A-MYB (MYBL1) stimulates murine testis-specific Ldhc expression via the cAMP-responsive element (CRE) site

Abstract

Generally, knowledge of the mechanism regulating gene expression in primary spermatocytes is incomplete. We have used the lactate dehydrogenase gene (Ldhc) as a model to explore these mechanisms during spermatogenesis. Its 100-bp core promoter contained two essential elements common to many genes, a GC box and a CRE site. Here we report results that support a model in which transcription factor MYBL1 acts as a coactivator directing tissue-specific expression via the CRE cis element. We hypothesize that this is a common mechanism involving activation of multiple genes in the primary spermatocyte. MYBL1 is expressed predominantly as a tissue-specific transcription factor in spermatocytes and breast epithelial cells. Our finding that LDHC expression is lost in 21-day testes of MYBL1 mutant mice supports our hypothesis. In the GC1-spg germ cell line exogenous MYBL1 induces activity 4- to 8-fold, although extracts from these cells do not show MYBL1 binding activity for the Myb consensus sequences in the Ldhc promoter by EMSA. Rather, MYBL1 stimulates expression from a synthetic promoter containing only CRE elements, suggesting MYBL1 activates the promoter by interacting with protein that binds to a CRE element. Mutation of three Myb sites does not affect Ldhc promoter activity significantly (P > 0.05). CREB-binding protein (CBP) is a coactivator that interacts with CRE-binding protein CREB. We show that the transactivation domain (TAD) in MYBL1 interacts with the KIX domain in CBP, and the TAD domain and DNA binding domain in MYBL1 each interact with the CREB N-terminal domain. MYBL1 also stimulated expression from testis-specific genes Pgk2 (phosphoglycerate kinase 2) and Pdha2 (pyruvate dehydrogenase alpha 2) promoters, each of which contains CRE promoter elements and is expressed in primary spermatocytes. We propose that MYBL1 directs germ cell-specific activation via the CRE site of certain genes that are activated specifically in the primary spermatocyte, although other, more indirect effects of MYBL1 remain a possible explanation for our results.

FIG. 1.

Sequence alignment: murine and human Ldhc gene promoters. Sequence segments containing conserved elements are displayed. The GC box, CRE sites, and TATA box are annotated by rectangles. Three Myb binding sites are labeled as Myb1, Myb2, and Myb3, but the consensus sequences for Myb1 and Myb3 are located in the complementary strand. The labels “Start” and “End” specify the mouse core promoter region. The only conserved fragment between the murine and human promoters contains a GC box and a CRE site.

FIG. 2.

MYBL1 regulates Ldhc gene expression. A) By Western blot, LDHC expression was not detected in 21-day male testis of MYBL1 mutant mouse. B) MYBL1 activates both mouse and human Ldhc promoters. Long (−425/+10) and short (−88/+10) murine and human promoter sequences (−502/+22) were tested for activation by cotransfection of promoter-reporter constructs and MYBL1 expression vector in a GC1-spg germ cell line. A corresponding 7-, 4-, and 8-fold increase with mouse long and short promoter and human promoter was observed in the MYBL1 treatment groups (**P < 0.01). Values are shown as the mean ± SEM.

FIG. 3.

Testis sections showing immunohistochemical distribution of MYBL1 and of a β-gal reporter transgene. In transgenic animals, the β-gal reporter was driven by a 100-bp Ldhc promoter. A positive cell is marked with an arrowhead. Both proteins are localized in the same cell type (spermatocytes). Consecutive testis sections from the transgenic animal were used for immunohistochemical analyses. Original magnification ×400.

FIG. 4.

