Colony-stimulating factor 1 (CSF1) is a key regulator of monocyte/macrophage differentiation that sustains the protumorigenic functions of tumor-associated macrophages (TAMs). We show that CSF1 is expressed in human melanoma, and patients with metastatic melanoma have increased CSF1 in blood compared to healthy subjects. In tumors, CSF1 expression correlated with the abundance of CD8 + T cells and CD163 + TAMs. Human melanoma cell lines consistently produced CSF1 after exposure to melanoma-specific CD8 + T cells or T cell-derived cytokines in vitro, reflecting a broadly conserved mechanism of CSF1 induction by activated CD8 + T cells. Mining of publicly available transcriptomic data sets suggested co-enrichment of CD8 + T cells with CSF1 or various TAM-specific markers in human melanoma, which was associated with nonresponsiveness to programmed cell death protein 1 (PD1) checkpoint blockade in a smaller patient cohort. Combination of anti-PD1 and anti-CSF1 receptor (CSF1R) antibodies induced the regression of BRAF V600E-driven, transplant mouse melanomas, a result that was dependent on the effective elimination of TAMs. Collectively, these data implicate CSF1 induction as a CD8 + T cell-dependent adaptive resistance mechanism and show that simultaneous CSF1R targeting may be beneficial in melanomas refractory to immune checkpoint blockade and, possibly, other T cell-based therapies.

Highlights d T cell-induced IFN-g correlates the best to immune checkpoint blockade (ICB) therapy d Immune signatures increase in on-therapy patients regardless of response to therapy d ICB therapy decreases MYC and WNT signaling in patients with clinical response d IFN-g exposure of melanoma cells leads to a conserved transcriptome signature

No abstract

Despite the success of immunotherapy using checkpoint blockade, many patients with solid tumors remain refractory to these treatments. In human cancer, the experimental options to investigate the specific effects of antibodies blocking inhibitory receptors are limited and it is still unclear which cell types are involved. We addressed the question whether the direct interaction between T cells and tumor cells can be enforced through blocking a set of inhibitory receptors including PD-1, TIM-3, BTLA and LAG-3, blocked either individually or in dual combinations with the anti-PD-1 antibody, and to determine the condition that induces maximal T cell function preventing tumor cell proliferation. Using short-term Melan-A-specific or autologous re-stimulations, checkpoint blockade did not consistently increase cytokine production by tumor-derived expanded T cells. We next set up a 5-day co-culture assay with autologous melanoma cell lines and expanded tumor infiltrating T cells, originating from tumor specimens obtained from 6 different patients. Amongst all combos tested, we observed that blockade of LAG-3 alone, and more strongly when combined with PD-1 blockade, enforced T cell responses and tumor cell growth control. The combination of anti-LAG-3 plus anti-PD-1 acted through CD8 T cells and led to increased IFNγ production and cytotoxic capacity. Our results show that LAG-3 and PD-1 are regulating the direct interaction between tumor cells and autologous T cells, suggesting that therapy effects may be promoted by enhanced access of the corresponding blocking reagents to the tumor microenvironment. ARTICLE HISTORY

INTRODUCTION: Melanoma-acquired resistance against T cell-based immunotherapy may be mediated by alterations of their responses to inflammatory cytokines. However, the cytokine signaling pathways in cancer cells remain poorly understood. METHODS: 8 melanoma cell lines were exposed to interferon gamma (IFNγ) and tumor necrosis factor alpha (TNFα) over a time course. Signaling molecules and their phosphorylated state were subsequently analyzed using mass cytometry. Briefly, cells were incubated with IFNγ (220 IU/mL), TNFα (1000 IU/mL), and both cytokines (220 IU/mL IFNγ + 200 IU/mL TNFα) for up to 48 hours. Staining was done with antibodies specific for (1) phosphorylated - STAT1(Y701), STAT3(Y707), AKT(S437), ERK1/2(T202/Y204), IKK(S176/180), NK-kbetap65 (S536), JNK (T183/T22), P38 (T180/Y182), Rb(S807/Y182), cMYC (S62), S6 (S235/S236), Histone3 (S28) and (2) non-phosphorylated - cPARP, β-catenin, CyclinB1, Ki67. Cells were read on a Helios Mass cytometer and de-barcoded using PREMESSA (R package). FlowJo was used to normalize time and eliminate calibration beads. Viable cell files were introduced to Cytobank where fold change (FC) with respect to time 0 of each protein was calculated with median intensity. The FC was exported to R for final analysis. RESULTS: All cell lines were sensitive to IFNγ/TNFα, with detectable (phosphorylated) signaling molecules. No synergistic effect was observed at any dose for IFNγ+TNFα. For the following data, numbers contained within parentheses represent the mean range of FC control vs treatment or mean standard error. pSTAT1 was significantly activated at all time points for IFNγ ([0.8, 1.0] vs [1.8, 3.6] p<0.02) and for IFNγ+TNFα ([1.9, 3.5] p<0.03). For TNFα it was activated at 48h ([3.1±0.9 vs 1.9±0.2] p=0.003). pJNK was significantly activated for IFNγ at time 0.25h and 0.5h ([0.9, 1.0] vs [1, 5] p<0.01); for TNFα at 0.25h and 1h ([0.9, 1.0] vs [1.5, 1.6] p<0.02); and for IFNγ+TNFα at 0.25h and 1h ([0.9, 1.0] vs 1.5 p<0.007). pNF-KBp65 was only significantly different for IFNγ+TNFα at 24h ([0.9±0.0 vs 11.6±0.2] p=0.025). The other time points followed a similar trend to activation. pSTAT3 was significantly activated for IFNγ at time 0.25h and 24h ([0.6, 0.9] vs [1.7, 2.1] p<0.04); for IFNγ+TNFα at time 0.25h and 24h ([0.6, 0.9] vs [1.9, 2.0] p<0.02); and for TNFα at 0.25h and 0.5h ([0.5, 0.6] vs [1.3, 1.5] p<0.02). pERK was significantly activated for IFNγ at 0.25h ([0.9±0.2 vs 1.6±0.3] p=0.04). β-catenin presented no difference in FC across cell lines. CONCLUSIONS: Our data show evidence for functional IFNγ/TNFα signaling in most melanoma cells. However, unexpected signaling was observed for some cell lines, possibly reflecting interpatient heterogeneity. Additional studies are necessary to understand the roles of cytokine signaling in cancer immunotherapy. Citation Format: Michael Cerniglia, Katie M. Campbell, Julianne Schwerdtfeger, Christophe Martignier, Connie B. Gilfillan, Karin Schaeuble, Gardenia Cheung-Lau, Yaroslav Teper, David Berger Maneiro, Miriam Guemes-Aragon, Alejandro J. Garcia, Daniel E. Speiser, Antoni Ribas, Begoña Comin-Anduix. Signaling pathways involved in melanoma cell functions and dedifferentiation induced by inflammatory cytokines [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1697.

No abstract