Abstract

Results

To analyze the capacity of human T cells to recognize OCT4, mononuclear cells from healthy blood donors (n = 30) were cultured with an overlapping peptide library spanning the entire OCT4 protein (see Table S1 for the sequences of peptides in this library), and culture supernatants assayed for the presence of IFN-γ inducible protein-10 (IP10). Positive reactivity in this assay was defined as ≥ 2-fold increase in IFN-γ IP10 production in the reactive mix versus control, with a minimal absolute measurement of 100 pg/mL predetermined to be positive for the presence of antigen-specific T cells. Using this cutoff, reactivity against peptide pools derived from this library was detectable in 25 of 30 (83%) donors tested (Fig. 1 A and B). Peptide reactive IP10 production was abrogated upon depletion of CD3⁺ T cells before stimulation (Fig. 1C), as well as blockade of IFN-γ (Fig. 1D), demonstrating that the reactivity in this assay was dependent on peptide reactive production of IFNγ by T cells. Analysis of peptide reactive proliferation indicated that the proliferating T cells were predominantly CD4⁺ T cells, and antigen-dependent proliferation correlated with reactivity in the IP10 assay (Fig. 2A). The rapidity of cytokine production in response to antigen stimulation suggested that the observed immune response was a memory T-cell response. To further evaluate this finding, bulk PBMCs were depleted of CD45RO⁺ T cells before stimulation with OCT4 peptide library [or phytoagglutinin (PHA) as a control]. Depletion of CD45RO⁺ T cells led to abrogation of OCT4 reactivity, suggesting that the OCT4-specific response as measured in this assay resided predominantly in the CD45RO⁺ subpopulation (Fig. 2B). Together, these data indicate the presence of OCT4-specific memory T-cell response, which can be readily detected in freshly isolated PBMCs from healthy donors.

Open in a separate window

Fig. 1.

Immunity to OCT4 in freshly isolated PBMCs from healthy blood donors. (A and B) Reactivity of freshly isolated human PBMCs from healthy donors to OCT4 peptide library: 2 × 10⁵ PBMCs were cultured alone or in the presence of 3 μg/mL of OCT4 peptide mixes, or PHA, as a control. After 48 h, the culture supernatant was harvested and analyzed for the presence of IP10 by Luminex assay. (A) Data for representative healthy donors with or without detectable reactivity to OCT4 peptide library. (B) Data for reactivity to the positive OCT4 mix (represented as fold-change in reactive mix compared with control) in all 30 donors tested. Based on interassay and intra-assay variance, ≥2-fold increase in IP10 production (in reactive mix versus control), with a minimal absolute measurement of 100 pg/mL, was predetermined to be positive for the presence of antigen-specific T cells, as discussed under Materials and Methods. (C) Bulk PBMCs or those depleted of CD3⁺ T cells were cultured alone (control) or with OCT4 peptide library. After 48 h, the culture supernatant was harvested and analyzed for the presence of IP10 by Luminex assay. Data are representative of similar experiments on three donors. (D) PBMCs were cultured alone or OCT4 peptide mix in the presence of either anti-IFN-γ blocking antibody or isotype control antibody. After 48 h, the culture supernatant was harvested and analyzed for the presence of IP10 by Luminex assay. Data are representative of similar experiments on three donors.

Open in a separate window

Fig. 2.

Antigen dependent proliferation of OCT4-specific memory T cells. (A) Proliferative response to OCT4 peptide library: 2 × 10⁵ PBMCs were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE) and cultured with 1 μg/mL of anti-CD28 and anti-CD49d antibody alone (control) or in the presence of OCT4-derived peptides (3 μg/mL). (Left) Proliferation of CD3⁺-gated cells in a representative donor. Note the predominant proliferation of CD4⁺ T cells. (Right) Antigen-dependent proliferation in six donors with or without reactivity in the IP10 assay, as in Fig 1A. (B) PBMCs were depleted of CD45RO⁺ T cells using immunomagnetic beads (or left undepleted) before stimulation with OCT4 peptides (Left), or PHA as a control (Right).

