Regulation of 17,20 Lyase Activity by Cytochrome b5 and by Serine Phosphorylation of P450c17*

Cytochrome P450c17 catalyzes the 17α-hydroxylase activity required for glucocorticoid synthesis and the 17,20 lyase activity required for sex steroid synthesis. Most P450 enzymes have fixed ratios of their various activities, but the ratio of these two activities of P450c17 is regulated post-translationally. We have shown that serine phosphorylation of P450c17 and the allosteric action of cytochrome b5 increase 17,20 lyase activity, but it has not been apparent whether these two post-translational mechanisms interact. Using purified enzyme systems, we now show that the actions of cytochrome b5 are independent of the state of P450c17 phosphorylation. Suppressing cytochrome b5 expression in human adrenal NCI-H295A cells by >85% with RNA interference had no effect on 17α-hydroxylase activity but reduced 17,20 lyase activity by 30%. Increasing P450c17 phosphorylation could compensate for this reduced activity. When expressed in bacteria, human P450c17 required either cytochrome b5 or phosphorylation for 17,20 lyase activity. The combination of cytochrome b5 and phosphorylation was not additive. Cytochrome b5 and phosphorylation enhance 17,20 lyase activity independently of each other, probably by increasing the interaction between P450c17 and NADPH-cytochrome P450 oxidoreductase.

Cytochrome P450c17 catalyzes the 17␣-hydroxylase activity required for glucocorticoid synthesis and the 17,20 lyase activity required for sex steroid synthesis. Most P450 enzymes have fixed ratios of their various activities, but the ratio of these two activities of P450c17 is regulated post-translationally. We have shown that serine phosphorylation of P450c17 and the allosteric action of cytochrome b 5 increase 17,20 lyase activity, but it has not been apparent whether these two post-translational mechanisms interact. Using purified enzyme systems, we now show that the actions of cytochrome b 5 are independent of the state of P450c17 phosphorylation. Suppressing cytochrome b 5 expression in human adrenal NCI-H295A cells by >85% with RNA interference had no effect on 17␣-hydroxylase activity but reduced 17,20 lyase activity by 30%. Increasing P450c17 phosphorylation could compensate for this reduced activity. When expressed in bacteria, human P450c17 required either cytochrome b 5 or phosphorylation for 17,20 lyase activity. The combination of cytochrome b 5 and phosphorylation was not additive. Cytochrome b 5 and phosphorylation enhance 17,20 lyase activity independently of each other, probably by increasing the interaction between P450c17 and NADPH-cytochrome P450 oxidoreductase.
Similar to all other microsomal forms of cytochrome P450, the catalytic activities of P450c17 require electron donation from NADPH through the intermediacy of a membrane-bound flavoprotein termed P450 oxidoreductase (POR) (15)(16)(17). All cytochrome P450 enzymes catalyze multiple reactions, but the ratios of these multiple reactions are determined primarily by substrate concentrations. By contrast, P450c17 appears to be unique in that the ratio of its two principal activities is developmentally regulated. Human serum concentrations of cortisol, an index of 17␣-hydroxylase activity, remain constant as a function of age, but concentrations of DHEA, an index of 17,20 lyase activity, rise over 100-fold during adrenarche (18), an event that is contemporaneous with, but independent of puberty (19 -22). The ratio of 17,20 lyase activity to 17␣-hydroxylase activity can be regulated by three distinct post-translational mechanisms. First, the presence of high molar ratios of POR to P450c17 favor the 17,20 lyase reaction (14,23). Second, 17,20 lyase activity can be enhanced by the presence of cytochrome b 5 (24 -26), which acts allosterically to foster interactions with POR (13,27). Third, serine/threonine phosphorylation of P450c17 increases 17,20 lyase activity but does not affect 17␣-hydroxylase activity (28,29). In some P450-mediated drug metabolism reactions, cytochrome b 5 appears to act as an alternative electron donor that can substitute for POR in the donation of second electron in the P450 cycle (30 -32) but it does not function in this fashion to foster 17,20 lyase activity, as apo b 5 , which is devoid of heme, is as effective as holo b 5 (13,32). Mutations in the POR binding site of P450c17 selectively reduce 17,20 lyase activity (33); when cytochrome b 5 is added, 17,20 lyase activity is partially restored, increasing the V max but not influencing the K m of the mutants (27), thus confirming the allosteric action of cytochrome b 5 . The mechanism by which P450c17 phosphorylation augments 17,20 lyase activity is not known, nor is it known whether the state of P450c17 phosphorylation influences the allosteric action of cytochrome b 5 . Using purified, catalytically active, bacterially-expressed human P450c17, POR, and b 5 in vitro and RNA interference knockdown of cytochrome b 5 in human adrenal NCI-H295A cells, we now show that these two mechanisms of augmenting 17,20 lyase activity function independently.

