ABSTRACT: Cytochrome P450c17 (P450c17) is the single enzyme that catalyzes steroid 17alpha-hydroxylase and 17,20 lyase activities and hence is the crucial decision-making step that determines the class of steroid made in a steroidogenic cell. Although both activities are catalyzed on a single active site, the ratio of these activities is regulated by posttranslational events. Serine phosphorylation of P450c17 increases 17,20 lyase activity by increasing the enzyme's affinity for its redox partner, P450 oxidoreductase. We searched for the relevant kinase(s) that phosphorylates P450c17 by microarray studies and by testing of kinase inhibitors. Microarrays show that 145 of the 278 known serine/threonine kinases are expressed in human adrenal NCI-H295A cells, only six of which were induced more than 2-fold by treatment with 8-Br-cAMP. Key components of the ERK1/2 and MAPK/ERK kinase (MEK)1/2 pathways, which have been implicated in the insulin resistance of PCOS, were not found in NCI-H295A cells, implying that these pathways do not participate in P450c17 phosphorylation. Treatment with various kinase inhibitors that probe the protein kinase A/phosphatidylinositol 3-kinase/Akt pathway and the calcium/calmodulin/MAPK kinase pathway had no effect on the ratio of 17,20 lyase activity to 17alpha-hydroxylase activity, appearing to eliminate these pathways as candidates leading to the phosphorylation of P450c17. Two inhibitors that target the Rho-associated, coiled-coil containing protein kinase (ROCK)/Rho pathway suppressed 17,20 lyase activity and P450c17 phosphorylation, both in NCI-H295A cells and in COS-1 cells transfected with a P450c17 expression vector. ROCK1 phosphorylated P450c17 in vitro, but that phosphorylation did not affect 17,20 lyase activity. We conclude that members of the ROCK/Rho pathway act upstream from the kinase that phosphorylates P450c17 in a fashion that augments 17,20 lyase activity, possibly acting to catalyze a priming phosphorylation. NCI-H295A adrenocortical cells were grown on 10-cm plates in RPMI 1640 with 2% fetal calf serum, gentamicin, insulin, and selenium. Three nearly confluent plates were incubated in the presence and absence of 1 mM 8-Br-cAMP for 18 h; the cells were washed with PBS and lysed with 1 ml Trizol (Invitrogen, Carlsbad, CA). Total RNA was purified further using RNeasy columns (QIAGEN, Valencia, CA), and 4 ?g RNA was used to synthesize double-stranded cDNA using the MessageAmp II-Biotin Enhanced Kit (Ambion, Austin, TX) with oligo-dT primer [5?-GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG-(dT)24] to incorporate a T7 RNA polymerase promoter. Antisense cRNA was labeled with biotin-UTP using T7 RNA polymerase (Ambion), and the cRNA was fragmented to 35–200 nucleotides (nt) at 94 C for 35 min in 40 ?l 40 mm Tris-acetate (pH 8.1), 100 mm KOAc, 30 mm MgOAc. The integrities of total RNAs, cRNAs, and fragmented cRNAs were assessed on a Bio-Rad (Hercules, CA) Experion BioAnalyzer. Each sample of fragmented cRNA was initially hybridized with Affymetrix (Santa Clara, CA) Test3 arrays to determine the quality of the labeled probe before hybridization to Affymetrix U133A plus 2.0 arrays. The chips were stained and washed using an Affymetrix GeneChip Fluidics Station 450, and the arrays were scanned using a GCS 3000 scanner with auto-loader in the Genomics Core laboratory. Hybridization intensity signals from the scanned images were analyzed using Affymetrix GeneChip Operating Software (GCOS, version 1.3) to provide qualitative detection calls and quantitative estimates of gene expression levels (signal values). Raw data were exported to Microsoft Excel for analysis. Transcripts displaying a cAMP-induced increased or decreased signal were selected for further analysis only if they had a mean ± 2.0-fold change and were statistically significant (P < 0.05) in three separate experiments; transcripts displaying no signal change relative to controls in at least three experiments were excluded from further analysis.
