The augmentation of high-titer antibodies to ATP6S1 is associated with favorable clinical outcomes in patients who received vaccination with autologous, irradiated tumor cells engineered to secrete GM-CSF and allogeneic bone marrow transplantation. Cellular immune responses to ATP6S1 are unknown. To define its role as an immune target, examination of cellular responses to ATP6S1 and immunity related to current therapies such as checkpoint blockade is needed. We used an overlapping peptide library representing the full-length ATP6S1 protein to screen for cellular responses from the peripheral blood of patients with stage III and IV melanoma. Reactive peptide pools were used to determine the individual peptide activity and epitopes. Recombinant ATP6S1 protein was used in an ELISA to assess potential correlation with humoral immune responses and changes in immunity related to CTLA-4 blockade with ipilimumab in these patients. We observed a broad array of CD4+ and CD8+ cellular responses against ATP6S1, including the identification of several MHC class I and II ATP6S1 epitopes. The generation of specific CD4+ and cytotoxic T cells revealed potent functional capability elicited by ipilimumab treatment in patients with metastatic melanoma, which revealed potent functional capability, including cytokine production, proliferation responsiveness to melanoma cell lines, and tumor-cell killing. Furthermore, the augmented humoral immune responses to ATP6S1 as a function of ipilimumab treatment were associated with beneficial clinical outcomes. These results support the continued development of ATP6S1 as a biomarker and therapeutic target. Cancer Immunol Res; 3(1); 59–67. ©2014 AACR.
ATP6S1 is an accessory unit of the Vacuolar–ATPase complex. Vacuolar-ATPase (v-ATPase) functions as an important regulator of intracellular acidification by translocating protons. As a result, it is critical in the processes of cell signaling, trafficking, enzyme degradation, and endocytosis. These intracellular activities are associated with tissue and organ functions such as renal acidification, bone resorption, and eye pigmentation (1–3). In cancer cells, v-ATPase upregulates glycolysis and acidifies the tumor microenvironment (4). Decrease in ATPase activity by specific inhibitors leads to reduction of tumor metastasis, suppression of tumor growth and survival, and prevention of chemoresistance (5).
We have previously demonstrated that patients with a variety of malignancies develop high-titer IgG antibodies against ATP6S1 following vaccination with irradiated, autologous GM-CSF–secreting tumor cells or allogeneic bone marrow transplantation. Furthermore, these humoral responses to ATP6S1 are associated with intense tumor infiltration by dendritic cells, macrophages, eosinophils, CD4+ and CD8+ T cells, along with tumor destruction, fibrosis, and edema in long-term surviving patients (6). The biologic function of ATP6S1 and immunogenicity associated with favorable pathologic and clinical outcomes suggest the need for better understanding its potential immunotherapeutic role.
Materials and Methods
PBMCs and sera
Peripheral blood samples were obtained from patients with melanoma and healthy donors on the protocols approved by the Dana-Farber Cancer Center Institutional Review Board. Peripheral blood was collected in heparinized tubes, and peripheral blood mononuclear cells (PBMC) were isolated by gradient centrifugation using Ficoll-Paque Plus (GE Healthcare Bio-Sciences).
To evaluate T-cell responses to ATP6S1 in patients with melanoma, a series of 15- to 18-mer peptides were generated to cover the full-length protein sequence with 10-amino acid overlap between each individual peptide (Fig. 1). Screening with the library of overlapping peptides was performed as described previously (7). Briefly, peptides were synthesized by New England Peptide. Cells were maintained in RPMI-1640 supplemented with 10% FCS. PBMCs of patients with melanoma were treated with the peptide pools for 1 week, and then restimulated with the same peptide pools for readout of the ELISPOT assay. PBMCs (1 × 106) were sensitized with 1 μg/mL of each peptide in overlapping peptide pools in the presence of 10 U/mL IL2 (BD Biosciences) for 1 week. Cells (5 × 104 per well) were restimulated with the same peptides at 10 μg/mL and autologous PBMCs overnight in a 96-well plate (Millipore) precoated with anti-IFNγ Ab (Mabtech). The plates were washed, probed with biotinylated anti-IFNγ Ab and streptavidin-ALP (Mabtech), developed with TMB (3,3′-5,5′ tetramethylbenzidine) + substrate + chromogen and BCIP (Promega), and counted using an Immunospot reader (C.T.L. Cellular Technology Ltd.). CD4+ and CD8+ T-cell responses to individual peptides (>90% purity) were further determined by depletion of CD4+ and CD8+ T cells through MACS separation columns (Miltenyi Biotec). Samples were performed in duplicate. The HIV peptide L11 (LLFGYPVYV) was used as a negative control.
