Introduction
Some of the most abundant proteins in the cell belong to the well-conserved family of proteins known as heat shock proteins (HSPs), or glucose-regulated proteins (GRPs). HSPs are present in all living cells; they can exist in an unbound state or a state bound to specific client proteins. HSPs function as molecular chaperones in numerous processes, such as protein folding, assembly and transport, peptide trafficking, and antigen processing under physiologic and stress conditions [
1,
2]. Levels of HSPs are elevated in many cancers [
3,
4]. One of the first identified HSP subtypes, Gp96, can reject tumors [
5]. HSP as a natural adjuvant can elicit in cancer patients a specific and active autoimmune response to a tumor [
6]. During tumor formation, HSPs increase and bind to exposed hydrophobic tumor polypeptides. HSP-chaperoned peptides enter antigen-presenting cells through specific receptors and prime T cells by increasing major histocompatibility complex (MHC) class I and II-mediated antigen presentation [
7‐
9]. The relevance of the peptides associated with HSPs for inducing specific immune responses is demonstrated by numerous studies, and GRP96, HSP70, HSP110 and GRP170 purified from diverse tumors and functioning as tumor vaccines have shown to cause tumor regression in animal models [
10‐
13]. The factor is successful in CD8
+ T cell-dependent tumor clearance. The immune recognition does not come from HSPs themselves but from binding to peptides [
14]. Some HSPs, such as HSP60 and HSP70, augment natural killer (NK) cell activity, which can also elicit innate immune responses [
15,
16].
As an alternative to selecting a single antigen for tumor vaccine development, random mutations in cancer cells generate antigens unique to an individual. Purification of chaperone HSP from a cancer is believed to co-purify an antigenic peptide "fingerprint" of the cell of origin [
17]. Thus, a vaccine comprising HSP/peptide (HSP/P) complexes derived from a tumor, which would include a full repertoire of patient-specific tumor antigens, obviates the need to identify cytotoxic T-lymphocyte (CTL) epitopes from individual cancers. This advantage extends the use of chaperone-based immunotherapy to cancers for which specific tumor antigens have not yet been characterized [
18].
After an extensive study, HSPs were found to augment tumor antigen presentation and NK cell activity leading to tumor lysis. Autologous patient-specific tumor vaccines have been generated by purifying HSP-antigen complexes from tumor specimens and are currently being evaluated in clinical trials. Preliminary clinical trials with Gp96 used as a personalized vaccine for immunotherapy in melanoma, renal, colon, ovarian cancer and non-Hodgkin lymphoma have reported results [
19‐
23]. HSP70 as a vaccine for leukemia was studied in a clinical trial [
24]. Although various immunotherapeutic approaches have been examined for the treatment of cancer, no such therapy has entered into the clinical standard of care, and the therapeutic effects was not satisfactory. Several challenges still need to be overcome.
Until now, all clinical trials have used the single subtype of HSPs, Gp96 or HSP70, whereas in a few animal tumor models, the combination of Gp96 and HSP70 has been shown to possess antitumor activity superior to the that of each type alone [
25]. These results suggest that the mixture of several HSP subtypes may be more effective in a broad range of tumor models. We used the mixture of HSP/Ps (mHSP/Ps) that include HSP60, HSP70, HSP110 and GRP96 as a vaccine and found an effective prophylactic antitumor effect of the mHSP/Ps in a mouse sarcoma model [
26,
27]. The effect protected against tumor challenge in 50% of immunized mice, but this strategy for the therapeutic treatment in already established tumors were not satisfactory, so enhancing the therapeutic immunity is needed.
Using cytokines to enhance immune reactivity has been reported both in experimental and clinical trials [
28]. Interleukin 12 (IL-12) is still the most important single cytokine in inducing antitumor immunity. In experimental tumor models, recombinant IL-12 has demonstrated marked antitumor effects through mechanisms of both innate and adaptive immunity [
29,
30]. The most unique antitumor activity of IL-12 is its ability to eradicate established tumors [
31,
32]. However, the significant antitumor activity of IL-12 in these models requires the presence of pre-existing immunity in tumor-bearing hosts [
33]. Thus, further improvement of IL-12-based immunotherapy also depends on the combination of vaccine-based modalities to establish pre-existing immunity in tumor-bearing hosts.
When patients are diagnosed with cancer, by definition, the tumor has "escaped" the immune system, having passed the phases of "elimination" and "equilibrium." The generation of immune response against these antigens is likely unproductive in the late stage because of multiple immune tolerance mechanisms such as Treg infiltration in the tumor bed, general immune suppression from immunosuppressive cytokines producing by tumor cells, and downregulation of MHC class I molecules on the tumor cells. Also, myeloid-derived suppressor cells (MDSCs) and tumor-associated macrophages (TAMs) create an immunosuppressive environment that leads to suppression of T-cell responses [
34,
35]. Thus, multiple immunological "brakes" need to be lifted to augment a productive immune response. Combined immunotherapeutic modalities need to be seriously considered. The use of combination therapy with more than one agent or modality is needed. To overcome the multiple immune tolerance mechanisms, combinations of anticancer drugs and immunotherapy have been shown to enhance tumor immunotherapy [
36,
37]. Treating mice with low-dose cyclophosphamide (CY) decreased the number of Tregs and enhanced the immunostimulatory and antitumor effects [
38‐
40].
To improve the efficacy of tumor immunotherapy, we used the mHSP/P vaccine as an agent to induce pre-existing immunity in a tumor-bearing mouse host, and combined with CY plus IL-12 to eradicate established large tumors in a therapeutic antitumor mouse model.
