The Mo-Se active site of nicotinate dehydrogenaseNadine Wagenera, Antonio J. Pierikb, Abdellatif Ibdahc, Russ Hillec, and Holger Dobbeka,1

aBioanorganische Chemie, Universitat Bayreuth, 95440 Bayreuth, Germany; bInstitut fur Zytobiologie, Philipps-Universitat Marburg, 35037 Marburg,Germany; and cDepartment of Biochemistry, University of California, Riverside, CA 92521

Edited by Douglas C. Rees, California Institute of Technology, Pasadena, CA, and approved May 19, 2009 (received for review February 27, 2009)

Nicotinate dehydrogenase (NDH) from Eubacterium barkeri is amolybdoenzyme catalyzing the hydroxylation of nicotinate to6-hydroxynicotinate. Reactivity of NDH critically depends on thepresence of labile (nonselenocysteine) selenium with an as-yet-unidentified form and function. We have determined the crystalstructure of NDH and analyzed its active site by multiple wave-lengths anomalous dispersion methods. We show that selenium isbound as a terminal MoASe ligand to molybdenum and that itoccupies the position of the terminal sulfido ligand in othermolybdenum hydroxylases. The role of selenium in catalysis hasbeen assessed by model calculations, which indicate an accelera-tion of the critical hydride transfer from the substrate to theselenido ligand in the course of substrate hydroxylation whencompared with an active site containing a sulfido ligand. TheMoO(OH)Se active site of NDH shows a novel type of utilizationand reactivity of selenium in nature.

Eubacterium barkeri � hydroxylase � molybdoprotein �molybdopterin � selenium

Selenium is an essential component of several enzymes andhas a key role in various biological redox processes. Usually

selenium occurs in proteins as selenocysteine, which is cotrans-lationally inserted as the 21st amino acid (1) and is found in avariety of proteins in all 3 kingdoms of life (2). Selenium alsofinds a natural use as 5-methylaminomethyl-2-selenouridine inthe ‘‘wobble’’ position of some tRNAs (3). The ionic radii andelectronegativities of selenium and sulfur are similar, but se-lenide is a stronger reducing agent than sulfide. Because of thelower pKa value of selenols compared with thiols selenocysteineis deprotonated under physiological conditions, whereas cysteineis mostly protonated (4). The ionization state together with thebetter polarizability of selenium makes selenocysteine a goodnucleophile. Several molybdenum- and tungsten-containing en-zymes have been shown to contain selenium (5), and a recentcomprehensive genomic analysis has revealed a clear relation-ship between selenium and molybdenum utilization across all 3domains of life (6). Selenium is found as selenocysteine in someprokaryotic molybdenum-containing oxotransferases like for-mate dehydrogenase H from Escherichia coli, where it coordi-nates molybdenum in the oxidized state of the enzyme (7–9). Inother enzymes, notably members of the molybdenum hydroxy-lase family (10, 11), like nicotinate dehydrogenase (NDH) fromEubacterium barkeri (12–14) and the xanthine oxidoreductases(XORs) of Clostridium purinolyticum (15), Clostridium acidiurici(16) and Eubacterium barkeri (17) and the purine hydroxylasefrom C. purinolyticum (15) contain a labile (nonselenocysteine)selenium in an unidentified form essential for their reactivity (18).

Molybdenum hydroxylases catalyze the hydroxylation of var-ious organic molecules following the general scheme:

RH � H2O3 ROH � 2e� � 2H�.

This hydroxylation reaction is unique in biology as it uses wateras the source for the hydroxyl oxygen and not dioxygen (10). Theactive site of molybdenum hydroxylases contains molybdenumcoordinated by the enedithiolate group of a pyranopterin co-factor, commonly referred to as molybdopterin. The crystalstructures of different molybdenum hydroxylases in complex

with substrates and inhibitors are known (19–27). These struc-tures demonstrate that the molybdenum coordination spheretypically consists of 1 oxo, 1 hydroxo, and 1 labile sulfido ligandin a square-pyramidal coordination sphere, with the MoAOoccupying the apical position. The sulfido ligand is prone tocyanolysis, and its replacement with a second oxo ligand in theprocess leads to inactivation of the enzymes (10). An extensionof the active site is found in aerobic carbon monoxide dehydro-genases. While in an earlier report on the structure the presenceof selenium in the active site was suggested (25), a latter studyusing analytical multiple wavelength anomalous dispersionmethods showed that the active site does not contain seleniumbut a linearly-coordinated Cu(I) ion bridged by a �-sulfido ligandto the molybdenum (28). This �-sulfido bridge occupies the positionof the MoAS seen in other molybdenum hydroxylases.

