RNF185 Control of COL3A1 Expression Limits Prostate Cancer Migration and Metastatic Potential

Abstract

RNF185 is a RING finger domain-containing ubiquitin ligase implicated in ER-associated degradation. Prostate tumor patient data analysis revealed a negative correlation between RNF185 expression and prostate cancer progression and metastasis. Likewise, several prostate cancer cell lines exhibited greater migration and invasion capabilities in culture upon RNF185 depletion. Subcutaneous inoculation of mouse prostate cancer MPC3 cells stably expressing short hairpin RNA against RNF185 into mice resulted in larger tumors and more frequent lung metastases. RNA-sequencing and Ingenuity Pathway Analysis identified wound-healing and cellular movement among the most significant pathways upregulated in RNF185-depleted lines, compared with control prostate cancer cells. Gene Set Enrichment Analyses performed in samples from patients harboring low RNF185 expression and in RNF185-depleted lines confirmed the deregulation of genes implicated in epithelial-to-mesenchymal transition. Among those, COL3A1 was identified as the primary mediator of RNF185's ability to impact migration phenotypes. Correspondingly, enhanced migration and metastasis of RNF185 knockdown (KD) prostate cancer cells were attenuated upon co-inhibition of COL3A1. Our results identify RNF185 as a gatekeeper of prostate cancer metastasis, partly via its control of COL3A1 availability.

Implications: RNF185 is identified as an important regulator of prostate cancer migration and metastasis, in part due to its regulation of COL3A1. Both RNF185 and COL3A1 may serve as novel markers for prostate tumors.

Fig. 1.. RNF185 expression correlates with metastatic load and PCa patient outcome.

(A) Disease-free survival in PCa samples stratified by RNF185 mRNA expression (“Low RNF185” = bottom quartile; “High RNF185” = top quartile). (B) Proportion of sample types in Low RNF185 vs. High RNF185. (C) RNA-seq analysis of RNF185 mRNA in primary tumors (n=131) and metastasis (n=19). (D) RNA-seq analysis of RNF185 mRNA in samples sorted by Gleason score (6: n=41; 7: n=76; 8: n=11; 9: n=11). (E) Disease-free survival in samples stratified by RNF185 mRNA expression using the same cutoffs as in (A). (F) RNA-seq analysis of RNF185 in samples sorted by Gleason score (6: n=45; 7: n=247; 8: n=64; 9: n=138; 10: n=4). (G) Overall survival in PCa samples stratified by RNF185 mRNA expression using the median as cutoffs. (H) RNA-seq analysis of RNF185 in samples sorted by Gleason score (5–7: n=29; 8–10: n=52). (I) RNA-seq analysis of RNF185 in patients with (n=33) and without (n=58) metastases. (J) RNA-seq analysis of RNF185 in samples sorted by pathological stages (2: n=30; 3: n=51; 4: n=7). (K) RNF185 IHC staining score in PCa TMA stratified by Gleason groups (3: n=127; 4: n=133; 5: n=69). (A-D) generated using datasets from Taylor et al. (46); (E-F) generated using The Cancer Genome Atals’ (TCGA) Pan-Cancer Atlas and Firehose legacy Prostate Adenocarcinoma (PRAD) datasets. (G-K) generated from VPC independent cohorts. Box plots show the median and whiskers (min to max), scatter plots show mean with SD. *P < 0.05, **P < 0.005, ***P < 0.001, ****P < 0.0001 by two-tailed t-test, Mann-Whitney u test, one-way ANOVA or Fisher’s exact test (B). Survival graphs used the Kaplan–Meier with a log-rank (Mantel-Cox).

Fig. 2.. RNF185 KD enhances PCa migration, tumor development and metastatic load.

(A) Western blot analysis of RNF185 protein levels in lysates of MPC3 cells stably expressing the indicated vectors. (B) Representative pictures of migrated MPC3 cells expressing the indicated vectors and quantification of migrated cells based on Crystal Violet (CV) absorbance. (C) Representative images of MPC3 cells expressing the indicated vectors that migrate through Matrigel and quantification of those migrated cells based on Crystal Violet absorbance. (D) Proliferation assay of the same cells shown in (A). (E) Growth of MPC3-pLKO and MPC3-shRNF185 cells after subcutaneous injection of 5×10⁵ cells into the flank of NSG mice (n=10/group). (F) Representative H&E staining of lung metastasis from mice inoculated with MPC3-shRNF185 cells (34 days post-injection). (G) Representative images of migrated PC3 cells transfected with siRNF185 or Universal negative control and quantification of migrated cells based on Crystal Violet (CV) staining absorbance. (H) Representative images of migrated C4–2B cells transfected with siRNF185 or Universal negative control and quantification of migrated cells based on Crystal Violet (CV) absorbance. Box plots show the median and whiskers (min to max). Growth curve shows the mean and whiskers (min to max). Box plots show median and whiskers. **P < 0.005, ***P < 0.001, ****P < 0.0001 by two-tailed t-test or two-way ANOVA.

