The HeyL-Aromatase Axis Promotes Cancer Stem Cell Properties by Endogenous Estrogen-Induced Autophagy in Castration-Resistant Prostate Cancer

Abstract

Treatment of patients with castration-resistant prostate cancer (CRPC) remains a major clinical challenge. We previously showed that estrogenic effects contribute to CRPC progression and are primarily caused by the increased endogenous estradiol produced via highly expressed aromatase. However, the mechanism of aromatase upregulation and its role in CRPC are poorly described. In this study, we report that HeyL is aberrantly upregulated in CRPC tissues, and its expression is positively correlated with aromatase levels. HeyL overexpression increased endogenous estradiol levels and estrogen receptor-α (ERα) transcriptional activity by upregulating CYP19A1 expression, which encodes aromatase, enhancing prostate cancer stem cell (PCSC) properties in PC3 cells. Mechanistically, HeyL bound to the CYP19A1 promoter and activated its transcription. HeyL overexpression significantly promoted bicalutamide resistance in LNCaP cells, which was reversed by the aromatase inhibitor letrozole. In PC3 cells, the HeyL-aromatase axis promoted the PCSC phenotype by upregulating autophagy-related genes, while the autophagy inhibitor chloroquine (CQ) suppressed the aromatase-induced PCSC phenotype. The activated HeyL-aromatase axis promoted PCSC autophagy via ERα-mediated estrogenic effects. Taken together, our results indicated that the HeyL-aromatase axis could increase endogenous estradiol levels and activate ERα to suppress PCSC apoptosis by promoting autophagy, which enhances the understanding of how endogenous estrogenic effects influence CRPC development.

Keywords: CRPC; HeyL; aromatase; autophagy; prostate cancer stem cell.

Figure 1

HeyL overexpression is associated with poor prognosis and estrogenic effects in patients with CRPC. (A) HeyL mRNA levels in normal prostate and PCa tissues from 6 publicly available datasets (Tomlins, n(normal)=23, n(PCa=29); Arredouani, n(normal)=8, n(PCa=13); Grasso, n(normal)=28, n(PCa=59); Luo Prostate 2, n(normal)=15, n(PCa=15); Varambally, n(normal)=6, n(PCa=7); Lapointe, n(normal)=40, n(PCa=60)) (t-test). (B) Kaplan-Meier curve of disease-free survival differences according to the HeyL expression level based on a TCGA cohort. (C) HeyL mRNA expression in primary PCa and CRPC tissues from 2 publicly available datasets (Chandran, n(primary PCa)=10, n(CRPC=21); Roudier, n(primary PCa)=11, n(CRPC=45)) (t-test). (D) Comparison of HeyL mRNA expression between recurrent and nonrecurrent PCa (n(nonrecurrent)=104), n(recurrent)=36) (t-test). (E) Representative images of HeyL IHC staining in benign prostate (n=9), low Gleason score PCa (Gleason score ≤ 6, n=11), high Gleason score PCa (Gleason score 8-10, n=12), and CRPC tissues (n=15); scale bar: 100 µm. (F) Average optical density of HeyL in low Gleason score vs. high Gleason score tissues and in primary PCa (low Gleason score and high Gleason score) vs. CRPC tissues (t-test). (G) GSEA plot of the association between HeyL and the estrogen response gene set in the TCGA-PRAD dataset. (H, I) ER transcriptional activity (ERE-Luc) (H) and level of intracellular 17β-estradiol (I) in control and HeyL knockdown PC3 cells. PRAD: prostate adenocarcinoma. The data are presented as the mean ± SD values (n=3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

HeyL promotes endogenous estrogenic effects by directly activating CYP19A1 transcription. (A) Mean mRNA expression levels of CYP19A1 in HeyL^High samples vs. HeyL^Low samples in the TCGA-PRAD dataset. (B) Kaplan-Meier analysis of disease-free survival for HeyL^HighCYP19A1^High patients vs. HeyL^LowCYP19A1^Low patients in the TCGA-PRAD dataset. (C) Representative images of HeyL and aromatase IHC staining in low Gleason score (Gleason score ≤ 6, n=5), high Gleason score PCa (Gleason score 8-10, n=5), and CRPC (n=5) tissues. Scale bar: 100 μm. The average optical densities of HeyL and aromatase were calculated, followed by correlation analysis of HeyL and aromatase expression in primary PCa and CRPC specimens. (D) Colocalization of HeyL and aromatase in PC3 cells was detected by IF staining. Scale bar: 20 μm. (E–I) Effects of HeyL on aromatase expression. (E–G) The mRNA (left panel) and protein (right panel) levels of aromatase in 22RV1, LNCaP-abl, and PC3 cells stably transfected with control or i_HeyL were analyzed (t-test). (H) The mRNA (left panel) and protein (right panel) levels of aromatase in ctrl or HeyL-overexpressing LNCaP cells were analyzed (t-test). (I) Relative mRNA level of aromatase in PCSCs transfected with siNC or siHeyL (t-test). (J, K) HeyL directly activates CYP19A1 transcription by occupying its promoter. (J) Left: Schematic diagram of the binding site between HeyL and the promoter of CYP19A1 promoter. Right: ChIP showed that HeyL binds directly to the CYP19A1 PII promoter region. (K) Relative aromatase promoter activity (AROM-pro-Luc) in LNCaP-abl and PC3 cells stably transfected with control or i_HeyL (t-test). (L–N) Relative ERE-Luc activity (L, N) and the concentration of intracellular 17β-estradiol (M) in PC3 cells after the indicated treatment (one-way ANOVA). PCSC, prostate cancer stem cell; ChIP, chromatin immunoprecipitation; E2, 17β-estradiol. The data are presented as the mean ± SD values (n=3). *p < 0.05 vs. ctrl. ^# p < 0.05 vs. i_HeyL or OE-HeyL. ***p < 0.001.

