Macrophage-Derived Cholesterol Contributes to Therapeutic Resistance in Prostate Cancer

Abstract

Castration-resistant prostate cancer (CRPC) is a lethal stage of disease in which androgen receptor (AR) signaling is persistent despite androgen deprivation therapy (ADT). Most studies have focused on investigating cell-autonomous alterations in CRPC, while the contributions of the tumor microenvironment are less well understood. Here we sought to determine the role of tumor-associated macrophages in CRPC, based upon their role in cancer progression and therapeutic resistance. In a syngeneic model that reflected the mutational landscape of CRPC, macrophage depletion resulted in a reduced transcriptional signature for steroid and bile acid synthesis, indicating potential perturbation of cholesterol metabolism. As cholesterol is the precursor of the five major types of steroid hormones, we hypothesized that macrophages were regulating androgen biosynthesis within the prostate tumor microenvironment. Macrophage depletion reduced androgen levels within prostate tumors and restricted AR nuclear localization in vitro and in vivo. Macrophages were also cholesterol-rich and were able to transfer cholesterol to tumor cells in vitro. AR nuclear translocation was inhibited by activation of liver X receptor (LXR)-β, the master regulator of cholesterol homeostasis. Consistent with these data, macrophage depletion extended survival during ADT and the presence of macrophages correlated with therapeutic resistance in patient-derived explants. Taken together, these findings support the therapeutic targeting of macrophages in CRPC. SIGNIFICANCE: These results suggest that macrophage-targeted therapies can be combined with androgen deprivation therapy to treat patients with prostate cancer by limiting cholesterol bioavailability and the production of intratumoral androgens.See related commentary by Al-Janabi and Lewis, p. 5399.

Figure 1.. Development of syngeneic mouse models representing CRPC.

A) Mouse model scheme. B) Tamoxifen induced GFP and Ki67 expression in prostate glands of a Pten^pce−/−/Trp53^pce−/−/TMPRSS2-ERG^pce+ mouse after 2 month of tamoxifen administration, n=1. C) Detection of prostate tumors and overall survival in Pten^pce−/−/Trp53^pce−/− and Pten^pce−/−/Trp53^pce−/−/TMPRSS2-ERG^pce+ mice. Days reflect time post tamoxifen administration. Tumor incidence and size were monitored by monthly MRI starting at 5-6 months. n=17-19 mice, pooled from two independent cohorts. Significance between the two groups was determined by log-rank. D) Hematoxylin and eosin (H&E) and Masson’s trichrome staining in Pten^pce−/−/Trp53^pce−/−/TMPRSS2-ERG^pce+ mice from C. Control prostate glands were harvested from mice at 6 and 12 months of age. E) Representative T2-weighted image by MRI of tumor lesion-bearing Pten^pce−/−/Trp53^pce−/− mice treated with vehicle or 0.6 mg Lupron subcutaneously every 28 days for two cycles, starting 4 months after tamoxifen administration. n=3-5 mice per group, data from a one cohort of mice. F) Expression of Krt5, Krt8, Krt14, Krt18 in the different cell lines generated from either Pten^pce−/−/Trp53^pce−/− (PT) or Pten^pce−/−/Trp53^pce−/−/TMPRSS2-ERG^pce+ (PTE) mice at end-stage. n=6 from one cohort, with data shown as the mean ± SEM. TRAMP-C2 cell line was used as a comparison. Representative confocal microscopy immunofluorescent images of the PTE-82 cell line stained for cytokeratin 14 and cytokeratin 18 are shown to the right. Similar results were obtained for the PT-09, PT-25, and PTE-24 lines. Images represent one of two independent experiments. G) Expression of Ar in the different cell lines. n=6 from one cohort, data shown as the mean ± SEM. H) Confocal microscopy immunofluorescent images of the PTE-82 cell line either treated with 10 nM DHT or left untreated in charcoal stripped serum (CSS). Androgen receptor (red) and DAPI (blue). Images are representative of one of three independent experiments. I) Dose response curve of enzalutamide in 5 different prostate cancer cell lines. Phase contrast images were acquired at 8 hr intervals using Incucyte, with confluence per well calculated per cell line. n=3, data shown as the mean ± SEM from one of at least three independent experiments.

