aotudui

fs

7

Archives of Biochemistry and BiophysicsVol. 395, No. 1, November 1, pp. 78–84, 2001doi:10.1006/abbi.2001.2566, available online at http://www.idealibrary.com on

Human Sterol 14a-Demethylase Activity Is Enhanced bythe Membrane-Bound State of Cytochrome b5

David C. Lamb, Naheed N. Kaderbhai, K. Venkateswarlu, Diane E. Kelly,Steven L. Kelly, and Mustak A. Kaderbhai1

Institute of Biological Sciences, University of Wales, Aberystwyth Ceredigion SY23 3DD, United Kingdom

Received June 21, 2001, and in revised form August 10, 2001

as

cttc

Human sterol 14a-demethylase (P45051; CYP51) cat-lyzes the oxidative removal of the C32 methyl groupf dihydrolanosterol, an essential step in the choles-erol biosynthetic pathway. The reaction is dependentpon NADPH cytochrome P450 reductase (CPR) thatonates the electrons for the catalytic cycle. Here wesed a recombinant yeast CPR to investigate the abil-

ties of four different forms of cytochrome b5 to sup-port sterol demethylation activity of CYP51. The cyto-chrome b5 derivatives were genetically engineeredorms of the native rat cytochrome b5 core-tail: theoluble globular b5 core (core), the core linked at its

N-terminus with the secretory signal sequence of al-kaline phosphatase (signal-core), and the signal se-quence linked to the native b5 (signal-core-tail). Therat core-tail enzyme greatly stimulated sterol demeth-ylation, whereas the signal-core-tail was only margin-ally active. In contrast, the core and signal-core con-structs were completely inactive in stimulating thedemethylation reaction. Additionally, cytochrome b5

enhanced sterol demethylation by more than three-fold by accepting electrons from soluble yeast CPRand in its ability to reduce P450. We show that thenature of transient linkage between the hemopro-teins and the redox partners is most likely broughtabout electrostatically, although productive interac-tion between cytochrome b5 and CYP51 is governed bythe membrane-insertable hydrophobic region in thecytochrome b5 which in turn determines the correctspatial orientation of the core. This is the first reportshowing the stimulation of CYP51 by cytochrome b5.© 2001 Academic Press

Key Words: cytochrome P450; cytochrome b5; sterol14a-demethylase; steroidogenesis; electron transport.

1 To whom correspondence should be addressed. l

8

Cholesterol is the predominant sterol in human be-ings as it is an integral and structural constituent of allcellular membranes and is the precursor molecule forsteroid hormone and bile acid biosynthesis. Sterol 14a-demethylase (also known as P45014DM or CYP51)2 ca-talyses a key step in cholesterol biosynthesis pathwayinvolving the cleavage of C32 (14a-methyl) from dihy-drolanosterol (1). The ubiquitous distribution of CYP51in humans suggests that all tissues can synthesizecholesterol (2). However, this distribution varies sig-nificantly, suggesting that CYP51 may also have otherfunctions (3) apart from cholesterol biosynthesis. Forexample, the product of the CYP51-catalyzed reaction,4,4-dimethylcholesta-8,14,24-trienol, has meiosis-in-ducing activity in mouse oocytes and is present inhuman follicular fluid (4, 5). Processed pseudogenes ofCYP51 have been identified in the human genome sug-gesting the expression of CYP51 in germ-line cells,since processed pseudogenes are considered to beformed by reverse transcription of mRNAs in germ-linecells (6, 7). These observations suggest that CYP51play a fundamental and essential role in human phys-iology. There is also considerable interest in the dis-covery of selective lanosterol 14 a-demethylase inhibi-tors as these can provide novel hypolipidemic drugs (8).

