Biochemical and Biophysical Research Communications 289, 712–717 (2001)

doi:10.1006/bbrc.2001.6032, available online at http://www.idealibrary.com on

Structural Role of Conserved Asn179 in the Short-ChainDehydrogenase/Reductase Scaffold

Charlotta Filling,* Erik Nordling,*,† Jordi Benach,‡,1 Kurt D. Berndt,‡,§ Rudolf Ladenstein,‡Hans Jornvall,* and Udo Oppermann*,2

*Department of Medical Biochemistry and Biophysics, Karolinska Institutet, SE-171 77 Stockholm, Sweden;†Stockholm Bioinformatics Center, Karolinska Institutet, SE-171 77, Stockholm, Sweden; ‡Center for StructuralBiochemistry, NOVUM, Karolinska Institutet, SE-141 57 Huddinge, Sweden; and §Departmentof Natural Sciences, Sodertorns Hogskola, Box 4101, SE-141 04 Huddinge, Sweden

Received October 29, 2001

terized from all forms of life (1–4). These enzymes

Short-chain dehydrogenases/reductases (SDR)

constitute a large family of enzymes found in allforms of life. Despite a low level of sequence iden-tity, the three-dimensional structures determineddisplay a nearly superimposable a/b folding pattern.We identified a conserved asparagine residue lo-cated within strand bF and analyzed its role in theshort-chain dehydrogenase/reductase architecture.Mutagenetic replacement of Asn179 by Ala in bacte-rial 3b/17b-hydroxysteroid dehydrogenase yields afolded, but enzymatically inactive enzyme, which issignificantly more resistant to denaturation by gua-nidinium hydrochloride. Crystallographic analysisof the wild-type enzyme at 1.2-Å resolution reveals ahydrogen bonding network, including a buried andwell-ordered water molecule connecting strands bEto bF, a common feature found in 16 of 21 knownthree-dimensional structures of the family. Based onthese results, we hypothesize that in mammalian11b-hydroxysteroid dehydrogenase the essentialAsn-linked glycosylation site, which corresponds tothe conserved segment, displays similar structuralfeatures and has a central role to maintain the SDRscaffold. © 2001 Elsevier Science

Key Words: short-chain dehydrogenase/reductase;hydroxysteroid dehydrogenase; 11b-hydroxysteroiddehydrogenase; site-directed mutagenesis; three-dimensional structure; glycosylation.

Short-chain dehydrogenases/reductases (SDR) con-stitute a large, functionally heterogeneous family ofenzymes with close to 2000 different members charac-

Abbreviations used: SDR, short-chain dehydrogenase/reductase;HSD, hydroxysteroid dehydrogenase.

1 Present address: Department of Biological Sciences, ColumbiaUniversity, 702 Fairchild Center, New York, NY 10027.

2 To whom correspondence and reprint requests should be ad-dressed. Fax: 146 8 33 74 62. E-mail: Udo.Oppermann@mbb.ki.se.

7120006-291X/01 $35.00© 2001 Elsevier ScienceAll rights reserved.

fulfill essential cellular functions through regulation ofhormone or mediator metabolism, by, e.g., catalysis ofthe “switch” between active and inactive hormone re-ceptor ligands. This principle is exemplified by thehormone pairs estradiol/estrone or cortisol/cortisone,mediated through 17b- and 11b-hydroxysteroid dehy-drogenases, respectively (5–7). Accordingly, NAD(P)(H)-dependent oxidoreductases acting on diverse sub-strates such as steroids, retinoids, lipid mediators andxenobiotics, form the largest subgroup within the pro-tein family, but SDRs also include epimerase and hy-dratase activities (1, 2, 8).

Despite a highly similar a/b folding pattern with acentral Rossmann-fold nucleotide binding motif, SDRsusually display a low level of sequence identity, typi-cally ranging between 10 and 30% (1–4, 9). Theseconserved residues are dispersed over a few recognizedmotifs, necessary to obtain the general fold and togenerate the catalytic center, with a Ser-Tyr-Lys triadof residues (1).

