Optimization of a self antigen for presentation of multiple epitopes in cancer immunity

doi:10.1172/JCI25591

J Clin Invest. 2006 May 1; 116(5): 1382–1390.

Published online 2006 April 13. doi: 10.1172/JCI25591

PMCID: PMC1435720

Optimization of a self antigen for presentation of multiple epitopes in cancer immunity

José A. Guevara-Patiño,¹ Manuel E. Engelhorn,¹ Mary Jo Turk,¹ Cailian Liu,¹ Fei Duan,^1,² Gabrielle Rizzuto,^1,² Adam D. Cohen,^1,² Taha Merghoub,¹ Jedd D. Wolchok,^1,² and Alan N. Houghton^1,²

¹Swim Across America Laboratory of Tumor Immunology, Memorial Sloan-Kettering Cancer Center, New York, New York, USA. ²Weill Medical and Graduate Schools of Cornell University, New York, New York, USA.

Address correspondence to: José A. Guevara-Patiño, University of Chicago, Section of General Surgery, 5841 S. Maryland Avenue, MC 7114, Chicago, Illinois 60637, USA. Phone: (773) 702-5438. Fax: (773) 834-8140. E-mail: jguevara/at/surgery.bsd.uchicago.edu .

Received May 9, 2005; Accepted February 21, 2006.

Abstract

T cells recognizing self antigens expressed by cancer cells are prevalent in the immune repertoire. However, activation of these autoreactive T cells is limited by weak signals that are incapable of fully priming naive T cells, creating a state of tolerance or ignorance. Even if T cell activation occurs, immunity can be further restricted by a dominant response directed at only a single epitope. Enhanced antigen presentation of multiple epitopes was investigated as a strategy to overcome these barriers. Specific point mutations that create altered peptide ligands were introduced into the gene encoding a nonimmunogenic tissue self antigen expressed by melanoma, tyrosinase-related protein-1 (Tyrp1). Deficient asparagine-linked glycosylation, which was caused by additional mutations, produced altered protein trafficking and fate that increased antigen processing. Immunization of mice with mutated Tyrp1 DNA elicited cross-reactive CD8⁺ T cell responses against multiple nonmutated epitopes of syngeneic Tyrp1 and against melanoma cells. These multispecific anti-Tyrp1 CD8⁺ T cell responses led to rejection of poorly immunogenic melanoma and prolonged survival when immunization was started after tumor challenge. These studies demonstrate how rationally designed DNA vaccines directed against self antigens for enhanced antigen processing and presentation reveal novel self epitopes and elicit multispecific T cell responses to nonimmunogenic, nonmutated self antigens, enhancing immunity against cancer self antigens.

Introduction

T cells play a central role in immunity against cancer. Cancer antigens recognized by T cells are encoded by 2 broad categories of genes: those expressed by normal somatic and germ cells and those expressed only by cancer cells (e.g., due to mutations acquired during or after malignant transformation) (1, 2).

A question had been whether T cells capable of recognizing self antigens on cancer cells are present in the repertoire. Evidence shows that T cells are positively selected for survival during development in the thymus by signals elicited by self peptides complexed with MHC molecules (pMHCs) and maintained in the periphery by self pMHCs (3–7). Thus the repertoire of T cells develops and is maintained through self reactivity. To avoid autoimmunity, T cells with high avidity for self pMHCs are deleted, but more weakly self-reactive T cells survive.

This pool of T cells discriminates nonself antigens of pathogens by cross-reactivity to foreign peptides in the context of proinflammatory and costimulatory signals elicited by pathogens through activation of APCs. On the other hand, cognate self peptides are too weak to activate T cells, particularly in the absence of sufficient costimulation (8). A corollary to these observations is that, despite abundant self-reactive T cells in the peripheral immune system, initiation of T cell responses against self antigens on cancer are restricted by insufficient signals.

One strategy to overcome these limitations uses active immunization with orthologous gene products. Immunization with orthologs can overcome tolerance or ignorance to self antigens, including those expressed by cancer (9–17). This approach relies on small differences in amino acid sequences between orthologous proteins from mutations accumulated over tens of millions of years of evolution. One mechanism underlying increased immunogenicity of orthologous antigens relates to the augmented avidity of orthologous peptide for host MHC molecules relative to weak avidity of the cognate self peptide (13, 18, 19). Altered peptide ligands with increased avidity for MHC molecules can enhance interactions between pMHCs and T cell receptors to produce T cell activation (20).

The tyrosinase family represents prototypic tissue self antigens expressed by cancer. Tyrosinase-related protein-1 (Tyrp1) is the most abundant glycoprotein in melanocytes and pigmented melanomas and is recognized by autoantibodies in melanoma patients (21, 22). Mice are tolerant to Tyrp1, but tolerance is broken by orthologous antigen or syngeneic antigen expressed by vaccinia that results in weak tumor immunity (12, 14, 23). Notably, immunity against Tyrp1 is independent of CD8⁺ T cells and is instead mediated by CD4-dependent, Th2-type autoantibodies (12, 14, 24).

To avoid relying on mutations selected through the millennial time frame of evolution, in the present study rationally selected mutations were introduced into the Tyrp1 gene to enhance antigen processing and presentation of multiple epitopes in order to determine whether it was possible to generate multivalent CD8⁺ T cell responses. We were able to convert the inert syngeneic Tyrp1 cDNA to forms that induced multi-epitope, CD8⁺-dependent, Th1-type melanoma immunity in mice. Such rationally designed, multivalent immunogens could be useful to direct specificity of responses in the context of nonspecific immune modulation therapies (25).

Results

Tyrp1 DNA constructs.

Immunization against multiple epitopes is associated with improved immunity in infectious disease models (26, 27). We decided to investigate active immunization against multiple MHC class I epitopes using the melanocyte-specific self antigen mouse Tyrp1. This autoantigen has only been connected to antibody-dependent, CD8-independent immunity, and to our knowledge no MHC class I–restricted epitopes have been previously identified in Tyrp1 (12, 14, 24). The strategy was to induce a potentially broad CD8⁺ T cell response against Tyrp1. To do this, we introduced a series of mutations in mouse Tyrp1, creating multiple altered peptide ligands for increased MHC presentation. Antigen processing and presentation were further enhanced by creating deficient glycosylation sites. The panel of Tyrp1 DNA plasmid constructs included nonmutated syngeneic (wild-type) mouse Tyrp1 (Tyrp1); epitope-enriched Tyrp1 (Tyrp1ee), incorporating 10 point mutations to create potential altered peptide ligands (Table (Table1);1); Tyrp1 mutated at 6 positions to generate deficient Asn-linked glycosylation sites (Tyrp1ng), but not optimized for MHC class I binding; and combined epitope-enriched, glycosylation-deficient Tyrp1 (Tyrp1ee/ng). Construction and characterization of these vectors is described in the following paragraphs.

