Immunological validation of the EpitOptimizer program for streamlined design of heteroclitic epitopes

Combination of AFP vaccine and immune checkpoint inhibitors slows hepatocellular carcinoma progression in preclinical models

Many patients with hepatocellular carcinoma (HCC) do not respond to the first-line immune checkpoint inhibitor treatment. Immunization with effective cancer vaccines is an attractive alternative approach to immunotherapy. However, its efficacy remains insufficiently evaluated in preclinical studies. Here, we investigated HCC-associated self/tumor antigen, α-fetoprotein-based (AFP-based) vaccine immunization for treating AFP (+) HCC mouse models. We found that AFP immunization effectively induced AFP-specific CD8+ T cells in vivo. However, these CD8+ T cells expressed exhaustion markers, including PD1, LAG3, and Tim3. Furthermore, the AFP vaccine effectively prevented c-MYC/Mcl1 HCC initiation when administered before tumor formation, while it was ineffective against full-blown c-MYC/Mcl1 tumors. Similarly, anti-PD1 and anti-PD-L1 monotherapy showed no efficacy in this murine HCC model. In striking contrast, AFP immunization combined with anti-PD-L1 treatment triggered significant inhibition of HCC progression in most liver tumor nodules, while in combination with anti-PD1, it induced slower tumor progression. Mechanistically, we demonstrated that HCC-intrinsic PD-L1 expression was the primary target of anti-PD-L1 in this combination therapy. Notably, the combination therapy had a similar therapeutic effect in the cMet/β-catenin mouse HCC model. These findings suggest that combining the AFP vaccine and immune checkpoint inhibitors may be effective for AFP (+) HCC treatment.

Calreticulin mutant myeloproliferative neoplasms induce MHC-I skewing, which can be overcome by an optimized peptide cancer vaccine

The majority of JAK2V617F-negative myeloproliferative neoplasms (MPNs) have disease-initiating frameshift mutations in calreticulin (CALR), resulting in a common carboxyl-terminal mutant fragment (CALRMUT), representing an attractive source of neoantigens for cancer vaccines. However, studies have shown that CALRMUT-specific T cells are rare in patients with CALRMUT MPN for unknown reasons. We examined class I major histocompatibility complex (MHC-I) allele frequencies in patients with CALRMUT MPN from two independent cohorts. We observed that MHC-I alleles that present CALRMUT neoepitopes with high affinity are underrepresented in patients with CALRMUT MPN. We speculated that this was due to an increased chance of immune-mediated tumor rejection by individuals expressing one of these MHC-I alleles such that the disease never clinically manifested. As a consequence of this MHC-I allele restriction, we reasoned that patients with CALRMUT MPN would not efficiently respond to a CALRMUT fragment cancer vaccine but would when immunized with a modified CALRMUT heteroclitic peptide vaccine approach. We found that heteroclitic CALRMUT peptides specifically designed for the MHC-I alleles of patients with CALRMUT MPN efficiently elicited a CALRMUT cross-reactive CD8+ T cell response in human peripheral blood samples but not to the matched weakly immunogenic CALRMUT native peptides. We corroborated this effect in vivo in mice and observed that C57BL/6J mice can mount a CD8+ T cell response to the CALRMUT fragment upon immunization with a CALRMUT heteroclitic, but not native, peptide. Together, our data emphasize the therapeutic potential of heteroclitic peptide-based cancer vaccines in patients with CALRMUT MPN.

Machine learning optimization of peptides for presentation by class II MHCs

T cells play a critical role in cellular immune responses to pathogens and cancer and can be activated and expanded by MHC-presented antigens contained in peptide vaccines. We present a machine learning method to optimize the presentation of peptides by class II MHCs by modifying their anchor residues. Our method first learns a model of peptide affinity for a class II MHC using an ensemble of deep residual networks, and then uses the model to propose anchor residue changes to improve peptide affinity. We use a high throughput yeast display assay to show that anchor residue optimization improves peptide binding. Supplementary information: Supplementary data are available at Bioinformatics online.