The anaerobic soil bacterium E. barkeri is able to fermentnicotinate to propionate, acetate, carbon dioxide, and ammoniawith the gain of 1 mol of ATP per mol of nicotinate. Thefermentation of nicotinate is initiated by its hydroxylation to6-hydroxynicotinate catalyzed by NDH (12). NDH has a (����)2subunit structure and contains [2Fe-2S] clusters, FAD and amolybdenum center with a pyranopterin cofactor that has beenelaborated as the dinucleotide of cytosine and termed molyb-dopterin cytosine dinucleotide (MCD) (18, 29). The genesencoding NDH occur in the transcriptional order ndhFSLM andare part of a 23.3-kb gene cluster dedicated to the fermentationof nicotinate (30). The NdhF subunit (33 kDa) carries 1 FADmolecule and the NdhS subunit (23 kDa) contains 2 [2Fe-2S]clusters. Contrary to all structurally characterized hydroxylasesthe molybdenum cofactor appeared to be contained not in 1 but2 subunits: the NdhL subunit (50 kDa) and NdhM subunit (37kDa). The most remarkable feature of NDHs is the presence oflabile (nonselenocysteine) selenium (14), which is essential tocatalyze the hydroxylation reaction (18). Here, we report theX-ray crystal structure of NDH and its Se-containing active site,together with computational studies demonstrating the catalyticprofit of the natural selection of Se over its congeners S and O.

ResultsOverall Structure. NDH was crystallized by vapor diffusion meth-ods under anoxic conditions in an atmosphere of 95% N2/5% H2.Crystals belonged to the space group P21 and contained thecomplete dimer of heterotetramers in 1 asymmetric unit. Thestructure of NDH was determined with Patterson search tech-niques by using a substructure of 4-hydroxybenzoyl-CoA reduc-tase (23) as search model. The final model contains all residuesand was refined to 2.2-Å resolution (Table 1).

Author contributions: N.W., R.H., and H.D. designed research; N.W., A.J.P., and A.I. per-formed research; N.W., A.J.P., A.I., R.H., and H.D. analyzed data; and A.J.P., R.H. and H.D.wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The atomic coordinates have been deposited in the Protein Data Bank,www.pdb.org (PDB ID codes 3HRD).

1To whom correspondence should be addressed. E-mail: holger.dobbek@uni-bayreuth.de.

This article contains supporting information online at www.pnas.org/cgi/content/full/0902210106/DCSupplemental.

www.pnas.org�cgi�doi�10.1073�pnas.0902210106 PNAS � July 7, 2009 � vol. 106 � no. 27 � 11055–11060

BIO

CHEM

ISTR

Y

Dow

nloa

ded

by g

uest

on

Mar

ch 2

8, 2

021

NDH is a dimer of heterotetramers with (FSLM)2 subunitcomposition in solution (12) and all subunits are present in thecrystal structure (Fig. 1). The dimer has overall dimensions of148 � 100 � 70 Å3. The F subunit (296 residues) harbors theFAD cofactor, which interacts with the N terminal (residues1–57) and middle domain (residues 58–178) through its adenosinand ribityl moiety. The C-terminal domain (residues 179–291)only interacts with FAD through K187F, whose amino group isin hydrogen-bonding distance to O4 of the isoalloxazine ring. Asin other molybdenum hydroxylases, the C-terminal domaincontributes to shield the N5 position of the isoalloxazine ringfrom the solvent. In molybdenum hydroxylases unable to reduce

