Overcoming regulatory T-cell suppression by a lyophilized preparation of Streptococcus pyogenes

Authors

  • Michiko Hirayama,

    1. Department of Cancer Vaccine, Mie University Graduate School of Medicine, Mie, Japan
    2. Department of Immuno-Gene Therapy, Mie University Graduate School of Medicine, Mie, Japan
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Hiroyoshi Nishikawa,

    Corresponding author
    1. Experimental Immunology, Immunology Frontier Research Center, Osaka University, Osaka, Japan
    • Department of Cancer Vaccine, Mie University Graduate School of Medicine, Mie, Japan
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Yasuhiro Nagata,

    1. Department of Surgery, National Hospital Organization Nagasaki Medical Center, Nagasaki, Japan
    Search for more papers by this author
  • Takemasa Tsuji,

    1. Ludwig Institute for Cancer Research, New York Branch, Memorial Sloan-Kettering Cancer Center, New York, NY, USA
    Search for more papers by this author
  • Takuma Kato,

    1. Department of Cellular and Molecular Immunology, Mie University Graduate School of Medicine, Mie, Japan
    Search for more papers by this author
  • Shinichi Kageyama,

    1. Department of Immuno-Gene Therapy, Mie University Graduate School of Medicine, Mie, Japan
    Search for more papers by this author
  • Shugo Ueda,

    1. Department of Gastroenterological Surgery and Oncology, Kitano Hospital, Tazuke Kofukai Medical Research Institute, Osaka, Japan
    Search for more papers by this author
  • Daisuke Sugiyama,

    1. Experimental Immunology, Immunology Frontier Research Center, Osaka University, Osaka, Japan
    Search for more papers by this author
  • Sahoko Hori,

    1. Department of Cancer Vaccine, Mie University Graduate School of Medicine, Mie, Japan
    2. Department of Immuno-Gene Therapy, Mie University Graduate School of Medicine, Mie, Japan
    Search for more papers by this author
  • Shimon Sakaguchi,

    1. Experimental Immunology, Immunology Frontier Research Center, Osaka University, Osaka, Japan
    Search for more papers by this author
  • Gerd Ritter,

    1. Ludwig Institute for Cancer Research, New York Branch, Memorial Sloan-Kettering Cancer Center, New York, NY, USA
    Search for more papers by this author
  • Lloyd J. Old,

    1. Ludwig Institute for Cancer Research, New York Branch, Memorial Sloan-Kettering Cancer Center, New York, NY, USA
    Search for more papers by this author
  • Sacha Gnjatic,

    1. Ludwig Institute for Cancer Research, New York Branch, Memorial Sloan-Kettering Cancer Center, New York, NY, USA
    Search for more papers by this author
  • Hiroshi Shiku

    Corresponding author
    1. Department of Immuno-Gene Therapy, Mie University Graduate School of Medicine, Mie, Japan
    • Department of Cancer Vaccine, Mie University Graduate School of Medicine, Mie, Japan
    Search for more papers by this author

Full correspondence: Dr. Hiroyoshi Nishikawa, Experimental Immunology, Immunology Frontier Research Center, Osaka University, 3-1 Yamadaoka, Suita, Osaka 565-0871, Japan

Fax: +81-6-6879-4464

e-mail: nisihiro@ifrec.osaka-u.ac.jp

Additional correspondence: Sacha Gnjatic, Ludwig Institute for Cancer Research, New York Branch, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10065, USA.

e-mail: gnjatics@mskcc.org

Additional correspondence: Hiroshi Shiku, Departments of Cancer Vaccine and Immuno-Gene Therapy, Mie University Graduate School of Medicine, 2-174 Edobashi, Tsu, Mie 514-8507, Japan.

e-mail: shiku@clin.medic.mie-u.ac.jp

Abstract

Cancer vaccines have yet to yield clinical benefit, despite the measurable induction of humoral and cellular immune responses. As immunosuppression by CD4+CD25+ regulatory T (Treg) cells has been linked to the failure of cancer immunotherapy, blocking suppression is therefore critical for successful clinical strategies. Here, we addressed whether a lyophilized preparation of Streptococcus pyogenes (OK-432), which stimulates Toll-like receptors, could overcome Treg-cell suppression of CD4+ T-cell responses in vitro and in vivo. OK-432 significantly enhanced in vitro proliferation of CD4+ effector T cells by blocking Treg-cell suppression and this blocking effect depended on IL-12 derived from antigen-presenting cells. Direct administration of OK-432 into tumor-associated exudate fluids resulted in a reduction of the frequency and suppressive function of CD4+CD25+Foxp3+ Treg cells. Furthermore, when OK-432 was used as an adjuvant of vaccination with HER2 and NY-ESO-1 for esophageal cancer patients, NY-ESO-1–specific CD4+ T-cell precursors were activated, and NY-ESO-1–specific CD4+ T cells were detected within the effector/memory T-cell population. CD4+ T-cell clones from these patients had high-affinity TCRs and recognized naturally processed NY-ESO-1 protein presented by dendritic cells. OK-432 therefore inhibits Treg-cell function and contributes to the activation of high-avidity tumor antigen-specific naive T-cell precursors.