Sterol CYP51 is the only cytochrome P450 familyfound in animal, plant, and fungal kingdoms (and re-cently in the bacterium Mycobacterium tuberculosis)nd represents an ancient metabolic role for CYP interol biosynthesis, undertaking C32 demethylation

2 Abbreviations used: CYP51, sterol 14a-demethylase; CYP, cyto-hrome P450; CPR, native yeast NADPH cytochrome P450 reduc-ase; D33CPR, N-terminal truncated (33 residues) yeast CPR; core-ail cytochrome b5, native rat liver endoplasmic cytochrome b5; coreytochrome b5, globular, soluble form of native rat cytochrome b5

comprising the first 99 amino acid residues; LB, Luria-Bertani; sig-nal-core-tail cytochrome b , the core-tail cytochrome b N-terminaly

5 5

inked to the secretory signal sequence of alkaline phosphatase.

0003-9861/01 $35.00Copyright © 2001 by Academic Press

All rights of reproduction in any form reserved.

mys2savatsucnupdcaae

ss

rox

79STIMULATION OF STEROL 14a-DEMETHYLASE BY CYTOCHROME b5

via three sequential hydroxylations (9) (see Fig. 1). Themonooxygenation cycle carried out by CYP51 requiressequential input of two electrons, which is supplied byanother integral membrane protein, NADPH CPR.However, the second electron can be supplied by cyto-chrome b5 via cytochrome b5 reductase (10). Althoughseveral studies have shown that cytochrome b5 stimu-lates drug oxidations of human P450 forms, the role ofcytochrome b5 in sterol biosynthesis has not been stud-ied extensively. Previous work has shown that cyto-chrome b5 selectively augments the 17,20-lyase activ-ity of human CYP17 by acting as an allosteric modu-lator that interacts with the P450-CPR complex (11).The CYP17 catalyzed 17-hydroxylase reaction leads tothe glucocorticoid, cortisol, and the 17,20-lyase reac-tion leads to the precursors of sex steroids. The partic-ipation of cytochrome b5 in CYP17 cleavage reactionwas suspected from the earlier observations that Cush-ing’s syndrome sufferers, who had developed adreno-mas secreting excessive androgens, exhibited abnor-mally high levels of cytochrome b5 (12). This functionalassociation of cytochrome b5 with androgen productionthrough interaction with CYP17 has now been con-firmed (13) in a recent study of the immunochemicaldistribution of cytochrome b5 in human adrenal glandand in adrenocortical adrenomas of the sufferers.

In this study we have investigated the influence ofcytochrome b5 on human cytochrome P450 sterol 14a-demethylase/lyase activities in reconstituted systemsemploying highly purified enzyme preparations. Ofparticular focus in this study was the influence of thehydrophobic tail domain of cytochrome b5 in this reac-

FIG. 1. The four steps in the sequential oxidation and removal of CReactions 2–4 were investigated in this study using [32-3H]-3b-hyd

tion mechanism. Consequently, various mutant forms

of cytochrome b5 were utilized to examine the likelyinfluence of the key domains on CYP51 activity.

EXPERIMENTAL PROCEDURES

Isolation, heterologous expression, and purification of humanCYP51. Human CYP51 cDNA was heterologously expressed un-der the control of the GAL10 promoter using the yeast episomalexpression vector, YEP51, as previously described (14). Briefly,Saccharomyces cerevisiae strain GRF18 (MATa leu2-3,2-112 his3-11,3-15 kanr) transformant colonies were cultured in 500 ml yeast

inimal medium in 1-liter flasks containing 1.34% (w/v) Difcoeast nitrogen base without amino acids, 2% (w/v) glucose andupplemented with 20 mg/ml histidine. Cultures were incubated at8°C with orbital agitation at 250 rpm. Cytochrome P450 expres-ion was induced following glucose exhaustion at a cell density ofpproximately 108 cells/ml. After an additional 4-h glucose star-ation, galactose was added to a concentration of 3% (w/v). After20-h induction, the cells were harvested by centrifugation and

he P450 and protein contents determined in the isolated micro-omal fraction. Human CYP51 was purified to homogeneity (14)sing procedures as essentially described for the Candida albi-ans CYP51 (15). Heterologously expressed microsomal CYP51 (50mol) was solubilized with sodium cholate (2% (w/v)) and purifiedsing amino-octyl-Sepharose and hydroxyapatite chromatogra-hy. Purified human CYP51-containing fractions (assessed by re-uced carbon monoxide difference spectroscopy) were pooled andoncentrated using an Amicon Centricon 10 micro-concentratornd enzyme purity was assessed by sodium dodecyl sulfate-poly-crylamide gel electrophoresis and specific heme content. Purifiednzyme preparations were stored at 280°C until use.Cloning and overexpression of soluble and native yeast CPR. A