MATERIALS AND METHODS
Cell Culture and Microsome Preparation-NCI-H295A cells (34), an adherent population of human adrenocortical carcinoma NCI-H295 cells (35,36), were grown in 150-mm Petri dishes as described (34). HepG2, JEG3 (37), and HEK293 (38) cells were grown as described. NCI-H295A cells from several plates were collected by scraping, washed with chilled phosphate-buffered saline (PBS), suspended in 50 mM sodium phosphate containing 150 mM KCl, and lysed by sonication. Unbroken cells and mitochondria were removed by centrifugation at 10,000 ϫ g for 15 min, and microsomes were collected by ultracentrifugation at 100,000 ϫ g for 90 min in a Beckman T-100 rotor. Microsomes were resuspended in 50 mM potassium phosphate buffer containing 20% glycerol.
Expression and Purification of Human P450c17-Human P450c17 was expressed in Escherichia coli and purified as described (39). The plasmid pCWH17-mod(His)4 containing the cDNA for N-terminally modified human P450c17 (40) was transformed into E. coli JM109, colonies were selected on LB agar plates containing 100 g/ml carbenicillin, and a single colony was grown to A 600 of 0.4 -0.6 in LB broth containing 100 g/ml carbenicillin. From this culture, 1 ml of bacteria were seeded into terrific broth containing 40 M FeCl 3 , 4 M ZnCl 2 , 2 M CoCl 2 , 2 M Na 2 MoO 4 , 2 M CaCl 2 , 2 M CuCl 2 , 2 M H 3 BO 3 , and 10% v/v potassium phosphate solution (0.17 M KH 2 PO 4 , 0.72 M K 2 HPO 4 , pH 7.4) and 100 g/ml carbenicillin and grown at 30°C with shaking at 250 rpm. At A 600 of 0.5-0.7, 0.4 mM isopropyl-1-thio-␤-D-galactopyranoside (IPTG) and 0.4 mM ␦-amino levulinic acid were added, and the culture was shaken at 125 rpm for 36 h at 30°C. The bacteria were then chilled on ice, pelleted by centrifugation at 5000 ϫ g for 10 min at 4°C, washed once with PBS, and resuspended in 100 mM Tris-HCl, pH 7.8 containing 500 mM sucrose and 0.5 mM EDTA (10 ml/g of pellet). Lysozyme (0.1 mg/ml) was added to the bacterial suspension, and the cells were kept on ice for 30 min with occasional stirring. Spheroplasts were harvested by centrifugation at 12,000 ϫ g for 15 min at 4°C and resuspended in 100 mM potassium phosphate pH 7.6 containing 6 mM magnesium acetate, 20% glycerol, and 0.1 mM phenylmethylsulfonyl fluoride. The resulting spheroplasts were sonicated on ice with 15-20 cycles of 20-s pulses followed by 30 s cooling at 50% power using a Fisher scientific 550 sonicator with a microprobe. The lysate was cleared of unbroken cells and debris by centrifugation at 10,000 ϫ g for 10 min at 4°C, and membranes were pelleted by ultracentrifugation in a Beckman T-100 rotor at 100,000 ϫ g for 90 min. The resultant pellet containing human P450c17 was used for enzyme assays and purification of P450c17; cytochrome P450 content was measured by reduced carbon monoxide binding spectra (41). P450c17 was purified from the E. coli membranes by solubilization in 0.7% Triton X-114 followed by chromatography on Ni-NTA and hydrophobic interaction chromatography as described (29,39). Purified P450c17 was resuspended in 100 mM potassium phosphate, pH 7.4, containing 20% glycerol and 0.1% Triton X-100.
