ZFHX3 is indispensable for ERβ to inhibit cell proliferation via MYC downregulation in prostate cancer cells

Abstract

Both estrogen receptor 2 (ESR2, also known as estrogen receptor beta (ERβ)) and the zinc-finger homeobox 3 (ZFHX3, also known as ATBF1 for AT motif-binding factor 1) modulate prostate development and suppress prostatic tumorigenesis in mice. ZFHX3 is integral to proper functions of ESR1 (i.e., estrogen receptor alpha (ERα)), which belongs to the same family of proteins as ESR2, but is hardly expressed in prostate epithelial cells. It is not clear how ZFHX3 suppresses prostatic tumorigenesis. In this study, we investigated whether ZFHX3 and ERβ functionally interact with each other in the suppression of prostatic tumorigenesis. In two androgen receptor (AR)-positive prostate cancer cell lines, C4-2B and LNCaP, we first validated ERβ's tumor suppressor activity indicated by the inhibition of cell proliferation and repression of MYC expression. We found that loss of ZFHX3 increased cell proliferation and MYC expression, and downregulation of MYC was necessary for ZFHX3 to inhibit cell proliferation in the same cell lines. Importantly, loss of ZFHX3 prevented ERβ from suppressing cell proliferation and repressing MYC transcription. Biochemically, ERβ and ZFHX3 physically interacted with each other and they both occupied the same region of the common MYC promoter, even though ZFHX3 also bound to another region of the MYC promoter. Higher levels of ZFHX3 and ERβ in human prostate cancer tissue samples correlated with better patient survival. These findings establish MYC repression as a mechanism for ZFHX3's tumor suppressor activity and ZFHX3 as an indispensable factor for ERβ's tumor suppressor activity in prostate cancer cells. Our data also suggest that intact ZFHX3 function is required for using ERβ-selective agonists to effectively treat prostate cancer.

Fig. 1. Activation of ERβ inhibits cell proliferation while upregulating ZFHX3 but downregulating MYC and CCND1 in both C4-2B and LNCaP cell lines.

a Expression of ZFHX3, AR, ERα, and ERβ in human prostate epithelial cell lines, as determined by Western blotting. Breast cancer cell lines MCF-7 and MDA-MB-231 were used as positive and negative controls, respectively, for ERα. b Treatment with DPN, an ERβ activator, decreased cell proliferation in two-dimensional (2D) culture in both C4-2B (left) and LNCaP (right) cell lines. Cells were serum-starved for 48 h before indicated DPN treatments. Optical densities represent cell numbers. n = 4. c Knockdown of ERβ eliminated the inhibitory effect of DPN on colony formation in both C4-2B (left) and LNCaP (right) cell lines in plates pre-coated with 0.35% soft agar. Cells were transfected with siESR2-2, which showed the highest efficiency of knockdown among the three siRNAs against ESR2, for 24 h and treated with DPN (0.1 µM) for 2 weeks. Colonies with a diameter > 100 µm were counted. Knockdown of ERβ was validated by Western blotting. n = 3. d Inhibition of ERβ function by its antagonist PHTPP eliminated the inhibitory effect of DPN on colony formation. n = 3. e DPN upregulates ZFHX3 expression but downregulates MYC and CCND1 in C4-2B and/or LNCaP cells. DPN treatments were at 0, 0.01, 0.1, and 1 µM for 48 h, and protein expression was determined by Western blotting. f Inhibition of ERβ function by its antagonist PHTPP eliminated the inhibitory effect of DPN on MYC expression in C4-2B cells. DPN and PHTPP were at 0.5 μM. *P < 0.05; **P < 0.01; ns, not significant. ZFHX3, zinc-finger homeobox 3; AR, androgen receptor; ERα, estrogen receptor alpha; DPN, diarylpropionitrile; siRNA, small interfering RNA; PHTPP, 4-[2-phenyl-5,7-bis(trifluoromethyl)pyrazolo[1,5-a]pyrimidin-3-yl]phenol