Analysis of the DNA-protein interaction between three Myb consensus sequences in the Ldhc promoter and MYBL1 or truncated mutant (mutant1) proteins by electrophetic mobility shift assay (EMSA). A) Schematic depicting structure of wild-type MYBL1 and mutant1 construct; oligonucleotide sequences for EMSA shown below: DBD, DNA binding domain; TAD, transactivation domain; NRD, negative regulatory domain. Myb sites are underlined; mutated nucleotides were labeled with lowercase letters. B) EMSA showing DNA-protein interactions with the three Myb consensus sequences of the Ldhc promoter, MYBL1 and the truncated mutant1. Lanes 1–10 demonstrated a protein-DNA interaction between mutant1 and Myb consensus sequences; gel shifts in lanes 2, 5, 7, and 10 were marked with arrows or NS (nonspecific). Lanes 11–14 show the interaction for intact MYBL1 protein. Nuclear extracts were prepared from GC1-spg germ cells transfected with MYBL1 or mutant1. The intact MYBL1 protein did not yield a specific shifted band with the Myb probe from the Ldhc promoter, while mutant1 produced distinctive band shifts that could be competed off by Myb consensus but not mutated sequence.

FIG. 5.

A) Synergistic effects of coactivator CBP/p300 with MYBL1 and the interaction with Myb and CRE cis elements. Coactivator CBP/p300 enhances Ldhc promoter activity (mouse Ldhc −88/+12) when cotransfected with MYBL1. Ldhc promoter activity was significantly higher with CBP/p300 overexpression (**P < 0.01) in treatments 5 and 6 compared to treatments 1–4; activity in treatment 4 is significantly different than treatments 1–3 (**P < 0.01). B) A Ldhc promoter fragment containing CRE and Myb sites only (CRE-Myb). Ldhc was cloned into vector pGL4 with a 40-bp minimal promoter (miniP). The Myb sites are either wild type (Myb site WT) or mutated (Myb site Mu 2 or Mu 3; two Myb sites mutated, or all three Myb sites mutated). The promoter activity was significantly different between control and MYBL1 treatment (**P < 0.01) but not significantly different between the three constructs (WT, Myb site Mu 2, and Mu 3) when cotransfected with MYBL1 expression vector (P > 0.05). C) A synthetic promoter with three copies of Ldh CRE sites (3xCRE_ldhc) and a miniP was activated significantly by MYBL1 (**P < 0.01) with 7-fold increase of promoter activity. A similar construct with two nucleotides mutated in the CRE sites (3xCREmu) showed only background activity that was significantly lower than that of the (3xCRE_ldhc) with (**P < 0.01) or without MYBL1 (*P < 0.05). D) Dose-dependent upregulation by CBP on (3xCRE_ldhc) activity. endo, endogenous. Values are shown as the mean ± SEM, *P < 0.05, **P < 0.01.

FIG. 6.

Protein-protein interactions between MYBL1 and CBP and CREB with the two-hybrid assay system. A) Schematic depicting domain structure of CBP and CREB proteins. Numbers refer to amino acid position. B) Interactions of three MYBL1 domains with the CBP KIX domain. Vectors pDBD, pTAD, and pNRD contain the DBD, TAD, and NRD domain from MYBL1, while pKIX contains the KIX domain from CBP. Empty vector pACT and pBIND contains VP16 and GAL4 fusion protein and was used as control for transfection of GC1-spg germ cells. Interaction between TAD and KIX domain was detected (**P < 0.01). C) Interactions of MYBL1 and CREB domains. Vectors pCREB1-99, pKID, and pCREB161-327contain 3 CREB domains. The KID and CREB161-327 domains did not interact with any of MYBL1 domains (P > 0.05), but the CREB N-terminal CREB1-99 domain interacted with both DBD and TAD from MYBL1 (**P < 0.01) but not the NRD domain (P > 0.05). Values are shown as the mean ± SEM.

FIG. 7.

The 468-bp testis-specific Pgk2 promoter and 187-bp Pdha2 promoter were activated 3- to 4-fold when cotransfected with MYBL1 into GC1-spg cells. Both Pgk2 and Pdha2 promoters contain a CRE cis element. Values are shown as the mean ± SEM, **P < 0.01.