USF2 inhibits C/EBP-mediated transcriptional regulation of the RIIbeta subunit of cAMP-dependent protein kinase

Abstract

Background: Cyclic AMP-dependent protein kinase (PKA) plays a central role in regulation of energy metabolism. Upon stimulation of testicular Sertoli cells by follicle stimulating hormone (FSH), glycolysis is activated to increase the production of nutrients for the germ cells, and a new regulatory subunit of cAMP-dependent protein kinase, RIIbeta, is induced. We have previously shown that production of the transcription factor C/EBPbeta is rapidly increased by FSH and cAMP in primary Sertoli cell cultures, and that C/EBPbeta induces the RIIbeta promoter.

Results: In this work we show that USF1, USF2 and truncated USF isoforms bind to a conserved E-box in the RIIbeta gene. Interestingly, overexpression of USF2, but not USF1, led to inhibition of both cAMP- and C/EBPbeta-mediated induction of RIIbeta. Furthermore, Western blots show that a novel USF1 isoform is induced by cAMP in Sertoli cells.

Conclusions: These results indicate that the expression of various USF isoforms may be regulated by cAMP, and that the interplay between USF and C/EBPbeta is important for cAMP-mediated regulation of RIIbeta expression. The counteracting effects of USF2 and C/EBPbeta observed on the RIIbeta promoter is in accordance with the hypothesis that C/EBP and USF play opposite roles in regulation of glucose metabolism.

Figure 1

USF isoforms bind to the RIIβ E-box. A [³²P]-labelled oligonucleotide from the RIIβ promoter region (-305 to -268) was incubated with Sertoli cell nuclear extracts and subjected to EMSA. Supershift experiments were performed with 2 μl of antiserum against USF1 N-terminal region (lane 2), USF2a C-terminal sequences (lane 3), or USF2 N-terminal sequences (lane 4).

Figure 2

A novel USF1 isoform is induced by cAMP. Immunoblotting was performed with antibodies specific for the C-terminal regions of USF1 (Panel A) or USF2 (Panel B) on stimulated (+, 100 μM 8-CPT-cAMP, 28 h) or unstimulated (-) nuclear extracts. 20 μg of nuclear exracts was loaded in each lane, and equal loading was determined by coomassie staining.

Figure 3

USF 2A inhibits cAMP-mediated induction of RIIβ. A reporter construct containing the region (-395 to -123) from the RIIβ promoter was transfected into Sertoli cell primary cultures together with expression vectors for USF1, USF2a and USF2b isoforms or the corresponding empty expression vector. Empty expression vector was added to ensure a total of 2 μg of DNA transfected in each well. The cells were stimulated for 28 h with 100 μM of 8-CPT-cAMP (black and grey bars) or left untreated (open bars). Reporter activity is shown relative to the cAMP-stimulated level that is set to 100 (black bar). Data are normalized for expression of luciferase from the internal control plasmid. Panel A: Reporter expression levels in unstimulated (open bar) and stimulated cells in the absence (black bar) or presence of expression vectors for USF1 (1), USF2a (2a) or USF2b (2b)(grey bars). Panel B: Reporter levels in the absence (black bar) or presence of different concentrations of USF2a expression vector (0 to 750 ng, grey bars). Three separate transfections were performed in triplicate.

Figure 4

USF2 inhibits induction of the RIIβ promoter by C/EBP. The RIIβ promoter construct (-395 to -123) in the CAT reporter vector was co-transfected with 500 ng of expression vectors for USF1 (1), USF2a (2A) or C/EBPβ Lap (horisontally striped bars) or Lip (hached bars) isoforms or combinations of these factors. Empty expression vector was added to ensure a total of 2 μg of DNA transfected in each well. Data represent reporter activities (CAT/Luc) relative to transfection with the RIIβ promoter construct under unstimulated conditions, and were normalized for expression of luciferase from a cotransfected control vector (pGL3Control). Three separate transfections were performed in triplicate.