Figure 2.. Macrophages are associated with cholesterol transport and steroid hallmarks.

A) Immune infiltration in orthotopic models of prostate cancer, as determined by flow cytometry. n=5 mice per group, with results from one of three independent experiments shown as the mean. B) Treatment schematic for Lupron and α-CSF1. C) Gene set enrichment analysis (GSEA) following whole tumor RNA sequencing. The normalized enrichment score is shown comparing tumors from mice treated with α-CSF1 versus an IgG control. Data reflects 3 tumors per group from a single experiment. D) Enrichment plots of the Androgen Response Hallmark show for tumors from mice treated with α-CSF1 or Lupron, compared to the IgG control. E) LC-MS/MS quantification of pregnenolone, progesterone, androstenedione, DHEA, 5-androstenediol, testosterone and DHT in fresh tumor tissues. n=5-6 mice per group, from one of two independent experiments. Measurements were normalized by tissue weight. Significance determined by unpaired t test for DHT. *p<0.05.

Figure 3.. Macrophages directly regulate AR nuclear translocation.

A) Representative immunohistochemistry for tumor macrophages (F4/80) and AR within orthotopic prostate tumors from mice treated with IgG (control), α-CSF or Lupron. B) Quantification of the percent of F4/80⁺ cells and AR^hi nuclei in serial sections. n=5-6 mice per group from one of two independent experiments. 4-5 regions of interest (ROI) per each slide were selected in non-necrotic areas of the tumor. Significance determined by unpaired t test and shown as ***p<0.001. C) Experimental schematic for cancer cell and BMDM co-culture and quantification of nuclear AR by confocal microscopy. D) Kernel density estimation of the AR nuclear to cytoplasmic ratio on a cell-by-cell basis following incubation of PTE-82, PTE-24 and PTE-09 cancer cell lines alone, or in co-culture with BMDMs under androgen-deprived conditions (i.e., charcoal-stripped serum, CSS) for 48 hrs. Percent of AR^hi cells was computed per individual images using the 3^rd quartile of the nucleus/cytoplasm AR intensity ratio as a threshold (3^rd quartile of PTE-82, PTE-24 or PTE-09 + BMDM = 3, 3, 2.4, respectively). n=3, data from one of at least three independent experiments. Significance determined by Mann-Whitney. E) Representative confocal microscopy images of the GFP⁺ PTE-24 cell line. AR (red), F4/80 (white) and DAPI (blue) fluorescence is shown. F) Impact of BMDMs on the proliferation of PTE-24 cells in the presence of serial concentrations of enzalutamide. Cell proliferation was monitored using live imaging with phase contrast images acquired every 6 hr. Data shown as the mean ± SEM and reflects one of 2 independent experiments. Significance determined by two-way ANOVA.

Figure 4.. Macrophages transfer cholesterol to prostate cancer cells.