Fig. 2. Loss of ZFHX3 increases cell proliferation, colony formation, and sphere formation in androgen-responsive prostate cancer cells.

a, b Isolation of clones with successful CRISPR-Cas9-mediated truncating mutations of ZFHX3 in C4-2B cells, as indicated by sequencing analysis of targeted locus (a) and expression detection of ZFHX3 by Western blotting (b). Wild-type sequence in parental cells is shown at the top of a. Wt, vector control cells; KO3, KO4, KO7, KO8, KO9, and KO10 are different clones of C4-2B cells with ZFHX3 mutations. c Loss of ZFHX3 increased cell proliferation in two-dimensional (2D) culture in C4-2B cells. Optical densities represent cell numbers. n = 4. d–g Loss of ZFHX3 increased sphere formation in Matrigel (d—upper, e, f) and colony formation in soft agar (d—lower, g) in C4-2B cells. Cells were grown for 10–14 days in Matrigel or soft agar. Shown are bright field images of spheres (d, upper) and colonies (d, lower), range of sphere sizes (e), the average number of spheres with a diameter >75 µm per well (f), and the average number of colonies with a diameter >100 μm (g). The ImageJ program was used to determine sphere/colony sizes. h–j Knockdown of ZFHX3 in LNCaP cells also increased sphere formation. The knockdown effect was validated by Western blotting (h), and bright field images of colonies (i) and the number of colonies with a diameter >100 μm (j) is shown. Scale bars in d, 100 μm. The n of e–g, j is 3. *P < 0.05; **P < 0.01; ***P < 0.001. ZFHX3, zinc-finger homeobox 3

Fig. 3. Inhibitory effect of ZFHX3 on colony formation depends on the downregulation of MYC in prostate cancer cells.

a Expression of MYC in prostate cancer cell lines, as detected by Western blotting. b Knockout or knockdown of ZFHX3 upregulated MYC expression in C4-2B and LNCaP cells, as detected by Western blotting. Wt, KO3, and KO8 are vector control and two ZFHX3-null clones of C4-2B. c–e Knockdown of MYC, by using siRNAs against MYC at two concentrations (20 and 40 nM), eliminated the promoting effect of ZFHX3’s loss on cell proliferation and colony formation in soft agar in C4-2B cells. The SRB assay was used to measure cell proliferation (d), while the soft agar assay was used for colony formation (e). The n of both d and e is 4. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. ZFHX3, zinc-finger homeobox 3; siRNA, small interfering RNA; Wt, wild type; SRB, sulforhodamine B

Fig. 4. ZFHX3 and ERβ physically interact with each other independent of DPN treatment.

a Cell lysates from C4-2B cells, which express both ZFHX3 and ERβ, were subjected to immunoprecipitation (IP) with anti-ERβ antibody and subsequent immunoblotting (IB) with anti-ZFHX3 antibody. Input indicates cell lysate not subjected to IP. b C4-2B cells were grown in phenol red-free medium with 5% charcoal-stripped FBS for 48 h and then treated with DPN for 48 h. IP and IB were performed as in a. c Expression plasmids for HA-tagged ZFHX3 (HA-ZFHX3) and FLAG-tagged ERβ (FLAG-ERβ) were transfected into 293T cells. Lysates were subjected to IP with anti-FLAG or anti-HA affinity gel, and then to IB with anti-HA or anti-FLAG antibody. d Schematic of full ZFHX3 protein (3703 residues, horizontal bar) with 23 zinc fingers (gray ovals) and 4 homeodomains (black rectangles). The six shorter bars below indicate six overlapping fragments of ZFHX3, named A to F. Each of the six fragments was tagged with HA, expressed in 293T cells, and tested for their interactions with ERβ by IP and IB. The two confirmed interactions with ERβ, A and D, are shown in solid dark. e IP and IB results for the interaction of FLAG-tagged ERβ and each of the six HA-tagged ZFHX3 fragments. The same procedures as in c were used. Arrows indicate the two fragments that were pulled down by ERβ (i.e., a and d). ZFHX3, zinc-finger homeobox 3; ERβ, estrogen receptor beta; FBS, fetal bovine serum; DPN, diarylpropionitrile