Introduction

Many tumor-associated antigens recognized by the immune system are normal self-constituents, and tumor immunity is considered to be in part an autoimmune response [1-3]. Therefore, mechanisms for maintaining immunological self-tolerance hamper effective anticancer immunity. CD4+CD25+ Treg cells are one of the major components in maintaining immunological self-tolerance in hosts by suppressing a wide range of immune responses [4-7]. Indeed, depletion of Treg-cell populations enhances spontaneous and vaccine-induced antitumor immune responses [6, 8, 9], and the stimulation of CD4+CD25+ Treg cells by immunization with self-antigens induces enhanced chemically induced primary tumor development and increased numbers of pulmonary metastasis following injection of transplantable tumor cells [10-12]. In human cancers, the presence of high numbers of CD4+CD25+ Treg cells or low ratio of CD8+ T cells to CD4+CD25+ Treg cells in tumors is correlated with unfavorable prognosis [13, 14]. In addition, the depletion of CD4+CD25+ Treg cells in patients receiving a DC vaccine enhances the stimulation of tumor-specific T-cell responses, indicating a crucial role for Treg cells in the regulation of antitumor immune responses in humans [15].

NY-ESO-1, a germ cell protein, was found by SEREX (serological identification of antigens by recombinant expression cloning) using the serum of an esophageal cancer patient [16, 17]. We have previously shown that NY-ESO-1–specific CD4+ T cells are detectable in cancer patients with spontaneous NY-ESO-1 serum Ab responses [17, 18]. In addition, NY-ESO-1–specific CD4+ T-cell precursors can expand and become detectable in healthy individuals after in vitro antigenic stimulation of peripheral CD4+ T cells, but only following depletion of CD4+CD25+ T cells [19, 20]. These results suggested that NY-ESO-1–specific CD4+ T-cell precursors are actually present at relatively high frequencies in healthy individuals, and that the activation/expansion of NY-ESO-1–specific naive CD4+ T cells is suppressed by CD4+CD25+ Treg cells. In healthy donors and in cancer patients with NY-ESO-1–expressing tumors but without spontaneous anti-NY-ESO-1 Ab (seronegative), naturally arising NY-ESO-1–specific T-cell responses are susceptible to Treg-cell suppression and are exclusively detected from naive populations (CD4+CD25CD45RA+). In contrast, most NY-ESO-1–specific CD4+ T cells in cancer patients with spontaneous anti-NY-ESO-1 Ab (seropositive) are derived from memory populations (CD4+CD25CD45RO+) and are detectable even in the presence of CD4+CD25+ Treg cells [20, 21]. After vaccination with HLA-DPB1*0401/0402-restricted NY-ESO-1157170 peptide in incomplete Freund's adjuvant, ovarian cancer patients develop NY-ESO-1–specific CD4+ T cells with only low avidity to antigen and low sensitivity to Treg cells, even though they have an effector/memory phenotype (CD4+CD25CD45RO+) [21]. Still, high-avidity naive NY-ESO-1–specific T-cell precursors are present in the peripheral blood of vaccinated patients, but they are subjected to continuous CD4+CD25+ Treg-cell suppression throughout vaccination [21]. Thus, a strategy to overcome Treg-cell suppression on preexisting high-avidity naive T-cell precursors is an essential component for effective cancer vaccines.

Accumulating data shed light on recognition of pathogen-associated molecular patterns through TLRs to break the suppressive environment in tumors [22]. It has been reported that TLR stimulants, such as lipopolysaccharide or CpG, block the suppressive activity of CD4+CD25+ Treg cells partially by an IL-6–dependent mechanism [23]. TLR2 signaling was reported to stimulate the proliferation of CD4+CD25+ Treg cells and to induce temporal loss of suppressive activity of CD4+CD25+ Treg cells [24]. TLR2 signaling has also been shown to increase IL-2 secretion by effector T cells, thereby rendering them resistant to CD4+CD25+ Treg-cell–mediated suppression [25]. We and others have recently reported that vaccination of tumor antigens by TLR stimulating viral or bacterial vectors was able to not only inhibit the suppressive function of CD4+CD25+ Treg cells but also break tolerance or hyporesponsiveness of effector T cells to tumor antigens even in the presence of Treg cells [26-28].

OK-432 is a lyophilized preparation of Streptococcus pyogenes that binds TLR-2, TLR-4, and/or TLR-9 and activates APCs, making it attractive for potential use as an adjuvant of cancer vaccine [29-33]. OK-432–matured DCs effectively prime antigen-specific T cells in vitro [29, 34]. Importantly, OK-432 has already been used for many years as a direct anticancer agent, particularly in Japan, and has a well-established clinical safety profile. However, while it is considered that OK-432 may inhibit Treg-cell suppressive activity by stimulating several TLR signaling pathways, its influence on Treg cells has not yet been shown. In this study, we addressed whether OK-432 inhibits Treg-cell suppressive function and could be a promising adjuvant of cancer vaccines.

Results

OK-432 inhibits the suppressive activity of CD4+CD25+ Treg cells

To address whether OK-432 inhibited CD4+CD25+ Treg-cell suppression, we employed the standard in vitro suppression system. CD4+CD25 T cells and CD4+CD25high Treg cells (highest 3% of CD4+CD25+ cells) were isolated from PBMCs of healthy individuals. CD4+CD25 T cells were cultured with irradiated autologous APCs (CD4-depleted PBMCs) and anti-CD3 Ab in the presence or absence of CD4+CD25high Treg cells. CD4+CD25 T-cell proliferation was analyzed as described in the Materials and methods. In accordance with previous reports [7], CD4+CD25high Treg cells markedly suppressed the proliferation of CD4+CD25 T cells (Fig. 1A and B). In sharp contrast, when OK-432 was added in the culture, suppressive activity of CD4+CD25high T cells was significantly inhibited (Fig. 1A and B). In addition, OK-432 did not induce death of CD4+CD25high Treg cells as the frequency of Annexin V+ and 7-AAD+ cells was not significantly increased in the presence of OK-432 (data not shown). Instead, CD4+CD25high Treg cells exhibited marginal proliferation in the presence of OK-432 (Fig. 1A). These data indicate that addition of OK-432 impairs the suppressive activity of CD4+CD25high Treg cells and partially reverses anergy status of Treg cells.

