Reversion of immune tolerance in advanced malignancy: modulation of myeloid-derived suppressor cell development by blockade of stem-cell factor function

Abstract

Tumor growth induced a significant increase of myeloid-derived suppressor cells (MDSCs) in the tumor-bearing host. In our previous study, we showed that MDSCs induced tumor-specific T-cell tolerance and the development of T regulatory cells (Tregs). Tumor-derived factors have been implicated in the accumulation of MDSCs. We hypothesize that reduction of MDSC accumulation in tumor-bearing hosts, through the blockade of tumor factors, can prevent T-cell anergy and Treg development and thereby improve immune therapy for the treatment of advanced tumors. Several tumor-derived factors were identified by gene array analysis. Among the candidate factors, stem- cell factor (SCF) is expressed by various human and murine carcinomas and was selected for further study. Mice bearing tumor cells with SCF siRNA knockdown exhibited significantly reduced MDSC expansion and restored proliferative responses of tumor-infiltrating T cells. More importantly, blockade of SCF receptor (ckit)-SCF interaction by anti-ckit prevented tumor-specific T-cell anergy, Treg development, and tumor angiogenesis. Furthermore, the prevention of MDSC accumulation in conjunction with immune activation therapy showed synergistic therapeutic effect when treating mice bearing large tumors. This information supports the notion that modulation of MDSC development may be required to achieve effective immune-enhancing therapy for the treatment of advanced tumors.

Figure 1

Expression of SCF by various murine and human tumor cell lines from multiple tissue origins and establishment of SCF-silenced tumor clones. (A) SCF expression in various murine tumor cell lines (colon cancer: MCA26, MC38; breast cancer: JC, 4T1; melanoma: B16). (B) SCF expression in various human tumor cell lines (colon cancer: HCT15, SW620, DLD-1, Colo205; breast cancer: MDA-MB435). Total RNAs were prepared from the murine and human tumor cell lines and SCF mRNA expressions were analyzed by RT-PCR. (C) SCF expression in SCF-silenced MCA26 clone and in control mock-transfected clone. Multiple stable SCF knockdown and mock-transfected control MCA26 clones were established as described in “Methods.” RNAs were prepared from representative clones and SCF expressions were analyzed by RT-PCR. (D) The protein expression of SCF in mock-transfected control and SCF-silenced MCA26 clones. Cell lysates were prepared and SCF concentration was measured by ELISA.

Figure 2

Tumor-derived SCF is required for MDSC accumulation in the tumor-bearing mice. (A) Decreased percentage of Gr-1⁺CD115⁺ MDSCs in bone marrow of mice bearing SCF knockdown tumors. BM Percoll fraction II (Fr. II) cells from mice bearing SCF-silenced (4 mice) versus control mock-transfected (5 mice) MCA26 tumors were stained with anti–Gr-1-APC plus anti–CD115-PE or isotype control antibodies followed by flow cytometric analysis. The percentage of MDSCs in mice bearing SCF-silenced tumors is significantly lower than that in mice bearing control tumors (P < .001 by Student t test). (B) Decreased number of Gr-1⁺CD115⁺ cells in mice bearing SCF knockdown tumors. The numbers of Gr-1⁺CD115⁺CD11b⁺ and Gr-1⁺CD115^-CD11b⁺ cells in bone marrow Fr. II derived from mice with large ( > 10 × 10 mm²), medium (7 × 7 mm²), or small ( < 7 × 7 mm²) control versus SCF-silenced tumors were compared. The average numbers of MDSCs in BM are presented. A significantly lower number of MDSCs were observed in mice bearing SCF knockdown tumors with sizes larger than 7 × 7 mm² compared with mice bearing control wild-type (WT) tumors (*P < .05, Student t test). (C) Reduced suppressive activity of Percoll Fr. II cells isolated from mice bearing SCF knockdown tumors. The suppressive activities of Percoll Fr. II, which contains MDSCs, isolated from mice bearing large SCF knockdown or control wild-type tumors were assessed using HA peptide-mediated proliferation at various ratios of CD4 HA TCR transgenic splenocytes/irradiated Fr. II cells. The Fr. III cells were used as negative control. The Fr. II cells from mice with SCF knockdown tumors exhibited lower suppressive activity compared with those from mice bearing wild-type tumors. Con indicates control. (D) Proliferative response of T cells derived from tumor tissues. The anti-CD3/anti-CD28–mediated proliferative responses of the T cells derived from SCF-silenced or control tumor tissues were assessed in a standard [³H]-thymidine incorporation assay. Splenocytes from naive normal mice were used as a positive control. w/o indicates without.

Figure 3

Proliferative responses of T cells isolated from tumor tissues after the treatment of anti-ckit antibodies. (A) Proliferative response of T cells derived from tumor tissues of anti-ckit– or control Ig–treated mice. The anti-CD3/anti-CD28–mediated proliferative responses of the T cells derived from tumor tissues were assessed in a standard [³H]-thymidine incorporation assay. Splenic T cells purified from naive mice were used as positive control. Stimulation index (SI; cpm in the presence of anti-CD3/anti-CD28 divided by cpm in the absence of anti-CD3/anti-CD28) of the result is presented. (B) Decreased percentage of Gr-1⁺CD115⁺ MDSCs in bone marrow of tumor-bearing mice treated with anti-ckit. BM Percoll Fraction 2 cells from tumor-bearing mice treated with anti-ckit or control rat Ig were stained with anti–Gr-1-APC plus anti–CD115-PE or isotype control antibodies followed by flow cytometric analysis. A significantly lower percentage of MDSCs was observed in tumor-bearing mice treated with anti-ckit compared with those treated with control rat Ig (P < .001 by Student t test). Numbers in quadrants represent percent positive. (C) Reduced suppressive activity of MDSCs isolated from tumor-bearing mice treated with anti-ckit. The suppressive activities of MDSCs were assessed using HA peptide-mediated proliferation at various ratios of CD4 HA TCR transgenic splenocytes and MDSCs. The MDSCs isolated from mice treated with anti-ckit exhibited lower suppressive activity compared with those from mice receiving control rat Ig. Error bars represent standard deviation.