Several studies have highlighted the general impor-tance of the two most conserved motifs, comprising aTGxxxGxG sequence and the indicated active siteS-Y-K triad (1). Sequence comparisons, site-directedmutagenesis and tertiary structure analyses of SDRenzymes identified further regions of importance (10–12). In this study, we report on the essential role of aseemingly conserved central Asn residue, using bacte-rial 3b/17b-hydroxysteroid dehydrogenase (3b/17b-HSD) as a model system. This residue is located withinthe bF strand, and is in the homologous position of anasparagine in a conserved glycosylation motif in type 111b-hydroxysteroid dehydrogenases (13, 14), thus ex-plaining the essential character of the Asn residue inthese structures. Furthermore, we discuss several fea-tures of this segment to maintain the overall SDRarchitecture and function.

Vol. 289, No. 3, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

MATERIALS AND METHODS

Cloning, expression, and purification of recombinant 3b/17b-HSD.Molecular cloning of 3b/17b-HSD (EC 1.1.1.51) from Comamonastestosteroni ATCC 11996 was carried out by generation of a PCRproduct from genomic DNA, using primers with compatible restric-tion sites encompassing the complete gene (10). The fragment wascloned into the NdeI/BamHI sites of the pET15b vector (Novagen),transformed into BL21DE3 cells, and recombinant protein was ex-pressed by isopropylthiogalactoside induction, and purified by affin-ity chromatography as described (10, 15).

FIG. 1. Detail of the structure-based alignment of determined 3Ddistal to strand bF. The conserved Asn residue (N179 in 3b/17b-HScatalytic residues Tyr and Lys are indicated by boxing; further residumarked by boxing. The corresponding segment of human 11b-hydrcoincidence of alignment of the central Asn residue. Figure created1dhr, dihydropteridine reductase; 1a27, 17-b-hydroxysteroid-dehydrtrihydroxynaphthalene reductase; 1geg, acetoin reductase; 1i01, b-keglucose dehydrogenase; 1ahh, 7a-hydroxysteroid dehydrogenasehydroxysteroid dehydrogenase; 1cyd, carbonyl reductase mouse; 1ae12,3-dihydrodiol-2,3-dehydrogenase; 1a4u, Drosophila alcohol dehydprotein (acp) reductase; 1qsg, enoyl-reductase; 1d7o, enoyl-[acyl-carrreductase.

713

Purity was confirmed by SDS/PAGE, and protein concentrationwas determined spectrophotometrically. Site-directed mutagenesiswas carried out using a PCR-based method. Mutant and wild-typesequences were confirmed by cycle sequencing (Big Dye, AppliedBiosystems) and analyzed using an ABI 377 system.

Determination of kinetic and binding constants. Enzyme activitieswere measured as NAD(H)-dependent 3b- and 17b-oxidoreductase ac-tivities by determination of the change of absorbance at 340 nm andusing a molar extinction coefficient for NADH of 6.22 mM21 cm21.Recordings were carried out with a Cary 300Bio instrument. Reactions

uctures of SDR enzymes. Depicted is the region comprising helix aF1hxh) is marked by an asterisk. The active-site region comprisingiscussed are the seemingly conserved ProGly and Thr residues, alsosteroid dehydrogenase 1 (DHI1_human) is shown, illustrating theth the program PRALIN. Abbreviations used/PDB accession codes:nase; 1e3w, rat short-chain 3-hydroxyacyl-coa dehydrogenase; 1doh,cyl [acp] reductase; 1edo, b-keto acyl carrier protein reductase; 1gco,coli; 1hdc, 3a,20b-hydroxysteroid dehydrogenase; 1hxh, 3b/17b-opinone reductase-I; 2ae1, tropinone reductase-ii; 1bdb, cis-biphenyl-enase; 1nas, mouse sepiapterin reductase; 1bvr, enoyl-acyl carrierprotein] reductase; 1fjh, 3a-hydroxysteroid dehydrogenase/carbonyl

-strD,es doxywi

ogetoaE., trrogier

LE

Vol. 289, No. 3, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

were performed in 1.0-ml volumes at 298K. Conditions for dehydroge-nase activities were 20 mM Tris/HCl (pH 8.5) and 250 mM NAD1,varying the amount of iso-ursodeoxycholic acid (3b,17b-dihydroxy-5b-cholan-24-oic acid, isoUDCA) or dehydroisoandrosterone (5-androsten-3b-ol-17-one; DHEA; 3b-HSD activity) and testosterone (4-androsten-17b-ol-3-one; 17b-HSD activity). Conditions for reductase activitieswere 20 mM Tris/HCl (pH 7.0) and 250 mM NADH, varying the amountof 5a-dihydrotestosterone (5a-androstane-17b-ol-3-one; 3-ketoreduc-tase activity) and androsterone (5a-androstan-3a-ol-17-one; 17-keto-reductase activity). Enzymatic constants were calculated with theEnzPack software (Biosoft). Binding of NADH was determined by mon-itoring fluorescence energy transfer as a function of nucleotide concen-tration in 20 mM Tris–HCl, pH 7.0, at 298K using a Shimadzu RF5000spectrofluorometer (excitation at 286 nm, emission at 450 nm). Bindingof NAD1 was assessed in a similar manner by displacement of boundNADH (excitation at 345 nm, emission at 450 nm) at 20 mM Tris/HCl,pH 8.5. Dissociation constants were obtained by nonlinear regressionusing a 1:1 binding model.