oluble CPR expression vector (D33CPR:pET15b) was constructed byubcloning a 2.3-kb SalI-BclI fragment containing the D33CPR gene,

cleaved from D33CPR:YEp51 (16), into the unique XhoI-BamHI siteof the Escherichia coli expression vector, pET15b. Similarly full-length CPR expression vector (CPR:pET15b) was constructed bysubcloning a 2.1-kb SalI-HindIII fragment containing CPR gene,

in dihydrolanosterol catalyzed by sterol 14-demethylase (P45014DM).ylanost-7-en-32-ol.

32

excised from CPR:YEp51 (16), into the XhoI-BamHI site of pET15b

stusflbC

mpc

Trc

mVu

o

b

m

80 LAMB ET AL.

using a 0.35-kb HindIII-BamHI fragment, obtained from YEp51plasmid, as a linker. All recombinant DNA manipulations and bac-terial transformations were carried out as described (17). Proteinexpression was performed essentially as described previously (16).The expression vectors were transformed into BL21DE3 (pLysS)strain (Novagen). Transformed strains were grown overnight at 37°Cand with shaking at 150 rpm in Luria-Bertani (LB) broth containing100 mg ampicillin/ml and 34 mg chloramphenicol/ml. When the celldensity had reached an OD660 ;0.5–0.7, heterologous expression wasinduced by addition of 0.5 mM (final concentration) isopropyl-b-D-thiogalactopyranoside for 20 h at 30°C with shaking at 150 rpmafter. All subsequent experiments were performed at 4°C. Harvestedcells (following centrifugation, 5000g for 10 min) derived from 1 literof culture were resuspended in 40 ml of buffer A (20 mM Tris-HCl,pH 7.5) buffer containing 0.5 M NaCl) and lysed by freeze-thawing;the BL21DE3 strain harbors pLysS which expresses the T7 lysozymeand so induces rapid cell lysis by freeze-thaw treatment. To shear thegenomic DNA, the lysate was sonicated using an MSE Soniprep setat full power setting by applying 5 bursts of 1 min with 5 min coolingperiod in between. A low-speed supernatant fraction (5000g for 10min) was then subjected to ultracentrifugation at 100,000g for 1 h torecover the membrane fractions, which were resuspended in bufferA. The protein content was determined by the bicinchoninic acidmethod using bovine serum albumin as the standard.

Purification of CPR and D33CPR. By cloning into the XhoI-BamHI site of pET15b, we engineered six histidine residues at theN-terminus and a thrombin cleavage site at the amino-terminus ofthe recombinant CPRs. His6-tagged CPR and D33CPR were purifiedin a single-step using nickel-chelating affinity chromatography asrecommended the by supplier (Amersham-Pharmacia, UK). The sol-ubilized membrane fractions of E. coli containing CPR and the cyto-solic fraction containing D33CPR were applied, respectively, onto a 5ml Ni-agarose column preequilibrated with 5 vol of binding buffer(20 mM Tris-HCl, pH 7.5, 0.5 M NaCl) containing 5 mM imidazole.The column was washed with 10 vol of binding buffer and 10 vol ofwash buffer (binding buffer containing 20 mM imidazole). Batchelution of protein from the column using 4 vol of elution buffer(binding buffer containing 100 mM imidazole) yielded CPR andD33CPR that contained several low molecular weight proteins. How-ever, highly purified CPR and D33CPR preparations were obtainedwhen a linear gradient of 5 to 100 mM imidazole (10 ml each of washbuffer and elution buffer) was used for elution (Fig. 2); both forms

FIG. 2. SDS-PAGE showing the purity of isolated yeast DCPR(lane 2) and CPR (lane 3). Lane 1 shows the marker proteins.