Phosphatase Treatment of Bacterially Expressed Human P450c17-Purified preparations of P450c17 (5 g) were treated with 5 units of alkaline phosphatase or protein phosphatase 2A (PP2A) for 15 min at 25°C, and the reaction was stopped by adding 1 mM sodium orthovanadate, 50 mM NaF, and/or 50 M okadaic acid (Calbiochem) in the case of protein PP2A. P450c17 treated with phosphatases was repurified on a Ni-NTA spin column (Amersham Biosciences) and used for enzyme assays. The amount of phosphate released into the supernatant was monitored by malachite green reaction (BioMol Laboratories).
Expression and Purification of Human POR-Human POR was expressed in bacteria by transforming E. coli BL21(DE3)pLysS cells with a pET22b vector (Novagen) containing human POR cDNA (17). Freshly transformed E. coli were selected on a LB agar plate with 100 g/ml carbenicillin and 34 g/ml chloramphenicol, and a single colony was seeded into LB media containing 100 g/ml carbenicillin and 34 g/ml chloramphenicol and grown at 28°C to A 600 of about 0.5. From this culture, 1 ml of bacteria were seeded into 150 ml of terrific broth supplemented with 40 M FeCl 3 , 4 M ZnCl 2 , 2 M CoCl 2 , 2 M Na 2 MoO 4 , 2 M CaCl 2 , 2 M CuCl 2 , 2 M H 3 BO 3 , and 10% v/v potassium phosphate solution (0.17 M KH 2 PO 4 , 0.72 M K 2 HPO 4 , pH 7.4), 100 g/ml carbenicillin and 34 g/ml chloramphenicol and grown at 30°C to A 600 of 0.4 -0.7, 0.4 mM IPTG and 0.1 mg/ml riboflavin were added, and the bacteria were shaken at 125 rpm for another 16 h before harvesting by centrifugation at 5000 ϫ g for 10 min at 4°C. The bacteria were then washed with PBS, treated with lysozyme (0.1 mg/ml) and EDTA (0.1 mM, pH 8.0), and membranes were prepared as described for P450c17. Membranes were dissolved in 100 mM phosphate buffer containing 100 mM NaCl, 1.5% Triton X-100 and 0.1 mM phenylmethylsulfonyl fluoride, and POR was purified by affinity chromatography on 2Ј-5Ј ADP-agarose as described (42). The catalytic activity of this POR preparation was assessed by assays using either cytochrome c or NADPH as substrates as described (17).
Expression and Purification of Cytochrome b 5 -Human cytochrome b 5 was prepared in E. coli strain BL21(DE3) transformed with plasmid pLW01b5 containing the cDNA for human cytochrome b 5 and selected over several cycles, as described (43). A single colony of transformed E. coli was inoculated into LB media containing 100 g/ml carbenicillin, grown to an A 260 of 0.5-0.7, and IPTG was added to a final concentration of 0.4 mM. Cell density was measured every 30 min; when the A 600 nm started to rise again (90 -180-min postinduction), 100 l of cells diluted 1:50 were plated on LB agar plates containing 100 g/ml carbenicillin and 0.4 mM IPTG and allowed to grow overnight at 37°C. A single colony was grown the next day and the process of induction and selection with carbenicillin and IPTG was repeated three more times. After four cycles of selection, the E. coli were transformed with plasmid pLysS Rare, which encodes lysozyme and six rare tRNAs that assist the expression of mammalian proteins by complementing the codon bias of E. coli. The resultant E. coli were used for expression of cytochrome b 5 . Expression and purification was performed as described previously (44).