Circular dichroism spectroscopy. Circular dichroism (CD) spec-troscopy was performed on an AVIV Model 62DS spectropolarom-eter. Spectra were recorded by measurement of the ellipticity as afunction of wavelength at 0.5 nm increments between 260 and 190nm. Protein samples were adjusted to a concentration of 0.5 mg/ml in10 mM Tris/HCl (pH 8.5), 100 mM NaCl, and 2 mM DTT prior tomeasurement.

FIG. 2. Analysis of expression and folding for wild-type and muspectroscopy. (A) Purified protein was separated on 12% Tris/Glycinstandard (Novex); lane 2, recombinant 3b/17b-HSD wild-type; lane 3wild-type 3b/17b-HSD.

TAB

Kinetic Constants for 3b- and 17b-Dehydrogenase A

Mutation

3b-HSDisoUDCA

3b-HSDDHEA

Km k cat

Relativek cat/Km Km k cat

Relativek cat/Km K

Wild type 22.3a 1.4a 100a 8.17 0.62 100 16Asn 179 Ala na — — na — — n

Note. Km values are in 1026 M, k cat values in 103 min21, and k cat/Kma Data taken from Ref. (10).

714

Equilibrium unfolding experiments by guanidinium hydrochloridetitration. Equilibrium unfolding measurements for 3b/17b-HSDwild-type and N179A mutant proteins were performed using guani-dinium hydrochloride (GuHCl) titration. The extent of unfolding wasdetermined by measuring the change in ellipticity at 222 nm atconstant protein and variable GuHCl concentrations. Values for theunfolding transition Cmid were determined by nonlinear regressionanalysis as described earlier (10).

Structure determination of 3b/17b-HSD. The structure of thewild-type 3b/17b-HSD apoform was determined by X-ray crystallog-raphy to a final resolution of 1.2 Å. Crystallization, X-ray analysis,model building, and refinement have been described in detail else-where (Benach et al., submitted), and the coordinates obtained weredeposited in the Protein Database under PDB Accession Code 1hxh.

RESULTS

Using a structural alignment of crystallized SDRenzymes (Fig. 1) we identified a short conserved seg-ment centered around a seemingly conserved Asn (179in 3b/17b-HSD) residue located in strand bF. Thisstructural element shows in most cases the sequenceI-R-V-N, is located between helix aF, bearing the active

t 3b/17b-HSD forms assessed by SDS/PAGE and circular dichroismls and stained by Coomassie brilliant blue. Lane 1, molecular mass179A mutant. (B) Circular dichroism spectra of N179A mutant and

1

ivities and the Reversible Activities of 3b/17b-HSD

Ketoreductase5a-DHT

17b-HSDtestosterone

17-Ketoreductaseandrosterone

k cat

Relativek cat/Km Km k cat

Relativek cat/Km Km k cat

Relativek cat/Km

0.13a 100a 11.8 1.2 100 31.7 0.21 100— — na — — na — —

ues in 106 s21 M21. Number of experiments, n 5 3–5. na, not active.

tange, N

ct

3-

m

.1a

a

val

of N179A revealed a complete loss of enzymatic func-

Vol. 289, No. 3, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

site motif YxxxK and the conserved PG motif, termi-nating strand bF.

To investigate the function of the Asn residue in SDRarchitecture we used bacterial 3b/17b-HSD from Co-mamonas testosteroni as a model system. Site-directedmutagenetic replacement of Asn by Ala was performed,wild-type and mutant proteins were expressed, andbiochemical properties of the two forms were com-pared. Both recombinant wild-type and the N179A mu-tant could be expressed to a similar level in Esche-richia coli BL21 cells (about 10–20 mg protein per literculture) (Fig. 2A), and both forms were found to displaynearly superimposable CD spectra, indicating similarfold characteristics (Fig. 2B). However, kinetic analysis

FIG. 3. GuHCl titration of wild-type 3b/17b-HSD (solid circles)and its N179A mutant (open circles). Unfolding was analyzed bymeasurement of the ellipticity at 222 nm at fixed protein and vari-able GuHCl concentrations. Abbreviation used: fu, fraction unfolded.