Figure 1.

OK-432 overcomes Treg-cell suppression in vitro. (A, B) CD4+CD25 T cells and CD4+CD25high Treg cells were collected from PBMCs of healthy individuals as described in the Materials and Methods. 1 × 104 CD4+CD25 T cells were cultured with irradiated autologous CD4-depleted PBMCs and anti-CD3 Ab in the presence or absence of 5 × 103 CD4+CD25+ Treg cells with/without OK-432. Proliferation was analyzed by 3H-thymidine incorporation. Data of three donors are shown as mean + SD of two replicates/samples and are from one experiment representative of at least two performed. (B) Summary of percent suppression in five healthy individuals. Percent suppression was calculated as: (1 – (cpm in coculture)/(cpm in CD4+CD25 T cells)) × 100. Data shown are from one experiment representative of at least two performed.

Inhibition of the suppressive activity of CD4+CD25+ Treg cells by OK-432 is dependent on IL-12

Since OK-432 reportedly induces TLR-2, TLR-4, and/or TLR-9 activation and subsequent production of proinflammatory cytokines [29-33], we examined the involvement of cytokines in this inhibition of Treg-cell suppression. To this end, Abs against several candidate cytokines were added to cultures. Among cytokines tested, only blocking Ab against IL-12 significantly abrogated the inhibition of Treg-cell suppression by OK-432 (Fig. 2A).

Figure 2.

IL-12 is a critical cytokine to overcome Treg-cell suppression by OK-432. CD4+CD25 T cells and CD4+CD25high Treg cells were collected from PBMCs of healthy individuals. 1 × 104 CD4+CD25 T cells were cultured with 1 × 105 irradiated autologous CD4-depleted PBMCs and anti-CD3 Ab in the presence or absence of 2.5 × 103 CD4+CD25high Treg cells with/without OK-432. (A) Blocking Abs (10 μg/mL) against several cytokines as indicated were added in the culture and proliferation was measured. (B) Recombinant cytokines (IL-12; 5 ng/mL, IL-6; 5 ng/mL, IFN-γ; 100 U/mL) were added into the cultures and proliferation was analyzed. Data for one representative donor among three donors are expressed as mean + SD of two replicates/samples and are from one experiment representative of at least two performed. *p < 0.05 and **p < 0.01 as compared with control, Student's t-test.

To confirm the importance of IL-12, we next analyzed whether the addition of IL-12 could inhibit Treg-cell suppression as observed by OK-432. CD4+CD25 T cells were cultured with CD4+CD25high Treg cells, irradiated autologous APCs and anti-CD3 Ab in the presence of IL-12. Treg-cell suppressive activity was significantly inhibited by the addition of IL-12, but not IL-6 or IFN-γ (Fig. 2B). Again, IL-12 did not kill CD4+CD25high Treg cells as the frequency of Annexin V+ and 7-AAD+ cells was not significantly increased in the presence of IL-12 (data not shown). Instead, CD4+CD25high Treg cells slightly proliferated in the presence of OK-432 (Fig. 2B). These data suggest a critical role for IL-12 in the inhibition of Treg-cell suppression by OK-432.

OK-432 induces higher amounts of IL-12 but not IL-10 from APCs compared with other stimuli

To gain insight into the cellular target(s) of OK-432, we explored the origin of IL-12 after OK-432 treatment based on the essential role of IL-12 in the inhibition of Treg-cell suppression by OK-432. We then analyzed whether OK-432 stimulation indeed induced IL-12 production from APCs, such as CD3-depleted PBMCs used in the standard Treg-cell suppression assays. CD3-depleted PBMCs from healthy donors were stimulated with OK-432, LPS, or TNF-α, and cytokine production was examined. OK-432 induced significantly higher amounts of IL-12 from CD3-depleted PBMCs than LPS or TNF-α (Fig. 3A). In addition, CD3-depleted PBMCs stimulated with OK-432 induced much less IL-10 production than LPS (Fig. 3A). Similar results, i.e. IL-12 rather than IL-10 was dominantly produced by CD3-depleted PBMCs stimulated with OK-432, were obtained from four esophageal cancer patients (Fig. 3B).

Figure 3.

APCs stimulated with OK-432 produce significantly higher amounts of IL-12. (A) 1 × 105 CD3-depleted PBMCs from four healthy individuals were cultured with TNF-α (100 ng/mL), LPS (1 mg/mL), or OK-432 (1 μg/mL) and supernatant was collected 48 h later. IL-12p70 and IL-10 production was analyzed with ELISA. (B) 1 × 105 CD3-depleted PBMCs from four esophageal cancer patients were cultured as in (A) and supernatant was collected 48 h later. IL-12p70 and IL-10 production was analyzed with ELISA. (A, B) Each symbol represents an individual donor; data shown are from one experiment representative of at least two performed. (C) PBMCs from two healthy individuals were cultured with/without OK-432. Cells were subjected to staining with the indicated surface markers and then intracellular IL-12, and were analyzed by flow cytometry. Data of two donors from one experiment representative of at least two performed. *p < 0.05 and **p < 0.01 as compared with control, Student's t-test.

