Background
Cancer immunotherapy is considered a promising therapeutic approach in oncology, and the recent successes obtained by Provenge and Ipilimumab support this view [
1‐
3]. However, despite the discovery of a great number of tumor associated antigens (TAA) [
4] and the setting of a large variety of immunotherapy protocols [
5], their clinical efficacy remains dismal [
4,
6]. This is likely due to: (a) poor immunostimulatory efficacy of immunotherapy procedures, and (b) escaping mechanisms, as the accumulation of regulatory lymphocytes (Treg) within the tumor environment causing the impairment of anti-tumor cytotoxic cells [
7]. Therefore, in presence of a wide array of TAA [
8] and a variety of immunotherapy protocols [
5], a systematic analysis leading to the identification of reference parameters for the selection and application of each single TAA is mandatory. This is the daunting challenging effort run by the NCI [
8] and informative guidelines came from the CIC [
9]. Thus, in the setting of an immunotherapy protocol using a specific TAA, a preliminary comparative analysis is recommended in order to identify the most immunogenic strategy able to achieve an optimal anti-tumor response.
Interleukin 10 (IL-10) is a pleiotropic cytokine secreted by a wide array of immune cells, including monocytes, macrophages, T cells, dendritic cells (DC), B cells, natural killer (NK) cells, mast cells, neutrophilic and eosinophilic granulocytes, and by several tumor cells [
10,
11]. This cytokine signals mainly via STAT3 regulating immunomodulating activities [
12]. In particular, it decreases the antigen presenting activity of macrophages and DC mainly through HLA and co-stimulatory molecules down-regulation. Moreover, IL-10 suppresses the production of pro-inflammatory cytokines (i.e. IL-1α, IL-1β, TNF-α, IL-6, IL-8, IL-12, IL-18, granulocyte–macrophage colony-stimulating factor [GM-CSF], macrophage inflammatory protein-1, RANTES, leukemia inhibiting factor, IFNγ), and inhibits nuclear translocation of NF-kB [
13,
14]. At the same time, this cytokine is a NK cell activator [
15] and a co-stimulator of mast cell proliferation [
16,
17]. Finally, IL-10 promotes B lymphocyte differentiation and immunoglobulin production [
18], and plays a relevant role in immune regulation, mediating the activity of different regulatory T cell (Treg) subsets [
19‐
21]. This pivotal role, associated with the finding that the tumor environment is usually rich in IL-10 secreted by tumor cells [
22,
23] and/or by tumor infiltrating elements [
24], suggests that this cytokine has a critical function in tumor immune escaping. Thus, the inhibition of IL-10 has been proposed as a useful strategy in anti-cancer therapy [
25,
26].
In this study, we focused on a single TAA, gp100, in an established melanoma model, analyzing different gp-100 centred vaccination strategies. Our main aim was to identify a protocol able to consistently provide 100% tumor protection, and this aim was achieved only when the most efficacious effector strategy was associated with IL-10 blockade to counteract Treg activity.
Materials and methods
Cell lines
The cell lines used in this study were: a) B16F10 melanoma cells (a spontaneous C57BL/6J-derived melanoma widely used for the evaluation of therapy [
27]); b) RMA/S cells, a murine T-lymphoma cell line deficient in the presentation of endogenously synthesized antigens by MHC class I molecules [
28] (kindly provided by Dr. G. Pietra from the Advanced Biotechnology Center of Genoa); c) VAC cell line, which is an immortalized line from primary lymphoma derived from a NPMALK–transgenic mouse backcrossed onto the BALB/c background (CD4+/ALK+ lymphoblastic lymphoma) [
29], and used for generating syngenic s.c. tumorgrafts of Anaplastic Large Cell Lymphoma (ALCL).
All lines were cultured in RPMI 1640 medium (GIBCO, Life Technologies Ltd., Paisley, UK) containing 10% heat-inactivated fetal bovine serum (FCS, Euroclone, Wetherby, UK), 2 mM/L L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin (GIBCO) in a humidified atmosphere at 37°C and 5% CO2.
Mice
C57BL/6J, 6- to 8-week old, female mice were purchased from Harlan Laboratories (S.Pietro al Natisone, Udine, Italy). BALB/c mice were purchased from Charles River Laboratories (Calco, Lecco, Italy).