Suppression of Cytochrome b 5 by Short Hairpin RNA (shRNA) in NCI-H295A Cells-NCI-H295A cells were transfected with the retroviral vector pSUPERretro (45,46) expressing shRNA targeted against both types 1 and 2 human cytochrome b 5 . The pSUPERretro vector was digested with BglI and HindIII and ligated to the 64-mer oligonucleotide duplexes (sense 5Ј-GATCCCCCAAGCTGGAGGTGACGCTATTCA-AGAGATAGCGTCACCTCCAGCTTGTTTTTGGAAA-3Јand antisense 3Ј-GGGGTTCGACCTCCACTGCGATAAGTTCTCTATCGCAGTGGAG-GTCGAACAAAAACCTTTTCGA-5Ј) encoding small interfering RNA (siRNA) against cytochrome b 5 and a 9-nucleotide loop region. The recombinant plasmid was purified from bacterial cultures and digested with EcoRI and HindIII to check the inserts. Purified plasmid was transformed into NCI-H295A, and cells containing the virus were selected by puromycin and used for cytochrome b 5 measurement and P450c17 enzyme assays.
Measurement of Cytochrome b 5 Protein and mRNA-The cytochrome b 5 content of NCI-H295A cells was measured by differential spectroscopy (47). Cells were homogenized in sodium phosphate buffer (50 mM, pH 7.4) containing 150 mM KCl; the homogenate was cleared by centrifugation at 5000 ϫ g for 10 min, placed into reference and sample cuvettes and baseline spectra were recorded between 400 and 500 nm at a speed of 120 nm/min. Cytochrome b 5 in the sample cuvette was reduced with 1 mg sodium dithionite, and the spectra were recorded again. Cytochrome b 5 content was estimated by the difference in absorbance at 423 and 490 nm using a millimolar extinction coefficient of FIG. 1. Early steps of sex steroid biosynthesis. P450scc converts cholesterol to pregnenolone, a C21 ⌬5-steroid. Human P450c17 performs the 17␣-hydroxylase reaction equally well using pregnenolone and progesterone as substrates, but the 17,20-lyase reaction occurs 50 -100 times more efficiently using 17OH-Preg as substrate rather than 17OH-Prog. Thus conversion of 17OH-Prog to androstenedione is insignificant, and DHEA is the major precursor of human sex steroid synthesis. 181 mmol cm Ϫ1 (47). Cytochrome b 5 mRNA was estimated by RT-PCR.
In Vitro Phosphorylation of P450c17-To prepare a cytoplasmic fraction enriched for the kinases that phosphorylate P450c17 and devoid of phosphatase activity, NCI-H295A cells were lysed, and cytosol was passed over a ␥-ATP-Sepharose column (Upstate Biotechnologies); the column was washed with NAD, NADP, ADP, and AMP, and ATPbinding proteins were eluted with 10 mM ATP (29). Kinase activity was checked by phosphorylation of microsomes or purified bacterially expressed recombinant human P450c17 by incorporation of [␥-32 P]ATP. Purified P450c17 was phosphorylated in vitro using 6 g of purified P450c17 incubated with 10 g of kinase fraction in the presence of 50 mM HEPES buffer, 10 mM ATP, 60 mM MgCl 2 , and 10 M okadaic acid at 25°C for 30 min and phosphorylated P450c17 was purified on mini Ni-NTA columns (29).
P450c17 Enzyme Assays-The 17␣-hydroxylase and 17,20 lyase activities of P450c17 were assayed as described (13,39). Purified P450c17 (10 pmol) and POR (20 pmol) were incubated with 100 mM potassium phosphate, 6 mM potassium acetate, 10 mM MgCl 2 , 1 mM reduced glutathione, 20% glycerol, 20 g phosphatidylcholine, 3 units of glucose-6-phosphate dehydrogenase, 0.1 mM glucose-6-phosphate and radiolabeled substrate ([ 14 C]progesterone for hydroxylase assay and [ 3 H]17OH-Preg for lyase assay) with or without 10 pmol of cytochrome b 5 for 5 min at 25°C in a total volume of 200 l. The reaction was started by addition of 20 l of 10 mM NADPH and incubations at 37°C were carried out for various times and stopped by adding ethyl acetate/ iso-octane (3:1) to extract the steroids. Steroids from different reactions were spotted on silica gel 60 F-254 thin layer chromatography plates (Merck) and developed with ethyl acetate/chloroform (3:1) (4). Plates were dried, and steroids were quantitated by autoradiography on a Storm 860 phosphorimager using Image Quant software. Kinetic behavior was approximated as a Michaelis-Menten system. Curve fitting and calculations of maximum velocity (V max ) and apparent Michaelis constant (K m ) values were performed using LEONORA (48).

