Androgen-induced Rhox homeobox genes modulate the expression of AR-regulated genes

Abstract

Rhox5, the founding member of the reproductive homeobox on the X chromosome (Rhox) gene cluster, encodes a homeodomain-containing transcription factor that is selectively expressed in Sertoli cells, where it promotes the survival of male germ cells. To identify Rhox5-regulated genes, we generated 15P-1 Sertoli cell clones expressing physiological levels of Rhox5 from a stably transfected expression vector. Microarray analysis identified many genes altered in expression in response to Rhox5, including those encoding proteins controlling cell cycle regulation, apoptosis, metabolism, and cell-cell interactions. Fifteen of these Rhox5-regulated genes were chosen for further analysis. Analysis of Rhox5-null male mice indicated that at least nine of these are Rhox5-regulated in the testes in vivo. Many of them have distinct postnatal expression patterns and are regulated by Rhox5 at different postnatal time points. Most of them are expressed in Sertoli cells, indicating that they are candidates to be directly regulated by Rhox5. Transfection analysis with expression vectors encoding different mouse and human Rhox family members revealed that the regulatory response of a subset of these Rhox5-regulated genes is both conserved and redundant. Given that Rhox5 depends on androgen receptor (AR) for expression in Sertoli cells, we examined whether some Rhox5-regulated genes are also regulated by AR. We provide several lines of evidence that this is the case, leading us to propose that RHOX5 serves as a key intermediate transcription factor that directs some of the actions of AR in the testes.

Figure 1

Generation of 15P-1 Sertoli cell clones expressing physiological levels of Rhox5. A, Western blot analysis of RHOX5 protein expression in 15P-1 stable clones stably transfected with a RHOX5-expression vector. SL12.4 cells served as positive control for RHOX5 expression. After probing with RHOX5 antiserum, the blot was stripped and reprobed with a β-actin antiserum as an internal control for loading. B, qPCR analysis of total cellular RNA from the 15P-1 Sertoli cell clones. Rhox5 mRNA expression levels in the 15P-1 cell clones relative to parental 15P-1 cells (which was arbitrarily given a value of 1), all normalized against ribosomal protein L19 transcript levels. Adult testis served as a positive control. Values denote the mean fold change ± sem.

Figure 2

Genes regulated by Rhox5 in the adult testis. qPCR analysis of total cellular RNA from adult Rhox5-null (KO) and wild-type (WT) littermate testes. mRNA levels were quantified as described in Fig. 1B. Statistical analysis was performed by use the two-tailed Student’s t test: *, P < 0.05; **, P < 0.01.

Figure 3

Cell subset expression pattern of Rhox5-regulated genes. qPCR analysis of total cellular RNA from purified Sertoli and interstitial cells (prepared by use of hypotonic shock to remove germ cells) from adult Rhox5-null (KO) and wild-type (WT) littermate mice testes. The value shown for each gene is the average WT/KO mRNA ratio from two independent cell preparations, quantified as described in Fig. 1B. a, mRNA ratios of either ≥2 or ≤0.5 (i.e. >2-fold change).

Figure 4

Developmental expression pattern of Rhox5-regulated genes. qPCR analysis of testes RNA from Rhox5-null (KO) and wild-type (WT) littermate mice of the indicated postnatal ages (six mice per time point). All values were quantified as described in Fig. 1B. Postnatal time points at which Rhox5 regulates the indicated genes are denoted by arrows (upward and downward arrows indicate genes that are up-regulated and down-regulated, respectively). A value of 1 was arbitrarily given to the lowest expression level among the time points examined for a given gene. Error bars indicate standard error.

Figure 5

Identification of Ar- and Rhox5-regulated genes. A, qPCR analysis of testes RNA from adult Tfm and littermate control mice. B, qPCR analysis of testes RNA from adult Scarko and littermate control mice. All values were quantified as described in Fig. 1B. Control mRNA levels were arbitrarily given a value of 1. Statistical analysis was performed by use the two-tailed Student’s t test: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 6

The candidate SAR gene Unc5c responds to Rhox5 but not Ar and androgen in vitro. A, Schematic diagram of Unc5c promoter and Rhox5 proximal promoter (Pp) constructs harboring the firefly luciferase (Luc) reporter. B, Luc expression from MSC1 cells transiently cotransfected with pGL3-Unc5c (100 ng), a Renilla luciferase vector (pRL-Tk) (25 ng) for an internal control, and either a Rhox5-expression vector or the corresponding empty expression vector (300 ng). C, Luc expression from MSC1 cells transiently transfected with the constructs shown in panel A (100 ng) and pRL-Tk (25 ng). Where indicated with +AR, the cells were cotransfected with an AR expression vector (100 ng). Where indicated with +T, the cells were incubated with the synthetic androgen R1881. The relative Luc activities shown are relative to nontreated cells (which were arbitrarily given a value of 1) from three independent transfection experiments. Error bars indicate standard error. Note that AR + T did not induce endogenous Rhox5 expression in these cells (data not shown), explaining why Unc5c expression was not regulated.

Figure 7

Regulation of candidate SAR genes by Ar-inducible Rhox genes. A, qPCR analysis of total cellular RNA from 15P-1 Sertoli cells transiently transfected with expression vectors encoding the indicated mouse Rhox genes or the empty expression vector (800 ng). A value of 1 was arbitrarily given to the expression of the indicated genes in cells transfected with the empty expression vector. All values were quantified as described in Fig. 1B. B, Same as panel A, except the cells were transiently transfected with expression vectors encoding the human RHOXF1 and RHOXF2 proteins. Statistical analysis was performed by using the two-tailed student’s t-test (*, P < 0.05).