FIG. 4. Structural details of strands bE (E) and bF (F) as deter(Benach et al., submitted). Electron densities and hydrogen bondingstructure elements are depicted.

715

tion (Table 1), as none of the activities measured withwild-type 3b/17b-HSD can be observed with the mu-tant form. This loss of activity is not due to alteredcoenzyme binding since kD values for NAD1 and NADHfor wild-type and the N179A mutant determined byfluorescence spectroscopy are largely similar (data notshown).

Equilibrium unfolding experiments by guanidiniumhydrochloride titration reveal essential differences inunfolding characteristics between wild-type andN179A 3b/17b-HSD. The midpoint of unfolding for thewild-type is 0.65 M, whereas a shift to 1.6 M GuHCl forN179A is observed (Fig. 3). Analysis of the apostruc-ture of wild-type 3b/17b-HSD by crystallography atatomic resolution reveals an essential hydrogen bond-ing network mediated by the N179 side chain (Fig. 4).This includes nitrogen amide hydrogen bondingthrough a buried and well-ordered water molecule (wa-ter molecule B-factor: 9.6 Å2; average protein main-chain B-factor: 15.3 Å2) to main-chain carbonyl groupsof residues Ser132 and Ile139 of strand bE, and inter-actions of the carbonyl and amide nitrogen of N179 tomain-chain and side-chain atoms of the substrate bind-ing segment (Met237 and Ser240). Inspection of theavailable 21 crystal structures of SDR enzymes reveala similar pattern of hydrogen bonding interactions in16 of the structures (Table 2), involving the Asn resi-due, whereas in 4 structures hydrophobic interactions

ed by crystallography of wild-type 3b/17b-HSD at 1.2-Å resolutioneractions of the central Asn179 residue with neighboring secondary

minint

TABLE 2

Vol. 289, No. 3, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

predominate (1a4u, 1dhr, 1e3w, 1nas), and in 1 struc-ture hydrogen bonding is mediated through a Ser res-idue (1a27). The altered unfolding characteristics ofthe N179A mutant point to an apparent increase instability, possibly due to loss of the buried water moleculeresulting in subsequent stabilization by hydrophobic in-teractions. However, the exact change in conformationalstability cannot be determined from this analysis as theunfolding of both enzymes is not fully reversible.

DISCUSSION

In this study, we identified an unknown, largelyconserved site in SDR enzymes, and have deduced itsfunction through mutational, enzymological and struc-tural analysis. Since SDRs form a functionally hetero-geneous protein family with considerable sequencevariation and possible insertions between conservedsegments and motifs, the elucidation of functional con-sequences of conserved sites is not a straightforwardtask. In this study, however, we used a structuralalignment to identify a short stretch of conserved I-R-V-N residues within strand bF in the 21 crystallizedSDR forms. This strand is part of the central b-sheet ofthe Rossmann-nucleotide binding fold and has severalimportant structural functions within the SDR scaf-fold. This segment interacts with relevant parts of theenzyme, the active site and the substrate binding re-

Structure Interacting residues

bF-substrate binding segmen

Distance(Å)

Distance(Å)

Hydrogen bon

1hxh Asn 179 ND2-Met 237O 3.00 Asn 179 OD1-Ser 240 N 2.87 AsnAsn 179 OD1-Ser 240 OG 3.03

1doh Asn 204 ND2-Val 270 O 2.90 Asn 204 OD1-Lys 273 N 3.18 Asn1geg Asn 178 ND2-Met 242 O 2.92 Asn 178 OD1-Gln 245 N 2.92 Asn

Wa1i01 Asn 177 ND2-Ile 230 O 3.03 Asn 177 OD1-Glu 233 N 3.05 Asn1edo Asn 193 ND2-Ile 247 O 2.79 Asn 193 OD1-Gln 250 N 3.22 Asn1gco Asn 184 ND2-Val 239 O 3.12 Asn 184 OD1-Ile 242 N 3.10 Asn1ahh Asn 185 ND2-Val 239 O 2.64 Asn 185 OD1-Gln 242 N 2.89 Asn

