Progestin and antiprogestin responsiveness in breast cancer is driven by the PRA/PRB ratio via AIB1 or SMRT recruitment to the CCND1 and MYC promoters

Abstract

There is emerging interest in understanding the role of progesterone receptors (PRs) in breast cancer. The aim of this study was to investigate the proliferative effect of progestins and antiprogestins depending on the relative expression of the A (PRA) and B (PRB) isoforms of PR. In mifepristone (MFP)-resistant murine carcinomas antiprogestin responsiveness was restored by re-expressing PRA using demethylating agents and histone deacetylase inhibitors. Consistently, in two human breast cancer xenograft models, one manipulated to overexpress PRA or PRB (IBH-6 cells), and the other expressing only PRA (T47D-YA) or PRB (T47D-YB), MFP selectively inhibited the growth of PRA-overexpressing tumors and stimulated IBH-6-PRB xenograft growth. Furthermore, in cells with high or equimolar PRA/PRB ratios, which are stimulated to proliferate in vitro by progestins, and are inhibited by MFP, MPA increased the interaction between PR and the coactivator AIB1, and MFP favored the interaction between PR and the corepressor SMRT. In a PRB-dominant context in which MFP stimulates and MPA inhibits cell proliferation, the opposite interactions were observed. Chromatin immunoprecipitation assays in T47D cells in the presence of MPA or MFP confirmed the interactions between PR and the coregulators at the CCND1 and MYC promoters. SMRT downregulation by siRNA abolished the inhibitory effect of MFP on MYC expression and cell proliferation. Our results indicate that antiprogestins are therapeutic tools that selectively inhibit PRA-overexpressing tumors by increasing the SMRT/AIB1 balance at the CCND1 and MYC promoters.

Keywords: AIB1; MYC; SMRT; antiprogestin therapy; antiprogestins; breast cancer; cyclin D1; hormone responsiveness; mifepristone; progesterone receptor isoforms.

© 2014 UICC.

Figures

Figure 1. Antiprogestins selectively inhibit the growth of mammary carcinomas with high PRA/PRB ratios

A, Growth curves of mammary carcinomas from the MPA-induced breast cancer model classified according to their MFP responsiveness as responsive, acquired resistant or constitutively resistant tumors. Agle and Proellex inhibited the growth of tumors that expressed higher levels of PRA than PRB. No inhibitory effects were observed in the MFP-resistant variants with higher levels of PRB than PRA. Only MFP stimulated the growth of the C4-2-HI tumors. Drugs were administered sc in doses of 10 mg/kg/day. In the case of 59-2-HI tumors as treatment began when animals had a size near 100 mm2, MFP or Proellex were administered within silastic pellets (6 mg). A representative curve of at least other two is shown (n=5/group). B, A representative WB showing the different PR ratios observed in the tumors shown in (A). C, Ten mice were implanted simultaneously with C4-HI and C4-2-HI tumors in opposite flanks. When C4-2-HI tumors reached approximately 25 mm2, MFP pellets (6 mg) or control silastic pellets were sc implanted into the mice. MFP inhibited C4-HI and stimulated C4-2-HI tumor growth in the same mouse. The arrows indicate treatment initiation. D, CCND1 and MYC expression in tumors from the experiment shown in C, treated for 72 h with MFP. Nuclear MYC and CCND1 staining decreased in C4-HI tumors, while an increase in MYC nuclear staining was observed in C4-2-HI tumors. Hematoxylin was used for nuclear staining. **, p<0.01 and ***, p<0.001 experimental vs. control group. Scale bar = 50 μm.