Table 1

Tyrp1 peptide optimization for MHC class I binding

Epitope-enriched Tyrp1.

The Tyrp1ee DNA constructs incorporated multiple amino acid mutations that were predicted to increase epitope binding to MHC class I molecules H2-K^b and H2-D^b. To design this construct, we identified 10 self peptides in Tyrp1 with predicted low-level binding to MHC class I molecules using existing matrices (28). T cells recognizing low-affinity, tissue-specific self peptides are expected to escape thymic deletion. Mutated peptides with single amino acid changes at candidate MHC anchor sites that were predicted to create altered peptide ligands with enhanced MHC class I binding were compared with nonmutated wild-type peptides. Eight of 10 mutated peptides had enhanced MHC class I binding compared with their wild-type counterparts by MHC class I stabilization assays (Figure (Figure1),1), using the transporter associated with antigen processing 2–deficient (TAP2–deficient) RMA-S cell line (29, 30). The Tyrp1ee cDNA variant was constructed by introducing single nucleotide mutations by site-directed mutagenesis into full-length mouse Tyrp1 cDNA (22, 31) (Table (Table1,1, H2-D^b– and H2-K^b–optimized peptides).

Figure 1

Rational optimization of peptide anchor residues in mouse Tyrp1 enhances MHC class I binding.

Tyrp1 with deficient Asn-linked glycosylation.

Tyrp1 is a transmembrane protein that, in its mature state, contains complex Asn-linked sialylated sugars. Tyrp1 carbohydrates are processed from high mannose to intermediate and complex glycans during movement from the ER through the Golgi complex (32–34). Tyrp1 contains 6 potential Asn-linked glycosylation sites in the lumenal domain (33–35). Our prior studies have shown that these glycans differentially determine egress from the ER, movement through the Golgi to the endosomal pathway, stability in the endosomal pathway, and overall stability of Tyrp1 (34).

Six mutations (Asn→Gln) were introduced into the potential glycosylation sites of Tyrp1 and Tyrp1ee to produce deficient Asn-linked sites Tyrp1ng and Tyrp1ee/ng (see Methods), the latter protein incorporating both sets of mutations. We predicted that these mutations would promote ER retention, enhanced susceptibility to proteolytic digestion, and decreased stability, thereby increasing MHC class I antigen processing, perhaps guided by an ER misfolded protein response (34, 36–38).

Biosynthesis of epitope-enriched and glycosylation-deficient Tyrp1 mutants.

To assess the stability of Tyrp1ee and Tyrp1ee/ng proteins compared with wild-type Tyrp1, Cos-7 cells were transiently transfected with DNA encoding each of the Tyrp1 constructs containing a Flag tag at the carboxyl terminus. New protein synthesis was blocked with muconomycin A at 24 hours. At this time point (designated 0), wild-type Tyrp1 and Tyrp1ee were detected as single bands of approximately 75 kDa (Figure (Figure2A).2A). Interestingly, Tyrp1ee/ng was consistently observed as a ladder of bands ranging from approximately 75 kDa to approximately 59–60 kDa, always with a weak top 75-kDa band that increased in intensity over the 6-hour chase (Figure (Figure2A).2A). These results were reproduced in 3 different experiments for wild-type Tyrp1 and Tyrp1ee and 4 experiments for Tyrp1ee/ng.

Figure 2

Biosynthesis and stability of wild-type Tyrp1, Tyrp1ee, and Tyrp1ee/ng.

Deficient maturation of carbohydrate moieties on Tyrp1ee and Tyrp1ee/ng.

Although the Tyrp1ee/ng ladder might represent intermediate degradation products, particularly because glycosylation sites had been mutated, we suspected that the ladder instead represented different unstable glycosylated forms of Tyrp1 based on a close resemblance to our prior results with partial digestion of Tyrp1 by N-glycanase (33). Digestion with the enzymes endoglycosidase H (Endo H) and N-glycanase collapsed the ladder to a single band of approximately 59–60 kDa (Figure (Figure2B).2B). These results show that sugars on Tyrp1ee/ng were not processed beyond the ER (based on Endo H sensitivity), with the ladder representing different glycosylated polypeptides of Tyrp1ee/ng. Thus Asn→Gln conversion did not completely abrogate glycosylation, but rather created a series of ER-retained, differentially glycosylated polypeptides (with glycans at 0, 1, 2, 3, or 4 Asn sites). The highest-mass 75-kDa band, which increased over 6 hours, was consistent with further posttranslational processing. In addition, Tyrp1ee glycans were also sensitive to Endo H, showing that the mutations in Tyrp1ee led to ER retention — probably through misfolding — with no further processing in the Golgi, followed by proteasome degradation (shown below; Figure Figure2B). 2B).

Tyrp1ee/ng mutants are degraded in a proteasome-dependent pathway.

Protein stability and degradation of wild-type Tyrp1, as assessed over a 24-hour chase, were nearly identical to what we have previously reported in mouse melanocytic cells, consistent with little or no effect of the Flag tag or expression in Cos-7 cells on stability or degradation (33, 35, 39) (Figure (Figure2A).2A). We investigated the effects of chloroquine (to inhibit degradation in acidic compartments) and the proteasome inhibitor MG132. The fate of mature wild-type Tyrp1 was minimally affected by either inhibitor (Figure (Figure2A).2A). A short-lived 17-kDa carboxyterminal degradation fragment of wild-type Tyrp1 was detected in the presence of inhibitors but was not detected for Tyrp1ee and Tyrp1ng.

The stability of Tyrp1ee was essentially the same as for wild-type Tyrp1, despite ER retention. The majority of Tyrp1ee underwent proteasome degradation, with a minor proportion reaching acidic compartments for delayed degradation (Figure (Figure2A).2A). Thus the mutations in Tyrp1ee altered its trafficking, leading to ER retention and proteasome degradation.

