Distinct transcriptional mechanisms direct expression of the rat Dmrt1 promoter in sertoli cells and germ cells of transgenic mice

Abstract

DMRT1 is a transcription factor expressed only in Sertoli cells and undifferentiated spermatogonia of the postnatal testis, where it is required for proper cellular differentiation and fertility. To elucidate the transcriptional regulatory regions that provide DMRT1's cell-specific expression, transgenic mice containing a LacZ reporter gene driven by variable amounts of rat Dmrt1 5' flanking sequence, 9 kb and smaller, were evaluated. Examination of transgene expression by RT-PCR indicated that multiple promoter regions direct Dmrt1 to the testis and that sequences upstream of 2.8 kb are needed for both Sertoli cell expression and limiting transcriptional influence imposed by surrounding chromatin. Thus, whereas many of the transgenes were expressed in the testis, the ones with smaller promoters were significantly more prone to expression at ectopic sites or to complete silencing. Transgene expression in Sertoli cells and germ cells was assessed by immunohistochemistry and RT-PCR following busulfan treatment to remove germ cells. Both evaluations indicated expression of the 9- and 3.2-kb promoters in Sertoli cells and germ cells, whereas activity of smaller promoters was largely restricted to germ cells. In all, the present study provides in vivo evidence that distinct promoter sequences participate in Dmrt1 regulation in somatic cells and germ cells, with the -3.2 kb/-2.8 kb region directing expression in Sertoli cells and downstream sequences (< or =1.3 kb) directing it in germ cells. Further exploration of the mechanisms restricting Dmrt1 expression to the testis revealed that FOXL2, a transcription factor required for differentiation of the ovary, repressed Dmrt1 promoter through the -3.2 kb/-2.8 kb regulatory region, offering a potential mechanism for Dmrt1 transcriptional silencing in granulosa cells.

FIG. 1.

Schematic representation of Dmrt1-LacZ transgenes. Dmrt1-LacZ transgenes contain the indicated LacZ/Hprt1 reporter cassette driven by various lengths of rat Dmrt1 5′ flanking sequence, denoted on the figure as −150bp, −400bp, −1.3kb, −2.8kb, −3.2kb, and −9kb, that ends 74 bp (+74) 3′ to the transcriptional start site (bent arrow). Hybridization positions and amplification directions for the RT-PCR primers LacZ3 and Hprt1 are indicated by arrows below the depicted transgene.

FIG. 2.

Observed profiles for Dmrt1-LacZ transgenes. RNA extracted from tissues derived from 15-day-old transgenic mice was evaluated for transgene expression by RT-PCR. Complementary DNA synthesis was confirmed by evaluating cDNA amplification of the ribosomal protein L7, and if required, templates for LacZ amplification were adjusted to equalize template levels based on that of Rpl7. RT-PCR was performed in the presence (left lane) and absence (right lane) of added reverse transcriptase, and amplified products were examined by agarose gel electrophoresis (product of LacZ mRNA is 757 bp; genomic DNA is 1412 bp). Representative expression profiles are shown for each transgenic construct, with the promoter length denoted on the right together with the evaluated transgenic line number in parenthesis. For transgenic constructs that exhibited more than one expression pattern, representative profiles of each distinct pattern are shown. B, brain; L, liver; h, heart; Ln, lung; S, spleen; K, kidney; St, stomach; T, testis; G, genomic DNA from positive transgenic animal; +, positive control (cDNA generated from testis of Dmrt1 9kb-LacZ transgenic mice). Results for all transgenic lines are summarized in Table 1, which represents data derived from two or more independent PCRs from a minimum of three animals for each line of a specific transgene.

FIG. 3.

Cell-specificity of Dmrt1-LacZ transgenes. Transgene expression in specific testis cell populations was evaluated in testes from 15-day-old transgenic mice born to mothers treated with busulfan (+Busulfan) or without busulfan (−Busulfan) to eliminate germ cells. Efficacy of the busulfan treatment was evaluated by DMRT1 immunohistochemistry (red) to assess the DMRT1-expressing germ cell population (A and B). DMRT1-positive (arrow) and -negative (arrowhead) germ cells are indicated (B, inset). Note the paucity of germ cells in the busulfan-treated group (A). Transgene expression was evaluated in testes from both treatment groups by both RT-PCR (C) and immunohistochemistry for β-galactosidase (D). RT-PCR analysis (+RT, −RT) was used to evaluate testis expression of transgenic (T+) and wild type (T−) littermates that represent a total of five different transgenic lines (Dmrt1–2.8kb-LacZ, lines 49 and 59; Dmrt1–3.2kb-LacZ, lines 31 and 20 and data not shown; Dmrt1 9kb-LacZ, line 35). β-Galactosidase expression in testes from males with 9- and 3.2-kb transgenes (denoted to left) and born to mothers of both treatment groups (+Busulfan, right; -Busulfan, left) was evaluated by immunohistochemistry (D). β-Galactosidase expression in a wild-type testis from the treated group (+Busulfan) and in a 1.3-kb transgenic testis from the untreated group (−Busulfan) also are shown. Original magnification ×4 (A and B), ×20 (insets in A and B), and ×40 (D).

FIG. 4.

FOXL2 represses Dmrt1 promoter activity via the distal regulatory region. Depicted is the promoter sequence encompassing the distal regulatory region from −3.2 kb to −2.8 kb of rat Dmrt1 (A). Arrows mark sequences matching the consensus binding site for FOXL2, and their direction indicates the putative element's orientation [–30]. Transient transfection analysis of primary rat Sertoli cells was used to evaluate activities of the 3.2-kb [black bars, Dmrt1(−3280/+75)Luc] and 2.8-kb [gray bars, Dmrt1(−2800/+75)Luc] rat Dmrt1 promoters in the presence of increasing amounts of FOXL2 expression vector (B). Graphed is the ratio of firefly luciferase to Renilla luciferase (RL-TK control) activity, for the indicated promoter and dose of transfected Foxl2 vector, relative to the same ratio (firefly/Renilla) for pGL3-control transfected with the same amount of Foxl2 vector. Transfections were done a minimum of three times, and error bars represent the SEM (*P = 0.0605, by Student t-test: two samples assuming unequal variances).