Docetaxel remodels prostate cancer immune microenvironment and enhances checkpoint inhibitor-based immunotherapy

Background: Prostate cancer is usually considered as immune “cold” tumor with poor immunogenic response and low density of tumor-infiltrating immune cells, highlighting the need to explore clinically actionable strategies to sensitize prostate cancer to immunotherapy. In this study, we investigated whether docetaxel-based chemohormonal therapy induces immunologic changes and potentiates checkpoint blockade immunotherapy in prostate cancer. Methods: We performed transcriptome and histopathology analysis to characterize the changes of prostate cancer immune microenvironment before and after docetaxel-based chemohormonal therapy. Furthermore, we investigated the therapeutic benefits and underlying mechanisms of chemohormonal therapy combined with anti-PD1 blockade using cellular experiments and xenograft prostate cancer models. Finally, we performed a retrospective cohort analysis to evaluate the antitumor efficacy of anti-PD1 blockade alone or in combination with docetaxel-based chemotherapy. Results: Histopathology assessments on patient samples confirmed the enrichment of tumor-infiltrating T cells after chemohormonal therapy. Moreover, we found that docetaxel activated the cGAS/STING pathway in prostate cancer, subsequently induced IFN signaling, resulting in lymphocytes infiltration. In a xenograft mouse model, docetaxel-based chemohormonal therapy prompted the intratumoral infiltration of T cells and upregulated the abundance of PD1 and PD-L1, thereby sensitizing mouse tumors to the anti-PD1 blockade. To determine the clinical significance of these results, we retrospectively analyzed a cohort of 30 metastatic castration-resistant prostate cancer patients and found that docetaxel combined with anti-PD1 blockade resulted in better prostate-specific antigen progression-free survival when compared with anti-PD1 blockade alone. Conclusions: Our study demonstrates that docetaxel activates the antitumoral immune response and facilitates T cell infiltration in a cGAS/STING-dependent manner, providing a combination immunotherapy strategy that would improve the clinical benefits of immunotherapy.

embedded sections or tissue microarrays were unmasked in 1 × Tris-EDTA buffer (pH 9.0) for 20 minutes at 95 ℃ and then incubated with specific antibodies for overnight at 4 ℃. For paraffin-embedded sections, digitalized images were taken using Nikon-80i microscope under 40× objective. For tissue microarray, slides were scanned using Leica Aperio AT2 under 40× objective.
For quantification of CD20 and CD56, two independent researchers calculated the average number of membrane-positive cells in five to six random 40× fields. For quantification of CD3, CD4 and CD8, two independent researchers considered the whole filed and evaluated number of membrane-positive cells.

Immunofluorescence staining
For staining of intratumoral immune cell subsets, we used CD3 (

Transcriptome data analysis
Differential expression analysis was performed using the DESeq (2012) R package. False discovery rate (FDR) value < 0.05 and foldchange > 2 or foldchange < 0.5 were set as the threshold for significantly differential expression. We used the Immunology Database and Analysis Portal database to identify immune-related differentially expressed genes (DEGs). Functional enrichment analysis of upregulated DEGs was performed using the Metascape online tool (http://metascape.org).
Single-sample Gene Set Enrichment Analysis (ssGSEA) algorithm was performed using the gsva R package. The Gaussian distribution was chosen as kcdf argument, with a minimum and maximum geneset size of 5 and 500, respectively.
Antigen presentation score, CD8+ effector T cell score, and T cell inflamed score were defined as the ssGSEA scores of relevant gene sets. Batf3-dendritic cell (DC) score was defined as the mean expression levels of genes included in the relevant gene set. CYT score was defined as the geometric mean of expression levels of GZMA and PRF1 (as expressed in TPM). The gene sets associated with the above scores were described in Supplemental Table S3.
We used CIBERSORT to estimate the relative fractions of intratumoral immune cell subsets according to the gene expression profiles. The LM22 file, which is a leukocyte gene signature matrix consisting of 547 genes, was used to define 22 immune cell types. The sum of fractions of all 22 intratumoral immune cell subsets is equal to 1 in each sample.
We used MiXCR to extract TCR and BCR CDR3 repertoires from RNA-Seq data.

QRT-PCR from Cell Lines
RNA isolation was performed using Trizol® method according to the manufacturer's instructions (ThermoFisher; 15596026). RNA concentration and quality were evaluated using a NanoDrop apparatus (NaNodrop Technologies). cDNA was synthesized using HiScript® III RT SuperMix for qPCR kit (Vazyme; 7E402G0).
Images of protein bands were taken by Tanon 5200 system. For quantification of the western blot results, densitometry intensity of western blot bands was analyzed using ImageJ software.

Flow cytometry
LNCaP and PC3 cells were firstly treated with bicalutamide, docetaxel, or bicalutamide plus docetaxel for 24 hours. Then, cells were fixed and permeabilized with eBiosience™ Foxp3/Transcription Factor Staining Buffer Set (Thermo Fisher; 00-5523-00) according to the provided manufacturer's instructions. After fix and permeabilization, cells were incubated with 10% goat serum (Beyotime; C0265) for 15 minutes at room temperature. After blocking, the cells were incubated with the following primary antibodies for 30 minutes at room temperature: Phospho-STING