We next examined which cell types in PBMCs produced IL-12 after OK-432 stimulation. The major sources of IL-12 in PBMCs after OK-432 stimulation were CD11c+ and CD14+ cells, and neither NK cells nor T cells produced IL-12 (Fig. 3C). Taken together, APCs, such as monocytes, macrophages, and DCs are considered to be the cellular targets of OK-432 to induce IL-12 which is a crucial component for the inhibition of Treg-cell suppression by OK-432.

OK-432 administration to tumor-associated exudates reduces local Treg-cell accumulation and function

As OK-432 is available as an anticancer agent in Japan and has been used for controlling tumor-associated exudate fluids by direct injection to the cavity, we next investigated its influence on Treg cells following in vivo treatment of OK-432. We analyzed the local Treg-cell accumulation and function of tumor-associated sites before and 2–3 days after local OK-432 administration. Cells were isolated from tumor-associated exudate fluids, such as pleural effusions and ascites. The frequency of Treg cells before and after treatment with OK-432 was examined by staining with Abs for CD4, CD25, and Foxp3. The Foxp3+ T-cell population in CD4+ T cells was markedly reduced (Fig. 4A). Furthermore, the proportion of Foxp3+ T cells in CD4+CD25+ T cells was also significantly reduced after OK-432 administration (Fig. 4A and B), indicating that the balance of helper T cells to Treg cells had changed.

Figure 4.

OK-432 administration into tumor-associated exudate fluids elicits reduction of local Treg-cell accumulation and function. Cells were isolated from tumor-associated exudate fluids (two pleural effusions or two ascites) and peripheral blood before and 2 days after OK-432 administration. (A) Cells were stained with anti-CD4, anti-CD25, and anti-Foxp3 Abs and were analyzed with flow cytometry. The staining pattern and gating method of a representative patient is shown (left); solid line histogram, Foxp3 expression in CD4+ CD25+ T cells; dotted line histogram, Foxp3 expression in CD4+ T cells. Foxp3 expression in CD4+CD25+ T cells of two representative patients is shown (right); filled histogram, control staining; gray line histogram, Foxp3. (B) The percentage of CD4+CD25+Foxp3+ cells in CD4+ T cells is shown for n = 4 patients’ samples. Statistical significance determined by Student's t-test. (C) 1 × 105 CD4+CD25 T cells were isolated from PBMCs before OK-432 administration and cultured with irradiated autologous CD4-depleted PBMCs and anti-CD3 Ab in the presence or absence of 1 × 105 CD4+CD25high T cells isolated from tumor-associated exudate fluids before and after OK-432 administration (n = 2). Data shown are from one experiment representative of two independent experiments.

We next addressed the suppressive activity of CD4+CD25high T cells in tumor-associated exudate fluids. CD4+CD25high T cells (highest 3% gate of CD4+CD25+ cells defined with peripheral blood was applied) were isolated from tumor-associated exudate fluids and cultured with CD4+CD25 T cells from PBMCs with irradiated autologous APCs and anti-CD3 Ab. After OK-432 administration, as the volume of tumor-associated exudate fluids decreased, sufficient amounts of CD4+CD25high T cells for proliferation assays were available only from two patients. CD4+CD25 T-cell proliferation was analyzed as described in the Materials and Methods. There was a trend, albeit not significant, toward a decrease in Treg-cell function after OK-432 administration (Fig. 4C). In contrast, we did not observe any differences in frequency and function of Treg cells in PBMCs before and after OK-432 administration (data not shown). These data propose that in vivo injection of OK-432 decreases the local Treg-cell accumulation and function.

Origin of the repertoire of CD4+ T-cell effectors elicited by vaccination with NY-ESO-1 and OK-432

To further explore the effect of OK-432 on the inhibition of in vivo Treg-cell activity, we also examined the potential of OK-432 as an adjuvant in a cancer vaccine. We have reported that high-avidity NY-ESO-1–specific CD4+ T-cell precursors are present in naive CD45RA+ populations and that their activation is rigorously suppressed by CD4+CD25+ Treg cells [20, 21]. We also found that synthetic peptide vaccination with incomplete Freund's adjuvant induces only peptide-specific CD4+ T cells with low-avidity TCRs (recognition of >1 μM peptide but not naturally processed NY-ESO-1 protein), but not high-avidity CD4+ T cells (recognition of naturally processed NY-ESO-1 protein or <0.1 μM peptide) that are susceptible to Treg-cell suppression [21]. Together, these data highlight the importance of blocking Treg-cell activity to allow activation/expansion of high-avidity NY-ESO-1–specific CD4+ T-cell precursors. For this reason, we investigated whether high-avidity NY-ESO-1–specific CD4+ T-cell precursors were activated by NY-ESO-1 protein vaccination with OK-432 as an adjuvant and were present in memory CD45RO+ populations.

