Tumor suppressor function of androgen receptor coactivator ARA70alpha in prostate cancer

Abstract

Androgen receptor (AR), a member of the steroid receptor family, is a transcription factor that has an important role in the regulation of both prostate cell proliferation and growth suppression. AR coactivators may influence the transition between cell growth and growth suppression. We have shown previously that the internally spliced ARA70 isoform, ARA70beta, promotes prostate cancer cell growth and invasion. Here we report that the full length ARA70alpha, in contrast, represses prostate cancer cell proliferation and anchorage-independent growth in vitro and inhibits tumor growth in nude mice xenograft experiments in vivo. Further, the growth inhibition by ARA70alpha is AR-dependent and mediated through induction of apoptosis rather than cell cycle arrest. Interestingly, AR with T877A mutation in LNCaP cells decreased its physical and functional interaction with ARA70alpha, facilitating the growth of LNCaP cells. The tumor suppressor function of ARA70alpha is consistent with our previous findings that ARA70alpha expression is decreased in prostate cancer cells compared with benign prostate. ARA70alpha also reduced the invasion ability of LNCaP cells. Although growth inhibition by ARA70alpha is AR-dependent, the inhibition of cell invasion is an androgen-independent process. These results strongly suggest that ARA70alpha functions as a tumor suppressor gene.

Figure 1

ARA70α inhibits cell growth in an androgen and AR-dependent manner. A: ARA70α inhibits growth of LNCaP cells in the presence of androgen R1881 (empty squares). Without R1881 (empty circles), LNCaP cells overexpressing ARA70α grow at the rate similar to that of control cells, both in the presence (black squares) and absence (black circles) of R1881. The expression of ARA70α was confirmed by using western blot (lower panels). B: Endogenous ARA70α limits the growth rate of LNCaP cells. LNCaP cells treated with ARA70α-specific siRNA (squares) and untreated LNCaP cells (circles) were grown in the presence (empty circles and squares) or absence (black circles and squares) of R1881. C: Growth-inhibitory effect of ARA70α depends on the presence of AR. AR levels were decreased by using siRNA (lower panels). AR knockdown with siRNA reversed the growth inhibition by ARA70α overexpression (empty circles and squares) in LNCaP cells, and the cells were grown in the presence (squares) and absence (circles) of R1881. Included for comparison are LNCaP-ARA70α cells not treated with AR siRNA in the presence of R1881 (crosses). D: ARA70α does not influence the growth of PC3 cells. PC3 cells treated with ARA70α-specific siRNA (squares) and untreated PC3 cells (circles) were grown in the presence (empty circles and squares) or absence (black circles and squares) of R1881.

Figure 2

ARA70α inhibits growth of LNCaP cells in anchorage-independent assays. Cells were grown suspended in agarose medium in the presence or absence of R1881. A: In the presence of R1881, ARA70α-overexpressing cell line produced fewer and smaller colonies than in the absence of R1881 compared with the control cells. B: Two clonal LNCaP cell lines overexpressing ARA70α yielded the same number of colonies as the control cell line in the absence of R1881 (black columns), but this number decreased by 50% in the presence of R1881 (white columns).

Figure 3

ARA70α reduced tumor growth in nude mice xenografts. A: Growth suppression by ARA70α on subcutaneous tumor xenografts, 8 weeks after injection, is shown. B: Tumor volume showing LNCaP-ARA70α cells have greatly reduced potential in tumorigenesis. Nude mice were injected either with 7 × 10⁶ LNCaP-ARA70α cells or LNCaP-pBabe cells. Tumor size was measured twice a week. Black bars, control LNCaP-pBabe cells; white bars, LNCaP-ARA70α−4 cells; gray bars, LNCaP-ARA70α−8 cells; Error bars represent SE. Inset: Increased cells positive for cleaved caspase 3 in LNCaP-ARA70α (right) compared with LNCaP-pBabe (left) cells.

Figure 4

Increased number of cells undergoing apoptosis in LNCaP cells overexpressing ARA70α. A: LNCaP-ARA70α and LNCaP-pBabe cells were grown in androgen medium, harvested, stained with dye DiOC₂(3), and analyzed by flow cytometry (488 nm excitation, 530/30 nm bandpass and 650 nm longpass filters). B: Androgen-dependent changes in apoptotic regulators in ARA70α overexpressing cells. LNCaP-pBabe and LNCaP-ARA70α cells were grown either in the presence or absence of androgen and the levels of Bcl-XL, Bax, and cleaved caspase 3 were analyzed using western blot. C: Effect of mutated AR on the extent of apoptosis caused by overexpression of ARA70α. PC3 cells stably overexpressing ARA70α were transiently transfected either with wild type AR or the AR^T877A mutant. Harvested cells were stained with DiOC₂(3) and analyzed as above.

Figure 5

Strong interaction between AR and ARA70α requires both AR DBD and LBD domains. A: Schematic domains of two-hybrid constructs containing the nuclear localization sequence (NLS), the B42 transcription activation domain, hemaglutinin tag (HA), and individual domains of AR (NTD, N-terminal domain; DBD-L, full length DNA-binding domain; DBD-S, short version of the DNA-binding domain; LBD, ligand-binding domain; DBD+LBD, DNA-binding domain and ligand-binding domain). B: Interaction between AR and ARA70α in a two-hybrid overlay test. The yeast patches were grown on the selective medium containing galactose in the presence or absence of R1881. The expression of the reporter gene was visualized by lysing the cells in situ in the presence of the chromogenic substrate x-gal. Yeast expressing two-hybrid constructs of FKBP1a and transforming growth factor β1 were used as a positive control. The tests were performed with ARA70α fused to the LexA DNA-binding domain and individual AR domains fused to the B42 transactivation domain. C: Co-immunoprecipitation of LBD+DBD domains of AR with ARA70α. Lysate was incubated with flag M2 antibody and immunoprecipitated AR domains were detected by western blot using HA antibody.

Figure 6

The T877A mutation in AR weakens the physical and functional interactions between AR and ARA70α. A: The expression of the β-galactosidase reporter gene was measured in the liquid assay in cells expressing ARA70α and either wild-type AR or AR^T877A in the presence or absence of the synthetic androgen R1881. B: Impaired ARA70α transcriptional activity with mutated AR in dual luciferase assays. LNCaP cells were transiently transfected with expression vectors carrying ARA70α, the reporter gene (under the control of promoter containing four androgen-responsive elements), and either wild-type AR or AR^T877A. The reporter gene activity was expressed as a ratio between the firefly and renilla luciferase luminescence (M1/M2). C: Growth kinetics revealed that T877A mutation in AR undermined the growth suppressive function of ARA70α.

Figure 7

ARA70α limits invasion potential of LNCaP cells independently of AR in Matrigel invasion assays. A: ARA70α greatly reduced the invasion ability of LNCaP cells. B: siRNA-mediated knockdown of AR in LNCaP-ARA70α cells did not alter the invasion ability. C: Androgen did not affect the invasion ability of LNCaP cells overexpressing ARA70α. LNCaP-pBabe or LNCaP-ARA70α cells, either treated with siRNA or untreated, were seeded into the Matrigel chamber by using fetal bovine serum as chemoattractant. Invading cells were counted and averaged from three fields after Diff-Quik stain.