eluted at approximately 60 mM imidazole. The fractions containing c

purified CPR or D33CPR (monitored by cytochrome c reduction as-ay, see below) were pooled, dialyzed overnight against 10 mM po-assium phosphate buffer (pH 7.5), concentrated by ultrafiltrationsing 30-kDa-cutoff Whatman membrane to 3 mg protein/ml, andtored at 280°C. All steps were carried out at 4°C and at a columnow rate of 1 ml/min. The extent of purification was also monitoredy sodium dodecyl sulfate-polyacrylamide gel electrophoresis andoomassie blue staining.Expression and purification of recombinant rat cytochrome b5 and

its derivatives. The recombinant rat core-tail cytochrome b5 and itscore, signal-core, and signal-core-tail derivatives were expressed inE. coli and purified (Fig. 3) (18–21). Cytochrome b5 reduction was

onitored as follows. A 1-ml volume containing 370 mM potassiumhosphate buffer (pH 7.7), 0.15%, w/v, sodium cholate, 200 pmolytochrome b5, and either 0.8 or 400 pmol of NADPH-CPR was

equally divided into a reference and sample cuvette and the assayinitiated by the addition of 200 mM NADPH to the sample cuvette.

he reduced minus oxidized absorption difference spectrum wasecorded after 10 s and every 30 s thereafter, and the rate of cyto-hrome b5 reduction was determined using an absorption coefficient

of 185 mM21 cm21 (424max and 409min). Sodium cholate was includedin the assay buffer to prevent aggregation of the hydrophobic pro-teins and to maintain them in solution.

Assays. The activities of the purified CPR and D33CPR wereeasured by their abilities to reduce cytochrome c as described byermilion and Coon (22), and the specific activity was determinedsing a millimolar extinction coefficient of 21 mM21 cm21.Cytochrome P450 was estimated by reduced carbon monoxide dif-

ference spectroscopy according to the procedures of Omura and Sato(23) using a millimolar extinction coefficient of 91 mM21 cm21.

To monitor CYP51 catalytic activity, each reaction mixture con-taining 50 pmol of purified CYP51, and 400 pmol CPR or D33CPR,200 pmol cytochrome b5, or its respective derivative in a total volumef 50 ml was dispersed in 50 mg dilauroylphosphatidylcholine. The

volume was adjusted to 950 ml with 100 mM potassium phosphateuffer (pH 7.4) and [32-3H]-3b-hydroxylanost-7-en-32-ol (52 mg, 1.62

mCi in 10 ml dimethylformamide) was added prior to ultrasonicationof the mixture. The reaction was started by addition of NADPH at afinal concentration of 1 mM to the mixture. The assays were con-ducted at 37°C and terminated after a 20-min incubation. The ex-

FIG. 3. SDS-PAGE of purified cytochrome b5 variants. Lane 1,arker proteins; lanes 2–5, purified preparations of signal-core-tail,

ore-tail, signal-core, and core, respectively.

Cpc(wpHgp

g

81STIMULATION OF STEROL 14a-DEMETHYLASE BY CYTOCHROME b5

traction and determination of radioactivity released in the form offormic acid were carried out as described elsewhere (1).

RESULTS AND DISCUSSION

Expression, Purification, and Characterizationof CPR, D33CPR, CYP51, and Cytochromeb5 Derivatives

We used an E. coli/pET expression system to obtainlarge quantities of the recombinant proteins for struc-tural and functional studies. In E. coli the full-length

PR was localized in the membrane fraction after ex-ression whereas the N-terminal 33 amino acid-trun-ated D33CPR was retained in the cytosolic fractiondata not presented). The expressed D33CPR proteinas accumulated at a significantly higher level in com-arison with its full-length counterpart, CPR protein.is6-tagged CPR and D33CPR were purified by a sin-

le step using nickel-chelating affinity chromatogra-hy. Highly purified CPR and D33CPR proteins were

obtained when a linear gradient of 5 to 100 mM imi-dazole was applied for elution. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by Coo-massie blue staining of both the purified CPR andD33CPR proteins revealed a single band of the ex-pected sizes (Fig. 2). About 30 mg of approximately95% purified D33CPR was obtained from 6 liters ofculture. The purified CPR and D33CPR were yellow incolor, indicating the presence of flavin. Spectra of pu-rified CPR and D33CPR showed maximal absorption at455 and 380 nm. Addition of NADPH to the purifiedforms of the CPR followed by air reoxidation resultedin a decrease in the absorption at 455 nm and theappearance of a broad absorption band at 550–650 nm,characteristic of the air-stable semiquinone. The oxi-dized and reduced spectra of CPR and D33CPR were