Total degradation of Tyrp1ee/ng was modestly delayed by proteasome inhibition over 12 hours (Figure (Figure2A).2A). When higher and lower Tyrp1ee/ng bands were examined, the higher-mass glycoforms (~75 and 69 kDa) were less stable than wild-type Tyrp1 and were degraded largely via acidic compartments, presumably the endosomal pathway (Figure (Figure2A).2A). The 3 lower-mass forms (~65–66, 61–63, and 59–60 kDa, representing glycoforms with 1 or 2 carbohydrate moieties and the naked polypeptide) were very unstable, retained in the ER, and not rescued by either proteasome inhibition or chloroquine (Figure (Figure2A).2A). In summary, Tyrp1ee/ng generates multiple unstable products that are degraded both in the endosomal pathway (high-mass forms) and through proteasomes (low-mass forms).

Increased immunogenicity of Tyrp1ee mutants.

Although candidate altered peptide ligands of Tyrp1ee had enhanced binding to MHC class I (Figure (Figure1),1), we needed to ascertain whether these epitopes could be naturally processed and presented. To increase immunogenicity further, Tyrp1 mutants were ligated to the translocation domain of Pseudomonas aeruginosa’s exotoxin A, which enhances CD8⁺ T cell responses in other immunization models (40–42). The exotoxin A domain was fused upstream of Tyrp1 (exo-Tyrp1), Tyrp1ee (exo-Tyrp1ee), and Tyrp1ee/ng (exo-Tyrp1ee/ng); alternatively, exotoxin A alone was noncovalently mixed with Tyrp1ee. Immunogenicity was initially addressed by assessing the ability of immunization with DNA constructs encoding full-length wild-type or mutated Tyrp1 constructs to induce CD8⁺ T cell responses against wild-type Tyrp1 peptides in C57BL/6 mice. Mice were immunized with 4 μg DNA 4 times by ballistic bombardment (14). Purified CD8⁺ T cells were tested for IFN-γ secretion in ELISPOT assays after 18-hour stimulations to detect responses to wild-type Tyrp1 epitopes and to intact syngeneic melanoma cells.

Immunization with wild-type Tyrp1, exo-Tyrp1, and Tyrp1ng did not induce any detectable T cell response, but CD8⁺ T cell responses were seen in mice immunized with Tyrp1ee, Tyrp1ee/ng, exo-Tyrp1ee, and exo-Tyrp1ee/ng (Figure (Figure3,3, A–C). Immunization with Tyrp1ee induced responses against 3 wild-type Tyrp1 peptides, 455, 481, and 522 (Figure (Figure3,3, A and C). We believe these to be novel MHC class I–restricted Tyrp1 epitopes revealed by this DNA immunization strategy. Immunization with Tyrp1ee/ng induced a substantially higher CD8⁺ T cell response against peptides 455, 481, and 522, increasing responses by 70%, 90%, and 95%, respectively, compared with Tyrp1ee (Figure (Figure3A). 3A).

Figure 3

Immunization with optimized Tyrp1 induces epitope-specific CD8⁺ T cell responses.

Exo-Tyrp1ee elicited higher CD8⁺ responses than did Tyrp1ee against peptide 455 (Figure (Figure3,3, B and C). However, fusion of the exotoxin A domain into Tyrp1ee/ng substantially diminished its potency down to Tyrp1ee levels (Figure (Figure3C).3C). Notably, CD8⁺ T cells primed by immunization with Tyrp1ee, Tyrp1ee/ng, exo-Tyrp1ee, and exo-Tyrp1ee/ng were able to respond to intact B16 melanoma cells, with Tyrp1ee/ng and Tyrp1ee eliciting the strongest response and exo-Tyrp1ee/ng the lowest (Figure (Figure3C).3C). Immunization with DNA encoding the exotoxin A fragment mixed with Tyrp1ee DNA showed no enhancement of CD8⁺ T cell responses compared with Tyrp1ee immunization alone (data not shown), indicating that the effect of exotoxin A is not mediated simply through the expression of exotoxin A (e.g., exotoxin helper epitopes). These results demonstrate that Tyrp1ee, Tyrp1ee/ng, exo-Tyrp1ee, and exo-Tyrp1ee/ng generated strong functional CD8⁺ T cell responses that recognized naturally processed and presented Tyrp1 epitopes expressed by B16 melanoma. Through this approach, we identified what we believe to be novel CD8⁺ T cell epitopes in a tissue self antigen, demonstrating a proof of concept for discovery of subdominant and cryptic epitopes in self antigens that can be processed and presented.

Antitumor effects following immunization with optimized DNA.

To investigate the in vivo effects of these rationally designed Tyrp1 DNA constructs, we assessed their ability to protect mice against lethal challenge with B16 melanoma cells. Palpable intradermal tumors were detected between days 4 and 8 following B16 challenge in naive mice with a tumor challenge that was 50- to 100-fold greater than the minimum lethal dose. No delay was observed in mice immunized with Tyrp1, Tyrp1ng, or exo-Tyrp1 (Figure (Figure4A;4A; Tyrp1ng data not shown). Notably, tumor protection was observed in approximately 40–45% of mice immunized with Tyrp1ee (P < 0.05, log-rank analysis) and 75–90% of mice immunized with Tyrp1ee/ng (P < 0.001; Figure Figure4A),4A), with the same results observed in 4 independent experiments. Immunization with exo-Tyrp1ee and Tyrp1ee/ng led to greater tumor protection compared with Tyrp1ee (~40–45% to ~75–90%; P < 0.001; Figure Figure4,4, A and B). Ligating exotoxin A to Tyrp1ee/ng substantially diminished antitumor effects to below 20% tumor protection (Figure (Figure4A),4A), and mixing Tyrp1ee with exotoxin A showed no enhancement of tumor immunity (Figure (Figure4B).4B). These results demonstrated the effectiveness of immunization with epitope-enriched self antigens and revealed increased potency when the self antigen was further destabilized by modifying glycosylation and conjugating to exotoxin A.

Figure 4

Immunization with optimized Tyrp1 rejects tumors and induces autoimmunity.

Autoimmunity manifested as patchy coat hypopigmentation was observed in 42 of the 45 mice immunized with Tyrp1ee, Tyrp1ee/ng, exo-Tyrp1ee, and exo-Tyrp1ee/ng, but in none of the 30 mice immunized with wild-type Tyrp1, in 3 experiments (data not shown). The most marked hypopigmentation was observed in mice immunized with Tyrp1ee/ng, with a representative mouse shown in Figure Figure4C.4C. Mice with autoimmune hypopigmentation after surviving tumor challenge remained otherwise healthy for up to 12 months or more.

Immunity induced by optimized Tyrp1 vaccines is CD8 dependent.