NAD�/NADP� access to the N5 position is usually blocked bya tyrosine (carbon monoxide dehydrogenase from Oligotrophacarboxydovorans, quinoline 2-oxidoreductase) or tryptophanside chain (carbon monoxide dehydrogenase from Hydrog-enophaga pseudoflava). In contrast, the less bulky side chain ofisoleucine is found in the NAD�/NADP�-reducing NDH andbovine and bacterial XOR (21, 22). All 3 domains of the Fsubunit show a mixed �/�-fold. The S subunit (157 residues)comprises 2 domains, each coordinating 1 [2Fe-2S] cluster. TheN-terminal domain (residues 1–79) is similar to ‘‘plant-typeferredoxins’’ and binds the [2Fe-2S] cluster close to the FAD,which is designated as Fe/S II or type II on the basis of its EPRcharacteristics in homologous molybdenum hydroxylases (31–33). The C-terminal domain (residues 80–157) displays a 4-helixbundle with 2-fold symmetry unique to molybdenum hydroxy-lases (19) and coordinates the [2Fe-2S] cluster (type I) closest tothe molybdopterin. The L subunit (425 residues) and M subunit(330 residues) harbor the Mo-bound pyranopterin cofactor.Both subunits have an extended structure and lie approximatelyperpendicular on top of each other. The L subunit interacts withthe M and S subunits and has an N-terminal extension (residues1–28L) that wraps around the C-terminal domain of the Ssubunit. A middle domain (residues 29–129L and 181–277L) with2 antiparallel �-sheets of 2 and 7 strands and 3 �-helices followsthe N-terminal extension. The C-terminal domain of the Lsubunit is dominated by a mixed 5-stranded �-sheet flanked on1 side by 2 �-helices that continue into a 2-stranded antiparallel�-sheet and a C-terminal �-helix. The M subunit can be dividedinto 2 domains, both containing mixed 4-stranded �-sheets and3 �-helices (Fig. 1).

Both monomers of NDH show the same arrangement ofsubunits and cofactors building 2 independently working elec-tron transfer chains (Fig. 1), similar to other structurally char-acterized molybdenum hydroxylases. However, NDH is unusualin containing 4 subunits per monomer, whereas other molybde-num hydroxylases of known structure, contain 1, 2, or 3 subunits.Compared with all structurally characterized molybdenum hy-droxylases, the Mo-pyranopterin binding subunit/domain has

Table 1. Statistics on diffraction data and structure refinement

Statistic

Dataset

Above Se edge Below Se edge Native

Data collectionWavelength, Å 0.9780 0.9803 1.5418Space group P21 P21

Cell dimensions (a,b,c in Å, b in °) 100.00, 72.15, 217.10, 90.56 97.08, 71.70, 214.49, 90.23Total/unique reflections 396,539/208,277 360,921/196,213 519,267/148,165Rs, % 7.8 (31.0) 8.7 (44.8) 9.7 (53.6)Resolution, Å 30–2.5 (2.6–2.5) 30–2.5 (2.6–2.5) 30–2.2 (2.3–2.2)Completeness, % 99.2 (97.6) 93.5 (74.9) 98.9 (96.8)(I)/(�I) 6.8 (2.4) 6.5 (1.5) 10.8 (2.3)

RefinementModel Rwork/Rfree factor, % 21.2/25.0No. atoms

Protein 17,180Ligand 249Water 954

B factorsProtein 36.3Ligand 26.1Water 29.3

rmsdBond lengths, Å 0.006Bond angles, ° 1.3

In the datasets at the Se edge the values given are for unmerged Friedel mates. The values in parentheses indicate the highest resolution.

Fig. 1. Overall structure of NDH. Ribbon plot representation of the NDHdimer. In the left monomer each subunit has its own color with green and redfor the MCD coordinating L and M subunits, blue for the [2Fe-2S] clusterscontaining S subunit, and yellow for the FAD containing F subunit. The rightmonomer is colored in different shades of gray. Cofactors are labeled, and FeSI

and FeSII indicate the position of the type I and type II [2Fe-2S] clusters of theS subunit. The shortest distances between the cofactors are 5.4 Å (MCD-FeSI),11.5 Å (FeSI-FeSII), and 6.4 Å (FeSII-FAD), with the shortest metal-to-metaldistances of 14.7 Å (MCD-FeSI) and 12.4 Å (FeSI-FeSII). All pictures were pre-pared by using PyMol (48).

11056 � www.pnas.org�cgi�doi�10.1073�pnas.0902210106 Wagener et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

8, 2

021

been split into 2 polypeptides in NDH: the L subunit correspondsto the N-terminal domain and the M subunit is homologous tothe C-terminal domain. Bioinformatic analysis has revealed splitmolybdopterin subunits in 1 of 3 other bacterial NDHs (SI Textand Fig. S1). Moorella thermoacetica, Carboxydothermus hydrog-enoformans, Petrotoga mobilis, Clostridium asparagiforme, 5�-proteobacterial species, Alkaliphilus oremlandii, and Alkaliphi-lus metalliredigens have split subunits in NDH homologs ofhitherto unknown catalytic activity (SI Text and Fig. S2).