Identification and characterization of heteroclitic peptides in TCR-binding positions with improved HLA-binding efficacy

The antigenicity as well as the immunogenicity of tumor associated antigens (TAAs) may need to be potentiated in order to break the immunological tolerance. To this aim, heteroclitic peptides were designed introducing specific substitutions in the residue at position 4 (p4) binding to TCR. The effect of such modifications also on the affinity to the major histocompatibility class I (MHC-I) molecule was assessed. The Trp2 antigen, specific for the mouse melanoma B16F10 cells, as well as the HPV-E7 antigen, specific for the TC1 tumor cell lines, were used as models. Affinity of such heteroclitic peptides to HLA was predicted by bioinformatics tools and the most promising ones were validated by structural conformational and HLA binding analyses. Overall, we demonstrated that TAAs modified at the TCR-binding p4 residue are predicted to have higher affinity to MHC-I molecules. Experimental evaluation confirms the stronger binding, suggesting that this strategy may be very effective for designing new vaccines with improved antigenic efficacy.

Engineering α-fetoprotein-based gene vaccines to prevent and treat hepatocellular carcinoma: review and future prospects

Activation of a patient's immune system offers an attractive approach to prevent and treat hepatocellular carcinoma (HCC). However, the antitumor efficacy of current HCC vaccines was weak owing to insufficient immune activation of targeting self/tumor antigens. We recently found that epitope-optimized α-fetoprotein effectively activated CD8 T cells and generated potent antitumor effects in the carcinogen-induced autochthonous HCC mouse model. We predict that the same antigen engineering approach of epitope-optimization will enable us to develop effective human vaccines to prevent HCC recurrence after liver resection. The engineered human HCC vaccines may also allow us to identify high-affinity T-cell receptors and antibodies that can be used to reprogram T cells to treat HCC tumors via adoptive transfer.

T-cell receptors: tugging on the anchor for a tighter hold on the tumor-associated peptide

Although it has been shown that human tumor-associated, HLA anchor residue modified "heteroclitic" peptides may induce stronger immune responses than wild-type peptides in cancer vaccine trials, it has also been shown that some T cells primed with these heteroclitic peptides subsequently fail to recognize the natural, tumor-expressed peptide efficiently. This may provide a molecular reason for why clinical trials of these peptides have been thus far unsuccessful. In this issue of the European Journal of Immunology, Madura et al. [Eur. J. Immunol. 2015. 45: 584-591] highlight a novel twist on T-cell receptor (TCR) recognition of HLA-peptide complexes. Tumor-associated peptides often lack canonical anchor residues, which can be substituted for the optimal residue to improve their antigenicity. T-cell cross-reactivity between the natural and modified (heteroclitic) peptides is essential for this approach to work and depends on whether the anchor residue substitution influences peptide conformation. The Melan-A/MART-126-35 peptide epitope is an example where T cells can make this distinction, with the natural peptide stimulating higher affinity CD8(+) T cells than the heteroclitic peptide, despite the heteroclitic peptide's more stable association with HLA-A2. The molecular basis for peptide discrimination is identified through the structure of the TCR bound to the natural peptide; TCR engagement of the natural peptide "lifts" its amino-terminus partly away from the HLA peptide binding groove, forming a higher affinity interface with the TCR than is formed with the anchor residue "optimized" heteroclitic peptide, which cannot be "pulled" from the HLA groove.

Structural basis for ineffective T-cell responses to MHC anchor residue-improved "heteroclitic" peptides

MHC anchor residue-modified "heteroclitic" peptides have been used in many cancer vaccine trials and often induce greater immune responses than the wild-type peptide. The best-studied system to date is the decamer MART-1/Melan-A26-35 peptide, EAAGIGILTV, where the natural alanine at position 2 has been modified to leucine to improve human leukocyte antigen (HLA)-A*0201 anchoring. The resulting ELAGIGILTV peptide has been used in many studies. We recently showed that T cells primed with the ELAGIGILTV peptide can fail to recognize the natural tumor-expressed peptide efficiently, thereby providing a potential molecular reason for why clinical trials of this peptide have been unsuccessful. Here, we solved the structure of a TCR in complex with HLA-A*0201-EAAGIGILTV peptide and compared it with its heteroclitic counterpart , HLA-A*0201-ELAGIGILTV. The data demonstrate that a suboptimal anchor residue at position 2 enables the TCR to "pull" the peptide away from the MHC binding groove, facilitating extra contacts with both the peptide and MHC surface. These data explain how a TCR can distinguish between two epitopes that differ by only a single MHC anchor residue and demonstrate how weak MHC anchoring can enable an induced-fit interaction with the TCR. Our findings constitute a novel demonstration of the extreme sensitivity of the TCR to minor alterations in peptide conformation.

