β-Catenin Activation Promotes Immune Escape and Resistance to Anti-PD-1 Therapy in Hepatocellular Carcinoma

Abstract

PD-1 immune checkpoint inhibitors have produced encouraging results in patients with hepatocellular carcinoma (HCC). However, what determines resistance to anti-PD-1 therapies is unclear. We created a novel genetically engineered mouse model of HCC that enables interrogation of how different genetic alterations affect immune surveillance and response to immunotherapies. Expression of exogenous antigens in MYC;Trp53 -/- HCCs led to T cell-mediated immune surveillance, which was accompanied by decreased tumor formation and increased survival. Some antigen-expressing MYC;Trp53 -/- HCCs escaped the immune system by upregulating the β-catenin (CTNNB1) pathway. Accordingly, expression of exogenous antigens in MYC;CTNNB1 HCCs had no effect, demonstrating that β-catenin promoted immune escape, which involved defective recruitment of dendritic cells and consequently impaired T-cell activity. Expression of chemokine CCL5 in antigen-expressing MYC;CTNNB1 HCCs restored immune surveillance. Finally, β-catenin-driven tumors were resistant to anti-PD-1. In summary, β-catenin activation promotes immune escape and resistance to anti-PD-1 and could represent a novel biomarker for HCC patient exclusion. SIGNIFICANCE: Determinants of response to anti-PD-1 immunotherapies in HCC are poorly understood. Using a novel mouse model of HCC, we show that β-catenin activation promotes immune evasion and resistance to anti-PD-1 therapy and could potentially represent a novel biomarker for HCC patient exclusion.See related commentary by Berraondo et al., p. 1003.This article is highlighted in the In This Issue feature, p. 983.

Conflict of interest statement

Disclosure of Potential Conflicts of Interest:

A. Lujambio has received grant support from Pfizer and Genentech, and lecture fees from Exelixis for unrelated projects. P. Molina-Sánchez has received grant support from Pfizer for unrelated projects. S.P. Monga had grant funding from Abbvie Pharmaceuticals for unrelated projects. A. Villanueva has received consulting fees from Guidepoint and Fujifilm, advisory board fees from Exact Sciences, and lecture fees from Exelixis for unrelated projects. J.M. Llovet has received research support from Bayer HealthCare Pharmaceuticals, Eisai Inc, Bristol-Myers Squibb, and Ipsen, and consulting fees from Bayer HealthCare Pharmaceuticals, Bristol-Myers Squibb, Eisai Inc, Celsion Corporation, Eli Lilly, Exelixis, Merck, Ipsen, Glycotest, Navigant, Leerink Swann LLC, Midatech Ltd, Fortress Biotech, Sprink Pharmaceuticals, and Nucleix. No potential conflicts of interest were disclosed by the rest of the other authors.

©2019 American Association for Cancer Research.

Figures

Fig. 1.. Expression of exogenous antigens in murine MYC;p53−/− HCCs leads to tumor delay.

A) Oncoprint of TP53 and MYC alterations in 366 HCC patients (TCGA, provisional, December 2018, cbio portal (62)). The percentage of patients harboring the alteration is shown. Amp, amplification; del, deletion; trun, truncating; mut, mutation; infr, inframe; miss, missense. B) Schematic of vectors injected into mice. The transposon-based vector overexpressing MYC can also express luciferase (luc) or a luciferase fused to model antigens (lucOS). C) Bioluminescence imaging six and 25 days after injection of vectors into representative mice. The color code for the luciferase signal is shown. D) Quantification of normalized luciferase signal six and 25 days after injection of vectors (n = 5 per group). Mean and standard deviation (SD) are shown. Mann-Whitney test. E-F) Survival curves in C57BL/6 WT (wild-type) females (E) and males (F). Number of mice per group is shown as well as median survival. Undef, undefined. Log-rank Mantel-Cox test. G) Pictures of representative livers from E and F. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm.

Fig. 2.. CD8+ T cells eliminate antigen-expressing MYC;p53−/− HCCs.