Asn 178 ND2-Wat 2.40 Ser 132 OG-Wat 2.41 Ile1cyd Asn 175 ND2-Thr 230 O 2.86 Asn 175 OD1-Glu 233 N 3.83 Asn1ae1 Asn 197 ND2-Ile 256 O 3.07 Asn 197 OD1-Gln 259 N 3.04 Asn2ae1 Asn 185 ND2-Val 243 O 3.04 Asn 185 OD1-Gln 246 N 2.77 Asn1bdb Asn 180 ND2-Ala 244 O 3.04 Asn 180 OD1-Ala 247 N 2.89 Asn1bvr Asn 187 ND2-Thr 244 O 2.73 Asn 187 OD1-Asp 247 N 3.06 Asn1qsg Asn 185 ND2-Ile 240 O 3.19 Asn 185 OD1-Glu 243 N 2.93 Asn1d7o Asn 229 ND2-Ile 284 O 2.85 Asn 229 OD1-Ala 287 N 2.88 Asn1fjh Asn 181 ND2-Val 237 O 2.93 Asn 181 OD1-Ala 240 N 2.82 Asn1a27 Ser 181 OG-Wat 8 2.91 Ser 181 OG-Thr 250 O 2.53 Val

1hdc Asn 178 ND2-Val 232 O 3.09 Asn 178 OD1-Ala 235 3.08 Asn

Hydrophob

1a4u Tyr 194 CE2-Ile 133 CG2 3.71 Tyr 194 CE1-Ala 134 CB 3.46 Tyr1dhr Ile 174 CD1-Leu 127 CD1 3.59 Ile 174 CD1-Pro 218 CG 3.77 Ile1e3w Val 194 CG2-Val 149 CG1 5.68 Val 194 CG2-Leu 246 CG 4.62 Val1nas Leu 195 CD2-Thr 152 CG2 6.04 Leu 195 CD1-Phe 251 CD1 4.01 Leu

716

gion, which could explain the loss of function in theN179A mutant. Connections to the active site are madeby Asn179 through H-bonds via a water molecule tostrand bE, which itself contains Ser139 as part of thecatalytic triad. Interactions to the substrate bindingloop comprising the most variable, C-terminal part ofSDR enzymes are obtained by various side-chain/main-chain interactions. It should be stated, that in thissystem, except for Asn179 and the water molecule, theinteracting residues are not conserved (Table 2). Thisarrangement, side-chain interactions of a conservedresidue mainly with peptide main chain atoms, cantolerate a high degree of sequence variation within thesubstrate binding and active site region, and stillmaintain the general scaffold. Accordingly, flexibilityand dynamics of the substrate binding region appear toplay an important role in SDR catalysis. The substrateloop acquires a defined structure from an unorderedstate upon substrate binding as determined in severalcrystallographic analyses (16). In line with these ob-servations, the altered unfolding characteristics andloss-of-function observed in this study point to an es-sential interaction of the substrate loop with strand bF.

Previous studies with mammalian 11b-HSD haveidentified an essential Asn-linked glycosylation site(207 in the human form, cf. Fig. 1) by site directedmutagenesis (14). By sequence comparison we identi-fied the Asn of the 11b-HSD glycosylation motif to

bE-bF

Distance(Å)

Distance(Å)

Distance(Å)

interactions

ND2-Wat1 2.92 Ser 132 OG-Wat 1 2.92 Ile 133 O-Wat1 2.90

ND2-Wat 3.08 Leu 159 O-Wat 2.94ND2-Wat 64 2.90 Ile 134 O-Wat 97 2.96 Wat 64-Wat 97 2.77

Leu 234 O 2.59 Wat 97-Thr 176 2.76ND2-Wat 2.99 Ile 133 O-Wat 3.17 Wat-Thr 175 OG1 2.93ND2-Wat 35 3.01 Ile 135 O-Wat 35 3.20 Wat 35-Wat 15 2.80ND2-Wat 2.92 Thr 139 OG-Wat 2.18 Val 140 O-Wat 4.23ND2-Wat 2.61 Ile 141 O-Wat 3.58 Wat-Leu 231 O 3.08