FIG. 4. N- and C-terminal sequences of the four variant cytochromlobular heme-binding domain.

indistinguishable from those of other microsomal CPR

(24, 25). The specific activities of yeast CPR andD33CPR were 8.3 and 60 mmol/min/mg, respectively.This is comparable to the previously reported specificactivities of 15 to 180 mmol/min/mg for human forms(26, 27). The specific content of reductase was calcu-lated as 17 nmol/mg protein (1 nmol CPR reduces 3mmol of cytochrome c). Human CYP51 was purified tohomogeneity as described previously (14) with a yieldof 20 nmol P450/liter of culture and final specific con-tent of 16.1 nmol CYP51/mg protein. The protein gavea typical reduced CO difference spectral maximum at448 nm.

The native rat cytochrome b5 comprises 133 aminoacids and is organized in two domains. The residues 1to 99 are folded to create a compact globular domain,whereas the C-terminal sequence, comprising residues100 onward, forms a hydrophobic tail (28), which pro-trudes through the membrane bilayer (29, 30). Theglobular domain, which can be obtained by the proteo-lytic digestion of the membrane-embedded protein (31,32), was one of the first proteins to be examined byX-ray diffraction (33); the three-dimensional structureof the complete molecule awaits determination.

In this study, the native rat cytochrome b5 (core-tail)and the various derivatives, core, signal-core, and sig-nal-core-tail (Fig. 4), were expressed in E. coli andpurified to near homogeneity (Fig. 3) as previouslydescribed (18–21, 34). All four rat cytochrome b5 deriv-atives (i) were spectrally indistinguishable as b-typecytochrome (ii) to have their predicted amino acid se-quence for the first five N-terminal residues and (iii)yielded masses by MALDI-TOF MS that were in com-plete agreement with their theoretical masses. Thespecific hemoprotein content of these cytochrome b5

5 used in this study. The core (residues 1–99) consists of the soluble,

e b

derivatives ranged from 40 to 45 nmol of hemopro-

fc

nlsp

e

o

wca

baoottta

CCCCCCC

82 LAMB ET AL.

tein/mg protein; the heme content ranged from 15 to 20nmol/mg of the purified protein.

Cytochrome b5 Stimulates CYP51 Hydroxylation andLyase Activities

[32-3H]-3b-Hydroxylanost-7-en-32-ol was aerobicallydemethylated by a reconstituted monooxygenase sys-tem containing human CYP51 and CPR or DCPR (Ta-ble I). Both CPR and the soluble D33CPR drove CYP51-mediated sterol 14a-demethylation with a turnover of0.42 nmol product formed/min/nmol CYP51. These val-ues compare well with a turnover of ;0.5 nmol productormed/min/nmol CYP51 for sterol 14a-demethylationatalyzed by full-length CPR and using [32-3H]-3b-

hydroxylanost-7-en-32-ol as substrate (14) and ;0.5mol product formed/min/nmol CYP51 using dihydro-

anosterol as substrate (35). The authenticity of theterol product was verified by gas chromatography cou-led to mass spectroscopy.The abilities of four different forms of genetically

ngineered cytochrome b5 to support CYP51-mediatedsterol 14a-demethylation were investigated using thestandard assay described under Experimental Proce-dures. The four different forms were native rat core-tail, the soluble core, signal-core, and the signal-core-tail (Fig. 4). At a 4:1 molar ratio of core-tail cytochromeb5 to CYP51, at a saturating point, the reaction showeda threefold increase in activity (Table I). Interestingly,D33CPR showed a greater stimulation in excess of 3times. In comparison, the signal-core-tail cytochromeb5 derivative containing two potential membrane-spanning segments also increased demethylation ac-tivity by approximately twofold. Signal-core and corederivatives of cytochrome b5 exerted no significant ef-fect on CYP51 activity. Thus, the native form of cyto-chrome b5 appeared to be important in the stimulationf CYP51-mediated sterol 14a-demethylation reactions

