Melanoma cells treated with GGTI and IFN-gamma allow murine vaccination and enhance cytotoxic response against human melanoma cells

Abstract

Background: Suboptimal activation of T lymphocytes by melanoma cells is often due to the defective expression of class I major histocompatibility antigens (MHC-I) and costimulatory molecules. We have previously shown that geranylgeranyl transferase inhibition (done with GGTI-298) stimulates anti-melanoma immune response through MHC-I and costimulatory molecule expression in the B16F10 murine model [1].

Methodology/principal findings: In this study, it is shown that vaccination with mIFN-gand GGTI-298 pretreated B16F10 cells induces a protection against untreated tumor growth and pulmonary metastases implantation. Furthermore, using a human melanoma model (LB1319-MEL), we demonstrated that in vitro treatment with hIFN-gamma and GGTI-298 led to the up regulation of MHC-I and a costimulatory molecule CD86 and down regulation of an inhibitory molecule PD-1L. Co-culture experiments with peripheral blood mononuclear cells (PBMC) revealed that modifications induced by hIFN-gamma and GGTI-298 on the selected melanoma cells, enables the stimulation of lymphocytes from HLA compatible healthy donors. Indeed, as compared with untreated melanoma cells, pretreatment with hIFN-gamma and GGTI-298 together rendered the melanoma cells more efficient at inducing the: i) activation of CD8 T lymphocytes (CD8+/CD69+); ii) proliferation of tumor-specific CD8 T cells (MelanA-MART1/TCR+); iii) secretion of hIFN-gamma; and iv) anti-melanoma specific cytotoxic cells.

Conclusions/significance: These data indicate that pharmacological treatment of melanoma cell lines with IFN-gamma and GGTI-298 stimulates their immunogenicity and could be a novel approach to produce tumor cells suitable for vaccination and for stimulation of anti-melanoma effector cells.

Figure 1. Vaccination of syngeneic mice with mIFN-γ+GGTI-298 pretreated B16F10 cells induces resistance to untreated tumor growth.

C57BL/6 mice were vaccinated with untreated B16F10 cells ; mIFN-γ (50 IU/mL); GGTI-298 (10 µM); or mIFN-γ (50 IU/mL) and GGTI-298 (10 µM) pretreated B16F10 cells. (A) One month after vaccination tumor growth was measured after s.c. injection of 1×10⁵ live B16F10 cells, by Vernier caliper every 3 days. Groups of 6 or 7 C57Bl/6 mice were tested. Data are representative of three independent experiments. (B) The number of tumor free mice were calculated for each kind of vaccination. Results are pooled for the three independent experiments.

Figure 2. Syngeneic mice vaccination with mIFN-γ+GGTI-298 pretreated B16F10 cells induces resistance to pulmonary metastases implantation.

Four groups of C57Bl/6 mice were vaccinated with untreated B16F10 cells (NT); mIFN-γ (50 IU/mL) (IFN); GGTI-298 (10 µM) (GGTI); or mIFN-γ (50 IU/mL) and GGTI-298 (10 µM) (IFN+GGTI). One month later, they were challenged intravenously with untreated B16F10 cells. (A) Metastases were screened macroscopically and microscopically as illustrated. (B) Lungs from the 4 groups of mice were slidded vertically and micro-metastases were counted and reported to the lung surfaces (mm²).

Figure 3. GGTI-298 enhances hIFN-γ-induced MHC-I expression on human melanoma cells.

LB1319-MEL cells were grown for 4 days in culture medium (filled profiles) or in the presence of hIFN-γ alone (50 IU/ml) (thin lines) or GGTI-298 alone (10 µM) (dotted lines) or IFN-γ and GGTI-298 (thick lines). A) HLA-A,B,C ; HLA-A0201; HLA-DP,DQ,DR and isotypic controls expressions were measured by flow cytometry after staining with PE-conjugated mAbs. B) HLA-A0201 (White Column) HLA-A,B,C (Grey Column) and HLA-DP,DQ,DR (Black Column) membrane expressions were tested by cytofluorometry on LB1319-MEL cells after 4 days treatment with either increasing doses of hIFN-γ (0, 25, 50 and 100 IU/ml) or with the combination of hIFN-γ (50 IU/ml) and GGTI-298 (10 µM). Results are expressed in ISF, related to isotype controls, as indicated in Material and Methods. Data are representative of 3 independent experiments.

Figure 4. On LB1319-MEL cells, GGTI-298 enhances hIFN-γ-induced TAP1 and TAP2 expressions.