Animal procedures were conducted in accordance with the institutional guidelines and experiments have been reviewed and approved by Ethics Committee for Experimentation on Animals (CSEA) of Genoa and Turin.
Synthesis of the highly immunogenic peptide encompassing residues 25–33 of mouse and human gp100
Mouse (EGSRNQDWL) and human (KVPRNQDWL) gp100
25–33 peptides, as well as influenza virus nucleoprotein A NP
366-374 (ASNENMETM) peptide, were in house synthesized at the biochemical facility of the Centre of Excellence for Biomedical Research by standard methods of solid phase peptide synthesis, which follows a 9-fluorenylmethoxycarbonyl (Fmoc) strategy with minor modifications [
30]. Synthesized compounds were purified by reverse-phase high performance liquid chromatography (HPLC) and molecular weights confirmed by electrospray ion-trap mass spectrometry. The purification of individual compounds was obtained on a Shimadzu LC-9A preparative HPLC (Shimadzu, Kyoto, Japan) equipped with a Waters C18 μBondapack column (19 x 300 mm; Waters, Milford, MA).
Culture of bone marrow derived DC and activation for the expression of gp100 on its surface
Bone marrow (BM) derived DC were generated as described [
31]. BM cells were harvested from thigh-bones and tibias of naïve C57BL/6J mice and washed with phosphate-buffered saline (PBS) (Invitrogen, Life Technologies). Cells (1x10
6/ml) were suspended in DC medium consisting in RPMI 1640 supplemented with 10% heat-inactivated FCS (HyClone, Logan, UT), 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 100 mM sodium pyruvate, 10 mM Hepes buffer solution, 100 mM non essential amino acids solution, 50 μM β-mercaptoetanol (GIBCO), and plated plate with 20 ng/ml murine GM-CSF (6 well plates, PeproTech, Rocky Hill, NJ). On day 3, fresh medium was added plus 5 ng/ml murine GM-CSF (PeproTech). On day 7, the DC phenotype was checked using the following mAbs: fluorescein isothiocyanate (FITC) conjugated anti-MHC class II (HLA DR), phycoerythrin (Pe) cyanin 7 conjugated anti-CD11c, allophycocyanin (APC) conjugated anti-CD86 (Biolegend, San Diego, CA). More than 90% of the cells showed high expression of all DC markers. Before administering DC to animals, DC (1x10
6/ml) were activated with CpG (1 μg/ml) for 8 hours; then, activated DC (2x10
6/ml) were pulsed by incubation with mouse or human gp100
25-33 peptide (2 μM) at 37°C for 2 hours.
gp100 vaccination of C57BL/6J mice against B16F10 melanoma
In vivo experiments were performed analyzing at least 5 animals per group; each experiment was repeated for 3 times.
C57BL/6J mice were immunized against gp100 molecule following three different protocols: 1) peptide plus adjuvant; 2) gene immunization; 3) peptide-pulsed DC immunization (summarized in Additional file
1: Figure S1 in Additional files).
Peptides plus adjuvant protocol: peptides were administered by injecting subcutaneously either mouse (mgp100
25-33) or human (hgp100
25-33) gp100
25-33 preparations (100 μg/mouse) in association with CpG (30 μg/mouse), as adjuvant. The unrelated NP
366-374 peptide from influenza virus nucleprotein A was used as negative control in preliminary experiments.
Gene immunization: this protocol was executed administering intramuscularly plasmids expressing either mouse or human gp100 gene sequence (100 μg/mouse). A pCMVscript plasmid (Stratagene) was used as gene vector. The gene coding for mouse gp100 molecule was amplified from cDNA obtained after extraction of RNAs from B16F10 melanoma cells. The construct coding for human gp100 was kindly provided by Prof. N. Restifo (NIH, Bethesda). PCR amplified products were first cloned into a sequencing vector (TOPO TA cloning from Invitrogen) and then Sanger sequenced. Both mouse and human gp100 cloned genes were subsequently transferred in the eukaryotic expressing vector pCMVscript.The empty pCMVscript plasmid was used as negative control in preliminary experiments.
