Identification of DNA Sequences that are Necessary and Sufficient for Activin Induction of Ovine Follicle Stimulating Hormone Beta-Subunit.

Abstract

Follicle stimulating hormone (FSH) is an alpha/beta heterodimer central to vertebrate reproduction, and its synthesis depends on expression of its beta-subunit (FSHB). Activin is the primary inducer of FSHB, and previous work from this laboratory showed that induction depends on sequences between -169/-158 bp of the ovine FSHB promoter. Deletion and promoter substitution studies suggested, however, that other 5’ and 3’ sequences might also alter activin induction. The work described here was designed to identify all sequences required for activin induction of ovine FSHB. These studies used mutant or wild type ovine FSHB promoter/reporter constructs (wild type = oFHSBLuc; -4741 bp of 5’ promoter plus 3’ exon/intron 1 linked to luciferase) which were analyzed using transient expression in transformed gonadotropes (LBT2 cells) plus expression in transgenic mice in some cases. First, eleven successive 5’ deletions were made to -195 bp which progressively decreased induction by activin from 9.5-fold to 1-fold (no induction). Also five more deletions internal to this region were made to -175 bp, but these deletions were replaced with exogenous DNA to maintain the original spacing. Also 3’ exon/intron deletions with replacement sequences were made. When correct spacing was maintained, none of the changes in the 5’ or 3’ exon/intron regions altered induction of FSHB by activin. Deletions between -90 bp and the TATA box revealed one important site (-68/-58 bp) that was necessary for normal basal expression and activin induction in LBT2 cells. This region contained a Pitx1 binding site and partially overlapping putative Runx binding site. Follow-up transgenic studies showed that this Pitx1 site was responsible for 99 % of oFSHBLuc expression in vivo, but had no effect on activin action. The putative Runx binding site had no effect on oFSHBLuc expression in vivo, but significantly degraded gonadotrope-specific expression of oFSHBLuc. Finally, to determine if the oFSHBLuc TATA box was important for activin induction, it was replaced with a minimal rat prolactin promoter (contains a TATA box), and full activin induction was maintained. Substitution with a thymidine kinase minimal promoter (no TATA box) prevented activin induction. IN SUMMARY, these studies plus previous results from our laboratory show that sequences between -169/-58 bp plus a minimal TATA box promoter are necessary and sufficient for robust activin-induced expression of oFSHBLuc in LBT2 cells. They also show the Pitx1 site between -68/-63 bp is necessary for 99 % of ovine FSHB expression in vivo but has no effect on activin induction. Finally, the putative Runx site that overlaps Pitx1 was found important for gonadotrope-specific expression of oFSHBLuc.