Generation of antigen-specific T cells
HLA A*0201 and DRB1 alleles were genotyped by PCR amplification and sequencing (8, 9) for each patient or normal donor evaluated. HLA binding prediction programs (http://www-bimas.cit.nih.gov/molbio/hla_bind/) were used to identify peptides for further analyses of specific epitope reactivity. To determine the binding of HLA A*0201 epitopes, the peptide-induced HLA A*0201 stabilization assay was performed as described previously (10). Briefly, T2 cells (ATCC) were incubated with 50 μmol/L peptide and 1 μmol/L β2Microglobulin (Sigma) in RPMI serum-free medium overnight at 37°C. The expression of HLA A*0201 on T cells was analyzed with anti-HLA A*2 FITC antibody (AbD Serotec). Fluorescence index (FI) = (mean FITC fluorescence with the peptides − mean FITC fluorescence without peptide)/(mean FITC fluorescence without peptide).
To generate antigen-specific CD4+ T cells, naïve CD4+ T cells were isolated from PBMCs using CD4+ microbeads (Miltenyi Biotec), and stimulated with irradiated autologous PBMCs in the presence of 10 μg/mL peptide. After two to four rounds of stimulation, the antigen-specific T cells were enriched with IFNγ selection (Miltenyi Biotec), and further expanded with 10 μg/mL PHA, irradiated allogeneic/autologous PBMCs, and 10 U/mL IL2. To test for haplotype specificity, 10 μg/mL HLA-DR antibody (L243; BD Biosciences) was used to block the interactions between CD4+ T cells and APCs in the presence of 10 μg/mL peptide.
To generate antigen-specific CD8+ T-cell clones, naïve CD8+ T cells were isolated from PBMCs with CD8+ microbeads (Miltenyi Biotec), and stimulated with irradiated CD40-activated autologous B cells in the presence of 10 μg/mL peptide and 10 U/mL IL2. After two to four rounds of stimulation, antigen-specific T cells were plated in limiting dilution, and expanded with 10 μg/mL PHA (Sigma), irradiated allogeneic PBMCs, and 10 U/mL IL2. To test for specificity, 10 μg/mL HLA class I antibody (W6/32; AbD Serotec) was used to block the interactions between T cells and T2 cells in the presence of peptides.
Cytotoxic T-cell Cr51 release assay
Functional activities of antigen-specific cytotoxic T lymphocytes (CTL) were further analyzed in a standard 4-hour Cr51 release assay. Briefly, 1 × 107 T2 cells or melanoma cell lines (K008, KO28, and K029 were generated from resected melanoma tumors from patients treated at the Dana-Farber Cancer Institute per approved DFCI IRB protocol) as target cells were labeled with 200 μCi Cr51 for 60 minutes at 37°C. After washing, 5 × 103 per well labeled cells were mixed with antigen-specific CTLs at the indicated ratio in the presence of 5 μg/mL peptide (L11 control, ATP78, ATP136, and ATP247). Supernatants were counted in a gamma counter (COBRA; PerkinElmer). All assays were performed in duplicate. Percent cytotoxicity was defined as (sample-spontaneous release)/(total release − spontaneous release) × 100. All cell lines used were tested to be Mycoplasma free. No other authentication assay was performed.