A-B) Survival curves in C57BL/6 Rag2−/− females (A) and males (B). Number of mice per group is shown as well as median survival. Log-rank Mantel-Cox test. C) Pictures of representative livers from A and B. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm. D-E) Survival curves in C57BL/6 WT (wild-type) with combined CD4+ and CD8+ T cell depletion (D) or separate CD4+ and CD8+ T cell depletion (E). Number of mice per group is shown as well as median survival. Undef, undefined. Log-rank Mantel-Cox test. Comparisons are to control mice injected with isotype control antibodies (IgG). F) Pictures of representative livers from D and E. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm.

Fig. 3.. β-catenin signaling is activated in immune-escaped HCC tumors.

A) GSEA of an HCC CTNNB1 gene signature in escaped MYC-lucOS;sg-p53 tumors (n = 3 per group). NES, normalized enrichment score. B) Volcano plot representing genes according to their fold change and p value in MYC-luc;sg-p53 escaped tumors when compared to MYC-lucOS;sg-p53 tumors. Axin2 is highlighted in red. C) Relative levels of Axin2 in MYC-luc;sg-p53 and escaped MYC-lucOS;sg-p53 tumors by qRT-PCR. Each dot represents one tumor coming from one independent mouse. Number of samples is shown. The samples used in the RNA-sequencing (A and B) are highlighted in blue. Mean and standard deviation (SD) are shown. Mann-Whitney test. D) Number of cases with Axin2 levels higher (blue) or lower (black) than 2.5 (which is the mean value in MYC-lucOS;sg-p53 tumors in C). Number of samples is shown. Fisher exact test. E-F) GSEA of an HCC CTNNB1 gene signature (E) and an adaptive immune response signature (F) in Axin2-escaped MYC-lucOS;sg-p53 tumors (n = 3) or low-Axin2-escaped MYC-lucOS;sg-p53 tumors (n = 4). NES, normalized enrichment score.

Fig. 4.. β-catenin signaling activation promotes immune escape in HCC.

A) Oncoprint of CTNNB1 and MYC alterations in 366 HCC patients (TCGA, provisional, December 2018, cbio portal (62)). The percentage of patients harboring the alteration is shown. Amp, amplification; del, deletion; trun, truncating; mut, mutation; infr, inframe; miss, missense. B) Schematic of vectors injected into mice. The transposon-based vector overexpressing MYC can also express luciferase (luc) or a luciferase fused to model antigens (lucOS). C) Bioluminescence imaging six and 27 days after injection of vectors into representative mice. The color code for the luciferase signal is shown. D) Quantification of normalized luciferase signal six and 27 days after injection of vectors (n = 5 per group). Mean and standard deviation (SD) are shown. Mann-Whitney test. E-F) Survival curves in C57BL/6 WT (wild-type) females (E) and males (F). Number of mice per group is shown as well as median survival. Log-rank Mantel-Cox test. G) Pictures of representative livers from E and F. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm.

Fig. 5.. β-catenin impairs dendritic cell recruitment in the context of HCC.