Samples from two patients who received vaccination with cholesteryl hydrophobized pullulan (CHP)-HER2 and NY-ESO-1 with OK-432 (Supporting Information Fig. 1) were available for this analysis. Whole CD4+ T cells or CD4+CD25CD45RO+ (effector/memory) T cells before and after vaccination were presensitized with NY-ESO-1–overlapping peptides covering the entire sequence of NY-ESO-1 and specific CD4+ T-cell induction was analyzed with ELISPOT assays. As the sample size was not sufficient to analyze specific CD4+ T-cell induction within CD4+CD25CD45RA+ (naive) T cells, we analyzed whether NY-ESO-1–specific high-avidity CD4+ T cells were induced from the CD4+CD25CD45RO+ (effector/memory) T-cell population after vaccination in Pt #1 (HLA-DR 4, 12 and HLA -DQ 4, 8) and #2 (HLA-DR 9, 15 and HLA-DQ 6, 9). Pt #1 exhibited spontaneously induced CD4+ T-cell responses against NY-ESO-191110 before vaccination and the responses were maintained after extensive vaccination (Fig. 5A). These spontaneously induced NY-ESO-191110–specific CD4+ T cells were detected in the CD4+CD25CD45RO+ (effector/memory) T-cell population before and after vaccination. Following vaccination with NY-ESO-1 protein in the presence of OK-432, CD4+ T-cell immune responses against NY-ESO-1111130 were newly elicited (Fig. 5A). These vaccine-induced NY-ESO-1111130–specific CD4+ T cells were detected in the CD4+CD25CD45RO+ (effector/memory) T-cell population only after vaccination (Fig. 5A). In Pt #2, while specific CD4+ T cells were not observed before vaccination, NY-ESO-1119141–specific CD4+ T cells were elicited after vaccination. The vaccine-induced NY-ESO-1119141–specific CD4+ T cells were also detected in the CD4+CD25CD45RO+ (effector/memory) T-cell population, as observed in Pt #1 (Fig. 5B).

Figure 5.

Vaccination with OK-432 elicits NY-ESO-1–specific effector/memory CD4+ T cells

CD4+CD25CD45RO+ T cells or unfractionated CD4+ T cells were prepared from PBMCs of (A) Pt #1 (HLA-DR 4, 12 -DQ 4, 8) and (B) Pt #2 (HLA-DR 9, 15 -DQ 6, 9) before and after vaccination as described in the Materials and Methods; 3–5 × 105 CD4+CD25CD45RO+ T cells or unfractionated CD4+ T cells were cultured with 3–5 × 105 CD4-depleted PBMCs pulsed with 10 μM pooled peptides covering the entire sequence of NY-ESO-1 for 3 weeks. Induction of NY-ESO-1–specific IFN-γ production was analyzed by ELISPOT. Data shown are from one experiment representative of at least two experiments performed.

NY-ESO-1 vaccination with OK-432 activates high-avidity preexisting NY-ESO-1–specific CD4+ T-cells

We then asked whether vaccine-induced T cells had a high-affinity TCR that recognized naturally processed antigens [21, 28]. We established NY-ESO-1–specific CD4+ T-cell clones. Four clones and a single clone that recognized different epitopes were generated from Pt #1 and Pt #2, respectively. Four minimal epitopes (NY-ESO-18396, 94–109, 119–130,121–134) were defined from CD4+ T-cell clones derived from Pt #1 (Fig. 6A and data not shown). Both spontaneously induced (#2–11) and vaccine-induced (#3–1) CD4+ T-cell clones recognized naturally processed NY-ESO-1 protein and as little as 0.1 nM of peptide (Fig. 6A). One minimal epitope defined from Pt #2 was NY-ESO-1122133 and the vaccine-induced CD4+ T-cell clone (#1–1) again recognized both the naturally processed NY-ESO-1 protein and as little as 0.1 nM of peptide (Fig. 6B), indicating that these T-cell clones had high-affinity TCRs against NY-ESO-1. Together, OK-432 as an adjuvant could overcome Treg-cell suppression and activate high-affinity preexisting NY-ESO-1–specific CD4+ T-cell precursors.

Figure 6.

Vaccination with OK-432 activates high-avidity NY-ESO-1–specific CD4+ T cells. NY-ESO-1–specific CD4+ T-cell clones were generated from PBMCs of patients harboring NY-ESO-1–specific CD4+ T cells. CD4+ T-cell clones from (A) Pt. #1 and (B) Pt. #2 were stimulated and expanded by 30 ng/mL anti-CD3 Ab in the presence of 20 U/mL IL-2 and irradiated PBMCs and EBV-transformed human B lymphocytes as feeder cells. A couple of weeks later, these NY-ESO-1–specific CD4+ T cell clones (1 × 104/well) were cultured with 1 × 104 NY-ESO-1 protein-pulsed DCs or 2 × 104 EBV-B cells pulsed with graded amounts of peptides and NY-ESO-1–specific IFN-γ production was analyzed by ELISPOT. Data are shown as mean ± SD of two replicates and are from one experiments representative of at least two experiments performed.

Discussion

While a subset of patients treated with immunotherapy has been shown to experience objective and durable clinical responses, it is becoming increasingly clear that several mechanisms downregulate antitumor immunity during the course of the immune response and play a major role in limiting the effectiveness of cancer immunity [6, 35, 36]. A plethora of cell types, cell surface molecules, and soluble factors mediate this suppressive activity [3, 6, 35, 36]. Among them, CD4+CD25+Foxp3+ Treg cells play a crucial role by suppressing a wide variety of immune responses, and finding ways to control Treg-cell suppression is a major priority in this field [6, 7]. In this study, we showed the potential of OK-432 (a penicillin-inactivated and lyophilized preparation of Streptococcus pyrogenes) which stimulates TLR signals [30, 33, 34] to control Treg-cell suppression, supporting the idea that OK-432 may be a promising adjuvant for cancer vaccines by inhibiting Treg-cell suppression and by augmenting induction of tumor-specific T cells against coadministered protein antigens.

