Article Figures & Data
Figures
- Figure 1.
AI-sensitive MCF7 and AI-resistant MCF7-Aro-LetR cells exhibit decreased cell viability when treated with BEZ235; however, only MCF7-Aro-LetR cells display decreased cell viability when treated with anti-AR therapy. A, Tabulated details on two sets of paired isogenic breast cancer cells lines (endocrine-sensitive MCF7, ZR75.1 and their AI-resistant derivatives MCF7-Aro-LetR and ZR-Aro-LetR) including luminal type, AR protein expression and mutations impacting PI3K signaling. B, (i) Traces from the Sequenom MassArray show that endocrine-sensitive MCF7 cells have a heterozygous PIK3CA E545K mutation and (ii) endocrine-resistant MCF7-Aro-LetR cells have a heterozygous PIK3CA E545K mutation. C, Western blot analysis was used to evaluate AR and ER expression across five breast cancer cell lines (MCF7, MCF7-Aro-LetR, ZR75.1, and ZR-Aro-LetR). D and E, The effect of pan class PI3K/mTOR inhibitor (BEZ235) and anti-AR (enzalutamide) on cell viability was assessed using MTS and colony formation assay in endocrine-sensitive and endocrine-resistant cells. D, MTS of MCF7 response to BEZ235 or to a combination of BEZ235 plus enzalutamide compared with vehicle control. F, (i) Colony formation assays reflect MTS results in MCF7 cells. E, MTS of MCF7-Aro-LetR cells in response to single agents BEZ235 and enzalutamide and also to a combination treatment. F, (ii) Colony formation assays reflect MTS results in MCF7-Aro-LetR cells. Graphs representative of n = 3. Error bars are representative of mean ± SEM. Two-way ANOVA with Bonferroni posttest to compare replicate means was used in MTS day 1 to day 9 readings to determine significance. Student paired, two-tailed t test established significance to compare means: **, P < 0.01; ***, P < 0.001.
- Figure 2.
4AD upregulates genes associated with steroid and PI3K signaling in two models of AI resistance but not in the parental cell lines. A, RNA-sequencing was performed following androstenedione (4AD 10−7 mol/L) treatment plus letrozole for 24 hours in AI-resistant MCF7-Aro-LetR cells versus vehicle. These data were then compared with array data for MCF7aro-treated with 4AD versus vehicle to identify 4AD-specific transcripts. These genes were then integrated with the ‘Group 4′ gene set of endocrine-resistant genes associated with ER+, HER2− disease yielding a list of 8 genes. B, (i–viii) qRT-PCR validated differentially expressed genes (GREB1, SGK3, PKIB, and MYBL1) in AI-resistant cell models MCF7-Aro-LetR and ZR-Aro-LetR cells. C, (i–viii) qRT-PCR showed expression of these genes in parental MCF7 and ZR75.1 cells on exposure to 4AD. Graphs representative of n = 3. Error bars are representative of mean ± SEM. Student paired, two-tailed t test established significance: *, P < 0.05; ***, P < 0.001. ¶, ref. (32), $, ref. (36).
- Figure 3.
Culturing AI-resistant cells (MCF7-Aro-LetR) in estradiol for 15 weeks reverts 4AD-driven gene expression. A, (i) and (ii) qRT-PCR of SGK3 mRNA displays a very modest increase, whereas GREB1 is significantly decreased compared with vehicle control. A, (iii) and (iv) PKIB and MYBL1 expression is unchanged in the presence of 4AD. Graphs representative of n = 3. Error bars are representative of mean ± SEM. B, Gene expression (SGK3, GREB1, PKIB, and MYBL1) after 24 hours 10−7 mol/L 4AD treatment in MCF7-Aro-LetR cells long term cultured in 4AD compared with MCF7-Aro-LetR cells long term cultured in 10−8 mol/L estradiol (MCF7aroLetR-Est). Student paired, two-tailed t test was carried out to compare individual gene expression across both cell lines. *, P < 0.05.
- Figure 4.
Androstenedione-stimulated transcript levels of SGK3 and PKIB are reduced in response to siRNA knockdown of either AR or ESR1; in addition, AR and/or ER are recruited to these target genes in vivo and in vitro. A, qRT-PCR of MCF7-Aro-LetR transfected with siRNA-AR treated with 4AD shows a significant decrease in SGK3 and PKIB transcript levels (i–ii) but does not affect transcript levels of GREB1 or MYBL1 (iii–iv). B, qRT-PCR of MCF7-Aro-LetR transfected with siRNA–ESR1 in MCF7-Aro-LetR cells treated with 4AD results in a significant decrease in transcript levels of the SGK3, GREB1, PKIB, and MYBL1 (i–iv). C, Recruitment of AR and ER to target gene promoters was evaluated using publicly available data from a breast cancer patient ChIP-sequencing study, and ER recruitment in the MCF7-Aro-LetR cell line via ChIP-sequencing. ARϕ recruitment to the SGK3 and MYBL1 promoter was validated via ChIP in MCF7-Aro-LetR. D, (i) Validation of the putative AR target gene SGK3 was confirmed by performing ChIP to determine recruitment of AR to the promoter of SGK3 in MCF7-Aro-LetR cells in the presence of 4AD. (ii) ChIP experiments were performed to determine recruitment of ER to the promoter of SGK3 in MCF7-Aro-LetR cells treated with 4AD. All graphs representative of three experimental replicates. Error bars are representative of mean ± SEM of n = 3. Student paired, two-tailed t test established significance: *, P < 0.05; **, P < 0.01; ***, P < 0.001.