Epitope-optimized alpha-fetoprotein genetic vaccines prevent carcinogen-induced murine autochthonous hepatocellular carcinoma

Immunization with effective cancer vaccines can offer a much needed adjuvant therapy to fill the treatment gap after liver resection to prevent relapse of hepatocellular carcinoma (HCC). However, current HCC cancer vaccines are mostly based on native shared-self/tumor antigens that are only able to induce weak immune responses. In this study we investigated whether the HCC-associated self/tumor antigen of alpha-fetoprotein (AFP) could be engineered to create an effective vaccine to break immune tolerance and potently activate CD8 T cells to prevent clinically relevant carcinogen-induced autochthonous HCC in mice. We found that the approach of computer-guided methodical epitope-optimization created a highly immunogenic AFP and that immunization with lentivector expressing the epitope-optimized AFP, but not wild-type AFP, potently activated CD8 T cells. Critically, the activated CD8 T cells not only cross-recognized short synthetic wild-type AFP peptides, but also recognized and killed tumor cells expressing wild-type AFP protein. Immunization with lentivector expressing optimized AFP, but not native AFP, completely protected mice from tumor challenge and reduced the incidence of carcinogen-induced autochthonous HCC. In addition, prime-boost immunization with the optimized AFP significantly increased the frequency of AFP-specific memory CD8 T cells in the liver that were highly effective against emerging HCC tumor cells, further enhancing the tumor prevention of carcinogen-induced autochthonous HCC.

CONCLUSIONS

Epitope-optimization is required to break immune tolerance and potently activate AFP-specific CD8 T cells, generating effective antitumor effect to prevent clinically relevant carcinogen-induced autochthonous HCC in mice. Our study provides a practical roadmap to develop effective human HCC vaccines that may result in an improved outcome compared to the current HCC vaccines based on wild-type AFP.

Engagement of NK receptor NKG2D, but not 2B4, results in self-reactive CD8+ T cells and autoimmune vitiligo

In this study, we demonstrate that engagement of two different natural killer receptors (NKRs) can lead to contrasting effects in the development of self-reactive CD8+T cells and autoimmune vitiligo. Specifically, using a mouse model, we show that CD8+T-cell targeting of a melanocyte antigen, tyrosinase-related protein-1 (TRP-1) in combination with delivery of the NKG2D ligands (Rae-1ϵ or H60), results in strong CD8+T-cell responses against TRP-1 and in the development of autoimmune vitiligo. In contrast, targeting of TRP-1 in combination with delivery of CD48, the natural ligand for the NKR 2B4, leads to reduced formation of TRP-1-reactive CD8+T-cell responses and decreased development of vitiligo. These data indicate that autoimmune vitiligo is limited by insufficient signals, despite plentiful self-reactive T cells in the peripheral immune system. To our knowledge, this is the first experimental evidence supporting the role of NKRs in modulating CD8+T-cell autoimmune vitiligo. This study supports the utilization of NKR signaling as a therapeutic avenue toward prevention of vitiligo and other autoimmune diseases.

Combination of epitope-optimized DNA vaccination and passive infusion of monoclonal antibody against HER2/neu leads to breast tumor regression in mice

HER2/neu is an oncogene amplified and over-expressed in 20-30% of breast adenocarcinomas. Treatment with the humanized monoclonal antibody trastuzumab has shown efficacy in combination with cytotoxic agents, although resistance occurs over time. Novel approaches are needed to further increase antibody efficacy. In this study, we provide evidence in a mouse breast cancer therapeutic tumor model that the combination of active immunization with a modified HER2/neu DNA vaccine and passive infusion of an anti-HER2/neu monoclonal antibody leads to significant regression of established tumors. Our data indicate that combination therapy with a HER2/neu DNA vaccine and trastuzumab may have clinical activity in breast cancer patients.

