Delivery ofcholesterol to inner mitochondrial membranes is rate-limiting for steroidogenesis in the zona fasciculata of adrenal cortex. A protein that stimulates this process was isolated to homogeneity from bovine adrenal tissue. This protein's primary structure has been determined in its entirety by a combination of automated Edman microsequencing, fast-atom bombardment mass spectrometry (FAB-MS). The sequence was identical to that previously reported for bovine brain endozepine, except that it lacks the last two residues, -Gly-Ile, at the C terminus. To our knowledge, isolation of an endozepine-related protein from a tissue other than brain has not been reported previously. Endozepine competes with benzodiazepines for saturable binding sites in synaptosomes and in mitochondria of specific peripheral tissues. Previous reports have localized the adrenal benzodiazepine receptor to the outer mitochondrial membrane. In this report, we show that the prototypic benzoiazepine, diazepam, effects a stimulation of adrenal mitochondrial cholesterol delivery similar to that observed for endozepine. The effective diazepam concentration was consistent with that previously shown to displace a high-affinity ligand of the mitochondrial benzodiazepine receptor. The action of diazepam in adrenal mitochondria suggests that the mediation of corticotropininduced steroidogenesis may be the physiological function of the peripheral-type benzodiazepine receptor. These studies provide new insights into the previously unknown function of peripheral benzodiazepine receptors and should allow new investigations into the stimulation of steroidogenesis by endozepines and benzodiazepines in the brain and in certain peripheral tissues.

The amino acid sequence of human placental aromatase was determined in part (about 40%) by microsequencing methods. Using a region of overlapping peptide sequences, synthetic oligonucleotide probes were constructed and used to screen a human placental lambda gt-11 cDNA library. Of a number of positive clones, one containing a 2.4-kb insert was characterized further by restriction mapping and determination of its nucleotide sequence. The cDNA-deduced amino acid sequence is in perfect agreement with the peptide sequence data, confirming that the clone encodes for aromatase. The sequence contains a 3' untranslated region of 1.2 kb, and an open-reading frame of 1.25 kb; approximately 0.3 kb is missing from the 5' end of the coding region. While exhibiting no more than 20-30% sequence homology with other mammalian cytochromes P450, it contains the highly conserved heme-binding domain, thus confirming the essential structural requirements for this class of protein. Two cDNA fragments containing sequences coding for the amino- and carboxy-portions of the protein were used to probe for the human aromatase gene by Southern blotting. The results of these studies suggest the existence of at least two human aromatase genes. The gene encoding the aromatase cDNA we cloned was assigned to human chromosome 15 using somatic cell hybrids. This gene was mapped to band 15q21.1 by in situ hybridization studies.

The present study addresses the site occupancy and the site-specific carbohydrate microheterogeneity of N-linked oligosaccharides in recombinant human factor VIII, expressed in Chinese hamster ovary cells. The four factor VIIIa polypeptides, formed upon incubation with human thrombin, were isolated and separately subjected to proteolysis with trypsin. These tryptic digests were analyzed by reversed-phase high-performance liquid chromatography/electrospray ionization mass spectrometry. Selected ion monitoring of diagnostic carbohydrate ions was utilized to identify glycopeptide-containing chromatographic peaks. Oligomannose and complex carbohydrates were detected at the glycosylation sites of the 50 and the 73 kDa polypeptides, while all the oligosaccharides identified on the B-domain were complex-type structures. Only the 43 kDa polypeptide was found nonglycosylated. These studies established a biantennary core-fucosylated carbohydrate as the major substituent, consistent with the conclusions of the analyses on the entire N-linked carbohydrate pool (Kumar, H. P. M.; Hague, C.; Haley, T.; Starr, C. M.; Besman, M. J.; Lundblad, R.; Baker, D. Biotechnol. Appl. Biochem. 1996, 24, 207-216.). In addition, this mass spectrometric investigation revealed the presence of a complex nonfucosylated oligosaccharide not reported previously for this glycoprotein.

The following study was undertaken to compare the content of aromatase cytochrome P450 (P450arom) mRNA with the content of the enzyme in rat ovarian tissues and to relate these changes with estradiol biosynthesis by follicles and corpora lutea isolated throughout pregnancy. A deoxyoligonucleotide (62 mer) probe derived from an amino acid sequence of purified human placental P450arom was used to screen a rat granulosa cell lambda gt11 cDNA expression library. Seven cDNA clones, ranging in size from 0.6-2.0 kilobases (kb), were identified and plaque purified. In vitro translation using mRNA that had been selected by hybridization to a 1.2-kb rat P450arom cDNA insert yielded an 35S-labeled translation product that bound antihuman aromatase immunoglobulin and comigrated with purified human placental aromatase on sodium dodecyl sulfate-polyacrylamide gel electrophoresis, thus verifying that the clones do encode for P450arom. Using the 1.2-kb cDNA insert as a radiolabeled probe, the hormonal regulation, tissue distribution, content, and size of mRNA for P450arom were analyzed. Filter hybridization assays demonstrated that P450arom mRNA was low in small antral (SA) follicles, increased 16-fold in preovulatory (PO) follicles, and reached a peak in granulosa cells within 1 h after an ovulatory dose of hCG. In the corpus luteum of pregnancy, P450arom mRNA content was low on day 4, and increased 3-fold on days 7-11 and 10-fold on days 15-19 of gestation. P450arom mRNA then decreased on days 21 and 23, the day of parturition. Northern analyses of RNA from PO follicles and corpora lutea revealed three bands of P450arom mRNA that were 3.3, 2.6, and 1.9 kb in size. Immunoblots of soluble cell extracts of SA, PO, and luteinizing (PO plus hCG) follicles and corpora lutea of pregnancy demonstrated that aromatase enzyme was low in SA follicles, increased 1.5- to 3-fold in PO follicles, and decreased within 3-5 h after an ovulatory dose of hCG. Changes in the content of P450arom enzyme in luteal cells during pregnancy exhibited a pattern similar to that observed for P450arom mRNA. In contrast, changes in estradiol biosynthesis by follicles and corpora lutea were not directly related to the contents of P450arom mRNA and enzyme. For example, although corpora lutea isolated on days 15-21 of gestation contain the highest amount of P450arom mRNA and enzyme, these tissues did not produce the most estradiol when incubated for 5 h at 37 C in the presence of aromatizable androgen substrate.(ABSTRACT TRUNCATED AT 400 WORDS)