Identification of aromatase activity in rodent pituitary cell strains.

To date, biochemical evidence has been presented for hypophysial aromatization in only one species, a teleost fish, although the pituitary glands of several mammals have been reported to be aromatase negative. To reinvestigate this problem, established clonal strains of rodent pituitary cells (GH3, GH4C1, and AtT20/D16) were incubated at 37 C for 6-48 h in serum-less medium containing [7-3H]androstenedione. Radiolabeled metabolites were isolated by solvent extraction, thin layer chromatography, and phenolic partition. The authenticity of the estrogenic products in both cells and incubation medium was verified by methylation and recrystallization to constant specific activity. Measurement of androgen metabolites was also validated by recrystallization of selected samples. Authentic estrone and 17 beta-estradiol were identified in cultures of the two PRL- and GH-secreting clones, and there were strain differences in the quantity of estrogen produced (GH3 greater than GH4C1). Under the same conditions, aromatization was not detectable in the ACTH-secreting line (AtT20/D16). A time-yield analysis of androgen metabolism in GH4C1 cells showed that aromatization was linear for 12 h after labeling, but that substrate was diverted mainly to 5 alpha-reducing pathways. Large amounts of highly polar metabolites accumulated 24 and 48 h after the addition of [3H]androgen, and subsequent hydrolysis revealed that these were sulfo- and glucuronoconjugates. The metabolic fate of estrogen in GH4C1 cultures was investigated indirectly by adding a radioinert estrone trap together with the radiolabeled androgen substrate and was also tested in separate cultures by adding [3H]estrone and [3H]estradiol directly. Although the two estrogens were interconverted, there was no evidence that formed or added estrogen was extensively metabolized or conjugated. We conclude that the expression of aromatase activity in hypophysial cells is not a property of all transformed lines but may be dictated by the secretory cell type. Although low relative to other metabolites, estrogen yields in cultured GH cells resemble production in primary cultures derived from other tissues known to be estrogen targets, including the hypothalamus.