TABLE I

Effect of Various Cytochrome b5 Derivatives on NADPHCPR-Driven Sterol 14a-Demethylase (Hydroxylation and

Lyase) Activities of Human CYP51a

Redox partners/combination

CYP51 activity(nmol of product/

min/nmol of P450)

YP51 1 CPR 0.42 6 0.04YP51 1 D33CPR 0.44 6 0.06YP51 1 CPR 1 (core-tail)b5 1.14 6 0.10YP51 1 D33CPR 1 (core-tail)b5 1.34 6 0.08YP51 1 CPR 1 (signal-core-tail)b5 0.78 6 0.12YP51 1 CPR 1 (signal-core)b5 0.40 6 0.06YP51 1 CPR 1 (core)b5 0.38 6 0.04

a The activities are averages of three determinations.

involving the hydroxylation and lyase activities. c

CYP51 Oxidations and Side-Chain Cleavage ActivitiesAre Strongly Influenced by Cytochrome b5

and Its Derivatives

We asked whether the different stimulatory patternsof the various cytochrome b5 derivatives in promotingthe side-chain cleavage activity of CYP51 were due toan altered mode of interaction with CYP51 or indi-rectly through altered electron coupling with CPR?Therefore, we measured the reduction of cytochrome b5

by CPR in the absence of CYP51. Using a CPR molarratio of 2:1, as used in the CYP51 side-chain cleavageactivity assay, all of the cytochrome b5 derivatives werefully reduced within 30 s. However, for a more criticalcomparison, reduction of the cytochrome b5 derivativeswas conducted by decreasing the amount of P450 re-ductase by 500-fold. Under these conditions (CPR mo-lar ratio of 1:250) the rat core-tail cytochrome b5 forms

ere reduced at the highest rates, whereas the signal-ore-tail, core, and signal-core species displayed 75, 35,nd 25% rates of reduction of the native cytochrome b5,

respectively. These results indicate partial impairmentof interaction between the engineered forms of cyto-chrome b5 and the CPR. However, this step is unlikelyto be rate-limiting in the side-chain cleavage reaction,since under the conditions used in this assay (400 pmolCPR) the rates of electron transfer from CPR to thevarious forms of cytochrome b5 were in a large excess,compared with an optimal side-chain cleavage activity.Furthermore, since the soluble CPR was fully func-tional in supporting demethylation reaction, we elimi-nate the possibility that the lack of side-chain cleavageby CYP51 in the presence of core and signal-core de-rivatives was due to their preferential association withCPR, so that the latter was not available for interac-tion with CYP51. These findings therefore allow us toconclude that the decreased stimulation of CYP51lyase activity by the modified cytochrome b5 deriva-tives is likely due to their altered interaction withCYP51 and not as a result of impaired electron cou-pling with CPR.

The findings from the present study show thatcytochrome b5 plays an important role in electrontransfer to human CYP51, a key enzymatic step ofcholesterol biosynthesis. Two possible modes of theaction of cytochrome b5 could be envisaged. One isthat the orientation of the cytochrome b5 in the mem-

rane determined by the nature of hydrophobic link-ge is crucially important. The native tail anchorageptimally determines the preferred mode of spatialrientation of the core with the human CYP51, al-hough a dual anchorage via both the N- and C-erminal is also possible to a lesser extent. In con-rast the soluble core or N-terminally membrane-nchored forms do not enhance the side-chain

leavage of C32. Thus, the stimulation of human

u

mt

sBd

ac

wR

83STIMULATION OF STEROL 14a-DEMETHYLASE BY CYTOCHROME b5

CYP51 activity by the native forms of cytochrome b5

but not by the core or signal core derivatives clearlydemonstrates that the C-terminal hydrophobic seg-ment of cytochrome b5 has a crucial role in mediatingthe interaction of the catalytically active globulardomain with human CYP51 (Fig. 4), and these inter-actions are most likely to be mediated via surfaceionic bridges as increasing the salt strength in thereconstituted assay system markedly inhibited core-tail-stimulated demethylation catalysis but not thatof cytochrome b5-independent reaction (data not pre-sented). The present findings are similar to thoseobserved for cytochrome b5 and its derivatives stim-