Fig. 3.. Transcriptional analysis of RNF185 KD MPC3 cells and comparison analysis with clinical transcriptomics reveals EMT pathway activation and collagen upregulation.

(A) Principal Component Analysis (PCA) plot of MPC3-pLKO and MPC3-shRNF185 #1 and #3 biological replicates. (B) Heatmap comparing significantly altered diseases and functions in MPC3-shRNF185 #1 and #3 compared to MPC3-pLKO. (C) GSEA enrichment plots showing enrichment of the EMT hallmark gene set in MPC3-shRNF185 #1 and #3 relative to MPC3-pLKO, and in TCGA’s PRAD patients with low RNF185 mRNA expression (bottom 25%) relative to those with high expression (top 25%). (D) Venn Diagrams of enriched genes from HALLMARK_EMT in MPC3-shRNF185 #1 and #3 cells and in low RNF185 expressing patients from TCGA dataset. (E) qRT-PCR analysis of COL5A2 and COL3A1 in MPC3-pLKO and MPC3-shRNF185 cells, and in the human prostate cancer line PC3 transfected with RNF185 siRNA. Box plots show median and whiskers.

Fig. 4.. COL3A1 depletion in RNF185-deficient PC3 cells rescues migration advantage conferred by RNF185 KD alone.

(A) Representative images of migrated PC3 cells transfected with the indicated siRNAs. (B) Quantification of migrated cells based on Crystal Violet (CV) absorbance. (C) Disease-free survival in samples stratified by COL3A1 mRNA expression (“Low COL3A1” = bottom quartile; “High COL3A1” = top quartile). (D) Scatter plot of RNF185 and COL3A1 mRNA expression in PCa patients (Pearson correlation=0.75, p-value=2.47e-53). (E) RNA-seq analysis of COL3A1 in samples sorted by Gleason score (5–7: n=29; 8–10: n=52). (F) RNA-seq analysis of COL3A1 in samples sorted by pathological stages (2: n=30; 3,4: n=58). (G) RNA-seq analysis of COL3A1 in patients with (n=33) and without (n=58) metastases. (H) RNA-seq analysis of COL3A1 in samples sorted by their chances of recurrence (High risk; n=57, Low risk: n=22). (I) PSA-recurrence survival in samples stratified by COL3A1 mRNA expression (“Low COL3A1” = bottom quartile; “High COL3A1” = top quartile). (C) generated using datasets from Taylor et al. (46) (D) generated using TCGA PRAD dataset. (E-I) generated using VPC dataset. Box plots show the median and whiskers (min to max), scatter plots show mean with SD. *P < 0.05, **P < 0.005, ***P < 0.001, ****P < 0.0001 by Mann-Whitney u test or one-way ANOVA. Survival graphs used the Kaplan–Meier with a log-rank (Mantel-Cox).

Fig. 5.. COL3A1 depletion in RNF185-deficient MPC3 cells rescues tumor growth advantage and metastatic propensity conferred by RNF185 KD alone in subcutaneous grafts.

(A) Representative images of MPC3 cells expressing the indicated vectors that migrated through Matrigel and quantification of migrated cells based on Crystal Violet (CV) absorbance. (B) Representative pictures of resected tumors originating from cells expressing the indicated vectors. (C) Volume of resected tumors originating from cells expressing the indicated vectors 28 days after subcutaneous injection (n=6–8/group). (D) Weights of resected tumors originating from cells expressing the indicated vectors 28 days after subcutaneous injection (n=6–8/group). (E) qRT-PCR analysis of RNF185 mRNA in tumors emerging following subcutaneous injection of either MPC3.plko, MPC3.shRNF185, MPC3.COL3A1 or MPC3.shRNF185+shCOL3A1 cells into flanks of NSG mice. (F) qRT-PCR analysis of COL3A1 mRNA in tumors emerging following subcutaneous injection of either MPC3.plko, MPC3.shRNF185, MPC3.COL3A1 or MPC3.shRNF185+shCOL3A1 cells into flanks of NSG mice. (G) Average metastatic area (mm³) per section, measured in 9 serial sections of the lungs for each mouse within each group (n=6–8 mice per group). Scatter plots show the mean with SD, box plots show median with whiskers (min to max). *P < 0.05, **P < 0.005, ***P < 0.001, ****P < 0.0001 by one-way ANOVA.