Tumor immunity was completely abrogated in mice immunized with Tyrp1ee, exo-Tyrp1ee, and Tyrp1ee/ng following depletion of CD8⁺ T cells, but not with depletion of CD4⁺ cells or NK cells (Figure (Figure5,5, A–C); to our knowledge, no difference in tumor-free survival in mice immunized with wild-type Tyrp1 with or without CD8⁺ depletion has been previously reported (14). These results showed that the immune responses induced by the optimized vaccines were CD8 dependent, without a requirement for CD4⁺ or NK cells for tumor rejection.

Figure 5

Immunization with Tyrp1ee, exo-Tyrp1ee, and Tyrp1ee/ng protects mice from tumor challenge through a mechanism dependent on CD8⁺ T cells.

Immunization with optimized Tyrp1 vaccines in a treatment model.

Immunization with optimized Tyrp1 constructs was examined in a treatment model by initiating immunization 4 days after challenge with B16 cells. Significant delays in tumor growth and prolonged survival were observed in mice treated with Tyrp1ee/ng (P < 0.05; Figure Figure6A)6A) and exo-Tyrp1ee (P < 0.01; Figure Figure6B).6B). No treatment effects were detected in mice immunized with Tyrp1ee, exo-Tyrp1, or Tyrp1.

Figure 6

Immunization with Tyrp1ee/ng and exo-Tyrp1ee significantly prolongs tumor-free survival in a treatment model.

Epitope optimization for HLA-A*0201.

To design a vaccine that might be clinically applied to people with melanoma, an epitope-enriched mouse Tyrp1 DNA construct (Tyrp1ee/A*0201) was created. This construct contained 6 mutations optimized for HLA-A*0201 binding within polypeptide regions that are highly conserved between mouse and human Tyrp1. HLA-A*0201/K^b transgenic mice (43) were immunized with plasmid DNA encoding wild-type Tyrp1 or Tyrp1ee/A*0201. Purified CD8⁺ T cells from immunized mice were assessed for responses to Tyrp1 epitopes by secretion of IFN-γ following stimulation with either TAP2-deficient T2 cells or HLA-negative K562 cells transfected with HLA-A*0201 (44) and pulsed with wild-type Tyrp1 peptides. Strong CD8⁺ T cell responses against 2 wild-type Tyrp1 peptides, 188–196 and 213–221, were detected only in mice immunized with Tyrp1ee/A*0201 (Figure (Figure7A).7A). CD8⁺ T cells against wild-type Tyrp1 peptide 213–221 responded to HLA-A*0201–positive, but not –negative, human melanoma cells, supporting processing and presentation of the 213–221 epitope by human melanoma cells (Figure (Figure7,7, B and C). CD8⁺ T cells against the Tyrp1 peptide 188–196 reacted with HLA-A*0201–positive melanoma cells but also responded weakly to HLA-A*0201–negative melanoma cells, making inferences about HLA-restricted processing and presentation for this epitope more difficult (data not shown). Of note, HLA-A*0201 transgenic mice express the mouse heavy chain α3 domain, while HLA-A*0201 molecules expressed by T2 and transfected K562 cells contain only the human heavy chain α3 domain, suggesting that newly generated Tyrp1-specific CD8⁺ T cells have high avidity because they respond independently of CD8 coengagement (45, 46).

Figure 7

Immunization with optimized Tyrp1 induces HLA-A*0201–restricted CD8⁺ T cells against Tyrp1.

Discussion

In this study we showed how altered ligands for MHC introduced into DNA encoding a tissue-specific self antigen led to multivalent T cell responses against that self antigen and also identified what we believe to be novel epitopes. We adapted the strategy to identify a HLA-A*0201–restricted Tyrp1 epitope that can be evaluated for clinical use in the future. Rationally altering individual MHC anchor residues not only can enhance MHC binding, but can also reposition solvent-facing residues for binding to CDR3 regions of T cell receptors for increased T cell activation (47, 48). We showed that these types of mutations also altered protein trafficking, leading to ER retention and enhancing availability of degradation products for processing by proteasomes.

Immunogenicity was further enhanced by creating different Tyrp1 glycoforms carrying mutations at Asn glycosylation sites to improve antigen processing and presentation. Sugars play roles both in protein folding and in protection from degradation by proteases. Notably, the Asn→Gln conversion still allowed inefficient glycosylation, creating isoforms with different numbers of carbohydrates. These glycosylation-deficient products included underglycosylated isoforms that were rapidly degraded for proteasome processing. In particular, glycoforms with 1 and 2 carbohydrate residues, along with the naked polypeptide, were rapidly degraded through a proteasome-dependent pathway. Immunity was greatest when mutations for altered peptides and glycosylation were combined.

Because Asn-linked glycosylation sites are highly conserved among members of the tyrosinase family (34), this optimization strategy should be applicable to other melanoma antigens. Furthermore, our studies have relevance to the design of optimized versions of other tissue-specific differentiation glycoproteins, overexpressed glycoproteins, and subdominant glycoprotein antigens of pathogens.

For self glycoproteins, immunization carries the risk of autoimmunity, which limits this approach when the target antigen is expressed by critical tissues. However, target antigens restricted to nonvital cell types and tissues such as melanocytes, prostate epithelium, or breast ductal epithelium are candidates for this strategy. A number of mouse studies have shown that autoimmune vitiligo is tolerated well when tumor immunity is induced against tyrosinase family antigens and that development of vitiligo is associated with improved response to therapy and outcome in patients with melanoma (reviewed in 1, 2, 13–16).

Prior reports in viral models and clinical observations have suggested that narrow T cell responses against individual epitopes can lead to persistence of virus, while broader responses are associated with control or resolution of infection (49–53). By using the strategy reported here, broader T cell responses against multiple epitopes can be generated with possible advantages, including a decreased likelihood of mutated tumor escape variants that might occur with altered ligands created with single amino acid substitutions (54). Thus we propose that the breadth and diversity of T cell responses is also important for cancer immunity (49).

Overall, our results indicate that it is the combined effect of epitope optimization, enrichment for multiple epitopes, and enhanced processing that leads to increased efficacy of immunization. The underlying mechanisms involved both facilitated processing and enhanced pMHC stability (with probable increased T cell activation). Thus multiple steps in the pathway to final antigen presentation can be manipulated to make multivalent vaccines. These observations provide broader strategies that can be applied to guide specificity of T cell responses in treatment with cytokines, cytotoxic T lymphocyte–associated antigen 4 blockade, recovering T cell homeostasis, and other nonspecific strategies known to generate cancer immunity and tumor rejection (25).