lating human CYP17 lyase activity (11).An intriguing insight into the mechanism reactionay also be revealed during these studies of electron

ransport to CYP51. Our previous work with C. al-bicans CYP51 revealed that the final oxidation of theterol aldehyde occurs between C-14 and C-32 in aayer-Villiger reaction or a peroxy enzyme interme-iate as has been suggested for aromatase (1). In

vitro reconstitution studies with purified CYP51show that only cytochrome P450 is necessary forcomplete demethylation, rather than other P450lyase forms or decarboxylating enzymes, thus thishemeoprotein must possess both the oxidative andlyase activities. It can thus be proposed that cyto-chrome b5 stimulates the lyase activity of CYP51 inan analogous fashion to the stimulation of lyase ac-tivity of human and bovine CYP17 (36, 37). Cyto-chrome b5 functions as an electron transfer proteinand readily accepts electrons from its native reduc-tase, NADH-dependent cytochrome b5 reductase,

nd also from NADPH-dependent CPR (38). Thusytochrome b5 mediates or enhances electron trans-

port via CPR to human CYP51. These findings arealso consistent with the proposed role of cytochromeb5 and its reductase in furnishing electron transportroles in a variety of P450-dependent transformationsfor ergosterol biosynthesis in a cpr-deleted yeaststrain (16). Lastly, although the engineered humanCPR has been overproduced in a soluble form, afunctional form capable of supporting any CYP-driven activities has not yet been observed. Intrigu-ingly, the yeast-soluble counterpart, which is func-tional, appears to drive the membrane-bound cyto-chrome b5 in the CYP51-catalyzed demethylationreaction.

ACKNOWLEDGMENTS

We are most grateful to Prof. M. Akhtar for generous donation ofthe CYP51 substrate [32-3H]-3b-hydroxylanost-7-en-32-ol. This work

as partly supported by the Biological and Biotechnology Science

esearch Council.

REFERENCES

1. Shyadehi, A. Z., Lamb, D. C., Kelly, S. L., Kelly, D. E., Schunck,W. H., Wright, J. N., Corina, D., and Akhtar, M. (1996) J. Biol.Chem. 271, 12445–12450.

2. Stromstedt, M., Rozman, D., and Waterman, M. R. (1996) Arch.Biochem. Biophys. 329, 73–81.

3. Rozman, D., and Waterman, M. R. (1998) Drug Metab. Dispos.26, 1199–1201.

4. Rozman, D. (2000) Pflugers Archiv. Eur. J. Physiol. 439, R56–R57.

5. Yoshida, Y., Yamashita, C., Noshiro, M., Fukuda, M., and Aoyama,Y. (1996) Biochem. Biophys. Res. Commun. 223, 534–538.

6. Rozman, D., Stromstedt, M., and Waterman, M. R. (1996) Arch.Biochem. Biophys. 333, 466–474.

7. Rozman, D., Stromstedt, M., Tsui, L. C., Scherer, S. W., andWaterman, M. R. (1996) Genomics 38, 371–381.

8. Swinney, D. C., So, O. Y., Watson, D. M., Berry, P. W., Webb,A. S., Kertesz, D. J., Shelton, E. J., Burton, P. M., and Walker,K. A. (1994) Biochemistry 33, 4702–4713.