Isoforms of Cytochrome b 5 in NCI-H295A Cells-Three
forms of cytochrome b 5 have been described: the 98-amino acid soluble and 134-amino acid microsomal form are encoded by one gene (49,50) and a 146-amino acid form associated with the outer mitochondrial membrane (OMb 5 ) is encoded by a second gene (51)(52)(53). To identify the forms of cytochrome b 5 found in NCI-H295A cells, we performed RT-PCR with two pairs of oligonucleotide primers that will amplify the three products of the two genes. In human liver HepG2 cells, the mRNAs for microsomal cytochrome b 5 and for OMb 5 were found in approximately equal amounts, but in human kidney HEK293, human placenta JEG-3 cells, and especially in human adrenal NCI-H295A cells, the mRNA for microsomal cytochrome b 5 was much more abundant than the mRNA for OMb 5 (Fig. 2A). The mRNA for the soluble form, generally associated with erythropoietic tissues, was not detected in the cell lines tested. As human P450c17 is a microsomal enzyme and as the mRNA for the microsomal form of cytochrome b 5 was most abundant in adrenal cells, we focused attention on microsomal cytochrome b 5 .
Knockdown of Cytochrome b 5 in NCI-H295A Cells-To determine the effect of cytochrome b 5 on the catalytic activities of P450c17 in steroidogenic cells, we used RNA interference to knock down cytochrome b 5 mRNA in human adrenal NCI-H295A cells (Fig. 2B). An shRNA targeted against cytochrome b 5 was transfected into human adrenal NCI-H295A cells, and cells were harvested 24, 48, 72, and 96 h later. Expression of this shRNA caused a gradual decrease in spectrally assayable cytochrome b 5 over time: cytochrome b 5 was reduced by 15% after 24 h, 65% after 48 h, and 85% after 72 h (Fig. 2C). Knockdown of cytochrome b 5 had no effect on 17␣-hydroxylase activity in NCI-H295A cells for up to 96 h (Fig. 2, D and E). There was no measurable change in the 17,20 lyase activity in the first 24 h, but 17,20 lyase activity was reduced 15% after 48 h and 30% after 72-96 h (Fig. 2, D and E). This decrease in 17,20 lyase activity was much less than the decrease in cytochrome b 5 protein, suggesting that another factor was promot-ing 17,20 lyase activity.
Increased Phosphorylation of P450c17 Counters the Effect of Reduced Cytochrome b 5 on 17,20 Lyase Activity-Because reducing cytochrome b 5 in intact cells had only a modest effect on 17,20 lyase activity, we sought to examine the potential interactions between the presence of cytochrome b 5 and the phosphorylation of P450c17 on 17,20 lyase activity. Okadaic acid, an inhibitor of serine/threonine phosphatases, specifically increases 17,20 lyase activity in NCI-H295A cells and increases P450c17 phosphorylation (29). Okadaic acid treatment of NCI-H295A cells expressing the shRNA against cytochrome b 5 increased the 17,20 lyase activity in cells expressing decreased amounts of cytochrome b 5 (Fig. 3). The increase in 17,20 lyase activity in control cells was similar to that in cells with a reduced content of cytochrome b 5 . Overexpression of cytochrome b 5 in NCI-H295A cells moderately increases 17,20 lyase activity. Okadaic acid treatment of NCI-H295A cells overexpressing cytochrome b 5 increased 17,20 lyase activity, but the increase was similar to control cells treated with okadaic acid. These studies suggest that cytochrome b 5 and P450c17 phosphorylation augment 17,20 lyase activity independently.