- Figure 5.
BEZ235 or combination of BEZ235 with the anti-AR bicalutamide decreases SGK3 mRNA expression. Knockdown of SGK3 inhibits AI-resistant cell proliferation only in the presence of androstenedione. A, (i) SGK3 protein expression is increased in the presence of 10−7 mol/L 4AD in MCF7-Aro-LetR cells. (ii) Densitometry analysis of 3 independent Western blots confirmed significant increase in SGK3 protein expression with 4AD treatment. B, qRT-PCR evaluation of SGK3 mRNA expression with BEZ235 ± anti-AR therapy bicalutamide in MCF7-Aro-LetR cells. C, (i) Western blot of SGK3 protein levels following siSGK3in MCF7-Aro-LetR. (ii) Densitometry analysis shows SGK3 protein levels following siSGK3. (iii) qRT-PCR of SGK3 mRNA following siSGK3 in MCF7aro-LetR; D, (i) MTS of MCF7Aro-LetR cell viability following knockdown of SGK3 in the absence of steroid. (ii) MTS assay of MCF7aro-LetR cell viability following siSGK3 combined with 4AD treatment. (iii) Cell counts confirmed decreased cell viability in MCF7Aro-LetR cells following siSGK3. Graphs representative of n = 3. Error bars are representative of mean ± SEM. Student two-tailed t test established significance for A, B, and D. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
- Figure 6.
SGK3, PKIB, and GREB1 are 4AD regulated transcripts mediated in part by AR in collaboration with ER. Associated outcome in clinical cohorts highlights a significant impact on therapeutic response to endocrine therapy. A, Schematic overview of key study findings highlights the morphologic and transcriptional changes driven by chronic exposure of MCF7 to 4AD in the presence of letrozole when stably overexpressing CYP19. These same cells, when maintained in estradiol, revert to their original morphology and no longer exhibit alterations in 4AD-mediated gene expression. B, Study summary: 1. In postmenopausal women, 100% of sex hormones are synthesized in peripheral tissues from circulating adrenal and ovarian precursors steroids. Clinical data have shown levels of 4AD to be elevated in patients whose disease progresses on AI therapy, and more recent data have shown 4AD dominates breast tumor intracrinology. Cyp 19 image: DOI: 10.2210/pdb4GL7/pdb. 2. In contrast, serum and tissue levels of estradiol are markedly reduced in patients treated with an aromatase inhibitor. 3. Our study identified SGK3 and PKIB as 4AD-regulated transcripts mediated in part by AR in collaboration with ER. C, (i) Meta-analysis of 4AD-regulated genes SGK3 and PKIB showed that there is no impact on RFS in the endocrine-untreated population [ER+ PR+, n = 379, HR = 1.23 (0.76–2.08), log-rank P = 0.37]. (ii) Meta-analysis of 4AD-regulated genes SGK3 and PKIB showed that SGK3 and PKIB is associated with poor RFS in the endocrine-treated population [ER+ PR+, n = 231, HR = 2.55 (1.34–4.85), log-rank P = 0.003].
- Figure 7.
Validation of SGK3 expression in clinical cohorts as an indicator of poor response to AI therapy. A, (i) Evaluation of GSE59515 shows SGK3 mRNA decreases significantly in patients who are responsive to AI therapy (mean expression: 16.75 → 0.563, one-tailed t test: P = 0.0015). (ii) Conversely, SGK3 mRNA in nonresponders is somewhat sustained (mean expression 16.64 → 6.44, one-tailed t test: not significant). (iii) Validation was performed in a second cohort of patients who were either responsive or nonresponsive to AI therapy. SGK3 mRNA was detectable only in nonresponders. B, (i) This was mirrored by levels of AR mRNA with a significant drop in responders (mean, 43.78→22.07 one-tailed t test: P = 0.0196) compared with sustained levels in (ii) nonresponders to AI therapy (mean 22.87→29.98, one-tailed t test: not significant). (iii) Validation was performed in a second cohort of patients who were either responsive or nonresponsive to AI therapy. AR mRNA was detectable only in nonresponders. C, UCSC Xena browser was used to interrogate TCGA breast cancer data. Kaplan–Meier survival curves showed that copy-number amplification of SGK3 significantly associates with poor survival in endocrine-treated, postmenopausal patients (P = 0.016). Analysis of the premenopausal patient cohort yielded no association; however, it should be noted that numbers in this cohort were <50.
Additional Files
Supplementary Data
- Supplementary Figure 1 - AR ER protein densitometry and dose response to PI3K/mTORi
- Supplementary Figure 2 - Characterisation of ZR75aro-LetR. Validation of ER AR siRNA
- Supplementary Figure 3 - Gene changes in MCF7aro. The impact of cyp19 overexpression on SGK3 expression in an independent study. Impact of siRNA-AR/ER on SGK3 mRNA
- Supplementary Figure 4 - AR and ER binding events in the proximal promoters of SGK3, MYBL1, GREB1 and PKIB.
- Supplementary Figure 5 - Impact of SiSGK3 in ZRaroLetR. MTS following siSGK3 under various hormonal stimuli.
- Supplemental Figure 6 - Correlation between CYP19 amplification and SGK3 expression.
- Supplemental Tables - Supplemental tables 1& 2
- Supplementary Methods - Supplemental methods
Volume 18, Issue 10