Flow cytometric analysis
Cells were stained with fluorescence-conjugated antibodies, and further analyzed using the BD FACSCanto II analyzer. CD4-FITC, CD8-FITC, and CD8PE (BD Biosciences) were used. Anti-IFNγ PE was purchased from Miltenyi Biotec.
T cells (1 × 104 cells/well) were cocultured with 2 × 104 irradiated melanoma cells for 2 days, and further pulsed with [3H]thymidine (0.5 μCi H3/well) overnight. The incorporated radioactivity was measured in a liquid scintillation counter (Wallac 1450 Microbeta Trilux; PerkinElmer).
To determine sera responses to ATP6S1, ELISA assays were conducted as previously described (7). Briefly, 127 pmol/well of GST-tagged recombinant ATP6S1 (Abnova) and recombinant GST as a control were coated on Nunc-Immno-plates overnight at 4°C. Plates were washed with PBS and blocked with 2% nonfat milk in PBS. Patient sera (100 μL; diluted 1:500 in 2% nonfat milk) was added per well and incubated overnight at 4°C. Plates were then washed with PBS/Tween-20, and incubated per well with 100 μL of horseradish peroxidase (HRP)–conjugated goat anti-human IgG (H+L) antibody (Life Technologies) diluted 1:2,000 in 2% nonfat milk for 2 hours at room temperature for further development with nitroblue tetrazolium substrate (Dako). Plates were read at an optical density (OD) of 450 nm. All samples were tested in duplicate. Values were reported as the mean OD of the sample wells minus the mean OD from wells coated with GST. Three-fold increases in mean OD between pre- and posttreatment of patient samples are considered positive.
Two-sample t tests were used to provide a uniform method for assessment of differences between peptide and negative control average frequencies based on ELISPOT. Peptides and negative control assays were conducted in duplicate. Differences between peptides and controls were evaluated in instances in which the average frequency of the peptide was at least 10 and was also more than 3 times the average frequency of the control. P values of the two-sample t tests were provided for screening purposes and are displayed according to “P < 1%” or “5% < P < 1%.” There was no adjustment for multiple comparisons. The lack of variability in the negative control was indicated on the output as “Cont. SD = 0.”
ATP6S1 protein and negative control assays were conducted in duplicate in ELISA. Three-fold increases between pre- and posttreatment were considered positive responses. Comparisons between negative and positive response groups were carried out with the two-tailed Fisher exact test. P < 0.05 was considered to be of statistical significance.
T-cell responses to ATP6S1 peptides
A total of 34 peptides were divided into five peptide pools (Fig. 1A). Several peptide pools for ATP6S1 induced T-cell responses in patients with stage III and IV melanoma (Fig. 1B and C; Supplementary Fig. S1A and Supplementary Table S1A). Individual peptides from reactive ATP6S1 peptide pools were further assessed by ELISPOT. T-cell responses to peptides ATP7, ATP13, ATP14, ATP21, ATP23, and ATP32 were observed (Fig. 1D; Supplementary Fig. S1B and Supplementary Table S1A). The frequency of T-cell responses to individual peptides in these patients was also examined (Fig. 1E). The relevant contribution of CD4+ or CD8+ T-cell responses was examined by CD4/CD8 depletion (Fig. 2A). CD4+ responses dominated T-cell responses to individual ATP6S1 peptides. One stage IV patient (P110) showed T-cell responses to ATP10 after anti–CTLA-4 treatment with ipilimumab (Fig. 2A).