Development of tumor-infiltrating CD8+ T cell memory precursor effector cells and antimelanoma memory responses are the result of vaccination and TGF-β blockade during the perioperative period of tumor resection

A main goal of cancer immunology research is the formation of Ag-specific memory T cell immunity capable of activation upon tumor re-encounter. The requirements necessary to overcome the inhibitory signals present in the tumor microenvironment and form such memory T cell responses are unknown. In contrast to previous studies targeting tumors expressing highly immunogenic model Ags, we demonstrate that alleviating tumor-induced suppression along with vaccination against authentic Ags during the perioperative period provides long-lasting protection against a highly suppressive and poorly immunogenic melanoma. In this study, we employed DNA vaccination with an immunologically optimized mouse melanoma-shared Ag, Trp1ee/ng, combined with systemic TGF-β blockade during the perioperative period of primary tumor resection, to confer protection against B16 melanoma, and against JBRH, an independently derived melanoma unrelated to B16. Importantly, we demonstrate that correlative to memory responses, perioperative immunotherapy increases the formation of tumor-infiltrating and tumor-reactive CD8(+) T cells expressing low levels of the transcription factor T-bet, defined as memory precursor effector cells. We show that conditions for an immunologically fertile environment are met when TGF-β blockade and vaccination are applied during the perioperative period of primary tumor resection. These findings address limitations of current CD8(+) T cell immunotherapies against cancer by generating effective CD8(+) T cell memory recall responses.

Modification of MHC anchor residues generates heteroclitic peptides that alter TCR binding and T cell recognition

Improving T cell Ags by altering MHC anchor residues is a common strategy used to enhance peptide vaccines, but there has been little assessment of how such modifications affect TCR binding and T cell recognition. In this study, we use surface plasmon resonance and peptide-MHC tetramer binding at the cell surface to demonstrate that changes in primary peptide anchor residues can substantially and unpredictably alter TCR binding. We also demonstrate that the ability of TCRs to differentiate between natural and anchor-modified heteroclitic peptides distinguishes T cells that exhibit a strong preference for either type of Ag. Furthermore, we show that anchor-modified heteroclitic peptides prime T cells with different TCRs compared with those primed with natural Ag. Thus, vaccination with heteroclitic peptides may elicit T cells that exhibit suboptimal recognition of the intended natural Ag and, consequently, impaired functional attributes in vivo. Heteroclitic peptide-based immune interventions therefore require careful evaluation to ensure efficacy in the clinic.

Identification and modification of an HLA-A*0201-restricted cytotoxic T lymphocyte epitope from Ran antigen

Ran is considered to be a promising target for tumor-specific immunotherapy because its protein is exclusively expressed in tumor tissues, though its mRNA can be expressed in most normal tissues. In our study, we obtained four candidate wild-type epitopes designated Ran1, Ran2, Ran3, and Ran4, derived from the Ran antigen with the highest predicted affinity with MHC-I, indicated by affinity prediction plots and molecular dynamics simulation. However, in vitro affinity assays of these epitopes showed only a moderate affinity with MHC-I. Thus, we designed altered peptide ligands (APLs) derived from Ran wild-type epitopes with preferred primary and auxiliary HLA-A*0201 molecule anchor residue replacement. Of the eight tested peptides, the 1Y analog had the strongest binding-affinity and lowest-dissociation rate to HLA-A*0201. Additionally, we investigated the CTLs activities induced by Ran wild-type peptides and the APLs in human PBMCs and in HLA-A*0201/K(b) transgenic mice. Ran1 1Y was superior to other APLs and wild-type peptides in eliciting epitope-specific CTL immune responses both in vitro and in vivo. In summary, a wild-type epitope of the tumor-specific antigen Ran, expressed broadly in many tumors, was identified and designated Ran1. An APL of Ran1, Ran1 1Y, was further designed and verified in vitro and in vivo and found to elicit a stronger Ran-specific CTL response, indicating a potential anti-tumor application in the future.