Peptide-pulsed DC immunization: this protocol was performed injecting in the right flank DC (2x10
6/mouse) pulsed with either mouse or human gp100
25-33 peptide. Unpulsed DC or DC pulsed with the unrelated NP
366-374 peptide were used as negative control in preliminary experiments. The schedule of each protocol consisted in three immunizations, administered weekly. At the time of the third immunization, mice received the tumor challenge, that consisted in sc. injection in the controlateral limb of B16F10 cells (1x10
5cells/mouse). Tumor masses were measured with calipers at 2–3 days intervals by measuring long and short axes. Area was calculated according to the formula: tumor area = length × width in mm
2. Mice were sacrificed when tumors reached >100 mm
2 or when ulceration and/or bleeding developed.
Immunotherapy of B16F10 melanoma in C57BL/6J mice or ALCL in BALB/c mice based on IL-10 neutralization
In vivo experiments were performed analyzing 5 animals per group; each experiment has been repeated 3 times.
C57BL/6J were s.c. challenged with B16F10 melanoma (1x105 cells/mouse) while BALB/c mice were injected s.c. withVAC (1x106 cells/mouse) cells. The neutralizing anti-IL-10 (Anti-Mouse IL-10 Functional Grade Purified, clone JES5-16E3) (eBioscience, San Diego, CA) mAb (150 μg/mouse) was administered subcutaneously at the site of tumor challenge immediately after tumor cell injection as well as after one and two weeks from baseline. A rat IgG2b,k control isotypic antibody (Biolegend, San Diego, CA) was administered using the identical schedule adopted for the anti-IL10 mAb. When IL-10 blockade was associated in combinatorial treatment to gp100 vaccination, the neutralizing anti-IL-10 mAb (150 μg/mouse), or its isotypic control, were subcutaneously administered as described above.
Purification of splenocytes and intratumor lymphocytes
Spleens or tumors, removed from sacrificed mice, were minced with a sharp sterile blade, placed in a 40-μm nylon cell strainer (BD Biosciences, Franklin Lakes, NJ), and pressed with the plunger of a syringe until cellular elements were released. Red blood cells were lysed with red blood cell lysing buffer (Sygma Aldrich), and washed. Tumor or splenocyte suspensions were collected in RPMI 1640 medium supplemented with 10% FCS. Splenocytes from treated and untreated mice (2.5x106/ml) were plated in flasks in the presence of irradiated (3000 rad) splenocytes (1x106/ml) and gp10025–33 peptide (10 μg/ml) and cultured for 5 days. After incubation, splenocytes were collected and separated by density gradient (Biochrom AG, Berlin, Germany). Splenocyte samples were used for phenotypical analysis, the remaining portion for ELISPOT assay.
In the ALCL model, we applied two overnight incubations with ALK inhibitor CEP28122 (20 ng/ml) followed by separation by density gradient before phenotype analysis.
Immunofluorescence and flow cytometric (FACS) analysis
Phenotypes of tumor infiltrating T lymphocytes (TIL) and whole splenocytes were analyzed by immunofluorescence incubating the cells (1×105 lymphocytes in 100 μl of PBS) with specific mAbs at 4°C for 30 minutes in the dark. The following mAbs were used: FITC conjugated anti-GranzymeB (eBioscience), Pe-conjugated anti-Foxp3 (eBioscience), Peridinin chlorophyl protein (PerCP)-cyanin 5.5-conjugated anti-CD8 (Biolegend), anti-CD28(Biolegend), PeCy 7-conjugated anti-CD3(BD Pharmingen), APC-conjugated anti- (BD Pharmingen), APCCy 7-conjugated anti-CD4 (BD Pharmingen). For intracellular staining, cells were permeabilized and fixed using FOXP3 Fix/Perm Buffer Set (Biolegend) according to manufacturer’s instructions and then incubated with fluorochrome conjugated anti-GranzymeB and anti-Foxp3mAb; fluorochrome-conjugated isotype matched Abs were also used as controls. After the staining, the analysis was performed by flow cytometry using a FACSCanto II flow cytometer equipped with FACS Diva software (BD).