Appropriate adjuvant combinations, such as those that are MyD88-dependent or MyD88-independent, or those that are TRIF-coupled and include endosomal signals, are known to synergistically activate DCs with regard to the production of inflammatory cytokines [37, 38]. As OK-432 is derived from bacterial components, its capacity to bind a combination of various TLRs makes it attractive. It has been shown that OK-432 exhibits antitumor effects through TLR-2, TLR-4, and TLR-9 using knockout mice for each TLR [30, 33, 34]. Alternatively, OK-432 reportedly stimulates DCs through the β2-integrin system rather than via TLR signals [29]. In the presence of OK-432, Treg cells slightly proliferated with TCR stimulation. TLR2 triggering results in a temporary loss of the anergic status of Treg cells and is associated with loss of Treg-cell suppressive function [24, 25]. The perturbation of Treg-cell anergy by OK-432 through TLR2 stimulation may play a role, at least in part, in the inhibition of Treg-cell suppressive function.

In accordance with previous reports [29, 34], we showed that APCs, including CD11c+ and CD14+ cells (monocytes, macrophage, and DCs), stimulated with OK-432 exhibited significantly higher production of IL-12 as compared with that of LPS- or TNF-α–matured APCs, and that OK-432–induced IL-12 from these APCs was a critical component for abrogating Treg-cell activity. Additionally, we found that monocyte-derived DCs stimulated with OK-432 produced significantly higher amounts of IL-12 compared with DCs stimulated with LPS or TNF-α (Supporting Information Fig. 2). It has been reported that IL-12 receptor expressed on effector T cells, but not on Treg cells has a critical role for abrogating Treg-cell suppression by IL-12 in mice [39, 40]. In accordance with this, downregulation of IL-12 receptors by siRNA on effector cells partially abrogated the OK-432–induced inhibition of Treg-cell suppressive activity (Supporting Information Fig. 3). IL-12 receptor was induced in both effector T cells and Treg cells after activation (Supporting Information Fig. 3). We attempted to downregulate the IL-12 receptor on Treg cells with siRNA to explore the exact target(s) of IL-12, however, the limitation in the availability of human materials hampered these analyses. Thus, IL-12 produced by APCs on the OK-432 stimulation could have two (or more) mutually compatible activities, (i) rendering effector cells resistant to Treg-cell suppression and (ii) inhibiting Treg-cell suppressive function directly, though the in vivo data argue against direct inhibition of Treg-cell suppression [39, 40].

Local administration of OK-432 reduced the number of CD4+CD25+Foxp3+ Treg cells in tumor-associated exudate fluids. After administration of OK-432, local chemokine gradient may be changed and infiltration of Treg cells may be blocked [6, 13]. Alternatively, the inflammatory environment after OK-432 administration may be suitable for effector T-cell activation and IL-2, that is critical for Treg-cell survival and function [41], may not be adequately provided, as observed during severe Toxoplasma gondii infection [42]. In addition, suppressive function of CD4+CD25high T cells in tumor-associated exudate fluids was reduced after OK-432 treatment in accordance with decreased expression of Foxp3 [43]. Considering the fact that IL-12, a main effector molecule induced by OK-432, renders effector cells resistant to Treg-cell suppression, direct administration of OK-432 may change the immunological balance in the local microenvironment from suppression by Treg cells to activation of helper T cells by augmenting helper T-cell activity. However, the sample size of patients analyzed in this study was relatively small and warrants cautious interpretation.

We have previously shown that while naive NY-ESO-1–specific CD4+ T-cell precursors are present in wide range of healthy individuals and cancer patients, their activation is kept under stringent CD4+CD25+ Treg-cell control [20, 21, 28]. Using OK-432 as an adjuvant, we detected high-affinity NY-ESO-1–specific CD4+ T cells in effector/memory population after vaccination in two esophageal cancer patients. In Pt #1, we found two responses; spontaneous and vaccine-induced NY-ESO-1–specific CD4+ T cells. Both of them exhibited a similar efficiency to recognize titrated peptide, indicating that these NY-ESO-1–specific CD4+ T cells had TCRs with similar affinity and were likely activated from naive high-affinity NY-ESO-1–specific CD4+ T-cell precursors. Vaccination with minimal peptide in incomplete Freund's adjuvant fails to activate high-affinity NY-ESO-1–specific CD4+ T-cell precursors, rather it dominantly expands low-avidity effector/memory CD4+ T cells that cannot recognize naturally processed antigens [21]. In addition, following DNA vaccination covering the entire sequence of NY-ESO-1, high-avidity NY-ESO-1–specific CD4+ T cells were not detected persistently because of rapid suppression by Treg cells [44]. While these data suggest a critical role for the inhibition of Treg-cell suppression by OK-432 in the activation of high-affinity NY-ESO-1–specific CD4+ T-cell precursors, it is still difficult to obtain conclusive evidence without direct in vivo Treg-cell inhibition/depletion. To formally address this issue, clinical trials using Treg-cell depletion reagents and another clinical trial having two arms of patients receiving NY-ESO-1 with/without OK-432 would be required.