Expression and Purification of Human P450c17, Cytochrome b 5 , and POR-To examine the role of the individual components in P450c17-POR-b 5 reaction, we used purified preparations of recombinant human P450c17, cytochrome b 5 , and POR expressed in E. coli. All three proteins were purified to apparent homogeneity, evidenced by Coomassie Blue staining on an SDS-PAGE gel (Fig. 4). The bacterially expressed P450c17 was not phosphorylated as shown in two fashions. First, assay with malachite green indicated that no phosphate was released when the bacterially expressed human P450c17 was treated with either alkaline phosphatase or PP2A (data not shown); by contrast, native human P450c17 isolated from human adrenal cells or NCI-H295A cells releases readily detected phosphate under these conditions. Second, mass spectrometric analysis of tryptic peptides of bacterially expressed human P450c17 showed no phosphopeptides.
Effect of Cytochrome b 5 and Phosphorylation on Bacterially Expressed Human P450c17-Using the purified, bacterially expressed human P450c17 and POR, we examined the effects of cytochrome b 5 and serine phosphorylation as independent variables on 17␣-hydroxylase and 17,20 lyase activities in vitro. The P450c17-POR-b 5 system was reconstituted using phosphatidylcholine (to provide a membrane environment) and an NADPH regeneration system consisting of glucose-6-phosphate and glucose-6-phosphate dehydrogenase. When recombinant human P450c17 was incubated with POR at a molar ratio of 1:2 and cytochrome b 5 was added in molar ratios of cytochrome b 5 to P450c17 ranging from 0.1 to 100, an effect of cytochrome b 5 on 17,20 lyase activity was first seen at a ratio of 0.5. A 1:1 ratio doubled the 17,20 lyase activity and a 3:1 ratio tripled it, but the maximal effect was seen until a molar ratio of about 30:1 (Fig. 5A). This high ratio for a maximal effect is consistent with previous results (13). Although the molar ratios of cytochrome b 5 to P450c17 in human androgen-producing cells are not known, it is clear that a substantial effect is achieved with low ratios.
To examine the potential interplay of cytochrome b 5 and phosphorylation of P450c17 on 17,20 lyase activity, bacterially expressed human P450c17 was phosphorylated in vitro using a kinase-enriched fraction from NCI-H295A cytosol that was devoid of 17␣-hydroxylase or 17,20 lyase activity (29). P450c17 activities were assayed in reconstituted systems containing 10 pmol of P450c17, 20 pmol of POR, and 20 g of phosphatidylcholine. In the absence of cytochrome b 5 or serine phosphorylation, bacterially expressed human P450c17 had 17␣-hydroxylase activity but very little 17,20 lyase activity. When cytochrome b 5 was added to bacterially expressed human P450c17 in vitro at a cytochrome b 5 to P450c17 ratio of 3:1, 17,20 lyase activity increased 3-4-fold. When bacterially expressed human P450c17 was phosphorylated with the kinase fraction from NCI-H295A cytosol, 17,20 lyase activity increased 4 -5-fold in the absence of cytochrome b 5 (Fig. 5B). Addition of cytochrome b 5 to phosphorylated P450c17 at a cytochrome b 5 to P450c17 ratio of 3:1, increased activity further, but the effects of cytochrome b 5 and phosphorylation were neither additive nor cooperative (Fig. 5B). To explore these findings further, we performed kinetic analysis of P450c17 activities from non-phosphorylated and phosphorylated P450c17 with and without cytochrome b 5 (Table I). Neither phosphorylation nor addition of cytochrome b 5 affected the kinetic parameters of the 17␣-hydroxylase reaction (p Ͼ 0.5). Addition of cytochrome b 5 increased the catalytic efficiency (V max /K m ) of the 17,20 lyase reaction 4-fold (p ϭ 0.013), and P450c17 phosphorylation increased catalytic efficiency 6-fold (p ϭ 0.003 compared with control). However, addition of cytochrome b 5 to phosphorylated P450c17 increased the 17,20 lyase activity only 7-fold. This effect was not significantly different from the action of phosphorylation alone (p ϭ 0.2). Thus the effects of cytochrome b 5 and P450c17 phosphorylation are not additive. Most of the effect seems to come from increased reaction velocities, as changes in K m were less than 2-fold.