O-Wat 3.11ND2-Wat 2.86 Ser 130 OG-Wat 2.30 Ile 131 O-Wat 2.83ND2-Wat 2.93 Asn 152 ND2-Wat 2.69 Val 153 O-Wat 3.52ND2-Wat 3.49 Asn 140 ND2-Wat 2.91 Val 141 O-Wat 3.02ND2-Wat 2.98 Asn 136 ND2-Wat 2.72 Val 137 O-Wat 2.85ND2-Wat 3.29 Ser 143 OG-Wat 2.57 Ile 144 O-Wat 3.17ND2-Wat 3.00 Leu 141 O-Wat 3.02 Wat-Leu 171 O 2.97ND2-Wat 3.26 Ser 183 O-Wat 2.91 Wat-Leu 232 O 2.97ND2-Wat 3.10 Ala 108 O-Wat 2.95 Wat-Leu 229 O 3.03O-Wat 8 2.91 Wat 8-Wat 12 2.93 Wat 12-Arg 136 nh1 2.80

Wat 12-Ala 243 O 2.84ND2-Wat 2.47 Wat-Ser 133 2.39 Ile 134 O-Wat 3.02

teractions

OH-Wat 2.54 Asn 228 ND2-Wat 2.79 Ile 225 O-Wat 2.73CD1-Lys 217 CG 3.73CB-Leu 151 CG1 4.82CD2-Val 154 CG2 4.76

t

ding

179

204178

t 64-177193184185

134175197185180187185229181137

178

ic in

194174194195

correspond to Asn179 of bacterial 3b/17b-HSD. In ab- 5. Duax, W. L., Griffin, J. F., and Ghosh, D. (1996) The fascinating

Vol. 289, No. 3, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

sence of any available 11b-HSD crystal structure wepredict similar essential interactions, e.g., Asn side-chain interactions to bE and the substrate loop. Thisnotion is further supported by recent observations thatthe carbohydrate moiety itself is dispensable for enzy-matic activity (17, 18) and likely necessary for correctfolding in the ER as determined in tunicamycin inhi-bition studies (9, 19).

Other motifs and residues immediately distal to bFhave been investigated in some detail, underliningtheir importance. The PG motif [(1, 9), correspondingto residues 189–190 in bacterial 7a-HSD, 1ahh] hasbeen noted early, and recently a functional role for thismotif in catalysis has been postulated by Ghosh andco-workers (20). Inferred from crystal structure analy-sis it is suggested that main-chain carbonyls of the Proand Gly residues allow interactions with the coenzyme,thus dictating the reaction direction. Four residuesaway from the PG motif a single, apparently conservedThr residue has been observed. Mutational studies us-ing NAD1-dependent prostaglandin dehydrogenaseand crystallographic analysis of available SDR en-zymes provided evidence for this Thr residue to main-tain interaction to the nicotinamide carboxamidestructure, important for catalysis (21).

Taken together, these results highlight the impor-tance of several less well studied structural motifs inSDR enzymes. Further research is required to deter-mine all essential structure-function relationships ofthis biomedically important family of enzymes. As partof this task, ongoing crystallization studies in our lab-oratories are directed to reveal the exact structuralalterations of mutants of SDR-type hydroxysteroid de-hydrogenases.

ACKNOWLEDGMENTS

Fruitful discussions with Lars Abrahmsen, Biovitrum AB, Stock-olm, are gratefully acknowledged. This study was supported by theEuropean Community (BIO 4 EC 97-2123), the NOVO Nordisk Foun-dation, the Swedish Medical Research Council, the Swedish Union ofPhysicians, and the Karolinska Institutet.

REFERENCES

1. Jornvall, H., Persson, B., Krook, M., Atrian, S., Gonzalez-Duarte,R., Jeffery, J., and Ghosh, D. (1995) Short-chain dehydrogenases/reductases (SDR). Biochemistry 34, 6003–6013.

2. Jornvall, H., Hoog, J. O., and Persson, B. (1999) SDR and MDR:Completed genome sequences show these protein families to belarge, of old origin, and of complex nature. FEBS Lett. 445,261–264.

3. Krozowski, Z. (1992) 11 Beta-hydroxysteroid dehydrogenase andthe short-chain alcohol dehydrogenase (SCAD) superfamily. Mol.Cell. Endocrinol. 84, C25–C31.

4. Oppermann, U. C., Filling, C., and Jornvall, H. (2001) Forms andfunctions of human SDR enzymes. Chem.–Biol. Interact. 130–132, 699–705.