Methods

Mice.

Female C57BL/6 mice were purchased from the National Cancer Institute breeding program. HLA-A*0201/K^b were obtained from L. Sherman (Scripps Institute, La Jolla, California, USA; ref. 55). These mice were intercrossed and phenotyped for HLA-A*0201 expression by flow cytometry. All mice used in experiments were between 8 and 10 weeks of age. Care and use of these mice was in accordance with institutional guidelines under a protocol reviewed and approved by the Institutional Animal Care and Use Committee of Memorial Sloan-Kettering Cancer Center (MSKCC).

Antibodies.

The anti-CD8 mAb 53.6-72 (rat IgG) was obtained from the American Type Culture Collection. The anti-CD4 mAb (Gk1.5) and anti-NK mAb (PK136) were obtained from the MSKCC Monoclonal Antibody Core Facility. Flow cytometry and the MHC class I stabilization assays were performed as previously described (29) using a FACScan instrument equipped with Lysis II software (BD).

Peptides.

The Tyrp1 peptides used in this study (Table (Table1),1), OVA_25–264(SIINFEKL, K^b-restricted) and influenza virus nucleoprotein_366–374(ASNENMETM, D^b-restricted), were synthesized by GeneMed Synthesis Inc. and purified by HPLC to >80% purity. They were resuspended in 100% DMSO at a concentration of 40 mg/ml, diluted with PBS (final concentration, 2 mg/ml), and then frozen in aliquots at –20°C.

Plasmid constructs and site-directed mutagenesis.

The mouse Tyrp1 cDNA was mutated using reiterations of the Quickchange protocol (Stratagene). Bulk plasmid miniprep from each round of mutagenesis was used as template for the next. Candidate clones for each construct were sequenced after the last round. A unique AscI restriction site was inserted into pCDNA3.1(+) (Invitrogen Corp.) by cloning an oligonucleotide duplex in the NotI site of the polylinker 5′-ggcctggcgcgccgtacgttaacatcgatgc-3′ (upper strand) and 5′-GGCCGCATCGATGTTAACGTACGGCGCGCCA -3′ (lower strand). The new polylinker was then subcloned into pCR3 (Invitrogen Corp.) as a HindIII-XbaI fragment, yielding the pCRAN expression vector. All inserts described in the following paragraphs were confirmed by DNA sequencing and subcloned into the pCRAN vector.

For construction of Tyrp1ee, the following oligonucleotides were used (mutant bases are underlined): L9Y, 5′-CGTCCTCCCCTACGCCTATATCTCC-3′; C113M, 5′-CTGTGGGACTATGCGTCCTGGGTGG-3′; I182L, 5′-CCAAATTTGAGAACCTGTCCGTTTATAACTACTTTG-3′; H224Y, 5′-GCTTTTCTCACATGGTACAGGTACCATCTGCTG-3′; V260M, 5′-GTCTGCGATATGTGCACTGATGACTTGATGGG-3′; S351L, 5′-CCTCCTTTTTATTCCAATCTGACAGACAGTTTTCGAAATACAGTG-3′; P398Y, 5′-CCCATTTGTCTCCCAATGATTACATTTTTGTCCTCCTGCACAC-3′; A463M, 5′-CCAGACAATCTGGGATATATGTATGAAGTTCAATGGCCAGG-3′; A485N, 5′-CCATTGCTGTAGTGAACGCGTTGTTACTTGTAGCTGCC-3′; E524Y, 5′-CACTATCAACGCTATGCTTACGACTATGAGGAGCTC-3′.

The Tyrp1ng mutant was created by Asn→Gln mutations at amino acid positions 96, 104, 181, 304, 350, and 385 using the following primers (only sense primers are shown): N96Q, CCTGGCCTCTGAGGTTCTTTCAGAGAACATGTCAGTGCAATG; N104Q, GAACATGTCAGTGCAATGATCAGTTCTCAGGACACAACTGTG; N181Q, GCAACACACCACAATTTGAGCAGATTTCCGTTTATAACTACTTTG; N304Q, CCCTGGGAACACTTTGTCAGAGCACTGAGGGTGGACC; N350Q, GACACACCTCCTTTTTATTCCCAGTCTACAGACAGTTTTCG; N385Q, CCTGGCCCACCTCTTCCTGCAGGGAACGGGAGG.

These same primers were used to create Asn→Gln mutations in Tyrp1ee to generate Tyrp1ee/ng, which contained both potential altered peptide ligands and deficient glycosylation sites. For creation of Tyrp1/HLA-A*0201, the following primers were used: Tyrp1–188-A2.1 CATTTCCGTTTATAACTACTTTCTGTGGACACACTATTATTCAG; Tyrp1–213-A2.1, CTTTGGGGATGTGGATTTCCTGCACGAAGGACCCGCTTTTC; Tyrp1–238-A2.1, GGAGAGAGACATGCAGGAGTACCTGCAGGAGCCTTCTTTC; Tyrp1–249-A2.1, CTCCCTTCCTTACTGGCTGTTTGCAACTGGGAAAAACGTC; Tyrp1–401-A2.1, GCACACTTTCACTGATGTGGTCTTTGACGAATGGC; Tyrp1–439-A2.1, GACATAACAGGCAATACTACATGGTGCCATTCTGG.

The P. aeruginosa exotoxin A fusion vector was created as follows: genomic DNA from P. aeruginosa (American Type Culture Collection) was extracted using the DNeasy kit (Qiagen). A fragment encoding the first 404 amino acids of exotoxin A (including signal peptide) was amplified by PCR using the following primers: sense (start codon underlined; HindIII site italicized), 5′-TTTAAGCTTCCACCATGCACCTGATACCCCATTG-3′; antisense (AscI site italicized), 5′-AGGCGCGCCCGCATTCACCGGCGGCGAC-3′. The PCR product was cloned as a HindIII-AscI fragment into pCRAN, yielding the pCT5 expression vector, in which exotoxin was upstream (aminoterminal) from the Tyrp1 sequence of interest. Wild-type and mutant versions of Tyrp1 cDNA were flanked by AscI (upstream) and NotI (downstream) sites. These restriction sites were used for subcloning into pCRAN, yielding the Tyrp1 and Tyrp1ee constructs, or pCT5, yielding the exo-Tyrp1 and exo-Tyrp1ee constructs. The amino acid sequence at the in-frame junction with exotoxin was AGECGRATM (Cys 404 of exotoxin underlined; Met 1 of Tyrp1 in bold).