To further assess the proliferative responses of reactive CD4+ T cells, ATP13-specific CD4+ T cells were generated from patient PBMCs. Antibody-blocking experiments with an HLA-DR–specific antibody confirmed the ATP13-induced CD4+ T cell–specific responses (Fig. 2B). Peptide binding prediction program (SYFPEITHI) for ATP13 showed broad binding potential to HLA-DRB1 alleles (data not shown). Genotyping for this patient with ATP13 reactivity revealed DRB1*09/15. As shown in Fig. 2C, the ATP13 CD4+ cells proliferated substantially in response to the HLA DRB1*15 melanoma cell line COO1. The addition of ATP13 peptide or an HLA DR–blocking antibody either enhanced or decreased this proliferation, respectively. In contrast, melanoma cell lines with alternate HLA DRB1 haplotypes did not induce significant proliferation of ATP13-specific CD4+ cells.
Generation and functional characterization of antigen-specific cytotoxic T lymphocytes
Results from screening with overlapping peptides and CD4+/CD8+ depletion analyses suggested that peptide ATP10 induced CD8+ T-cell responses in addition to CD4+ T-cell responses in patient 110 (Fig. 2A). Genotyping analysis revealed that the patient was HLA-A*0201 positive. There was a potential HLA-A*0201 binding epitope (e.g., ATP78) with a binding score of 2,537 (ILFWAQNFSV) within the ATP10 peptide (Fig. 1A). Potential HLA A*0201 epitopes from the full-length ATP6S1 protein were further analyzed by peptide prediction program. In addition to ATP78, several other potential HLA A*0201 epitopes were identified—ATP240, ATP247, and ATP136, with binding scores of 1,054, 510, and 374, respectively. The HLA A*0201 affinities of these four peptides were further examined by T2 binding assays. Figure 3A summarizes the increased fluorescent intensity (FI) of HLA A*0201 relevant to absent peptide control for peptides ATP78, ATP136, and ATP247. Peptide ATP240 was not sufficiently soluble in culture medium for use in the T2 binding analyses.
With the identification of HLA-A*0201 binding epitopes for ATP6S1, we next analyzed the abilities of these epitopes to induce CD8+ T-cell responses. The incidence of CD8+ T-cell responses to ATP78, ATP136, and ATP247 in patients with stage III and IV melanoma was first investigated by ELISPOT using PBMCs. No significant reactivity was detected for ATP78, ATP136, and ATP247 in these untreated patients. Given the binding affinities predicted by these epitopes, we next determined whether epitope-specific CD8+ T cells could be generated. CD8+ T cells isolated from HLA A*0201 healthy donors were stimulated with CD40L-activated autologous B cells in the presence of 10 μg/mL of peptide. After two to four rounds of stimulation, ATP78-, ATP136-, and ATP247-specific T cells were identified that were able to successfully stimulate IFNγ secretion in ELISPOT assays. ATP78-, ATP136-, and ATP247-specific CD8+ T-cell clones were generated by limiting dilution. Thirty-one clones specific for ATP78, 70 clones for ATP136, and 43 clones specific for ATP247 were selected. Six of the ATP78-, 12 of the ATP136-, and 9 of the ATP247-specific T-cell clones were found responsive to the corresponding peptides by ELISPOT. One ATP78-, three ATP136-, and one ATP247-specific clones were successfully expanded for further assessments. The expanded clonal T-cell lines were responsive by ELISPOT to T2 cells in the presence of relevant peptides, but not of control peptide (Fig. 3B). The presence of an HLA class I–blocking antibody abrogated IFNγ secretion, confirming the HLA A*0201–restricted CD8+ T-cell responses.
The cytotoxicity function of the epitope-specific CD8+ clones was assessed. The cytolytic activities of the CD8+ T-cell clones against T2 cells in the presence of 5 μg/mL of individual peptide indicate functional and specific activities of target killing (Fig. 4). In addition, these clones exhibited different capacities to lyse HLA-A*0201–positive melanoma cell lines—K028 and K029—in the absence of any cytokines. The addition of peptides or an HLA class I–blocking antibody enhanced or decreased the cytotoxic activities, respectively. In contrast, these clones did not exhibit cytolytic activity against the melanoma cell line K008, which is genotypically HLA A*0201–positive but fails to express HLA A*0201 molecules on its surface as determined by flow cytometry. Taken together, CD8+ T-cell clones to ATP6S1 epitopes are able to be generated and are able to successfully target HLA A*0201–expressing melanoma cells.