LB1319-MEL cells were cultivated for 4 days in the presence or in the absence of hIFN-γ (50 IU/ml) and/or GGTI-298 (10 µM) as indicated. A) The expression of molecules implicated in the MHC-I Ag processing pathway were tested in these treated tumor cells. β-actin was used as protein loaded control. The expressions were quantified by calculating the ratio between the protein of interest and the β-actin. We defined the ratio of relevant protein over β-actin for untreated cells equal to 1. B) Expressions of TAP1 and TAP2 proteins in these in vitro treated and permeabilized LB1319-MEL were tested by flow cytometry using TAP1 and TAP2 specific mAbs and a PE-conjugated secondary Ab. C) TAP1 membrane expression was also tested by cytofluorometry on LB1319-MEL cells after 4 days treatment with either increasing doses of hIFN-γ (0, 25, 50 and 100 IU/ml) or with the combination of hIFN-γ (50 IU/ml) and GGTI-298 (10 µM). Results are expressed in ISF, related to isotype controls, as indicated in Material and Methods. Data illustrated are representative of 3 independent experiments.

Figure 5. On LB1319-MEL cells, GGTI-298 induces enhancement of CD86 and reduction of IFN-γ-induced PD-1L expression.

Membrane expression of CD86 (A) and its isotype control (B) was determined by flow cytometry with PE-conjugated specific antibodies on LB1319-MEL cells after 4 days in vitro treatment with medium alone (filled profiles); hIFN-γ (50 IU/mL) alone (thin lines); GGTI-298 (10 µM) alone (dotted lines); or the combination of both (thick lines). Data are representative of three independent experiments. C) CD86 membrane expression was also tested by cytofluorometry on LB1319-MEL cells after 4 days treatment with either increasing doses of GGTI-298 (0, 5,10, 15 and 20 µM) or with the combination of hIFN-γ (50 IU/ml) and GGTI-298 (10 µM). Results are expressed in percentage of CD86 positive cells. D) Membrane expression of inhibitory molecule PD-1L was determined by flow cytometry with PE-conjugated specific Ab on LB1319-MEL cells after 4 days in vitro treatment with or without hIFN-γ (50 IU/ml) and/or GGTI-298 (10 µM). Results are expressed in ISF, related to isotype controls, as indicated in Material and Methods. Data are representative of 3 independent experiments.

Figure 6. Enhanced activation of CD8+ T lymphocytes in PBMC co-cultivated with LB1319-MEL cells pre-treated with hIFN-γ+GGTI-298.

PBMC from HD HLA-A0201 were stimulated twice with LB1319-MEL cells either untreated; or pretreated with hIFN-γ (50 IU/mL); GGTI-298 (10 µM); or hIFN-γ (50 IU/mL) plus GGTI-298 (10 µM). CD8+ activation was evaluated by CD69 membrane expression. Experiments were performed in triplicate with PBMC from one healthy donor representative of one other giving comparable results.

Figure 7. Enhanced specific cytotoxic activity in PBMC co-cultivated with LB1319-MEL cells pre-treated with hIFN-γ+GGTI-298.

Cytotoxic activities of in vitro stimulated PBMC were tested after two stimulations with LB1319-MEL cells either untreated or pretreated with hIFN-γ (50 IU/mL); GGTI-298 (10 µM); or hIFN-γ (50 IU/mL) plus GGTI-298 (10 µM). A) Cytotoxicity was evaluated by CD107a expression on CD8+ gated T lymphocytes. Cytotoxic activity was also tested on an unspecific target cell: B) Jurkat and the specific one: C) LB1319-MEL. Experiments were performed in triplicate with PBMC from one healthy donor representative of one other giving comparable results.

Figure 8. PBMC stimulation with hIFN-γ+GGTI-298 pretreated LB1319-MEL cells induces specific cells proliferation, labeled with MART-1/HLA-A2 tetramers.

A) HLA-A0201 positive PBMC from one representative HD were stimulated twice in vitro with LB1319-MEL cells either untreated (NT) or pretreated with 50 UI/mL hIFN-γ (IFN), or 10 µM GGTI-298 (GGTI), or with both hIFN-γ and GGTI-298 (IFN+GGTI). At the end of the culture period, specific TCR MelanA / HLA-A2 expression was evaluated on CD8+ gated T lymphocytes. B) HLA-A0201 positive PBMC from the same HD were stimulated twice in vitro with LB1319-MEL cells either untreated (NT) or pretreated with 50 UI/mL hIFN-γ (IFN), or 10 µM GGTI-298 (GGTI), or with both hIFN-γ and GGTI-298 (IFN+GGTI) or with irradiated (HLA-A0201) L1-EBV cells. At the end of the culture period, specific TCR EBV/HLA-A2 expression was evaluated on CD8+ gated T lymphocytes.