Transfection experiments.

To introduce a FLAG tag at the carboxyl terminus of Tyrp1 and its variants, cDNAs were reamplified by PCR using a downstream primer that omits the normal Stop codon and cloned between the EcoRV and SalI sites of the pCMV-Tag4A vector (Stratagene). The upstream primer was ggcgcgccaccatgaaatcttacaacgtcc, and the downstream primer was ATTCATAGTCGACGACCATGGAGTGGTTAGGATTC.

COS-7 cells (1–2 × 10⁶) were transfected with each FLAG-tagged Tyrp1 variant using the Fugene 6 reagent (Roche Diagnostics Corp.). At 24 hours after transfection, cells were treated for 1 hour with muconomyin A (0.6 μM; Sigma-Aldrich). Following washes, culture was prolonged either without further treatment or in the presence of the proteasome inhibitor MG132 (20 μM; Sigma-Aldrich) or chloroquine (50 μM). Cells were washed in PBS and lysed in 12.5 mM Tris (pH 7.5), 1.25% Triton X-100, and 190 mM NaCl for 30 minutes on ice. Cell lysates were normalized for protein content and analyzed by SDS-PAGE and Western blotting using peroxidase-conjugated M2 anti-FLAG mAb (Sigma-Aldrich).

Immunizations.

Genetic immunization using DNA-coated gold particles was performed using a gene gun provided by PowderMed Ltd. as previously described (14, 18). Four injections (400 pounds/inch²) were delivered to each mouse, one to each of the abdominal quadrants, for a total of 4 μg plasmid DNA per mouse per week for 4 weeks.

IFN-γ ELISPOT assay.

IP-Multiscreen plates (Millipore) were coated with 100 μl anti-mouse IFN-γ antibody (10 μg/ml; clone AN18; Mabtech) in PBS and incubated overnight at 4°C. After washing with PBS to remove unbound antibody, plates were blocked with RPMI containing 7% fetal bovine serum for 2 hours at 37°C. Purified CD8⁺ T cells were plated at a concentration of 10⁵ cells/well. For antigen presentation, 5 × 10⁴ irradiated EL-4 cells, T2 lymphoma cells, or HLA-A*0201/K562 cells pulsed with 10 μg/ml peptide were added to a final volume of 100 μl/well. After incubation for 20 hours (EL-4 or HLA-A*0201/K562 cells) or 7 days (T2 cells) at 37°C, plates were extensively washed with PBS plus 0.05% Tween and incubated with 100 μl/well of biotinylated antibody against mouse IFN-γ (2 μg/ml; clone R4-6A2; Mabtech). Plates were incubated for an additional 2 hours at 37°C, and spot development was performed as described (56). Spots were counted with an Automated ELISPOT Reader System with KS software (version 4.3; Carl Zeiss Inc.) by an independent investigator in a blinded fashion.

Tumor challenge.

These experiments were carried out with B16 melanoma cells as described previously (18). Prophylactic and therapeutic effects of the vaccine were determined by intradermally challenging mice with 10⁵ and 3 × 10⁴ B16F10 melanoma cells, respectively. Cells were injected into the shaved right flank of each mouse. Tumor growth was monitored at least 3 times per week, initially by palpation only and subsequently using Vernier calipers when tumors were palpated. Mice surviving tumor challenge were followed for a minimum of 40–100 days. If maximum tumor diameter reached 10 mm, the tumor became ulcerated, or mice showed discomfort, mice were euthanized. Tumor-free survival was plotted by Kaplan-Meier plots and compared by log-rank analysis.

Depletion of CD8+, CD4+, and NK cells.

Depletion of CD8⁺ T cells, CD4⁺ T cells, and NK cells in vivo was performed by intraperitoneal injections (3 days before starting immunization and then weekly for the remainder of the experiment) with 500 μg/injection of 53.6-72 rat anti-mouse mAb, Gk1.5 mAb, and PK136 mAb, respectively. Depletion of CD8⁺ T cells (≥93%), CD4⁺ T cells (≥94%), and NK cells (≥97%) was confirmed by fluorescence-activated cell sorting analysis of peripheral blood.

Endo H and N-glycanase analysis.

Lysates of 2 × 10⁴ COS-7 cells receiving no inhibitor treatment were digested 24 hours after transfection for 5 hours with 0.005 U Endo H or N-glycanase (Sigma-Aldrich) at 37°C. Samples were analyzed by SDS-PAGE as we have previously described (34, 35, 39).

Statistics.

Tumor-free survival was plotted using the Kaplan-Meier method, and statistical significance was determined by log-rank analysis. Differences were considered significant at P < 0.05.

Acknowledgments

Support for this work was provided by NIH grants R01 CA56821, P01 CA33049, and P01 CA59350 (to A.N. Houghton); Swim Across America; the Mr. William H. Goodwin and Mrs. Alice Goodwin and the Commonwealth Cancer Foundation for Research and the Experimental Therapeutics Center of MSKCC (to A.N. Houghton, J.A. Guevara-Patiño, M.E. Engelhorn); NIH training grant T32 CA09149 (to M.J. Turk); and a Fellowship award from the Cancer Research Institute (to M.E. Engelhorn). J.D. Wolchok is supported by a Damon Runyon/Eli Lilly Award, and A.N. Houghton has Damon Runyon/Eli Lilly mentorship support. We are extremely grateful to the Annenberg Hazen Foundation as well as to the Louis & Anne Abrons Foundation, the T.J. Martell Foundation, the Peter J. Sharp Foundation/Breast Cancer Research Foundation, and the Mr. and Mrs. Quentin J. Kennedy Fund. We thank David Posnett and Lisa Denzin for helpful comments, Rodica Stan for invaluable assistance in preparing the manuscript, and Linda Sherman for providing the HLA-A*0201/Kb mice.

Footnotes

Nonstandard abbreviations used: Endo H, endoglycosidase H; exo-, fusion with exotoxin A domain; pMHC, peptide complexed with MHC molecules; Tyrp1, tyrosinase-related protein-1; Tyrp1ee, epitope-enriched Tyrp1; Tyrp1ee/ng, epitope-enriched, glycosylation-deficient Tyrp1; Tyrp1ng, glycosylation-deficient Tyrp1.

Conflict of interest: The authors have declared that no conflict of interest exists.

Citation for this article: J. Clin. Invest. 116:1382–1390 (2006). doi:10.1172/JCI25591.

Mary Jo Turk’s present address is: Norris Cotton Cancer Center, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire, USA.