Humoral immune responses of patients with melanoma to ATP6S1
To further characterize patients' immune responses to ATP6S1 and determine whether coordinated cellular and humoral responses occur, we evaluated the antibody responses to recombinant ATP6S1 protein by ELISA. Consistent with previous studies, spontaneous immunity to ATP6S1 was observed in the sera of patients with stage IV melanoma (Supplementary Fig. S2).
Immune responses to ATP6S1 in patients with melanoma receiving CTLA-4 blockade treatment with ipilimumab
To better understand the clinical significance of immunity to ATP6S1, we first examined the T-cell responses of 10 patients with stage IV melanoma who received ipilimumab (10 mg/kg) treatment. As shown in Fig. 5A and B, and Supplementary Fig. S3A and S3B, and Supplementary Table S1B, 4 of 10 patients developed detectable T-cell responses. Furthermore, in 2 of the patients, ipilimumab treatment induced CD4+ T-cell response to ATP7 and CD8+ T-cell responses to ATP136 and ATP247 as a response to treatment, respectively (Fig. 5C and D). All four patients who had anti-ATP6S1 T-cell responses experienced clinical benefit from ipilimumab treatment with stable disease for greater than 6 months, or partial response, or complete response (Table 1). One of the 4 patients with T-cell responses was also seropositive for ATP6S1 (25%). In comparison, 2 of the 6 patients without anti-ATP6S1 T-cell responses experienced clinical benefit (Table 1). These 6 patients were seronegative for ATP6S1.
To further assess changes in ATP6S1 immunity in patients receiving ipilimumab, we used sera available from 27 treated patients who received ipilimumab treatment (11). Ipilimumab significantly augmented antibody responses against ATP6S1 in 5 patients (19%). Four patients experienced clinical benefit, and their serologic responses to ATP6S1 are shown by ELISA time courses (Fig. 6). T-cell responses in these 5 patients were then evaluated. Two of the 5 patients also developed T-cell response to ATP6S1 as a function of the treatment. In comparison, 6 of 22 patients who did not develop significant humoral immunity to ATP6S1 experienced clinical benefit. The differences in clinical benefit between patients who developed enhanced antibody responses to ATP6S1 and patients who did not are presented in Table 1.
Previous studies have demonstrated that the development of high-titer antibodies against ATP6S1 in patients receiving active immune therapy is associated with beneficial therapeutic outcomes (6). As both CD4+ and CD8+ T-cell infiltration occurs in metastases following vaccination and CTLA-4 blockade, it is important to characterize further the cellular responses to ATP6S1 in patients. In the present study, ATP6S1-specific CD4+ and CD8+ T cells were detected in patients with melanoma. HLA class I– and II–restricted T-cell antigenic determinants were identified and found to be presented by melanoma cells. Notably, enhanced immune responses to ATP6S1 by CTLA-4 blockade were observed and correlated with clinical benefit.
Immunodominant peptides differed between patients, likely due to differences in HLA class II. The ELISPOT analyses were based on a 7-day culture of T cells, which does not efficiently generate antigen-specific T cells from naïve cells (12). It is known that autoreactive lymphocytes for antigens such as MART-1, gp100, tyrosinase, vascular endothelial growth factor receptor, survivin, and ML-IAP circulate in healthy donors as well as in patients with tumors (7, 13–17). In line with previous studies, individuals can develop autoreactive antibodies to ATP6S1. It remains likely that the assessment of cellular responses via peripheral blood is more sensitive for class II based on the binding promiscuity of peptides in comparison with that of class I.
