The porcine SRY promoter is transactivated within a male genital ridge environment

In mammals the SRY gene functions as a dominant genetic switch for testis determination (Gubbay et al.: Nature 346:1128-1135, 1990; Koopman et al.: Nature 351:117-121, 1991; Sinclair et al.: Nature 346:240-244, 1990). To study SRY transcriptional regulation within an evolutionary context, we have generated transgenic mice that express green fluorescent protein (GFP) under the control of 4.5 kb of pig SRY 5' flanking sequences (pSRYp-GFP). Autofluorescence was observed in the genital ridges of e11.5 male embryos (18-21 tail somites), and by e12.5 (27 tail somites) autofluorescence was observed within the testes cords. The expression of the transgene did not display the abrupt termination characteristic of endogenous mouse SRY, but rather showed a gradual reduction in expression characteristic of human, pig and sheep SRY. Surprisingly, no autofluorescence was observed in normal XX genital ridges, although more sensitive RT-PCR analysis detected transgene transcription. When the transgene was bred into a constitutively male line of mice (Odsex; Bishop et al.: Nat Genet 26:490-494, 2000), autofluorescence was visible in genital ridges of XX animals, in the genetic absence of Sry protein. Via RT-PCR analysis, purified autofluorescent cells from e12.5 gonadal ridges expressed mouse SRY but not Oct4 transcripts, whereas autofluorescent cells from e14.5 gonadal ridges expressed MIS but not Oct4 transcripts, in each case consistent with a pre-Sertoli cell phenotype. In vitro expression studies performed in CV-1 cells demonstrated that pig SOX9 cDNA transactivated the pig SRY promoter but that pig SRY cDNA did not. When a SOX9 potential binding site identified at -205 of the pig SRY 5' flanking sequences was mutated, the SOX9 transactivation effect was reduced by 70%. This site is conserved in the 5' flanking sequences of bovine and human SRY genes but not in the mouse gene. Gel retardation assays using this binding site showed specific binding to SOX9-enriched nuclear extracts that was competed by excess unlabelled binding site but not by mutated binding site. We suggest that pig SRY gene is responsive to a testicular environment and propose a model of feedback amplification of pig SRY transcription by SOX9.