Elispot assay
Detection of IFNγ production by splenocytes in response to peptide stimulation was carried out using an enzyme-linked immunospot (Elispot) assay (Millipore, Merck KGaA, Darmstadt, Germany) according to the manufacturer’s instructions. Briefly, splenocytes (2x105/well) from treated and untreated mice, cultured for 5 days with irradiated splenocytes plus peptide (10 μg/ml) were harvested, and dead cells were removed by centrifugation on density gradient (Biochrom). Then, cells were collected and washed before incubating them overnight at 37°C with either mgp10025-33- or hgp10025-33-pulsed RMA/S cells (1x104 cells) in 96-well plates coated with 10 μg/ml of unlabeled anti-mouse IFNγ rat monoclonal capture antibody (clone AN18, Millipore). At the end of incubation, a biotinylated detection antibody (2 μg/ml) was added to the wells and reacted with alkaline phosphatase-streptavidin (100 μl/well) and the development with BCIP/NBT Phosphatase substrate (50 μl/well) (KPL, Kirkegaard & Perry Laboratories, Inc.). Frequencies of IFNγ producing cells were measured using a BioReader 3000 Elispot Reader (Bio-Sys GmbH). Data are expressed as the mean number of spots per duplicate.
Proliferation inhibition assay
The suppression activity was evaluated by monitoring the proliferation of splenocytes isolated from C57BL/6J and from BALB/c mice by dye dilution approach. For this assay, splenocytes (used as responder cells) were stained with carboxyfluorescein succinimidyl ester (CFSE) 5 μM (Molecular Probes, Life Technologies) and co-cultured with Concanavalin A (Con A) (Sigma Aldrich) at 5 μg/ml for 5 days in a 96 well flat bottomed plate with or without tumor infiltrating lymphocytes (TIL) from melanoma or lymphoma tumors (used as suppressor cells). Splenocytes:TIL cell ratio was 2:1. Co-cultures with TIL from both tumors were also performed in the presence or absence of the neutralizing anti-IL-10 mAb (clone JES5-16E3 eBioscience) (10 μg/ml) or of its isotypic control. After 5 days the cultures were washed in PBS and acquired by FACS Canto flow cytometer equipped with FACS Diva software (Becton Dickinson, BD). A total of 50000 CFSE positive responder cells were analyzed. The results were expressed as percentage of CFSE bright splenocytes.
Immuno-histochemical analysis
Tumors specimens were taken from either anti-IL-10 mAb treated mice or controls and fixed with nitrogen vapors. Tissue sections, obtained from melanoma specimens, were stained with monoclonal antibody against IL-10 (Purified Rat anti-mouse IL-10, clone JES5-16E3, BD, Milano, Italy) following the manufacturers’ instructions.
Statistical analysis
Comparisons between mean values were performed by unpaired or paired t test using the Graph-Pad Prism 5.0 software (Graph-Pad Software, Inc, San Diego, CA, USA). Data were considered statistically significant when p ≤ 0.05.
Discussion
Our data demonstrate that: a) the vaccination with DC pulsed with gp10025-33 peptide was the most effective in protecting C57BL/6 mice from B16F10 melanoma development compared to other gp100 vaccination protocols such as peptide plus adjuvant or gene vaccination; b) different protective effects were observed in syngeneic or xenogeneic settings in gp100 peptide-centred cancer vaccination protocols; c) the protective effects exerted by DC vaccination were related to increased intratumoral concentration of granzyme+ CD4+ T cells and decreased concentration of CD4+CD25+Foxp3+ Treg; d) counteraction of IL-10 by a neutralizing mAb lowered the rate of tumor development, inducing an increase of intratumoral concentration of granzyme+ CD4+ T cells and a decrease of CD4+CD25+Foxp3+ Treg; e) only the combinatorial treatment associating DC vaccination with IL-10 blockade was able to abrogate completely tumor growth.