Natural antigen presentation on the cell surface involves expression of relevant protein, proteasome processing, TAP transport, and peptide affinity to MHC molecules (18). The presence of spontaneous ATP6S1-specific T cells in patients with melanoma suggests that antigenic epitopes can be naturally processed and presented. Although we did not detect T-cell responses to the three ATP6S1 HLA A*0201 epitopes (ATP78, 136, and 247) in patients with melanoma, the precursor frequency of these T cells may be beyond the sensitivity of the established ELISPOT. CD8+ T-cell clones to the three ATP6S1 HLA A*0201 epitopes were successfully generated, and their ability to lyse HLA A*0201–matched melanoma cell lines demonstrated. This indicates the relevant natural tumor immunogenicity of these epitopes and provides therapeutic implications for targeting ATP6S1. Moreover, this observation builds upon recognition of the importance of functional tumor antigens that may not be expressed on the cell surface yet whose epitopes are processed for immune recognition.
The Functional development of CD8+ T cells requires the presence of CD4+ T cells (19). Moreover, CD4+ T cells play critical roles in tumor eradication and induction of a protective immune response (20, 21). Natural responses to both ATP6S1-specific CD4+ and CD8+ T cells with functional relevance to melanoma cells were observed; of particular interest are the increased responses of both ATP6S1-specific CD4+ and CD8+ T cells in patients with melanoma receiving CTLA-4 blockade. This association also provides insight into future therapeutic strategies targeting ATP6S1 alone or in conjunction with immune checkpoint blockade.
The broad expression of ATP6S1 in normal tissue raises a concern that its targeting as an antigen may lead to adverse effects such as autoimmunity. Previously, high-titer antibody responses against ATP6S1 via whole-cell vaccination did not provoke clinical autoimmune manifestations (6). In our present study, patients who developed humoral and cellular immune responses to ATP6S1 did not express clinical signs or symptoms of autoimmunity. These results indicate that therapeutically targeting ATP6S1 might be a safe approach to cancer treatment.
Understanding the targets to an effective antitumor immune response remains an important goal for clinical investigation. With the recent success of immune checkpoint blockade, it becomes relevant to discover which tumor antigens are involved in favorable clinical outcomes. As an example, studies of immunity to the cancer–testis antigen NY-ESO-1 resulting from ipilimumab treatment showed increases in cellular as well as humoral immune responses (22). The previous report of GM-CSF–secreting whole-cell tumor vaccination followed by CTLA-4 blockade and the current ipilimumab study demonstrated improvement in both ATP6S1 humoral and cellular immunity as a function of treatment. Taken together, these results suggest that ATP6S1 is an antigenic target associated with therapeutic favorable clinical outcomes. Given its immunogenicity and cellular function in tumors, further investigation with ATP6S1 in clinical studies, as well as its potential role as a marker of effective treatment, is worthy of pursuit.
Disclosure of Potential Conflicts of Interest
F.S. Hodi reports receiving commercial research support to his institution, has ownership interest (including patents) in, and is a consultant/advisory board member for Bristol-Myers Squibb. No potential conflicts of interest were disclosed by the other authors.
Conception and design: J. Zhou, F.S. Hodi
Development of methodology: J. Zhou, F.S. Hodi
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Zhou, M. Gupta, X. Wu, C. Yoon, F.S. Hodi
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J. Zhou, A. Giobbie-Hunder, F.S. Hodi
Writing, review, and/or revision of the manuscript: J. Zhou, X. Wu, A. Giobbie-Hunder, F.S. Hodi
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Zhou, F.S. Hodi
Study supervision: J. Zhou, F.S. Hodi
This work was supported in part by Sharon Crowley Martin Memorial Fund for Melanoma Research (to F.S. Hodi) and Malcolm and Emily Mac Naught Fund for Melanoma Research (to F.S. Hodi) at Dana-Farber Cancer Institute.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Supplementary data for this article are available at Cancer Immunology Research Online (http://cancerimmunolres.aacrjournals.org/).
- Received October 2, 2014.
- Accepted October 3, 2014.
- ©2014 American Association for Cancer Research.