717

complexities of steroid-binding enzymes. Curr. Opin. Struct.Biol. 6, 813–823.

6. Duax, W. L., and Ghosh, D. (1997) Structure and function ofsteroid dehydrogenases involved in hypertension, fertility, andcancer. Steroids 62, 95–100.

7. Nobel, S., Abrahmsen, L., and Oppermann, U. (2001) Metabolicconversion as a pre-receptor control mechanism for lipophilichormones. Eur. J. Biochem. 268, 4113–4125.

8. Simard, J., Sanchez, R., Durocher, F., Rheaume, E., Turgeon, C.,Labrie, Y., Luu-The, V., Mebarki, F., Morel, Y., de Launoit, Y., etal. (1995) Structure–function relationships and molecular genet-ics of the 3 beta-hydroxysteroid dehydrogenase gene family. J.Steroid Biochem. Mol. Biol. 55, 489–505.

9. Oppermann, U. C., Persson, B., Filling, C., and Jornvall, H.(1997) Structure–function relationships of SDR hydroxysteroiddehydrogenases. Adv. Exp. Med. Biol. 414, 403–415.

10. Oppermann, U. C., Filling, C., Berndt, K. D., Persson, B.,Benach, J., Ladenstein, R., and Jornvall, H. (1997) Active sitedirected mutagenesis of 3b/17b-hydroxysteroid dehydrogenaseestablishes differential effects on short-chain dehydrogenase/reductase reactions. Biochemistry 36, 34–40.

11. Cols, N., Atrian, S., Benach, J., Ladenstein, R., and Gonzalez-Duarte, R. (1997) Drosophila alcohol dehydrogenase: Evaluationof Ser139 site-directed mutants. FEBS Lett. 413, 191–193.

12. Benach, J., Atrian, S., Gonzalez-Duarte, R., and Ladenstein, R.(1998) The refined crystal structure of Drosophila lebanonensisalcohol dehydrogenase at 1.9 Å resolution. J. Mol. Biol. 282,383–399.

13. Stewart, P. M., and Krozowski, Z. S. (1999) 11b-Hydroxysteroiddehydrogenase. Vitam. Horm. 57, 249–324.

14. Agarwal, A. K., Mune, T., Monder, C., and White, P. C. (1995)Mutations in putative glycosylation sites of rat 11b-hydroxysteroid dehydrogenase affect enzymatic activity. Bio-chim. Biophys. Acta 1248, 70–74.

15. Benach, J., Knapp, S., Oppermann, U. C., Hagglund, O., Jorn-vall, H., and Ladenstein, R. (1996) Crystallization and crystalpacking of recombinant 3 (or 17) b-hydroxysteroid dehydroge-nase from Comamonas testosteroni ATTC 11996. Eur. J. Bio-chem. 236, 144–148.

16. Benach, J., Atrian, S., Gonzalez-Duarte, R., and Ladenstein, R.(1999) The catalytic reaction and inhibition mechanism of Dro-sophila alcohol dehydrogenase: Observation of an enzyme-boundNAD-ketone adduct at 1.4 Å resolution by X-ray crystallography.J. Mol. Biol. 289, 335–355.

17. Walker, E. A., Clark, A. M., Hewison, M., Ride, J. P., and Stew-art, P. M. (2001) Functional expression, characterization, andpurification of the catalytic domain of human 11-beta-hydroxysteroid dehydrogenase type 1. J. Biol. Chem. 276,21343–21350.

18. Blum, A., Martin, H. J., and Maser, E. (2000) Human 11beta-hydroxysteroid dehydrogenase type 1 is enzymatically active inits nonglycosylated form. Biochem. Biophys. Res. Commun. 276,428–434.

19. Agarwal, A. K., Tusie-Luna, M. T., Monder, C., and White, P. C.(1990) Expression of 11 beta-hydroxysteroid dehydrogenase us-ing recombinant vaccinia virus. Mol. Endocrinol. 4, 1827–1832.

20. Ghosh, D., and Vihko, P. (2001) Molecular mechanisms of estro-gen recognition and 17-keto reduction by human 17b-hydroxysteroid dehydrogenase 1. Chem.–Biol. Interact. 130–132, 637–650.

21. Zhou, H., and Tai, H. H. (1999) Threonine 188 is critical forinteraction with NAD1 in human NAD1-dependent 15-hydroxyprostaglandin dehydrogenase. Biochem. Biophys. Res.Commun. 257, 414–417.