Figure 4

Treatment of anti-ckit prevents the development of T-cell anergy in tumor-bearing mice. Thy1.2⁺ CD4 HA-specific TCR-transgenic T cells (5 × 10⁶/mouse) were injected via tail vein into congenic Thy1.1⁺ MCA26 or HA-MCA26 tumor-bearing mice 3 days after the first dose of anti-ckit or rat Ig injection (50 μg/mouse). At day 15 after transfer, Thy1.2⁺ splenocytes were recovered by sorting. (A) Proliferative responses of sorted Thy1.2⁺ CD4 HA-specific T cells to HA peptides. The culture was pulsed with [³H]-thymidine for the last 8 hours of a 72-hour culture. Stimulation index (SI) is calculated as the proliferation count (cpm) in the presence of HA peptide divided by that in the absence of HA peptide. Data shown are from a representative of 2 reproducible experiments (3 to 4 mice per group). *P <.001 (B) The residual tumor weight. The residual tumors were resected from the liver tissue and the tumor weight (gram) was measured. Error bars represent standard deviation. (C) The expression of Foxp3 in tumor (HA)–specific T cells. RNA was prepared from Thy1.2⁺ CD4 HA TCR transgenic T cells recovered from treated mice and Foxp3 expression was analyzed by one-step RT-PCR and real-time PCR. GAPDH expression was used as housekeeping gene control. (D) Intracellular staining of Foxp3 in tumor (HA)–specific T cells. Splenocytes were prepared from treated mice and stained with fluorochrome-conjugated anti-Thy1.2 (FITC) plus anti-CD4 (APC) plus anti-CD25 (PE-Cy7) plus anti-Foxp3 (PE). Thy1.2⁺CD4⁺ gated dot plots are presented. Naive mice with adoptively transferred T cells and rat Ig (left); tumor-bearing mice with adoptively transferred CD4 T cells and rat Ig control (middle); tumor-bearing mice with adoptively transferred T cells and anti-ckit (right). (E) Cytokine profile of tumor-specific T cells. Culture supernatants of recovered Thy1.2⁺ CD4 HA TCR transgenic T cells in the presence of HA peptide (5 μg/mL) and irradiated antigen-presenting cells (naive splenocytes) were collected. The naive CD4 HA TCR splenocytes cultured in the presence or absence of HA peptide were used as positive and negative controls, respectively. The cytokine concentrations were measured by ELISA kits (R&D Systems). The cytokine profile of supernatant collected from culture of tumor (HA)–specific T cells skewed toward a Th1 response with a higher concentration of IFN-γ and a lower concentration of IL-10 (*P < .05; Student t test).

Figure 5

Blockade of SCF/ckit pathway by the use of anti-ckit antibody or SCF knockdown of tumor cells results in decreased infiltration of MDSCs in tumor tissue and decreased tumor angiogenesis. (A) Immunostaining of Gr-1⁺ cells in wild-type versus SCF-silenced MCA26 tumor tissues. (B) Immunostaining of Gr-1⁺ cells in tumor tissues from HA-MCA26 tumor-bearing mice that were injected with rat Ig (left), 5 × 10⁶ HA-TCR T cells (middle), or 5 × 10⁶ HA-TCR T cells and anti-ckit (right). (C) Immunostaining of CD31⁺ cells in wild-type versus SCF-silenced MCA26 tumor tissues. (D) Immunostaining of CD31⁺ cells in tumor tissues from HA-MCA26 tumor-bearing mice that were injected with rat Ig (left), 5 × 10⁶ HA-TCR T cells (middle), or 5 × 10⁶ HA-TCR T cells and anti-ckit (right). Slides were viewed with a Leica DM RA2 fluorescent microscope (Leica Microsystems, Wetzlar, Germany) using an HC PLAN APO lens at 63×/1.32 (A,B) and 40×/0.85 (C,D) and Klear Mount medium (GBI, Mukilteo, WA). Images were acquired using a Hamamatsu ORCA-ER digital camera (Minneapolis, MN) Model C4742-80-12AG, and were processed with Openlab version 5.02 (Improvision, Waltham, MA) and Adobe Photoshop version 7.0 software (Adobe Systems, San Jose, CA).

Figure 6

Treatment of anti-ckit antibody significantly improves the long-term survival rate of mice treated with immune modulatory therapy of IL-12 plus 4-1BB activation. Mice bearing large MCA26 tumors (10 × 10 mm²) in the liver were divided into the following treatment groups: DL312 (control viral vector) plus control Ig; DL312 plus anti-ckit; Adv.mIL-12 plus anti–4-1BB plus rat Ig; Adv.mIL-12 plus anti–4-1BB plus anti-ckit; Adv.mIL-12 plus anti-ckit; and DL312 plus anti–4-1BB plus anti-ckit. The survival advantage for the mice treated with Adv.mIL-12 plus anti–4-1BB plus anti-ckit was statistically significant compared with those treated with Adv.mIL-12 plus anti–4-1BB plus rat Ig (P < .01, log-rank test).