Figure 2. DNMT and HDAC inhibitors induce PRA re-expression and restore MFP responsiveness in constitutively MFP-resistant tumors

A, Growth curves. Animals (n=5/group) with palpable C4-2-HI and 59-HI tumors were treated with vehicle, MFP, 5azadC or TSA, each individually and in all possible combinations. The tumor sizes of one of three representative experiments are shown. The maximal therapeutic effect was obtained with the 5azadC+TSA+MFP combination treatment. Representative images are shown in Supplementary Figure 2A. B, Quantification of Ki67 expression; representative IHC images are shown in Supplementary Figure 2B. The percentage of Ki67 positive cells in relation to the total number of cells is plotted. Tumors treated with 5azadC+TSA+MFP had the lowest Ki67 values. C, Western blots of C4-2-HI tumor extracts confirmed PRA re-expression after 5azadC and TSA treatment. ERK served as a loading control. D, IHC studies using the C-19 antibody showed an increase in PRA nuclear expression in tumors treated with 5azadC and TSA. This antibody stains mainly PRA of mouse tissues in IHC studies (cited in ; p<0.05). *, p<0.05 and ***, p<0.001 experimental vs. control group. Scale bar = 50 μm.

Figure 3. Antiprogestins inhibit the growth of human T47D xenografts expressing high levels of PRA

A, Western blots for PRA or PRB in T47D, T47D-YA and T47D-YB cell lysates. B, Relation between PRB/total PR mRNA determined by qPCR. In T47D-YB cells the ratio between PRB and total PRB is considered as 1. C, Schematic of the experimental protocol for the T47D-YA and –YB experiments. D, Growth curves. T47D cells were inoculated into NOD/LtSz-scid/IL-2Rgamma null female mice (n=6/group) bearing a 0.5 mg E2 pellet. T47D-YA and –YB cells were simultaneously inoculated into the right and left flanks, respectively, of the same mouse. When the tumors reached 30-40 mm2, MFP (10 mg/kg/day) or vehicle treatments were initiated. Only T47D and T47D-YA tumors regressed after 17 days of treatment. E, CK staining illustrating the remodeling induced by MFP in T47D and T47D-YA xenografts; scale bar = 100 μm. F, Seventy-two hours after treatment initiation, 2 mice in each group were euthanized, and the tumors were processed for CCND1 and MYC immunostaining. T47D-YA tumors showed nuclear CCND1 and MYC expression, whereas a decrease in nuclear staining was observed in MFP-treated tumors. No decrease in nuclear CCND1 and MYC staining was observed in MFP-treated T47D-YB tumors. Hematoxylin was used for nuclear staining. *, p<0.05; **, p<0.01; ***, p<0.001 experimental vs. control group. Scale bar = 50 μm.

Figure 4. Co-localization of PR with SMRT or AIB1 and the proliferative effects of MPA or MFP

A, Left: MPA stimulates and MFP inhibits 3H-thymidine uptake in C4-HI cells. Middle: C4-HI cells were incubated for 30 min with MPA (10 nM) or MFP (10 nM), and confocal immunofluorescence images were captured using a Nikon Eclipse E800 Laser Confocal Microscope. Right: The nuclear co-localization of the proteins was estimated using a Pearson’s correlation coefficient. IF studies were performed with the available combinations of mouse monoclonal and rabbit antibodies (Vector Laboratories, Burlingame, CA), and the label color refers to the particular secondary antibody (Texas red, red; FITC, green). Black and white images at the right of each picture represent the separate green and red channels for each merged image. At the top, the merged image, at the middle the pPR/PR stains and at the bottom AIB1/SMRT stains. MPA increased the co-localization of pPR and AIB1, and MFP increased the co-localization with SMRT. Similar studies were performed in C4-2-HI (B) T47D-YA (C) and T47D-YB (D). Increased PR/SMRT colocalization was observed in MPA-treated C4-2-HI cells, which are growth inhibited by MPA, and in MFP- treated T47D/YA cells, which are both growth inhibited by MFP. An increase in PR/AIB1 co-localization was observed in MFP-treated C4-2-HI cells which are stimulated to proliferate with MFP.*, p<0.05; **p<0.01; ***, p<0.001 experimental vs. control group. Scale bar = 15 μm.