Discussion
Vaccination with HSP/Ps is personalized, delivering tumor antigen as a fingerprint genome. The vaccine is polyvalence. Here we developed a vaccine with a mixture of HSP/Ps which, in addition to HSP70 or Gp96, also included HSp60 and HSP110. The antitumor effects of this mHSP/Ps vaccine were more potent than those of HSP70 or HSP60 alone and of tumor lysates used as vaccine in prophylactic immunization, Table
1. [
25]. When using this mHSP/P vaccine in mice after tumor transplantation (therapeutic immunization), the antitumor action was not effective, as we showed in this study. The efficacy of therapeutic immunization was effective only in the combination therapy that used immunotherapeutic mHSP/Ps combined with CY and IL-12.
Table 1
Comparison of antitumor effects of various HSPs
No. of animals tested | 10 | 10 | 10 | 10 | 10 |
Complete regression, no. (%) | 0 | 4 (40%) | 3 (33.3%) | 1 (10%) | 2 (20%) |
Tumor growth inhibition rate (%) | | 82.3 | 62.3 | 42.6 | 66.2 |
For specific immunotherapy, the identical MHC genetic molecules are important, We had no information about the MHC genetic molecules of S180 or MCA-207 when we selected the mouse sarcoma cell lines S180 and MCA-207 as models. However, from reported experimental information and our experiments, we knew that the S180 sarcoma cell lines can grow both in BALB/C and C57 mice, as in our control group, in which all the S180 tumors grew and were not rejected. This finding suggests S180 and BALB/C mice have the matched MHC locus even in allogenic transplantation. The MCA-207 only grew in C57 mice but was rejected in BALB/C mice, and this result suggests that the MHC of MCA-207 matched only with the MHC of C57 mice; therefore, in our animal models, the allogenic immune rejection did not occur, and the results of mHSP/P antitumor effects were not related to unmatched MHC.
To identify the specificity of mHSP/P vaccine, we compared the cytolysis ratio of mHSP/Ps isolated from liver and muscle of naïve mice in vitro and saw no cytolytic effect against S180 sarcoma. The cytolysis ratio was lower than 1%. Also, we compared the mHSP/p of S180 against rabbit liver cancer cell line vx2, and the cytolysis effect was lower than 10%, [data not shown]. In addition, we found that the mice vaccinated with mHSP/P of MCA207 were protected only against MCA207 but not S180 in vivo. Thus, the mHSP/P-induced immune reaction may be autologous tumor-specific, like individual vaccines.
IL-12 is highly effective against established immunogenic tumors. In our study, the combination of IL-12 and Cy eradicated tumors in 30% of mice, and in IL-12-treated mice, all tumor mass necrosis and an ulcer formed before tumor eradication, suggesting the anti-angiogenesis activity of IL-12 was involved [
41], When we combined mHSP/Ps with CY and IL-12 to enhance the immunization efficacy, the antitumor efficacy enhanced. However, with mHSP/Ps and CY alone or with mHSP/Ps and IL-12 alone, the antitumor efficacy was not improved. Our results suggested that one potential mechanism of mHSP/Ps and CY plus IL-12 in augmenting therapeutic immunotherapy strategies was that mHSP/P immunization activated the antitumor immunization, and at the same time, also induced the T-cell tolerance directed toward tumor-associated antigens and limited the repertoire of functional tumor-reactive T cells. Therefore, the ability of vaccines to elicit effective antitumor immunity was impaired. CY has immunomodulatory effects, and low-dose CY (20 mg/kg) was found to selectively deplete CD4+CD25+ T cells (Treg) and impede the tolerance [
42]. CY can preconditioning enhance the CD8+ T-cell response to peptide vaccination, thus leading to enhanced antitumor effects against pre-existing tumors [
43]. Cy markedly enhanced the magnitude of secondary but not primary CTL response induced by vaccines and synergized with vaccine in therapy but not in prophylaxis tumor models [
44].
With our enhanced vaccine, IFN-γ secretion was significantly increased. In addition, CD8+ and NK cells were triggered to release IFN-γ and mediate cytotoxic activity. The increased IFN-γ secretion may also be due to the combined effects of HSP60 in mHSP/P and IL-12. Hsp60-inducing IFN-γ depends strictly on the ability of the macrophages to produce IL-12 [
45].
Activation and expansion of tumor-specific T cells by HSP/Ps were identified [
46]. Our study showed that mHSP/Ps purified from S180 sarcoma cells activated tumor antigen-specific T cells
in vitro, and the induction of tumor-specific CTLs with enhanced vaccine was stronger than that with mHSP/Ps alone, possibly because of the combined effect of HSP60 and IL-12. HSP60 induces a strong non-specific immune reaction, but when it meets IL-12, it can activate cytotoxic T cells. HSP60 can mediate the activation of cytotoxic T cells, which depends on production of IL-12 [
47].
Our data showed that inflammatory cells infiltrated tumors with mHSP/P vaccination and that a preexisting antitumor immune response was elicited, which was required for an effective IL-12 response for tumor rejection.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
Q-YG The design of the study. MY Conceived and the design of the study, drafted the manuscript. JP Carried out the animal study and performed the statistical analysis. X-MC Preparation the HSP/P vaccine, carried out the immunoassays. GS Carried out the immunoassays. XS Carried out the animal study and the immunoassays. S-BL Conceived of the study. All authors read and approved the final manuscript.