Certain types of immunization methods or DC stimulations elicit/augment CD4+CD25+ Treg cells in vivo [10-12, 45]. As many tumor-associated antigens recognized by autologous tumor-reactive lymphocytes are antigenically normal self-constituents [1-3], they also could be recognized with CD4+CD25+ Treg cells. Given that a proportion of cancer/testis antigens are targets of Treg cells [46], it is necessary to avoid unwanted activation of cancer/testis antigen-specific CD4+CD25+ Treg cells. Though the sample size of patients analyzed in this study was small and warrants cautious interpretation, including OK-432 in vaccine components as an adjuvant would be a promising strategy to establish favorable circumstances for stimulating effector T cells by inhibiting Treg-cell activation. Furthermore, since this agent has a long history and is widely applied as an anticancer drug, particularly in Japan, its clinical safety profile has been already established. Our data provide a critical cue for effective cancer vaccines and immunotherapy during antigen priming through modulation of CD4+CD25+ Treg-cell function.

Materials and methods

Blood samples

All healthy donors were subjects with no history of autoimmune disease. PBMCs, pleural effusions, or ascites from cancer patients were collected before and after local administration of OK-432 based on the protocol approved by the Human Ethics Committees of Mie University Graduate School of Medicine and Nagasaki University Graduate School of Medicine. PBMCs from esophageal cancer patients enrolled in a clinical trial of CHP-NY-ESO-1 and CHP-HER2 vaccination with OK-432 [47] (Supporting Information Fig. 1) were collected based on the protocol approved by the Human Ethics Committees of Mie University Graduate School of Medicine and Kitano Hospital. The clinical trial was conducted in full conformity with the current version of the Declaration of Helsinki and was registered as NCT00291473 of Clinical Trial. gov, and 000001081 of UMIN Clinical Trial Registry. All samples were collected after written informed consent.

Abs and reagents

Synthetic peptides of NY-ESO-1120 (MQAEGRGTGGSTGDADGPGG), NY-ESO-11130 (STGDADGPGGPGIPDGPGGN), NY-ESO-12140 (PGIPDGPGGNAGGPGEAGAT), NY-ESO-13150 (AGGPGEAGATGGRGPRGAGA), NY-ESO-14160 (GGRGPRGAGAARASGPGGGA), NY-ESO-15170 (ARASGPGGGAPRGPHGGAAS), NY-ESO-16180 (PRGPHGGAASGLNGCCRCGA), NY-ESO-17190 (GLNGCCRCGARGPESRLLEF), NY-ESO-181100 (RGPESRLLEFYLAMPFATPM), NY-ESO-191110 (YLAMPFATPMEAELARRSLA), NY-ESO-1101120 (EAELARRSLAQDAPPLPVPG), NY-ESO-1111130 (QDAPPLPVPGVLLKEFTVSG), NY-ESO-1119143 (PGVLLKEFTVSGNILTIRLTAADHR), NY-ESO-1131150 (NILTIRLTAADHRQLQLSIS), NY-ESO-1139160 (AADHRQLQLSISSCLQQLSLLM), NY-ESO-1151170 (SCLQQLSLLMWITQCFLPVF), NY-ESO-1161180 (WITQCFLPVFLAQPPSGQRR), and HIV P173751 (ASRELERFAVNPGLL) [48] were obtained from Invitrogen (Carlsbad, CA, USA). Recombinant NY-ESO-1 protein was prepared using similar procedures as described previously [49]. OK-432 was purchased from Chugai Pharmaceutical (Tokyo, Japan). LPS (Escherichia coli 055:B5) was obtained from Sigma (St. Louis, MO, USA). Purified and FITC-conjugated anti-IL-12 (C8.6; mouse IgG1), purified anti-IL-6 (MQ2–13A5; rat IgG1), purified anti-IFN-γ (NIB42; mouse IgG1), purified anti-IL-23 (HNU2319; mouse IgG1), PE-conjugated anti-CD20 (2H7; mouse IgG2b) and PE-conjugated anti-CD56 (MEM188; mouse IgG2a) Abs were purchased from eBioscience (San Diego, CA, USA). Purified anti-IL-1β Ab (8516; mouse IgG1) was purchased from R&D Systems (Minneapolis, MN, USA). PE-conjugated anti-CD14 (MϕP9; mouse IgG2b), PE-conjugated anti-CD45RA (HI100; mouse IgG2b), PerCP-conjugated anti-CD4 (RPA-T4; mouse IgG1), and FITC-conjugated anti-CD4 (RPA-T4; mouse IgG1), Foxp3 (259D; mouse IgG1), and CD45RO (UCHL1; mouse IgG2a) Abs were purchased from BD Biosciences (Franklin Lakes, NJ, USA). PerCP-Cy5.5-conjugated anti-CD11c Ab (3.9; mouse IgG1) was obtained from Biolegend (San Diego CA, USA). PE-conjugated anti-CD25 Ab (4E3; mouse IgG2b) was obtained from Miltenyi Biotec (Bergisch Gladbach, Germany). Recombinant IL-6, IL-12, and TNF-α were purchased from PeproTech (Rocky Hill, NJ, USA).

Intracellular cytokine staining

PBMCs were cultured with/without OK-432 and GolgiStop reagent (BD Biosciences) for 20 h. Cells were stained for cell surface markers and then for intracellular cytokine (IL-12) after permeabilization. Results were analyzed by flow cytometry (FACSCanto; BD Biosciences).