Figure 3

HeyL functions to maintain the characteristics of PCSCs. (A) GSEA plot of the correlation between HeyL expression and gene signatures associated with stem cell proliferation in HeyL knockdown cells. (B) Relative mRNA levels of HeyL, CYP19A1, CD44, and the indicated stemness-associated genes in PCSCs enriched from LNCaP-abl cells (t-test). (C) Colocalization of HeyL and CD44 in PC3 cells was detected by IF staining. Scale bar: 20 μm. (D, E) Tumorsphere-forming ability of LNCaP and LNCaP-abl cells treated as indicated. The top panel shows representative micrographs (scale bar: 100 μm); the bottom panel shows the quantitative results [(D), LNCaP, t-test; (E), LNCaP-abl, t-test). (F, G) The mRNA levels of CD44 and the indicated stemness-associated genes in HeyL-overexpressing LNCaP cells (F) and HeyL knockdown LNCaP-abl cells (G) (t-test). (H) Flow cytometry analysis of the CD44⁺/CD24^- subpopulation in 22RV1, LNCaP-abl, and PC3 cells after HeyL knockdown. PCSC, prostate cancer stem cell. The data are presented as the mean ± SD values (n=3). *p < 0.05 vs. ctrl.

Figure 4

HeyL regulates PCSC properties by upregulating CYP19A1 expression. (A–C) Tumorsphere-forming ability of LNCaP-abl cells treated as indicated. The protein levels of aromatase, representative micrographs, and quantitative results (one-way ANOVA) are shown in (A), (B), and (C), respectively. Scale bar: 100 μm. (D, E) Flow cytometry analysis of the CD44⁺/CD24⁻ subpopulation in PC3 cells treated as indicated. (D) Protein levels of aromatase. (E) The results of flow cytometry analysis. (F–H) mRNA levels of CD44 and the indicated stemness-associated genes in PC3 (F, G) and LNCaP-abl (H) cells treated as indicated (one-way ANOVA). The data are presented as the mean ± SD values (n=3). *p < 0.05 vs. ctrl. ^# p < 0.05 vs. i_HeyL or OE-HeyL.

Figure 5

Blocking the HeyL-aromatase axis sensitizes PCa cells to bicalutamide treatment. (A, B) The mRNA (left panel) and protein (right panel) levels of HeyL, aromatase, SOX2, and CD44 were analyzed in LNCaP (A) and LNCaP-abl (B) cells treated with bicalutamide (t-test). (C) Cell viability was evaluated in LNCaP cells after the indicated treatments (one-way ANOVA). (D) The clonogenic ability of vector control and HeyL-overexpressing LNCaP cells treated with bicalutamide and/or letrozole was analyzed (one-way ANOVA). (E) Vector control LNCaP cells and HeyL-overexpressing LNCaP cells were treated with bicalutamide and/or letrozole. Apoptosis was evaluated by flow cytometry. (F) Cell viability was evaluated in LNCaP-abl cells after the indicated treatments (one-way ANOVA). The data are presented as the mean ± SD values (n=3). *p < 0.05 vs. ctrl. ^# p < 0.05 vs. OE-HeyL.

Figure 6

The activated HeyL-aromatase axis promotes autophagic survival of PCSCs. (A) Gene signatures associated with autophagy-associated signaling pathways identified by GSEA in HeyL^High samples vs. HeyL^Low samples from the TCGA-PRAD dataset and in HeyL knockdown cells vs. ctrl cells. (B, C) Western blot analysis of Beclin1, LC3, and CD44 in LNCaP (B) and PC3 (C) cells after the indicated treatments. (D) Flow cytometry analysis of the CD44⁺/CD24⁻ subpopulation in PC3 cells treated as indicated. (E, F) mRNA (E) and protein (F) levels of stemness-associated genes and LC3 in PC3 cells after the indicated treatment (one-way ANOVA). (G, H) mRNA and protein levels of stemness-associated genes and autophagy-related genes in PC3 cells treated with ICI182780 (G) and siERα (H) (t-test). (I, J) Cleaved caspase3 levels in LNCaP and PC3 cells after the indicated treatments (one-way ANOVA). (K, L) Cleaved caspase3 levels in PC3 cells treated with ICI182780 (K) or transfected with siERα (L) (t-test). Bic, bicalutamide; CQ, chloroquine. The data are presented as the mean ± SD values (n=3). *p < 0.05 vs. ctrl. ^# p < 0.05 vs. OE-HeyL, i_HeyL or OE-CYP19A1.

Figure 7

Schematic depiction of the proposed mechanism by which the HeyL-aromatase axis promotes stemness maintenance and apoptosis resistance in CD44⁺/CD24^- PCSCs. CRPC cells exhibit higher levels of HeyL and aromatase than androgen-dependent PCa cells. HeyL promotes CYP19A1 expression by directly binding to the CYP19A1 PII promoter region. The activated HeyL-aromatase axis enhances the stemness of the CD44⁺/CD24^- subpopulation and attenuates bicalutamide sensitivity. The residual testosterone derived from serum and intratumoral biosynthesis after castration is catalyzed to estradiol by the HeyL-aromatase axis in CRPC patients. Increased levels of intracellular estradiol enhance ERα-induced autophagy, thereby promoting the apoptotic resistance of the CD44⁺/CD24^- subpopulation. T, Testosterone; E2, Estrogen/Estradiol; ERα, Estrogen receptor α.