Effect of POR and Cytochrome b 5 on 17,20 Lyase Activity-POR is the obligate electron donor for both activities of P450c17, and high molar ratios of POR to P450c17 increase 17,20 lyase activity (14,23). To examine the role of POR in modulating 17,20 lyase activity, we fixed the molar ratio of cytochrome b 5 :P450c17 at 3:1 and varied the amount of POR. As expected, excess POR significantly increased 17,20 lyase activity (Fig. 6A). A dose response was seen for POR either in the absence or presence of cytochrome b 5 , but 17,20 lyase activity was always greater when cytochrome b 5 was present. Addition of POR increased the hydroxylase activity slightly but addition of cytochrome b 5 had no effect on the hydroxylase activity. These data support our previous findings that cytochrome b 5 augments 17,20 lyase activity by facilitating electron transfer from POR (13,29).
Effect of Cytochrome b 5 and Phosphorylation on NCI-H295A Microsomes-To determine whether the effects seen on recombinant human P450c17 in vitro also occurred in a native protein environment, we isolated microsomes from NCI-H295A cells and examined the effect of cytochrome b 5 and of in vitro protein phosphorylation on 17␣-hydroxylase and 17,20 lyase activities. Using the kinase preparation from NCI-H295A cytosol (29) we phosphorylated microsomes from NCI-H295A cells in vitro and assayed them for hydroxylase and lyase activities in the presence and absence of purified cytochrome b 5 (Fig. 6B). NCI-H295A cells normally have a low level of 17,20 lyase activity. Adding cytochrome b 5 increased 17,20 lyase activity 2-3-fold but had no effect on 17␣-hydroxylase activity. In vitro phosphorylation of microsomal P450c17 using the kinase fraction from NCI-H295A cells increased 17,20 lyase activity 3.5-fold but had no effect on 17␣-hydroxylase activity. Adding cytochrome b 5 to microsomes that had been phosphorylated in vitro increased 17,20 lyase activity 5-fold, but had no effect on 17␣-hydroxylase activity. Thus both protein phosphorylation and cytochrome b 5 can enhance 17,20 lyase activity independently of each other, but their effects are neither additive nor cooperative. DISCUSSION Steroidogenesis in the primate adrenal is divided into three morphologically and functionally distinct zones (54). The zona glomerulosa, located just below the adrenal capsule, does not express P450c17 and produces the 17-deoxy 21-carbon steroid aldosterone, the principal mineralocorticoid, under the regulation of angiotensin II. The zona fasciculata, which lies just below the glomerulosa, express P450c17 and has abundant 17␣-hydroxylase but very little 17,20 lyase activity, and produces the 17-hydroxy 21-carbon steroid cortisol, the principal glucocorticoid, under the regulation of corticotropin. The inner zona reticularis, which does not become morphologically identifiable until the onset of adrenarche, expresses P450c17 and has both 17␣-hydroxylase and 17,20 lyase activities and pro-duces 17-hydroxy 19-carbon precursors of sex steroids under ill-defined regulation. The event(s) triggering adrenarche remain unknown (21). Understanding the reticularis-specific activation of the 17,20 lyase activity of P450c17 has been a major challenge as there are no non-primate systems that recapitulate this biology (22). Several mechanisms contribute to the developmental and tissue-specific differential regulation of P450c17 activities (54). Serine phosphorylation of P450c17 increases 17,20 lyase activity without significantly affecting 17␣hydroxylase activity (28) and treatment with alkaline phosphatase or PP2A eliminates almost all 17,20 lyase activity without changing 17␣-hydroxylase activity (28,29). Thus the activities of P450c17 can be differentially regulated by protein phosphorylation based on differential expression of protein kinases and/or phosphatases in different cell types or times in development, thus determining the pattern of steroid hormones produced.