A-B) Quantification of the percentage of SIINFEKL-specific CD8+ T cells (A) or number of DC1 dendritic cells (B) in the livers of the corresponding mice (n = 5–6 per group). Representative of three independent experiments. N, normal liver; sg-p53, Myc-lucOS;sg-p53; CTNNB1, Myc-lucOS;CTNNB1. Mean and standard deviation (SD) are shown. ANOVA test. C) Survival curves in C57BL/6 WT (wild-type) or Batf3−/− females harboring MYC-lucOS;sg-p53 tumors. Batf3−/− mice are represented with a dotted line. Number of mice per group is shown as well as median survival. Undef, undefined. Log-rank Mantel-Cox test. D) Pictures of representative livers from C. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm. E-F) Quantification of the number of DC1 dendritic cells (E) or the percentage of SIINFEKL-specific CD8+ T cells (F) in the livers of the corresponding mice (n = 5 per group). sg-p53, Myc-lucOS;sg-p53; Control, non-injected; KO, Batf3−/−; WT, wild-type. Mean and standard deviation (SD) are shown. ANOVA test. G-H) Heatmap showing the average expression values of different genes in murine tumors (G) and human TCGA tumors (H). Colors are adjusted for each row and for each group comparison from high (red) to low (blue). luc p53, MYC-luc;sg-p53 (n = 3); lucOS p53, MYC-lucOS;sg-p53 (n = 3); low Axin2, low-Axin2-escaped MYC-lucOS;sg-p53 (n = 4); luc CTNNB1, MYC-luc;CTNNB1 (n = 7), lucOS CTNNB1, MYC-lucOS;CTNNB1 (n = 5); CTNNB1 WT, wild type (n = 263); CTNNB1 mut, mutant (n = 97); CTNNB1 low (n = 120); CTNNB1 inter, intermediate (n = 120); CTNNB1 high (n = 120). Samples were stratified depending on CTNNB1 status as WT or mutant, or CTNNB1-mutant HCC gene signature enrichment levels (in tertiles). I) Number of CD8+ T cells in tumor and peritumor areas in HCC patient samples (n = 59), which were classified as having membrane (mb, n = 37) or nuclear (nc, n = 22) staining for β-catenin protein. Mean and standard deviation (SD) are shown. Mann-Whitney test. J) Representative pictures of the stainings for CD8 and β-catenin summarized in I.

Fig. 6.. Ccl5 expression restores immune surveillance in β-catenin-driven HCCs.

A) Venn diagram displaying the chemokines differentially expressed in mice or human liver tumors. The number of samples for each dataset is included. Red, significantly upregulated; blue, significantly downregulated. The intersection shows the chemokines disregulated in both datasets. B) Expression of CCL5 in 360 human HCC samples (LIHC, liver hepatocellular carcinoma, from the TCGA). Box and whisker plot, with the central line representing the median, the ends of the box representing the upper and lower quartiles, and the whiskers extending to the highest and lowest observations. Mann-Whitney test. CTNNB1 signature low (n = 120), intermediate (inter, n = 120), and high (n = 120). C) Schematic of vectors injected into mice. The transposon-based vector overexpressing CTNNB1 also expresses Ccl5. D) Survival curves in C57BL/6 WT (wild-type) females. Number of mice per group is shown as well as median survival. Log-rank Mantel-Cox test. E) Pictures of representative livers from D. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm. F-G) Quantification of the number of DC1 dendritic cells (F) or percentage of SIINFEKL-specific CD8+ T cells (G) in the livers of the corresponding mice (n = 5–7 per group). N, normal liver; CTNNB1-Ccl5, Myc-lucOS;CTNNB1-Ccl5. Mean and standard deviation (SD) are shown. Mann-Whitney test. H) Survival curves in C57BL/6 WT (wild-type) females with combined CD4+ and CD8+ T cell depletion. Number of mice per group is shown as well as median survival. Undef, undefined. Log-rank Mantel-Cox test. I) Pictures of representative livers from Figure F. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm. J-K) Survival curves in C57BL/6 WT (wild-type, WT) or Batf3−/−males (J) and females (K). Number of mice per group is shown as well as median survival. Undef, undefined. Log-rank Mantel-Cox test. L) Pictures of representative livers from Figure K. The number indicates the number of days from injection to death for that particular mouse. Scale bar represents 1 cm.

Fig. 7.. β-catenin signaling activation confers resistance to anti-PD-1 therapy in HCC.

A-D) Survival curves of C57BL/6 WT (wild-type) females harboring MYC-lucOS;sg-p53 (A), MYC-luc;sg-p53 (B), MYC-lucOS;CTNNB1 (C), or MYC-luc;CTNNB1 (D) tumors treated with control antibodies (IgG) or anti-PD-1 (aPD-1). Number of mice per group is shown as well as median survival. Undef, undefined. Log-rank Mantel-Cox test. E) Schematic of HCC patients treated with nivolumab at Mount Sinai Hospital, including their response and CTNNB1 mutational status. F) Scans of a responder (CTNNB1 wild type, WT; upper panel, 6 months) and non-responder patient (CTNNB1 mutated, mut; lower panel, 3 months) to nivolumab, before and after treatment. Responder shows complete resolution. Non-responder shows tumor growth and new lesions.