Active-Site Structure and Selenium Detection. The substrate channelis formed by residues from the M and L subunits, which createa funnel leading to the active site. The active site contains themolybdenum ion with a distorted square pyramidal coordination(Fig. 2A). Two enedithiolate sulfurs and 1 hydroxo-ligand occupythe equatorial positions at the Mo-ion and an oxo-ligand is foundin the apical position. The activity of NDH is known to dependon the presence of selenium in the active site, and different Seoccupancies depending on preparation and specific activity havebeen observed (14, 18, 34). Therefore, detection of seleniumcould not solely rely on the strength of the observed electrondensities, which depend on the occupancy of the ligands. Tolocate Se in the structure of NDH we collected completecrystallographic datasets at 2 different X-ray wavelengths, one atthe low-energy side (� � 0.9803 Å) and one at the high-energyside (� � 0.9780 Å) of the X-ray absorption K-edge of selenium(Table 1). Whereas Bijvoet difference maps of the low-energydataset show only a signal at the Mo ion, Bijvoet difference mapscalculated from the high-energy dataset reveal strong additional

anomalous scattering from only 1 of the equatorial ligands of theMo ion (Fig. 2B). That this position is indeed occupied byselenium is supported by the stronger electron density observedat this position and from a Mo ligand bond length of 2.3 Å (Fig.2C), which is longer than the typical MoOS bond (Table S1). Seis a direct ligand and coordinates Mo in the equatorial plane. Thelack of additional density around Se indicates that it is a terminalligand. We estimated the occupancy of the Se ligand to be �80%by adjusting the Se occupancy of the model such that the Se atomrefines to a similar B factor as the neighboring atoms includingthe other Mo ligands.

The second coordination sphere around the Mo-ion is formedby E289M in trans to the apical oxo-ligand, Q208L, in hydrogen-bonding distance to the apical oxo-ligand and residues that maycontribute to the binding and stabilization of the substrate, likeY312L, R319L, F353L, and Y13M (Fig. 2 A). An additional smallligand bound near the active site has been modeled as nitrate,which is present in the crystallization buffer. Nitrate interactswith the main chain of residues forming the substrate-bindingpocket and is 8 Å away from the Mo ion and 6 Å away from theSe. Nitrate is found at the same position where an acetate moleculehas been modeled in the crystal structure of bXOR (21) and bothshow the same pattern of interactions with the surrounding aminoacids. A physiological role for this conserved anion-binding sitenear the active site has not been demonstrated.

Structure-Based Reaction Mechanism of NDH. Increasing evidenceon the chemistry of XORs supports a mechanism of substratehydroxylation involving a base-assisted nucleophilic attack of theequatorial Mo-OH group on the substrate, with a concomitanthydride transfer from the substrate to the Mo(�VI)AS group togive Mo(�IV)-SH. Two residues in the direct environment ofmolybdenum, Q208L in NDH (Q767 in bXOR) and E289M(E1261 in bXOR), are conserved among the molybdenumhydroxylases. E1261 serves as general base catalyst in acceptingthe proton from Mo-OH upon reaction in bXOR (11) (Fig. S3)and, accordingly, replacement of this residue in bacterial XORby alanine profoundly compromises the reactivity of the enzyme(35). Additionally, NDH and bXOR have common residues inthe active site like R319L (R880 in bXOR) and F353L (F914 inbXOR) (Fig. S3). R880 of bXOR was suggested to stabilize thedeveloping negative charge on the substrate during the hydroxy-lation step, and its mutation results in an increase of thedissociation constant, KD, for substrate binding and a decreasein the rate constant of enzyme reduction, kred (36). The conser-vation of amino acids essential for catalysis indicates commonways of substrate binding and transition state stabilization inNDH and XORs. Based on these similarities (Fig. S3), weconstructed a structural model for nicotinate binding in theactive site of NDH in analogy to the structure of bXOR incomplex with 2-hydroxy-6-methylpurine (27). It has recentlybeen shown that 2-hydroxy-6-methylpurine binds between phe-nylalanine residues in the active site of bXOR, with the carbonatom to be hydroxylated in close proximity to the equatorialhydroxyl ligand. To allow binding of nicotinate in the active siteof NDH the side chain of F353L has to rotate to assume a similarconformation as found for substrate/ligand-bound bXOR (21,27) (Fig. S3). Modeling of substrate binding in analogy to bXORproduces a mechanistically reasonable complex. The nitrogenatom and carboxylate of nicotinate are at hydrogen-bondingdistance to Y13M and R319M, respectively. C6, the carbon atomto be hydroxylated, is in a distance of �2 Å from the hydroxylligand and 2.5 Å from the Se ligand. This binding mode wouldfacilitate the nucleophilic attack of the hydroxyl ligand on C6 andthe concomitant hydride transfer from C6 to the Se ligand (Fig. S4).

