PPT1 inhibition enhances the antitumor activity of anti–PD-1 antibody in melanoma
Gaurav Sharma, … , Dmitry I. Gabrilovich, Ravi K. Amaravadi
Gaurav Sharma, … , Dmitry I. Gabrilovich, Ravi K. Amaravadi
Published August 11, 2020
Citation Information: JCI Insight. 2020;5(17):e133225. https://doi.org/10.1172/jci.insight.133225.
View: Text | PDF
Research Article Oncology Therapeutics

PPT1 inhibition enhances the antitumor activity of anti–PD-1 antibody in melanoma

Abstract

New strategies are needed to enhance the efficacy of anti–programmed cell death protein antibody (anti–PD-1 Ab) in cancer. Here, we report that inhibiting palmitoyl-protein thioesterase 1 (PPT1), a target of chloroquine derivatives like hydroxychloroquine (HCQ), enhances the antitumor efficacy of anti–PD-1 Ab in melanoma. The combination resulted in tumor growth impairment and improved survival in mouse models. Genetic suppression of core autophagy genes, but not Ppt1, in cancer cells reduced priming and cytotoxic capacity of primed T cells. Exposure of antigen-primed T cells to macrophage-conditioned medium derived from macrophages treated with PPT1 inhibitors enhanced melanoma-specific killing. Genetic or chemical Ppt1 inhibition resulted in M2 to M1 phenotype switching in macrophages. The combination was associated with a reduction in myeloid-derived suppressor cells in the tumor. Ppt1 inhibition by HCQ, or DC661, induced cyclic GMP-AMP synthase/stimulator of interferon genes/TANK binding kinase 1 pathway activation and the secretion of interferon-β in macrophages, the latter being a key component for augmented T cell–mediated cytotoxicity. Genetic Ppt1 inhibition produced similar findings. These data provide the rationale for this combination in melanoma clinical trials and further investigation in other cancers.

Authors

Gaurav Sharma, Rani Ojha, Estela Noguera-Ortega, Vito W. Rebecca, John Attanasio, Shujing Liu, Shengfu Piao, Jennifer J. Lee, Michael C. Nicastri, Sandra L. Harper, Amruta Ronghe, Vaibhav Jain, Jeffrey D. Winkler, David W. Speicher, Jerome Mastio, Phyllis A. Gimotty, Xiaowei Xu, E. John Wherry, Dmitry I. Gabrilovich, Ravi K. Amaravadi

×

Figure 3

PPT1 inhibition induces a change in macrophage polarization that favors antitumor immunity.

Options: View larger image (or click on image) Download as PowerPoint
PPT1 inhibition induces a change in macrophage polarization that favors ...
(A) Dot plot for the quantitative PCR (qPCR) expression in mouse macrophage RAW 264.7 cells polarized to an M2 phenotype following treatment with HCQ 10 μM or DC661 0.6 μM at the indicated time points (in hours) showing the results of 4 independent experiments. (B) Bright-field images of RAW 264.7 polarized to an M2 phenotype treated with HCQ or DC661 M1 phenotype control (IFN-γ + LPS). M2 cells have a round morphology whereas drug-treated M2 cells take on an elongated morphology with multiple pseudopodia typical of M1 macrophages (positive control). Images were taken at original magnification ×10. (C) Expression of M2 and M1 markers in mouse BMDMs treated with HCQ or DC661 as in A. Each result was reproduced by 3 independent experiments. (D) Dot plot showing qPCR expression of iNOS and RETNLA/FIZZ1 following Ppt1 KD in RAW 264.7 cells polarized to an M2 phenotype, and each result was reproduced by 5 independent experiments. Immunoblot showing the KD status of Ppt1 protein. (E) qPCR expression of M1 and M2 markers in Ulk1-, Pik3c3-, and Atg7-KD conditions. Results were reproduced with 3 independent experiments. Immunoblots showing the KD status of Ulk1, Pik3c3, Atg7 protein and expression of LC3 and p62. (F) Schema of experimental setup and cytotoxicity elicited by primed splenocytes with or without exposure to MCM collected from RAW 264.7 macrophages treated with control HCQ or DC661 for 24 hours. (G) Immunophenotyping for M1/M2 ratio of TAMs and percentage PMN-MDSCs in B16 melanoma tumors after 8 days of treatment. Mean and SEM are representative of 4 to 5 replicates, and the experiment was repeated at least 3 times. A P value is presented for the test of the hypothesis that the addition of HCQ to anti–PD-1 Ab is significantly different compared with anti–PD-1 Ab + Veh; * indicates an adjusted P < 0.05 testing the hypothesis that each experimental group is different from control. All t tests were 2 tailed (1 sample in A and C, 2 sample in F and G). One-way ANOVA and Dunnett’s procedure (D and E). Scale bar: 100 μm.