Sp1 transcription factor and GATA1 cis-acting elements modulate testis-specific expression of mouse cyclin A1

Abstract

Cyclin A1 is a male germ cell-specific cell cycle regulator that is essential for spermatogenesis. It is unique among the cyclins by virtue of its highly restricted expression in vivo, being present in pachytene and diplotene spermatocytes and not in earlier or later stages of spermatogenesis. To begin to understand the molecular mechanisms responsible for this narrow window of expression of the mouse cyclin A1 (Ccna1) gene, we carried out a detailed analysis of its promoter. We defined a 170-bp region within the promoter and showed that it is involved in repression of Ccna1 in cultured cells. Within this region we identified known cis-acting transcription factor binding sequences, including an Sp1-binding site and two GATA1-binding sites. Neither Sp1 nor GATA1 is expressed in pachytene spermatocytes and later stages of germ cell differentiation. Sp1 is readily detected at earlier stages of spermatogenesis. Site-directed mutagenesis demonstrated that neither factor alone was sufficient to significantly repress expression driven by the Ccna1 promoter, while concurrent binding of Sp1, and most likely GATA1 and possibly additional factors was inhibitory. Occupancy of Sp1 on the Ccna1 promoter and influence of GATA1-dependent cis-acting elements was confirmed by ChIP analysis in cell lines and most importantly, in spermatogonia. In contrast with many other testis-specific genes, the CpG island methylation status of the Ccna1 promoter was similar among various tissues examined, irrespective of whether Ccna1 was transcriptionally active, suggesting that this regulatory mechanism is not involved in the restricted expression of Ccna1.

Figure 1. 5′ deletion analysis of the Ccna1 promoter reveals a repressor element upstream of -120 bp with respect to the transcription start site (TSS).

Chimeric constructs of the Ccna1 promoter-luciferase reporter are schematically represented on the category-axis. The upstream and downstream endpoints of each fragment are indicated with respect to the TSS, which is indicated by a bent arrow. Constructs were tested for transcriptional activity in GC-4spc and NIH3T3 cells as described in the Methods section. The luciferase activity obtained for the pGL-Basic vector, in which a functional promoter was absent, was set at 1.0 and all other luciferase activities were expressed as fold of the values obtained for pGL-Basic. All experiments were performed in quadruplicate, and the means ± standard deviations of luciferase expression from independent experiments are shown. In all experiments the luciferase activity of pRL-CMV, Renilla luciferase expressing under CMV promoter was used as an internal control.

Figure 2. Identification of transcription factor consensus sequences in the Ccna1 promoter.

The −290/−100 region was screened for binding sites using TRANSFAC, a mammalian transcription factor databas, using a cut-off score of 85. Consensus sequences are underlined with dashed arrows and GC boxes are enclosed within rectangles.

Figure 3. The minimum region of the Ccna1 promoter necessary for protein binding resides within positions

−290/−120. (A) EMSA were performed in nuclear extracts (NE) from pnd10 testis and NIH3T3 and GC-4spc cells incubated with a DNA probe of the −290/−120 bp region of the of Ccna1 promoter. “No comp” designates lanes without cold competitor. For competition assays lysates were incubated with 100-fold excess of the indicated unlabelled DNA prior to addition of the labeled −290/−120 probe. Formation of DNA-protein complexes in reactions with labeled −290/−120 probe alone was abolished completely by the addition of the same unlabeled competitor (lanes −290/−120). Oligo spanning the −200/−120bp region, ds oligo containing GATA1 binding site or an oligo containing Sp1 binding site did not or only partially compete with the −290/−120 probe for binding to transcription factors. Arrows indicate protein-DNA complexes that lead to the electrophoretic mobility shift. (B) DNA probes from the Ccna 1 promoter (positions −290 to −120) were incubated with nuclear extract (NE) from pnd10 testis and GC-4spc cells. Addition of anti-Sp1 antibody is indicated with a “+”. Arrows point to supershift products of the Sp1-DNA probe complex bound to the antibody.

Figure 4. Sp1 and possibly GATA1 bind concomitantly to the Ccna1 promoter and suppress its activity in vivo.

(A) EMSA assay in nuclear extracts from indicated cell lines using radioactively labeled probe −290/−120 bp of the Ccna 1 promoter shows specific DNA-protein complex formation (arrows). Complex formation was significantly reduced by the addition of 100-fold molar excess of unlabeled DNA competitor containing an intact Sp1 site while cold competitor containing mutant Sp1 recognition site failed to disrupt the binding. (B) Firefly luciferase reporter assay reveals binding to both Sp1 and GATA1 sites suppresses activation of the Ccna1 promoter (constructs shown schematically to the left). Sp1 and GATA1 binding sites were marked as an open ellipse and black diamond, respectively. Crosses (“X”) indicate mutated binding sites. As an internal control, pRL-CMV expressing Renilla luciferase under the CMV promoter was co-transfected in each assay. Bars show the average luciferase activity of three independent experiments, normalized with respect to the Renilla luciferase activity.

Figure 5. Suppression of SP1 and GATA1 expression leads to up-regulation of Ccna1 promoter activity.

The reporter constructs used for transfection are shown schematically on the left. Sp1 and GATA1 binding sites are marked with white ellipses and black diamonds, respectively. Reporter plasmids and corresponding siRNA oligonucleotides (100 µM) were co-transfected into NIH3T3 and GC-4spc cells. In all experiments pRL-CMV was co-transfected as an internal control. Bars show the average luciferase activity of five independent experiments, normalized with respect to the Renilla luciferase activity.

Figure 6. Identification of protein complexes by ChIP analysis demonstrates cell-specific binding of Sp1 and active RNA polymerase II to the Ccna1 promoter as well as an enrichment of H3K27me3.

Antibodies to Sp1, H3K27me3 or the corresponding IgG control were used to immunoprecipitate protein/DNA complexes from sonicated lysates of indicated cells or a tissue equivalent to 10⁶ cells in each case. After reversing the cross-linking, bound DNA was isolated and PCR was performed using primers that amplify the −290 to -100 bp region of the Ccna1 promoter to detect bound Sp1 and H3K27me3 and the –120 to +90 bp region of the Ccna1 promoter to detect RNA polymerase II binding.

Figure 7. Tissue-specific Ccna1 expression is not modulated by differential promoter methylation.

(A) Gross methylation analysis of the Ccna1 promoter in tissues and cell lines as follows: mouse testis (T) at the indicated age, kidney (K), liver (L), brain (B), heart (H), pancreas (P), purified pachytene cells (Pac), round spermatids (Rnd) and cell lines GC-4spc (GC), NIH3T3 (NIH). For each analysis, 500 ng of total genomic DNA was divided into four and assigned to mock control, methylation-sensitive, methylation-dependent and double digestion respectively. After digestion with methylation sensitive and insensitive nucleases, the methylation profile was obtained from the Ct values of real-time PCR with Ccna1-CpG methylation specific primers. (B) The methylation status of individual CpG residues on the Ccna1 promoter was analyzed by bisulfite sequencing. One µg of DNA from the indicated samples was subjected to bisulfite conversion followed by amplification with primers specific for the Ccna1 CpG island. PCR products were cloned into a pGEMTeasy vector, sequenced, and sequences were analyzed with QUMA software. In vitro CpG methylated NIH3T3 DNA was used as a positive control and a PCR amplified DNA from unconverted sequence was used as a negative control. Non-methylated CpG are indicated as open circles. The Ccna1 promoter with CpG positions is depicted schematically.