Generation of NY-ESO-1–specific CD4+ T cells

NY-ESO-1–specific CD4+ T cells were elicited as described previously [20]. Briefly, CD4+ T cells and CD4+CD25 T cells were isolated from PBMCs using a CD4+CD25+ Treg Isolation Kit (Miltenyi Biotec). CD4+CD25 T cells were further separated into CD45RO+ T cells or CD45RA+ T cells by FACSAria (BD Bioscience) after staining with anti-CD45RO and CD45RA Abs. CD4 PBMCs pulsed with 10 μM of peptide overnight were used as APCs. After irradiation, 5 × 105 APCs were added to round-bottom 96-well plates (Nunc, Roskilde, Denmark) containing 1–5 × 105 unfractionated CD4+ or CD4+CD25CD45RO+ T cells and were fed with 10 U/mL IL-2 (Kindly provided by Takeda Pharmaceutical, Osaka, Japan) and 20 ng/mL IL-7 (R&D Systems). Subsequently, one-half of medium was replaced by fresh medium containing IL-2 (20 U/ml) and IL-7 (40 ng/mL) twice per week.

Generation of NY-ESO-1–specific CD4+ T-cell clones

Cloning was performed by limited dilution as described previously [50]. Briefly, NY-ESO-1–specific CD4+ T cell lines (0.3 cells/well) were stimulated and expanded in the presence of irradiated 5 × 104 cells/well PBMCs and 1 × 104 cells/well irradiated EBV-transformed human B lymphocytes with 10% AB serum, 20 U/ml IL-2, and 30 ng/mL anti-CD3 Ab (OKT3; eBioscience) in 96-well round-bottom plates.

Proliferation assay

CD4+CD25 T cells were cultured with 1 × 105 irradiated CD4-depleted PBMCs and stimulated with 0.5 μg/mL anti-CD3 Ab (OKT3, eBioscience) in round-bottom 96-well plates. CD4+CD25high Treg cells (highest 3% of CD4+CD25+ cells) were purified with FACSAria (BD Biosciences), and graded numbers of them added in the culture as indicated in figure legends. Proliferation was evaluated by 3H-thymidine with 1 μCi/well for the last 18 h of 6-day culture. 3H-thymidine incorporation was measured by a scintillation counter.

ELISPOT (enzyme-linked immunospot) assay

The number of IFN-γ secreting antigen-specific CD4+ T cells was assessed by ELISPOT assays as described [20, 21]. Briefly, flat-bottomed, 96-well nitrocellulose-coated microtiter plates (Millipore, Bedford, MA, USA) were coated with anti-IFN-γ Ab (1-D1K; MABTECH, Stockholm, Sweden). The presensitized T cells and phytohaemagglutinin (PHA HA15; Murex Diagnostics, Dartford, UK) activated CD4+ T cells, EBV-transformed human B lymphocytes or DCs pulsed with 10 μM of peptides or 25 μg/mL protein overnight were added to each well and incubated for 24 h. Spots were developed using biotinylated anti-IFN-γ Ab (7-B6–1-biotin; MABTECH), alkaline phosphatase conjugated streptavidin (Roche, Mannheim, Germany) and 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (Sigma) and counted with C.T.L. Immunospot analyzer and software (Cellular Technologies, Cleveland, OH, USA).

Preparation of monocyte-derived DCs

Monocyte-derived DCs were generated from PBMCs as previously described with some modifications [51]. Briefly, CD14+ monocytes were enriched by positive selection using CD14 Microbeads (Miltenyi Biotec). Monocytes were cultured in the presence of 20 ng/mL GM-CSF (Immunex, Seattle, WA, USA) and 20 ng/mL IL-4 (R&D systems) in RPMI1640 supplemented with 2.5% fetal calf serum. Medium was replaced by fresh medium containing cytokines 3 days later. On day 6, cells were harvested and used for subsequent experiments.

ELISA

The concentration of IL-12p70 and IL-10 was measured by ELISA Kit (eBioscicence) according to the instruction provided by the manufacturer.

Statistical analysis

Statistical significance was evaluated by Student's t-test; p values less than 0.05 are considered significant.

Acknowledgments

This article is dedicated to the memory of Lloyd J. Old, M.D. We thank Drs. T. Takahashi and J. B. Wing for critical reading of the manuscript, and L. Wang, C. Brooks, E. Krapavinsky, E. Ritter, and D. Santiago for technical support.

This study was supported by Grant-in-Aid for Scientific Research on Priority Areas (No. 17016031, H. Shiku, and No. 20015019, H. Nishikawa) and Grants-in-Aid for Scientific Research (B) (No. 23300354, H. Nishikawa), the Cancer Research Institute Investigator Award (H. Nishikawa) and Cancer Vaccine Collaborative Grant for Immunological Monitoring (S. Gnjatic, G. Ritter and L.J. Old), Cancer Research Grant from Foundation of Cancer Research Promotion (H. Nishikawa), Takeda Science Foundation (H. Nishikawa), Kato Memorial Bioscience Foundation (H. Nishikawa), the Sagawa Foundation for Promotion of Cancer Research (H. Nishikawa), and Senri Life Science Foundation (H. Nishikawa). MH is a research fellow of the Japan Society for the Promotion of Science.

Conflicts of interest

The authors declare no financial or commercial conflict of interest.

Abbreviation
CHP

cholesteryl hydrophobized pullulan

Ancillary