Prospects for developing an effective particle-mediated DNA vaccine against influenza

Vaccine strategies capable of conferring broad protection against both seasonal and pandemic strains of influenza are urgently needed. DNA vaccines are an attractive choice owing to their capacity to induce robust humoral and cellular immune responses at low doses and because they can be developed and manufactured rapidly to more effectively meet the threat of an influenza epidemic or pandemic. Particle-mediated epidermal delivery (PMED), or the gene gun, is a DNA vaccine delivery technology shown to induce protective levels of antibody and T-cell responses in animals and humans against a wide variety of diseases, including influenza. This review focuses on current advances toward the development of an effective PMED DNA vaccine against influenza, including strategies to enhance vaccine immunogenicity, the potential for PMED-based DNA vaccines to improve protection in the vulnerable elderly population, and the prospects for a vaccine capable of providing cross-protection against both seasonal and pandemic strains of influenza.

Towards in silico design of epitope-based vaccines

BACKGROUND

Epitope-based vaccines (EVs) make use of immunogenic peptides (epitopes) to trigger an immune response. Due to their manifold advantages, EVs have recently been attracting growing interest. The success of an EV is determined by the choice of epitopes used as a basis. However, the experimental discovery of candidate epitopes is expensive in terms of time and money. Furthermore, for the final choice of epitopes various immunological requirements have to be considered.

METHODS

Numerous in silico approaches exist that can guide the design of EVs. In particular, computational methods for MHC binding prediction have already become standard tools in immunology. Apart from binding prediction and prediction of antigen processing, methods for epitope design and selection have been suggested. We review these in silico approaches for epitope discovery and selection along with their strengths and weaknesses. Finally, we discuss some of the obvious problems in the design of EVs.

CONCLUSION

State-of-the-art in silico approaches to MHC binding prediction yield high accuracies. However, a more thorough understanding of the underlying biological processes and significant amounts of experimental data will be required for the validation and improvement of in silico approaches to the remaining aspects of EV design.

Modified peptides in anti-cancer vaccines: are we eventually improving anti-tumour immunity?

The discovery of tumour antigens recognized by T cells and the features of immune responses directed against them has paved the way to a multitude of clinical studies aimed at boosting anti-tumour T cell immunity as a therapeutic tool for cancer patients. One of the different strategies explored to ameliorate the immunogenicity of tumour antigens in vaccine protocols is represented by the use of optimized peptides or altered peptide ligands, whose amino acid sequence has been modified for improving HLA binding or TCR interaction with respect to native epitopes. However, despite the promising results achieved with preclinical studies, the clinical efficacy of this approach has not yet met the expectations. Although multiple reasons could explain the relative failure of altered peptide ligands as more effective cancer vaccines, the possibility that T cells primed by modified tumour peptides might may be unable to effectively cross-recognize tumour cells has not been sufficiently addressed. Indeed, the introduction of conservative amino acid substitutions may still produce diverse and unpredictable changes in the HLA/peptide interface, with consequent modifications of the TCR repertoire that can interact with the complex. This could lead to the expansion of a broad array of T cells whose TCRs may not necessarily react with equivalent affinity with the original antigenic epitope. Considering the results presently achieved with this vaccine approach, and the emerging availability of alternative strategies for boosting anti-tumour immunity, the use of modified tumour peptides could be reconsidered.

A modified epitope identified for generation and monitoring of PSA-specific T cells in patients on early phases of PSA-based immunotherapeutic protocols

