Isolation and characterization of a complementary deoxyribonucleic acid insert encoding bovine aromatase cytochrome P450

Abstract

Aromatase, an enzyme complex comprised of aromatase cytochrome P450 (P450arom; the product of the CYP19 gene) and the flavoprotein NADPH-cytochrome P450 reductase, catalyzes the conversion of androgens to estrogens. Three cDNA inserts encoding P450arom were isolated from a bovine placental cDNA library. These inserts were sequenced and found to correspond closely to human P450arom sequence from the internal EcoRI restriction site (exon III) through the termination codon (exon X) into the 3'-untranslated region. The rapid amplification of cDNA ends technique was used to generate the rest of the cDNA 5' of the internal EcoRI site, using mRNA obtained from bovine granulosa cells as a template. This insert was sequenced, and when aligned with the other inserts, an open reading frame was found which was predicted to encode a protein of 503 amino acid residues. The deduced polypeptide shares 84% identity with human P450arom and 79%, 76%, 71%, and 57% identity with mouse, rat, chicken, and trout P450arom, respectively. A full-length open reading frame was generated using the polymerase chain reaction and mRNA obtained from bovine granulosa cells as template. After this insert was ligated into the pCMV5 expression vector, it was transfected into COS-1 monkey kidney tumor cells. We were able to demonstrate aromatase activity by assaying the incorporation of tritium into [3H] water from [1 beta-3H]androstenedione. Northern analysis revealed a single transcript of approximately 6 kilobases in poly(A)+ RNA obtained from bovine placental tissue and granulosa cells. This indicated for the first time a correspondence between the pattern of estrogen biosynthesis throughout the bovine ovarian cycle and the levels of transcripts encoding P450arom. In addition, weak hybridization was noted to transcripts of the expected size, namely 3.4 and 2.9 kilobases, in poly(A)+ RNA obtained from human placental tissue. The large size of the bovine transcript is due to a long 3'-untranslated region, because, based on the rapid amplification of cDNA ends technique, there appeared to be approximately 150 basepairs 5' of the start site of translation, and we were never able to find a polyadenylation site, even in one clone that went well past the corresponding polyadenylation site in human P450arom.