References

1. Overwijk W.W., Restifo N.P. Autoimmunity and the immunotherapy of cancer: targeting the “self” to destroy the “other”. Crit. Rev. Immunol. 2000;20:433–450. [PMC free article] [PubMed]

2. Houghton A.N., Guevara-Patino J.A. Immune recognition of self in immunity against cancer. J. Clin. Invest. 2004;114:468–471. doi: 10.1172/JCI200422685. [PMC free article] [PubMed] [Cross Ref]

3. Nikolic-Zugic J., Bevan M.J. Role of self-peptides in positively selecting the T-cell repertoire. Nature. 1990;344:65–67. [PubMed]

4. Tanchot C., Lemonnier F.A., Perarnau B., Freitas A.A., Rocha B. Differential requirements for survival and proliferation of CD8 naive or memory T cells. Science. 1997;276:2057–2062. [PubMed]

5. Ernst B., Lee D.S., Chang J.M., Sprent J., Surh C.D. The peptide ligands mediating positive selection in the thymus control T cell survival and homeostatic proliferation in the periphery. . Immunity. 1999;11:173–181. [PubMed]

6. Kieper W.C., Burghardt J.T., Surh C.D. A role for TCR affinity in regulating naive T cell homeostasis. J. Immunol. 2004;172:40–44. [PubMed]

7. Stefanova I., Horejsi V., Ansotegui I.J., Knapp W., Stockinger H. GPI-anchored cell-surface molecules complexed to protein tyrosine kinases. Science. 1991;254:1016–1019. [PubMed]

8. Townsend S.E., Allison J.P. Tumor rejection after direct costimulation of CD8+ T cells by B7-transfected melanoma cells. Science. 1993;259:368–370. [PubMed]

9. Kaplan M.H., Craig J.M. Immunologic studies of heart tissue. VI. Cardiac lesions in rabbits associated with autoantibodies to heart induced by immunization with heterologous heart. J. Immunol. 1963;90:725–733. [PubMed]

10. Tanaka N., Nishimura T., Tada T., Okabayashi A. Autoimmune phenomenon occurred in the course of prolonged stimulation of heterologous protein. Jpn. J. Exp. Med. 1964;34:53–57. [PubMed]

11. Weigle W.O. The induction of autoimmunity in rabbits following injection of heterologous or altered homologous thyroglobulin. J. Exp. Med. 1965;121:289–308. [PMC free article] [PubMed]

12. Naftzger C., et al. Immune response to a differentiation antigen induced by altered antigen: a study of tumor rejection and autoimmunity. . Proc. Natl. Acad. Sci. U. S. A. 1996;93:14809–14814. [PubMed]

13. Overwijk W.W., et al. gp100/pmel 17 is a murine tumor rejection antigen: induction of “self”-reactive, tumoricidal T cells using high-affinity, altered peptide ligand. J. Exp. Med. 1998;188:277–286. [PMC free article] [PubMed]

14. Weber L.W., et al. Tumor immunity and autoimmunity induced by immunization with homologous DNA. J. Clin. Invest. 1998;102:1258–1264. [PMC free article] [PubMed]

15. Bowne W.B., et al. Coupling and uncoupling of tumor immunity and autoimmunity. J. Exp. Med. 1999;190:1717–1722. [PMC free article] [PubMed]

16. Hawkins W.G., et al. Immunization with DNA coding for gp100 results in CD4 T-cell independent antitumor immunity. Surgery. 2000;128:273–280. [PubMed]

17. Hawkins W.G., et al. Xenogeneic DNA immunization in melanoma models for minimal residual disease. J. Surg. Res. 2002;102:137–143. [PubMed]

18. Dyall R., et al. Heteroclitic immunization induces tumor immunity. J. Exp. Med. 1998;188:1553–1561. [PMC free article] [PubMed]

19. Gold J.S., et al. A single heteroclitic epitope determines cancer immunity after xenogeneic DNA immunization against a tumor differentiation antigen. J. Immunol. 2003;170:5188–5194. [PubMed]

20. Rudolph M.G., Wilson I.A. The specificity of TCR/pMHC interaction. Curr. Opin. Immunol. 2002;14:52–65. [PubMed]

21. Tai T., Eisinger M., Ogata S.-I., Lloyd K.O. Glycoproteins as differentiation markers in human malignant melanoma and melanocytes. Cancer Res. 1983;43:2773–2777. [PubMed]

22. Vijayasaradhi S., Bouchard B., Houghton A.N. The melanoma antigen gp75 is the human homologue of the mouse b (brown) locus gene product. J. Exp. Med. 1990;171:1375–1380. [PMC free article] [PubMed]

23. Overwijk W.W., et al. Vaccination with a recombinant vaccinia virus encoding a “self” antigen induces autoimmune vitiligo and tumor cell destruction in mice: requirement for CD4(+) T lymphocytes. Proc. Natl. Acad. Sci. U. S. A. 1999;96:2982–2987. [PubMed]

24. Wolchok J.D., et al. Alternative roles for interferon-gamma in the immune response to DNA vaccines encoding related melanosomal antigens. . Cancer Immun. 2001;1:9. [PubMed]

25. Overwijk W.W., et al. Tumor regression and autoimmunity after reversal of a functionally tolerant state of self-reactive CD8+ T cells. J. Exp. Med. 2003;198:569–580. [PMC free article] [PubMed]

26. Barouch D.H., et al. Elicitation of high-frequency cytotoxic T-lymphocyte responses against both dominant and subdominant simian-human immunodeficiency virus epitopes by DNA vaccination of rhesus monkeys. J. Virol. 2001;75:2462–2467. [PMC free article] [PubMed]

27. Subbramanian R.A., et al. Magnitude and diversity of cytotoxic-T-lymphocyte responses elicited by multiepitope DNA vaccination in rhesus monkeys. J. Virol. 2003;77:10113–10118. [PMC free article] [PubMed]

28. Parker K.C., Shields M., DiBrino M., Brooks A., Coligan J.E. Peptide binding to MHC class I molecules: implications for antigenic peptide prediction. Immunol. Res. 1995;14:34–57. [PubMed]

29. Huard R., Dyall R., Nikolic-Zugic J. The critical role of a solvent-exposed residue of an MHC class I-restricted peptide in MHC-peptide binding. Int. Immunol. 1997;9:1701–1707. [PubMed]