Efficacy of vaccination in cancer patients on immunotherapeutic protocols can be difficult to evaluate. The aim of this study was therefore to identify a single natural or modified epitope in prostate-specific antigen (PSA) with the ability to generate high levels of PSA-specific T cells to facilitate monitoring in patients after vaccination against prostate cancer. To the best of our knowledge, this study describes for the first time the peptide specificity of T cells stimulated by endogenously processed PSA antigen. The peptide specificity of HLA-A*0201-restricted CD8(+) T cells against human and rhesus PSA was investigated both in vivo after DNA vaccination in HLA-A*0201-transgenic mice and in vitro after repetitive stimulation of human T cells with DNA-transfected human dendritic cells (DCs). One of seven native PSA peptides, psa53-61, was able to activate high levels of PSA-specific CD8(+) T cells in HLA-A*0201-transgenic mice after PSA DNA vaccination. Psa53-61 was also the only peptide that induced human T cells to produce IFNgamma after stimulation with PSA transfected DCs, however not in all donors. Therefore, plasmids encoding modified epitopes in predicted HLA-A*0201 sequences were constructed. One of these modified PSA plasmids consistently induced IFNgamma producing CD8(+) T cells to the corresponding modified peptide as well as to the corresponding native peptide, in all murine and human T cell cultures. This study demonstrates a novel concept of introducing a modified epitope within a self-tumor antigen, with the purpose of eliciting a reliable T cell response from the non-tolerized immune repertoire, to facilitate monitoring of vaccine efficacy in cancer patients on immunotherapeutic protocols. The purpose of such a modified epitope is thus not to induce therapeutically relevant T cells but rather to, in case of weak or divergent T cell responses to self antigens/peptides, help answer questions about efficacy of vaccine delivery and about the possibility to induce immune responses in the selected and often immunosuppressed cancer patients.

Keratinocyte growth factor enhances DNA plasmid tumor vaccine responses after murine allogeneic bone marrow transplantation

Keratinocyte growth factor (KGF), which is given exogenously to allogeneic bone marrow transplantation (allo-BMT) recipients, supports thymic epithelial cells and increases thymic output of naive T cells. Here, we demonstrate that this improved T-cell reconstitution leads to enhanced responses to DNA plasmid tumor vaccination. Tumor-bearing mice treated with KGF and DNA vaccination have improved long-term survival and decreased tumor burden after allo-BMT. When assayed before vaccination, KGF-treated allo-BMT recipients have increased numbers of peripheral T cells, including CD8(+) T cells with vaccine-recognition potential. In response to vaccination, KGF-treated allo-BMT recipients, compared with control subjects, generate increased numbers of tumor-specific CD8(+) cells, as well as increased numbers of CD8(+) cells producing interferon-gamma (IFN-gamma) and tumor necrosis factor-alpha (TNF-alpha). We also found unanticipated benefits to antitumor immunity with the administration of KGF. KGF-treated allo-BMT recipients have an improved ratio of T effector cells to regulatory T cells, a larger fraction of effector cells that display a central memory phenotype, and effector cells that are derived from a broader T-cell-receptor repertoire. In conclusion, our data suggest that KGF can function as a potent vaccine adjuvant after allo-BMT through its effects on posttransplantation T-cell reconstitution.

Safety and immunogenicity of tyrosinase DNA vaccines in patients with melanoma

Immunity to self antigens on cancer is constrained by tolerance/ignorance. DNA vaccines encoding xenogeneic differentiation antigens, such as tyrosinase (TYR), mediate tumor protection and regression in implantable mouse models, and dogs with spontaneous melanoma. We conducted a trial of mouse and human TYR DNA vaccines in stage III/IV melanoma patients. Eighteen human leukocyte antigen (HLA)-A*0201(+) melanoma patients were randomized as follows: one group received three mouse TYR DNA injections followed by three human TYR DNA injections; the other group received the same vaccines in opposite sequence. The study was conducted at three dose levels: 100, 500, and 1,500 microg DNA/injection, administered intramuscularly (IM) every 3 weeks. Most toxicities were grade 1 injection site reactions. Seven patients developed CD8(+) T-cell responses, defined by a >3 SD increase in baseline reactivity to TYR peptide in tetramer or intracellular cytokine staining (ICS) assays. There was found to be no relationship between dose, assigned schedule, and T-cell response. At a median of 42 months follow-up, median survival has not been reached. Mouse and human TYR DNA vaccines were found safe and induced CD8(+) T-cell responses in 7 of 18 patients. T cells recognizing a native TYR peptide had a phenotype consistent with that of effector memory cells.