To analyze the fine specificity of the anti-OCT4 response and confirm reactivity to the peptide mix, PBMCs were cultured against individual peptides from the reactive peptide mix in 10 donors. Although individual donors had reactivity to different peptides within the pool, some hot-spot areas of promiscuous reactivity were also identified (Fig. 3A). Analysis of areas of hot spots of reactivity by BLAST (National Center for Biotechnology Information) demonstrated that these sequences represented highly conserved regions within OCT4 protein, without homology to known pathogen-derived sequences (Table S2). To further analyze the capacity of functional OCT4-specific T cells to expand in culture, T cells were stimulated with peptide-pulsed autologous dendritic cells (DCs). The presence of peptide-specific IFNγ-secreting CD4⁺ T cells was then analyzed by intracellular cytokine flow cytometry following peptide-specific restimulation (Fig. 3B). Together, these data demonstrate that T cells responding to peptide epitopes derived from OCT4 can be readily detected and expanded in culture using stimulation with peptide-loaded autologous DCs. Together, these data demonstrate the lack of immune tolerance to this critical ES antigen in humans, and show that these T cells can be readily harnessed for potential therapeutic utility after ex vivo stimulation.

Open in a separate window

Fig. 3.

Fine specificity of OCT4-specific T-cell immunity and activation of peptide-specific T cells. (A) Reactivity to individual peptides: PBMCs were cultured with individual peptides from submix 3 to determine the specific reactive peptide. After 48 h, culture supernatant was harvested and analyzed for the presence of IP10 by Luminex assay. The figure shows the reactivity against individual peptides for 10 healthy donors. (B) Expansion of peptide-reactive T cells with peptide-pulsed DCs. Monocyte-derived mature DCs loaded with active peptide (as identified in experiments in Fig 2A) were used to stimulate autologous T cells. The presence of peptide-specific IFNγ-producing T cells was analyzed by intracellular cytokine flow cytometry following restimulation with anti-CD28 and anti-CD49d, with or without stimulating peptide. Data shown are representative of four different donors.

OCT4 has emerged as an important biomarker and critical determinant of oncogenic fate in GCTs (12, 13). Therefore, we prospectively analyzed the presence of OCT4-specific T cells in patients with newly diagnosed GCT, at baseline and after the completion of therapy. Of 21 patients analyzed, 20 had pathologic evidence for predominant seminoma or embryonal carcinoma, subtypes typically associated with highest OCT4 expression (Table S3) (12). Reactivity to viral antigens (CMV, EBV, and influenza; CEF) and PHA was monitored as a control. In contrast to the finding with healthy donors, only 8 of 21 patients with GCT (38%) had measurable OCT4-specific T cell immunity at baseline (P < 0.05) (Fig. 4A). An intriguing aspect of GCTs is the high curability of these tumors, even in the setting of advanced cancers (14). Recent studies have suggested that the ability of some chemotherapies to induce immunogenic death of tumor cells may contribute to their antitumor effects in vivo (15). However, whether specific immunity to tumor-associated antigens is induced after curative therapy for GCT is not known. In contrast to analysis of baseline samples, the presence of OCT4-specific T cells was detected at the completion of chemotherapy in 10 of 12 patients (83%) tested, including 5 patients who lacked such responses at baseline (Fig. 4B). The induction of antigen-specific T cells after therapy was specific for OCT4, as there were no significant changes in virus-specific T-cell responses (Fig. 4B). There was also no correlation between reactivity to OCT4 and candida, as an additional control for human antigen-specific CD4⁺ T-cell responses (r² = 0.002) (Fig. S1). As testicular tumors commonly lack MHC II (16), we hypothesized that the induction of OCT4-specific T cells might involve the uptake of dying tumor cells by DCs. Stimulation of autologous T cells with DCs loaded with dying embryonal carcinoma cells led to the induction of OCT4-specific immune responses in culture (Fig. S2). Therefore, DCs can process and present OCT4 from dying GCT cells. Together, these data suggest that chemotherapy-induced cell death may lead to the induction of tumor immunity in vivo.

Open in a separate window

Fig. 4.

Immunity to OCT4 in patients with GCTs. (A) Reactivity of freshly isolated human PBMCs from newly diagnosed GCT patients (n = 21) or healthy controls (n = 30) to OCT4 peptide library (as in Fig 1B) or PHA as a control. Positive reactivity was predetermined to be ≥2-fold increase in IP10 production (in reactive mix versus control), with a minimal absolute measurement of 100 pg/mL, as discussed under Materials and Methods. (B) Effect of antitumor therapy on OCT4 immunity: PBMCs from GCT patients (n = 12) were analyzed at baseline and at the completion of therapy for reactivity to OCT4 and to viral antigens (CEF) as a control. Data shown are reactivity at baseline and at the completion of therapy. Bold lines represent mean reactivity.