Cytochrome b 5 also enhances the 17,20 lyase activity of human P450c17 (24,(55)(56)(57) by allosteric action that does not involve electron donation (13). Whereas P450c17 is expressed in both the human zona fasciculata and zona reticularis and shows little change as a function of age, the expression of cytochrome b 5   increases. The adult human (60) and rhesus monkey (61) zona reticularis contains abundant cytochrome b 5 as well as P450c17, and adrenal adenomas that produce high levels of DHEA also contain large amounts of cytochrome b 5 (62,63). Similarly, cytochrome b 5 is found in gonadal cells that produce sex steroids, including testicular Leydig cells, follicular theca cells, theca lutein cells, and ovarian stroma (60). Thus there is a strong association between the presence of 17,20 lyase activity and expression of cytochrome b 5 .
There are three forms of cytochrome b 5 expressed from two genes. A gene on chromosome 18q23 consists of 6 exons encoding two alternatively spliced mRNAs: exons 1-4 encode the 98-amino acid soluble form while exons 1-3, 5, and 6 encode the 134-amino acid form bound to the endoplasmic reticulum (49,50). A gene on chromosome 16q22.1 consists of 5 exons encoding a distinct 146-amino acid form of cytochrome b 5 bound to the outer mitochondrial membrane (OMb 5 ). The heme-binding N termini of the products of these two genes are about 70% identical. The C-terminal 10 residues of rat OMb 5 determine targeting to the endoplasmic reticulum or mitochondria (52). Rat OMb 5 can support 17,20 lyase activity (64,65), but unlike the soluble or endoplasmic reticulum form of cytochrome b 5 , OMb 5 preferentially supports 17␣-hydroxylase activity in studies with rat testicular Leydig cells (65). As the principal form of cytochrome b 5 found in the adrenal is the 136-amino acid membrane-bound form, and as OMb 5 fosters 17␣-hydroxylase as well as 17,20 lyase activities, it is likely that the major effects of cytochrome b 5 on human adrenal synthesis of 19carbon precursors of sex steroids is mediated by the microsomal form.
Overexpression of cytochrome b 5 in non-steroidogenic HEK-293 cells co-transfected with P450c17 and POR dramatically enhances 17,20 lyase activity (64). However, reducing cytochrome b 5 expression in NCI-H295A cells by 60% using RNA interference did not affect 17,20 lyase activity, and an 85% reduction in cytochrome b 5 only decreased 17,20 lyase activity by 30%. Similarly, overexpression of cytochrome b 5 in NCI-H295A cells increased 17,20 lyase activity only modestly. Consistent with this we have found that NCI-H295A cells contain about 5-fold more cytochrome b 5 protein than do HEK-293 cells or other non-steroidogenic cell lines. Our data indicate that cytochrome b 5 and protein phosphorylation enhance 17,20 lyase activity independently and that each mechanism is sufficient to achieve nearly maximal induction on its own. Their effects do not seem to be cooperative since we did not observe enhanced DHEA production in samples that were both phosphorylated and contained cytochrome b 5 when compared with the effect of each factor individually. These results also suggest that endogenous cytochrome b 5 in NCI-H295A cells is sufficient to maintain the 17,20 lyase activity and perhaps outer mitochondrial cytochrome b 5 supports some of the 17,20 lyase activity that could not be reduced even after knockdown of more than 85% cytochrome b 5 .
Irrespective of P450c17 phosphorylation or the presence of cytochrome b 5 , increasing the ratio of POR to P450c17 increases 17,20 lyase activity (14,23). As both the 17␣-hydroxylase and 17,20 lyase activities require the donation of electrons from P450 oxidoreductase, this observation thus suggests that both the presence of cytochrome b 5 and phosphorylation of P450c17 increase 17,20 lyase activity by facilitating the association of POR with P450c17, speeding the lyase reaction by increasing the efficiency of electron transfer. Thus the combination of these factors regulates the activities of P450c17 (Fig.  7). It is likely that the direction of steroid biosynthesis in a particular tissue or at a particular age is regulated by one or more of the factors involved in P450c17 activities. A change in 17,20 lyase activity may come from changes in cytochrome b 5 levels, extent and pattern of P450c17 phosphorylation or presence of other enzymes like 3␤-hydroxysteroid dehydrogenase. In a disease like polycystic ovary syndrome one or more of these factors might contribute to the overall change in DHEA production.