Fig. 5. ZFHX3 and ERβ bind to MYC promoter to coordinately repress its transcription.

a DPN treatment decreased MYC mRNA level in C4-2B cells. Hormone-deprived medium was used for DPN treatment. n = 4. b Inhibition of ERβ function by its antagonist PHTPP eliminated the inhibitory effect of DPN on MYC mRNA expression in C4-2B cells. DPN and PHTPP were at 0.5 µM. n = 4. c Knockout of ZFHX3 increased MYC mRNA level. Wt, control clone; KO3 and KO8, two ZFHX3-null clones of C4-2B cells. n = 4. d, e DPN decreased the activity of MYC promoter. Expression plasmids of pGL3 vector control (pGL3-basic), pGL3 with MYC full-length promoters (a, pGL3-MYC with bases −2455 to 309) or pGL3 with two smaller MYC promoter fragments (d, pGL3-MYC-1 with bases −2024 to −1193 and pGL3-MYC-2 with bases −1200 to −200), and the pRL-TR reporter were transfected into C4-2B cells in phenol red-free medium supplemented with 2% CS-FBS. Twenty-four hours later, DPN treatments (0.1 µM, 48 h) were applied, and relative luciferase activities were then determined. n = 4. f Schematic of the MYC promoter region from base −1200 to base −200 relative to the P2 transcriptional initiation site (TIS), with locations of all four TISs, the first 3 exons, and primers used to amplify promoter regions A–C. Arrows under the promoter indicate primer locations. g Detection of ZFHX3- and ERβ-bound MYC promoter DNA in parental C4-2B cells using ChIP and regular PCR. h, i Binding of ERβ to MYC promoter region A in the presence (Wt) and absence (KO8) of ZFHX3 (h), with or without DPN treatment (i), using ChIP and regular PCR (upper) or real-time PCR (lower) in Wt and KO8 clones of C4-2B cells. The n of h, i is 3. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. ZFHX3, zinc-finger homeobox 3; ERβ, estrogen receptor beta; mRNA, messenger RNA; PHTPP, 4-[2-phenyl-5,7-bis(trifluoromethyl)pyrazolo[1,5-a]pyrimidin-3-yl]phenol; Wt, wild type; CS-FBS, calf serum-fetal bovine serum; DPN, diarylpropionitrile, ChIP, chromatin immunoprecipitation

Fig. 6. Loss of ZFHX3 eliminated the inhibitory effect of ERβ on colony formation and MYC expression in prostate cancer cells and correlated with worse patient survival.

C4-2B (a, c) and LNCaP (b, d) cells were used for both colony formation assay (a, b) and MYC expression analysis (c, d). In colony formation assay, cells plated on 0.35% soft agar in phenol red-free medium were cultured, and colonies >100 µm were counted. MYC protein was detected by Western blotting. DPN was added to enhance the ERβ activity. e Transfection-mediated re-expression of ZFHX3 in the ZFHX3-null KO8 clone of C4-2B cells decreased MYC expression, as detected by Western blotting. f A model for how ZFHX3 is indispensable for ERβ to suppress cell proliferation and tumor growth in prostate cancer cells. In the presence of ZFHX3, ERβ interacts with ZFHX3 to repress the transcription of MYC and other oncogenes, but this repression is eliminated by the loss of ZFHX3. g, h Kaplan–Meier analysis of overall survival (g) and disease-free survival (h) of prostate cancer patients with different statuses of ZFHX3 and ESR2 expression. The n of a, b is 3. *P < 0.05; **P < 0.01; ns, not significant. ZFHX3, zinc-finger homeobox 3; ERβ, estrogen receptor beta; DPN, diarylpropionitrile