Discussion

In this article, we have demonstrated that most healthy humans have naturally occurring memory T-cell responses to specific epitopes derived from OCT4, indicating that most individuals lack immune tolerance to this critical pluripotency antigen. Several studies have examined the presence of self-antigen reactive T cells in healthy individuals and patients with cancer or autoimmunity. Detection of T cells specific for MelanA/MART1 (CD8⁺, naive), carcinoembryonic antigen (CD4, naive), wild-type p53 (CD8/CD4, naive), cyclin B1 (CD4/CD8, naive/memory), glutamic acid decarboxylase, and insulin (CD4/CD8, naive) was feasible in T cells from healthy individuals (17–21). However, none of these nor similar studies evaluated the immunogenicity of ES-restricted antigens in humans or mice, which may provide a potential mechanism for immune-mediated surveillance (22) against unrestrained growth of pluripotent cells, as well as GCT. The presence of rare OCT4-expressing cells with pluripotent potential in adult tissues has recently been suggested to play a role in tissue regeneration in mice (12, 23, 24). The possibility of immune surveillance against GCT is supported by the observation that the immunity to OCT4 is reduced in GCT patients compared with healthy donors and increased risk of these tumors in patients with immunodeficiency (25). In prior studies, we have shown that T-cell responses against another ES antigen, SOX2, are commonly detected in patients with preneoplastic gammopathy and correlate with reduced risk of progression to overt myeloma (26). However, in contrast to OCT4, anti-SOX2 responses were not detected in healthy donors using similar assays (26). The reason underlying the differences in immune reactivity between these two core ES factors in healthy individuals is not known, but may relate to differences in their immunogenicity, or patterns of aberrant expression (12, 23, 24). Further studies are needed to explore the full spectrum of immunogenic ES-specific antigens and understand the mechanism of induction and maintenance of these immune responses.

Our finding that anti-OCT4 T cells are specifically induced in vivo in response to therapy of GCTs demonstrates that these T cells are functional in vivo and can respond to stimulation. Most of the OCT4-specific T cells consist of CD4⁺ T cells. As most testicular tumors are typically MHC II-negative (16), the induction of OCT4 T cells may involve the uptake of dying tumor cells and presentation of tumor-derived antigens by DCs. Tumor-specific CD4⁺ Th1 cells can mediate antitumor effects against MHC II-negative cells by multiple mechanisms, including the activation of macrophage-mediated innate resistance and antiangiogenesis (27–29). These data also provide evidence that curative therapy of GCTs can lead to the induction of tumor antigen-specific T-cell responses in vivo. Further studies are needed to understand whether the observed induction of tumor immunity associated with chemotherapy is critical for the long-term cure of GCTs.

These findings also have several therapeutic implications. Cells derived from IPS cells are likely to reach the clinic in the near future, but carry a real risk of tumor formation, particularly if the ES genes were to be reactivated in vivo (9, 30). Boosting immunity to OCT4 may therefore be important to minimize the tumorigenicity of these cells in the clinic. A growing body of data points to the importance of shared pathways of “stemness” in the biology of human cancer and ES cells (5, 31). OCT4 is essential for pluripotency of ES and IPS cells. OCT4 has been shown to act as a dose-dependent oncogenic fate determinant in mice and thought to play a critical role in the pathogenesis of GCT (32). However, interactions between transcriptional networks regulating pluripotency/stemness may be cell type/context-dependent, particularly in the setting of transformed cells. For example, pluripotency in some variant ES lines was found to be independent of OCT4 (33). Therefore, further study is needed to elucidate the functional significance of OCT4 expression in putative cancer stem cells in different cancers. An important consideration when targeting putative cancer stem-cell genes is the potential impact on normal stem cells. It is therefore of interest that OCT4 is dispensable for the function of adult stem cells in mice, making it an attractive target on cancer stem cells (34). There is a long history of targeting embryonal tissues toward vaccines against cancer, as reviewed recently (35). Notably in these studies, the protective effect was most evident only with early but not late embryonic tissues and lacked precise definition of antigenic targets (35). Interestingly, a recent study demonstrated that vaccination of mice with ES/IPS cells mediated protection in a colon-cancer model (36). The immunogenic epitopes of OCT4 as identified here may therefore serve as the basis of a vaccine for prevention or therapy of several cancers, or adoptive T cell-based therapies, including in the setting of allogeneic stem cell transplantation. Immunity to stemness genes may be critical for harnessing the immune system against cancer (37).

Materials and Methods

Supplementary Material

Acknowledgments

Footnotes

References