Computational Study Comparing MoASe and MoAS. The functionaladvantage of selenium over sulfur as a ligand is not immediately

Fig. 2. The Mo-Se active site of NDH. (A) Stereoview of the active-siteenvironment. Transparent ribbons are colored as in Fig. 1 (left monomer). (B)Selenium identification by anomalous scattering detected at energies higher(� � 0.9780 Å in red) and smaller (� � 0.9803 Å in green) than the energy ofthe Se K edge. Both Bijvoet difference maps are contoured at the �5.0-� level.(C) Fobs � Fcalc map for the Mo and pyranopterin cofactor at a contour level of�5.0 �. For the calculation of the map all atoms of the Mo ion and thepyranopterin cofactor were omitted.

Wagener et al. PNAS � July 7, 2009 � vol. 106 � no. 27 � 11057

BIO

CHEM

ISTR

Y

Dow

nloa

ded

by g

uest

on

Mar

ch 2

8, 2

021

evident. The molybdenum sulfido ligand found in the active siteof molybdenum hydroxylases like XOR (MoAS) plays an im-portant role in the catalytic activity of these enzymes. Itsreplacement by a (second) MoAO group to give the so-calleddesulfo form of the enzyme leads to its complete inactivation.The hydride transfer taking place in the initial step of thereaction can be considered a nucleophilic attack of the hydrideon the MoAE antibonding orbital (*) (E � O, S). The stronger interaction leads to higher * anti-bonding energy, whichraises the barrier of the reaction. The strength of the MoAO bond relative to that of MoAS leads to a very high * anti-bonding orbital and is presumably the basis for the lack ofactivity seen in the desulfo form of the enzymes (Fig. 3).

A MoASe rather than MoAS is expected to yield an evenweaker bond with molybdenum than sulfur and oxygen (O �S � Se) (Fig. 3) and is thus expected to increase the reactivityof the enzyme in any mechanism involving hydride transfer. Inthe case of less reactive carbon centers, the increased reactivitymay be important in catalyzing the hydroxylation. Computa-tional studies have been carried out after the reaction of amolybdenum center such as that found in bXOR and NDH, but

with an equatorial MoASe rather than MoAS group, withethylaldehyde and formamide as simple substrates (Fig. 4 A andB, respectively). Ethylaldehyde and formamide were used assimple analogs for xanthine and nicotinate to reduce the nec-essary computational resources. Additionally, it is not necessaryto include alternative protonation states with these 2 substrates.The calculations indicate that the transition state for the firststep of the reaction is stabilized by 3.1–3.4 kcal/mol uponsubstitution of selenium for sulfur in the reaction with bothethylaldehyde (�H‡ is 3.47 kcal/mol for MoAS and 0.13 kcal/molfor MoASe) and formamide (�H‡ is 14.81 kcal/mol for MoASand 11.67 kcal/mol for MoASe). This amounts to a factor of200–300 in rate acceleration for this step.

The geometry of the optimized MoASe-containing structureis square pyramidal, with the MoAE (E � S, Se), the hydroxylgroup, and the enedithiolate ligand of the pyranopterin cofactorin the equatorial plane (Fig. S5). Table S1 gives the metricparameters for both MoAS and MoASe structures. It can beseen that substitution of selenium for sulfur does not have asignificant influence on the bond angles made by the ligands,although the MoASe is longer than the MoAS (2.31 vs. 2.18 Å),as expected given the larger selenium and weaker MoASe bond.The calculated geometry of the selenium-containing model isvery similar to what has been found in the structure of NDH andboth show a Mo-Se distance of �2.3 Å.

