PD-L1 is highly expressed in Enzalutamide resistant prostate cancer

Abstract

Efficacy of Enzalutamide (ENZ) in castration resistant prostate cancer (CRPC) patients is short-lived. Immunotherapy like T cell checkpoint blockade may improve patient survival. However, when and where checkpoint molecules are expressed in CRPC and whether immune evasion is a mechanism of ENZ resistance remains unclear. Thus, we investigated whether clinically relevant immunotherapy targets, specifically PD-L1/2 , PD-1 and CTLA-4, are upregulated in ENZ resistant (ENZR) patients and in a pre-clinical model of ENZ resistance. We show for the first time that patients progressing on ENZ had significantly increased PD-L1/2+ dendritic cells (DC) in blood compared to those naïve or responding to treatment, and a high frequency of PD-1+T cells. These data supported our pre-clinical results, in which we found significantly increased circulating PD-L1/2+ DCs in mice bearing ENZR tumors compared to CRPC, and ENZR tumors expressed significantly increased levels of tumor-intrinsic PD-L1. Importantly, the expression of PD-L1 on ENZR cells, or the ability to modulate PD-L1/2+ DC frequency, was unique to ENZR cell lines and xenografts that did not show classical activation of the androgen receptor. Overall, our results suggest that ENZ resistance is associated with the strong expression of anti-PD-1 therapy targets in circulating immune cells both in patients and in a pre-clinical model that is non-AR driven. Further evaluation of the contribution of tumor vs. immune cell PD-L1 expression in progression of CRPC to anti-androgen resistance and the utility of monitoring circulating cell PD-L1 pathway activity in CRPC patients to predict responsiveness to checkpoint immunotherapy, is warranted.

Figure 1. Progression on ENZ in CRPC patients is associated with increased frequency of PD-L1/2 DCs in circulation

(A) Evaluation of DCs in blood from CRPC patients: Whole blood was collected from CRPC patients defined as naïve (n=3) responding (resp, n=4) or progressing (prog, n=8) on ENZ at the time of collection and frequency of PD-L1/2⁺ DCs isolated from patient blood was assessed by flow cytometry. Frequency of PD-L1/2⁺ DC (Lin⁻CD11c⁺MHCII^hi) is shown. Contour plots show DC PD-L1 and PD-L2 expression in representative blood samples, graphs show mean frequency of positive cells +/− SD, ** P=<0.005. (B) Resistance to ENZ in progressing patients is associated with increased PD-L1/2+ DCs: Frequency of PD-L1/2⁺ DCs isolated from blood of progressing patients stratified by maximum PSA decline (% PSA reduction from start of ENZ treatment) is shown. <50% decline, n=5, >50% decline n=3. (C) Time on ENZ increases PD-L1/2+ DC frequency: Frequency of PD-L1/2⁺ DCs isolated from blood of progressing patients stratified by the duration of ENZ treatment is shown. 3.5 mo, n=5; 5.5 mo, n=2; 9 mo, n=1, *P=<0.05. All cell populations are downgated on live, CD45⁺ cells.

Figure 2. Differential expression of T cell checkpoint molecules in ENZ resistance

(A) Expression profile of checkpoint molecules in ENZ resistance: RNA sequencing (left) and microarray (right) data shows average fold change expression in checkpoint molecule genes in ENZ resistant (ENZR) cell lines 42D and 49F compared to CRPC (=1), n=2. (B) Reduced AR activity in ENZR cell lines correlates with PD-L1 expression: CRPC and ENZR cell lines were grown in vitro and assessed for AR and PSA expression by western blot, vinculin was used as a loading control. Representative blots from more than three independent experiments are shown. (C) Expression of PD-L1 in ENZ resistant cell lines: Surface expression of PD-L1 on CRPC, ENZR 42D, 42F, 49C and 49F cell lines grown in vitro was assessed by flow cytometry and shown as representative histograms from one of three independent experiments, or fold changes in mean fluorescence intensity (MFI) on ENZR 42D and 42F cell lines compared to CRPC (=1). Bar graph shows mean fold MFI changes pooled from three independent experiments, error bars represent SEM, *P=<0.05, ** P=<0.01.

Figure 3. non-AR driven ENZR 42D and 42F xenografts increase circulating PD-L1/2 DCs in vivo

Evaluation of DCs in blood from mice bearing ENZ resistant tumors: Blood was harvested from mice bearing ENZ resistant (ENZR) or CRPC subcutaneous xenografts when tumors reached 350-650mm³ and frequency of PD-L1, PD-L2 and PD-L1/2 double positive DCs isolated from blood was assessed by flow cytometry. Frequency of (A) PD-L1⁺ DC (CD11c⁺MHCII^hi), (B) PD-L2⁺ DC and (C) PD-L1/2⁺ DC is shown. All cell populations are downgated on live, CD45+ cells. ** P=<0.005, * P=<0.05, ***P=<0.001, error bars on graphs represent SD of representative data from of two independent experiments, n (mouse number)=5-8.

Figure 4. non-AR driven ENZR 42D and 42F xenografts decrease tumor infiltrating PD-L1/2⁺ DCs in vivo

Evaluation of tumor infiltrating leukocytes: Tumors were harvested from mice bearing ENZ resistant (ENZR) or CRPC subcutaneous xenografts when tumors reached 350-650mm³ and frequency of infiltrating PD-L1, PD-L2 and PD-L1/2 double positive DCs isolated from tumors was assessed by flow cytometry. Frequency of (A) PD-L1⁺ DC (CD11c⁺MHCII^hi), (B) PD-L2⁺ DC and (C) PD-L1/2⁺ DC is shown. All cell populations are downgated on live, CD45⁺ cells. * P=<0.05, ** P=<0.01, ***P=<0.001 error bars on graphs represent SEM of pooled data from two independent experiments, n (tumor number)=11-20.