Laboratory, Schepens Eye Research Institute, Harvard Medical School, Boston, MA 02114.
3 Address correspondence and reprint requests to Dr. Gilles Benichou at the current address: Cellular and Molecular Immunology Laboratory, Schepens Eye Research Institute, Harvard Medical School, 20 Staniford Street, Boston, MA 02114.
4 Abbreviations used in this paper: mut, mutated; wt, wild type; p53 KO, BALB/c J-Trp 53tm/tyj; Myo, sperm whale myoglobulin; HEL, hen eggwhite lysozyme.
Received for publication January 27, 2000. Accepted for publication March 15, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. P. Castano, P. Mroz, M. X. Wu, and M. R. Hamblin Photodynamic therapy plus low-dose cyclophosphamide generates antitumor immunity in a mouse model PNAS, April 8, 2008; 105(14): 5495 - 5500. [Abstract] [Full Text] [PDF]
|
|
|
|
) peptides were tested for their
ability to inhibit the binding of biotinylated peptide Myo106–118 to
purified Ad molecules. B, p53.277–291 (
)
and known Ed binder HEL104–120 (
) peptides were tested
for their ability to inhibit the binding of biotinylated peptide
HEL104–120 to purified Ed molecules. The data represent
the amount of bound biotinylated reporter peptide expressed as counts
per second (cps). The dashed lines represent the amount of bound
reporter peptide measured in the absence of competitor peptide.
C, Spleen T cell response of normal BALB/c
(
) and p53 KO BALB/c (
), p53.316–334 (
). In all experiments,
irrelevant OVA323–339 peptide was used as a control (
). Results are
expressed in pg/ml. Each figure represents results obtained from
spleens pooled from three mice. The data shown are representative of
three to six separate experiments.
) in the presence of irradiated splenocytes as APCs. After 48
h, culture supernatants were collected and analyzed for the presence of
IFN-