Cancer vaccination is based on the identification at the molecular level of antigens expressed preferentially by specific tumors. Although Sipuleucel-T is the first cancer vaccine that recently reached the clinic after receiving the authorization for clinical use in prostate cancer [
34], it is believed that immunological protocols will reach a full clinical efficacy and wide application in oncology only when well defined oncoantigens will be proved efficient by in vivo protective experiments. Hence, this process requires a systematic analysis of the protective effect of the different TAA and related protocols of administration, since several variables, including type of antigen, route of administration, molecular form of antigen administration, selection of adjuvants, may condition the outcome of the treatment [
35]. In this paper, we compare the protective efficacy of different protocols of cancer vaccination based on a classical, widely adopted model system constituted by gp100 vaccination in B16F10 melanoma [
27]. Intriguingly, the comparison of different vaccination protocols, such as peptide plus adjuvant, gene immunization and peptide-pulsed DC immunization, showed different levels of protective efficacy among the three protocols, being DC vaccination the most effective one. This is a remarkable observation since suggests that adequate stimulation of a single (or limited number of) T cell clone(s), specific for one specific TAA epitope, as induced by peptide-pulsed DC vaccination, may have more pronounced protective effects than multi-epitope mediated stimulation of a wide TAA-specific T cell repertoire, as induced by vaccination with a gene coding for the entire TAA molecule. Importantly, the protective effects of identical protocols performed in syngeneic or xenogeneic settings differed greatly. This is not surprising since the efficacy of xenogeneic immunization has been already reported in different models of experimental tumor and associated with the expression of a heteroclitic epitope by the immunizing agent [
36,
37]. The use of xenogeneic or modified tumor antigens, as a system for increasing the immunogenicity of the vaccine and breaking tolerance, has been suggested for long time [
4,
38]. However recent evidences from human trials seems to disprove the efficacy of this approach [
39]. Taken together, data reported here further demonstrate the need of experimental testing to assess the most efficient protocols. Indeed, our analyses deserve to be replicated in different tumor model systems in order to verify whether our findings represent a generalized phenomenon in cancer vaccination.
The protective effect exerted by a cancer vaccination protocol is likely related to a number of biological factors concerning the immunogenicity of the antigen, the competence of the T and B cell repertoires, the level of activation of innate immunity, the capacity of the antigen to elicit immune responses involving a wide scenario of lymphocyte subtypes (T helper, CTL, Th17, B lymphocytes) and, last but not least, the number of boosts [
40‐
44]. Hence, monitoring cancer vaccine-induced immune response, should take into consideration a variety of factors, likely becoming an impossible challenge for clinical trials.
Trying to simplify what underlies complex biological networks, the outcome of an immune response may be function of the ratio between the effector and the regulatory response [
45‐
47]. In this study, we selected the intratumoral concentration of granzyme+CD4+ or CD8+ T cells as a parameter representative of effector immune responses. Similarly, we analyzed intratumoral concentration of CD4+CD25+FoxP3+ Treg as a parameter representative for regulatory circuits. Strikingly, protective effects induced by the different protocols of cancer vaccination were related and dependent on increasing number of tumor infiltrating CD4+granzyme+ T cells and decreasing concentration of Treg. Whether the effector/regulatory ratio can be further exploited to assess the efficacy of a cancer vaccine-induced immune response needs to be analyzed in future studies.
Notwithstanding the clear protective efficacy of peptide-pulsed DC vaccination, this protocol was not able to provide abrogation of tumor growth in 100% of treated mice. This prompted us to test an alternative immunotherapy based on IL-10 counteraction. This choice derives from our previous observation related to the presence within tumors of IL-10-secreting Treg able to strongly impair anti-tumor immune response [
21]. Since presence of both IL-10 and IL-10-secreting Treg was demonstrated in B16F10 melanoma [our data and 24], we explored the possibility to counteract the effects of intratumoral IL-10 to unleash protective effector responses. In fact, the administration of a neutralizing anti-IL-10 mAb exerted protective effects similar to that demonstrated by peptide-pulsed DC vaccination, inducing intratumoral recruitment of CD4+granzyme+ T cells and deplenishing of CD4+CD25+FoxP3+ Treg. These effects were not specific for B16F10 melanoma and they could be replicated in the unrelated cancer model of ALCL. More importantly, when peptide-pulsed DC vaccination and IL-10 blockade were applied in a combinatorial protocol, complete abrogation of tumor growth was achieved in 100% of treated mice.
Competing interests
The authors have no competing interests to disclosure.
Author’s contributions
KF performed in vivo and in vitro experiments and drafted the manuscript, MR carried out immunization based on IL-10 neutralization in ALCL tumor model, BF performed in vivo and in vitro experiments, PA and FD carried outflow cytometric analysis, CG, FF and TS participated in experiments performed with plasmids, PM performed in vivo and in vitro experiments, SM performed tumor implantation, ME synthesized peptides, DPR carried out immunohistochemical analysis and participated in writing the paper, IG participated in designing the experimental flow chart and supervised the results, FG participated in designing the experimental flow chart, supervised the results and wrote the paper. All authors read and approved the final manuscript.