30. Dyall R., Fremont D.H., Jameson S.C., Nikolic-Zugic J. T cell receptor (TCR) recognition of MHC class I variants: intermolecular second-site reversion provides evidence for peptide/MHC conformational variation. J. Exp. Med. 1996;184:253–258. [PMC free article] [PubMed]

31. Vijayasaradhi S., Xu Y., Bouchard B., Houghton A.N. Intracellular sorting and targeting of melanosomal membrane proteins: identification of signals for sorting of the human brown locus protein, gp75. J. Cell Biol. 1995;130:807–820. [PMC free article] [PubMed]

32. Roux L., Lloyd K.O. Glycosylation characteristics of pigmentation-associated antigen (GP75): an intracellular glycoprotein of human melanocytes and malignant melanomas. Arch. Biochem. Biophys. 1986;251:87–96. [PubMed]

33. Vijayasaradhi S., Doskoch P.M., Houghton A.N. Biosynthesis and intracellular movement of the melanosomal membrane glycoprotein gp75, the human b (brown) locus product. Exp. Cell Res. 1991;196:233–240. [PubMed]

34. Xu Y., et al. Diverse roles of conserved asparagine-linked glycan sites on tyrosinase family glycoproteins. Exp. Cell Res. 2001;267:115–125. [PubMed]

35. Xu Y., Setaluri V., Takechi Y., Houghton A.N. Sorting and secretion of a melanosome membrane protein, gp75/TRP1. J. Invest. Dermatol. 1997;109:788–795. [PubMed]

36. Ostankovitch M., Robila V., Engelhard V.H. Regulated folding of tyrosinase in the endoplasmic reticulum demonstrates that misfolded full-length proteins are efficient substrates for class I processing and presentation. J. Immunol. 2005;174:2544–2551. [PubMed]

37. Dorner A.J., et al. The stress response in Chinese hamster ovary cells. Regulation of ERp72 and protein disulfide isomerase expression and secretion. . J. Biol. Chem. 1990;265:22029–22034. [PubMed]

38. Travers K.J., et al. Functional and genomic analyses reveal an essential coordination between the unfolded protein response and ER-associated degradation. Cell. 2000;101:249–258. [PubMed]

39. Xu Y., Vijayasaradhi S., Houghton A.N. The cytoplasmic tail of the mouse brown locus product determines intracellular stability and export from the endoplasmic reticulum. J. Invest. Dermatol. 1998;110:324–331. [PubMed]

40. Ulmer J.B., Donnelly J.J., Liu M.A. Presentation of an exogenous antigen by major histocompatibility complex class I molecules. Eur. J. Immunol. 1994;24:1590–1596. [PubMed]

41. Lippolis J.D., et al. Pseudomonas exotoxin-mediated delivery of exogenous antigens to MHC class I and class II processing pathways. Cell Immunol. 2000;203:75–83. [PubMed]

42. Hung C.F., et al. Cancer immunotherapy using a DNA vaccine encoding the translocation domain of a bacterial toxin linked to a tumor antigen. Cancer Res. 2001;61:3698–3703. [PubMed]

43. Theobald M., et al. Tolerance to p53 by A2.1-restricted cytotoxic T lymphocytes. J. Exp. Med. 1997;185:833–841. [PMC free article] [PubMed]

44. Britten C.M., et al. The use of HLA-A*0201-transfected K562 as standard antigen-presenting cells for CD8(+) T lymphocytes in IFN-gamma ELISPOT assays. J. Immunol. Methods. 2002;259:95–110. [PubMed]

45. LaFace D.M., et al. Human CD8 transgene regulation of HLA recognition by murine T cells. . J. Exp. Med. 1995;182:1315–1325. [PMC free article] [PubMed]

46. Sherman L.A., Hesse S.V., Irwin M.J., La Face D., Peterson P. Selecting T cell receptors with high affinity for self-MHC by decreasing the contribution of CD8. Science. 1992;258:815–818. [PubMed]

47. Borbulevych O.Y., Baxter T.K., Yu Z., Restifo N.P., Baker B.M. Increased immunogenicity of an anchor-modified tumor-associated antigen is due to the enhanced stability of the peptide/MHC complex: implications for vaccine design. J. Immunol. 2005;174:4812–4820. [PMC free article] [PubMed]

48. Chen J.L., et al. Structural and kinetic basis for heightened immunogenicity of T cell vaccines. J. Exp. Med. 2005;201:1243–1255. [PMC free article] [PubMed]

49. Messaoudi I., Guevara Patino J.A., Dyall R., LeMaoult J., Nikolich-Zugich J. Direct link between mhc polymorphism, T cell avidity, and diversity in immune defense. Science. 2002;298:1797–1800. [PubMed]

50. Dalod M., et al. Broad, intense anti-human immunodeficiency virus (HIV) ex vivo CD8(+) responses in HIV type 1-infected patients: comparison with anti-Epstein-Barr virus responses and changes during antiretroviral therapy. J. Virol. 1999;73:7108–7116. [PMC free article] [PubMed]

51. Pantaleo G., et al. The qualitative nature of the primary immune response to HIV infection is a prognosticator of disease progression independent of the initial level of plasma viremia. Proc. Natl. Acad. Sci. U. S. A. 1997;94:254–258. [PubMed]

52. Day C.L., et al. Broad specificity of virus-specific CD4+ T-helper-cell responses in resolved hepatitis C virus infection. J. Virol. 2002;76:12584–12595. [PMC free article] [PubMed]

53. Kim A.Y., et al. The magnitude and breadth of hepatitis C virus-specific CD8+ T cells depend on absolute CD4+ T-cell count in individuals coinfected with HIV-1. Blood. 2005;105:1170–1178. [PubMed]

54. Khong H.T., Restifo N.P. Natural selection of tumor variants in the generation of “tumor escape” phenotypes. Nat. Immunol. 2002;3:999–1005. [PMC free article] [PubMed]

55. Vitiello A., Marchesini D., Furze J., Sherman L.A., Chesnut R.W. Analysis of the HLA-restricted influenza-specific cytotoxic T lymphocyte response in transgenic mice carrying a chimeric human-mouse class I major histocompatibility complex. J. Exp. Med. 1991;173:1007–1015. [PMC free article] [PubMed]

56. Scheibenbogen C., et al. A sensitive ELISPOT assay for detection of CD8+ T lymphocytes specific for HLA class I-binding peptide epitopes derived from influenza proteins in the blood of healthy donors and melanoma patients. Clin. Cancer Res. 1997;3:221–226. [PubMed]

Articles from The Journal of Clinical Investigation are provided here courtesy of
American Society for Clinical Investigation