Figure 5. Changing the PR isoform ratio in T47D cells reverts the proliferative responses to MPA/MFP and the PR and AIB1/SMRT co-localization pattern

A, Left: MPA (10 nM) stimulates and MFP (10 nM) inhibits 3H-thymidine uptake in T47D cells. B, Cells were incubated for 30 min with MPA or MFP, and confocal immunofluorescence images of the pPR or PR and AIB1 or SMRT staining were captured using a Nikon Eclipse E800 Laser Confocal Microscope. The nuclear co-localization of the proteins was estimated using a Pearson’s correlation coefficient. IF studies were performed with available combinations of mouse monoclonal and rabbit antibodies (Vector Laboratories), and the label color refers to the particular secondary antibody (Texas red, red; FITC, green). MPA increased the co-localization of PR or pPR and AIB1, and MFP increased the co-localization with SMRT. C, Duolink assays were performed in cells incubated as described above. Pink spots represent interactions between both proteins. Increased AIB1/PR interactions were observed in MPA-treated cells, and increased PR/SMRT interactions in MFP-treated cells. D, Left: Western blot showing PR expression in T47D cells transfected with siRNA scrambled (siControl) or siRNA PRB (siPRB); Right: Cell counting after 5 days of incubation with MPA or MFP in the presence of 2% steroid stripped FCS. MPA did not increase cell proliferation of siPRB-transfected cells while MFP exerted a similar inhibitory effect in both cases. E, Left: Western blot showing PR expression in empty vector (pSG5) or in PRB (pSG5-PRB)-transfected T47D cells; Right: Cell counting experiments of the cells in E left. MPA did not increase cell proliferation whereas MFP increased the number of PRB-transfected cells after 5 days of incubation. F, PRB-transfected or control cells were incubated with MPA or MFP for 15 min and the co-localization of PR with AIB1 or SMRT was evaluated as explained in B. A change in the pattern of co-localization was observed in PRB transfected cells. *, p<0.05; **, p<0.01; ***, p<0.001 experimental vs. control group; a vs. b, p<0.001 and c vs. d, p<0.05. Scale bar = 15 μm.

Figure 6. MPA and MFP activate PR, which interacts with AIB1 or SMRT at the CCND1 and MYC promoters to increase or repress gene transcription, respectively

T47D cells treated with 10 nM MPA or MFP (30 min) were subjected to qPCR (A) and ChIP (B) assays. The presence of AIB1 or SMRT at the CCND1 and MYC promoters was evaluated in a region known to be bound by PR (right; Supplementary Table 1) . MPA increased the recruitment of AIB1 and PR at the MYC and CCND1 promoters, whereas MFP induced SMRT recruitment. No recruitment of PR, AIB1 or SMRT was observed using control primers (Supplementary Figure 4A). C, siRNA SMRT (mix of three siRNAs) or control scrambled siRNA were transfected into T47D cells. After 48 h, SMRT expression was evaluated by IF (left; FITC) or qPCR (right). Propidium iodide (PI; red) was used for nuclear counterstaining. Scale bar = 50 μm. D, 3H-thymidine uptake assay. Cells transfected with SMRT or control siRNAs were incubated with MFP for 48 h and 3H-thymidine for the last 18 h. The proliferation index was calculated as the cpm observed in MFP-treated cells as compared with their respective control groups. E, IF assays. MYC expression in siRNA SMRT-transfected or control cells incubated with MPA (10 nM) or MPA+MFP (10 nM) for 60 min. The percentage of MYC+ (FITC) cells in relation to the total number of cells was quantitated (PI stained cells). Scale bar = 50 μm. F, qPCR. MYC mRNA expression in siRNA SMRT-transfected or control cells incubated with MPA (10 nM) and/or MFP (10 nM) for 20 min. MFP was unable to decrease cell proliferation or MYC expression in siRNA SMRT-transfected cells. *, p<0.05; **, p<0.01; ***, p<0.001 experimental vs. control group.