9. Aoyama, Y., Yoshida, Y., and Sato, R. (1984) J. Biol. Chem. 259,1661–1666.

10. Pompon, D., and Coon, M. J. (1984) J. Biol. Chem. 259, 15377–15385.

11. Lee-Robichaud, P., Kaderbhai, M. A., Kaderbhai, N., Wright,J. N., and Akhtar, M. (1997) Biochem. J. 321, 857–863.

12. Sakai, Y., Yanase, T., Hara, T., Takayanagi, R., Haji, M., andNawata, H. (1994) Clin. Endocrinol. 40, 205–209.

13. Yanase, T., Sasano, H., Yubisui, T., Sakai, Y., Takayanagi, R.,and Nawata, H. (1998) Endocr. J. 45, 89–95.

14. Lamb, D. C., Kelly, D. E., Waterman, M. R., Stromstedt, M.,Rozman, D., and Kelly, S. L. (1999) Yeast 15, 755–763.

15. Lamb, D. C., Kelly, D. E., Schunck, W. H., Shyadehi, A. Z.,Akhtar, M., Lowe, D. J., Baldwin, B. C., and Kelly, S. L. (1997)J. Biol. Chem. 272, 5682–5688.

16. Venkateswarlu, K., Lamb, D. C., Kelly, D. E., Manning, N. J.,and Kelly, S. L. (1998) J. Biol. Chem. 273, 4492–4496.

17. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) MolecularCloning, A Laboratory Manual, 2nd ed. (Nolan, C., Ed.), ColdSpring Harbor Laboratory Press, Cold Spring Harbor, NY.

18. Kaderbhai, N., Karim, A., Hankey, W., Jenkins, G., Venning, J.,and Kaderbhai, M. A. (1997) Biotechnol. Appl. Biochem. 25,53–61.

19. Kaderbhai, N., and Kaderbhai, M. A. (1996) Protein ExpressionPurif. 7, 237–246.

20. Kaderbhai, N. N., Harding, V. J., and Kaderbhai, M. A. (1995)Biochem. Soc. Trans. 23, S58.

21. Gallagher, J., Kaderbhai, N., and Kaderbhai, M. A. (1992) Appl.Microbiol. Biotechnol. 38, 77–83.

22. Vermilion, J. L., and Coon, M. J. (1978) J. Biol. Chem. 253,2694–2704.

23. Omura, T., and Sato, R. (1964) J. Biol. Chem. 239, 2370–2378.24. Kurzban, G. P., Howarth, J., Palmer, G., and Strobel, H. W.

(1990) J. Biol. Chem. 265, 12272–12279.25. Oprian, D. D., and Coon, M. J. (1982) J. Biol. Chem. 257, 8935–

8944.26. Andersen, J. F., Utermohlen, J. G., and Feyereisen, R. (1994)

Biochemistry 33, 2171–2177.27. Shephard, E. A., Phillips, I. R., Bayney, R. M., Pike, S. F., and

Rabin, B. R. (1983) Biochem. J. 211, 333–340.28. Spatz, L., and Strittmatter, P. (1971) Proc. Natl. Acad. Sci. USA

68, 1042–1046.

84 LAMB ET AL.

29. Vergeres, G., Ramsden, J., and Waskell, L. (1995) J. Biol. Chem.270, 3414–3422.

30. Vergeres, G., and Waskell, L. (1995) Biochimie 77, 604–620.31. Ito, A., and Sato, R. (1968) J. Biol. Chem. 243, 4922–4923.32. Enomoto, K., and Sato, R. (1973) Biochem. Biophys. Res. Com-

mun. 51, 1–7.33. Kajihara, T., and Hagihara, B. (1968) J. Biochem. (Tokyo) 63,

453–461.34. Karim, A., Kaderbhai, N., Evans, A., Harding, V., and Kaderb-

hai, M. A. (1993) Bio/Technology 11, 612–617.

35. Lamb, D. C., Kelly, D. E., Manning, N. J., Hollomon, D. W., andKelly, S. L. (1998) FEMS Microbiol. Lett. 169, 369–373.

36. Shet, M. S., Faulkner, K. M., Holmans, P. L., Fisher, C. W.,and Estabrook, R. W. (1995) Arch. Biochem. Biophys. 318,314 –321.

37. Shet, M. S., Fisher, C. W., Arlotto, M. P., Shackleton, C. H.,Holmans, P. L., Martin Wixtrom, C. A., Saeki, Y., andEstabrook, R. W. (1994) Arch. Biochem. Biophys. 311, 402–417.

38. Lamb, D. C., Kelly, D. E., Manning, N. J., Kaderbhai, M. A., and

Kelly, S. L. (1999) FEBS Lett. 462, 283–288.