A) Intergene correlation analysis of genes associated with macrophage infiltration (CD86, CSF1R, MRC1, CD163) and bile acid metabolism from TCGA. B) Intergene correlation analysis of genes associated with macrophage infiltration and de novo cholesterol synthesis (ACAT2, HMGCR, HMGCS1, SQLE). Analysis was performed with the Corrplot package in R. Correlation coefficients are indicated by changes in circle color and size. C) Neutral lipid staining in immune cells within orthotopic PTE-82 tumors, as measured by BODIPY-493/503 staining in single cell suspensions. n=5 mice, data shown as the mean ± SEM from one of three independent experiments. D) Expression of cholesterol metabolism-related genes in prostate cancer cell lines and BMDMs. n=3, data from one of three independent experiments. E) RAW264.7 cells were loaded with human LDL covalently conjugated to pHrodo Red for 3 hrs, then co-cultured with prostate cancer cells for 24 hrs prior to flow cytometric analysis. Cells were cultured separately and then admixed just prior to data acquisition for the 0 min control. The acquisition of fluorescence by the CD45⁻ cancer cells is highlighted in the 3^rd quadrant in red. Quantification of percent LDL-positive tumor cells is shown to the right. n=3 biological replicates from one of two independent experiments. Data are presented as mean ± SEM. Student’s t test was utilized for statistical analysis and is shown as ***p < 0.001. F) Immunofluorescence staining of cholesterol rich lipid rafts in the plasma membrane of RAW264.7 cells using Cholera toxin B (red). Cells were cultured alone or with PTE-82 cells in CSS for 48 hrs. Only nuclear GFP⁻ RAW264.7 cells are shown. n=3, data from one of two independent experiments. G) Live imaging of uptake of pHrodo Red-LDL (0.5 μg/ml) by PTE-82 tumor cells in the presence or absence of RAW264.7 cells. Arrows indicate tumor cells with red fluorescence. n=3, data reflects one of three independent experiments.

Figure 5.. Perturbing cholesterol homeostasis inhibits macrophage-mediated AR translocation.

A) PTE-82 prostate cancer cells were orthotopically injected into LysM-Cre⁺/Abca1^fl/fl/Abcg1^fl/fl or LysM-Cre⁻/Abca1^fl/fl/Abcg1^fl/fl and nuclear expression of AR was evaluated after 19 days by IHC. Representative images are shown below. n=5-6 mice per group from one of two independent experiments, data shown as the mean ± SEM. Significance was determined by unpaired t test assuming Gaussian distribution and with Welch’s correction. B) Tumor volume in mice from A, as measured by MRI. Representative T2-weighted MRI images are shown. C) Nuclear to cytoplasmic AR ratio of GFP⁺ PTE-82 cells cultured in CSS, either alone or in the presence of BMDMs derived from LysM-Cre⁺/Abca1^fl/fl/Abcg1^fl/fl or LysM-Cre⁻/Abca1^fl/fl/Abcg1^fl/fl donors. 6-8 randomly selected images from two wells of the chamber slides were pooled for analysis. Data reflects one of three independent experiments. Significance was determined by Mann-Whitney. D) Expression of Nr1h2 (LXRβ) and Nr1h3 (LXRα) in prostate cancer cell lines and BMDMs. n=3, data shown as the mean ± SEM from one of two independent experiments. E) Nuclear to cytoplasmic AR ratio of GFP⁺ PTE-82 cells cultured in CSS or co-cultured with BMDMs. 5 μM SR9243 (LXRα/β inverse agonist) or 5 μM RGX-104 (LXRβ agonist) were added to co-cultures as indicated. 6-8 randomly selected images from two wells of the chamber slides were pooled for analysis. Data reflects one of three independent experiments. Significance determined by Mann-Whitney.

Figure 6.. Macrophage depletion sensitizes CRPC to Lupron.

A) Diagram demonstrating set up of PDEs and treatment with bicalutamide (BIC). B,C) 4 μM sections from PDEs were stained with CD68. Response to bicalutamide was determined by the ratio between the number of nuclei/mm² in sections obtained from bicalutamide PDEs and number of nuclei/mm² in sections prepared from control treated PDEs. D) Expression of macrophage-associated genes was correlated with ADT response score, with stratification by ETS status. F) PTE-82 tumor cells were implanted orthotopically into prostates of C57 mice, which were then treated with Lupron depot and/or αCSF-1 as indicated. Tumor volumes for individual mice are shown for day 21. n=13-15 mice per group, data pooled from two independent experiments. Significance was determined by one-way ANOVA and is show compared to the IgG control group, unless otherwise indicated. G) Tumor volume in treatment groups as measured by MRI. H) Survival as determine by tumor volume or organ obstruction. n=21-23 mice, data pooled from three experiments. Significance was determined via log-rank.