DiscussionThe nature of the selenium moiety of NDH has been investigatedover the last decade by using various approaches. After thediscovery that selenium is essential for hydroxylase activity (13),Dilworth (14) showed that selenium is a component of NDH,which could be released by heat treatment or the addition ofchaotrophic agents like urea. That selenium in NDH is not partof a stable organic molecule but that is instead present as aneutral selenol or an inorganic selenide has been indicated by theliberation of selenium from NDH by incubation with alkylatingagents, resulting in the formation of dialkylselenides (14). Theapproximate ratio of 1 mol of selenium per mol of NDH wasestablished by EPR studies of NDH with the 77Se isotope thatshowed the nuclear spin of 77Se couples to the Mo(V) electronspin, suggesting that the selenium could be a ligand to molyb-

Fig. 3. Molecular orbital view and the relative energy of MoAE and * (E �O, S, Se). Depicted are the bonding interactions between the dxy orbital of themolybdenum and the px orbital of the coordinated chalcogen in bonding(lower) and antibonding configurations, with the relative energies forMoAO, MoAS, and MoASe species indicated.

Fig. 4. Calculated reaction coordinates. (A) Reaction of ethylaldehyde with the MoO(OH)S (Left) and the MoO(OH)Se analogue (Right). (B) Reaction offormamide with the MoO(OH)S (Left) and the MoO(OH)Se analogue (Right).

11058 � www.pnas.org�cgi�doi�10.1073�pnas.0902210106 Wagener et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

8, 2

021

denum (18). This left the possibilities that selenium replaced thesulfido ligand at the molybdenum or that it could be weaklybound to a heteroatom of a separate cofactor adjacent tomolybdenum. The crystal structure of NDH presented hereshows that selenium indeed replaces sulfur as a molybdenumligand, present as MoASe. The refined bond length of 2.3 Åagrees well with the expected bond length for such a terminalselenido ligand and specifically is too short for a selenol ligand(37). No further stabilization of the selenium by other interac-tions is observed, nor do we find residues such as cysteines in thevicinity of the Mo ion, which might form a bond to the selenidoligand. Access to the selenido ligand is partly blocked by F353L(Fig. 2 A), which could explain why incubation of active NDHwith potassium cyanide does not lead to a rapid inactivation ofthe enzyme and inactive enzyme could not be reactivated by theaddition of sodium selenide or selenophosphate (18, 34).

Whether selenium is also a ligand to molybdenum in otherselenium-dependent molybdenum hydroxylases remains to beestablished in future structural studies. For the selenium-containing XOR from E. barkeri it has been shown that theenzyme can be inactivated with potassium cyanide and enzymethus inactivated can be reactivated by incubation with selenideunder reducing conditions (17), observations that could beexplained by the presence of a selenido ligand bound to molyb-denum such as is seen here with NDH. However, with the purinehydroxylase from C. purinolyticum no magnetic interaction be-tween the nuclear spin of 77Se and the Mo(V) electron spin couldbe detected, raising the possibility that the labile selenium is notcoordinated to Mo in all forms of this enzyme (38). The catalyticadvantage of the incorporation of selenium in a molybdenumhydroxylase is evident from comparing XORs of differentorganisms. While for bXORs with a sulfido ligand at themolybdenum turnover number of 2–25 s�1 has been reported(39), the selenium containing XOR from E. barkeri achievesturnover rates �400 s�1 (17). If the XOR from E. barkeri has aselenido ligand like NDH, the higher rates are likely caused byan acceleration of the hydride transfer step, as suggested by ourmodel calculations.

For the molybdenum hydroxylase family we now see that atleast 3 variations for the common theme MoO(OH)X coordi-nation sphere exist, where X is: (i) S for XORs (40), quinolineoxidoreductase (24), and 4-hydroxybenzoyl-CoA reductase (23);(ii) SCu for carbon monoxide dehydrogenase (28); and, asdetailed above (iii) Se for NDH. All active sites can be convertedto an inactive state in which X is O, to give a second MoAOligand. These variations demonstrate how nature adapts a pro-tein bound ligand-metal motif for different substrates andreactivities by the exchange of 1 ligand and indicate a flexiblebiological chemistry of molybdenum enzymes by ligand tuning.

Materials and MethodsPurification of NDH. As NDH is instable and looses activity with time (34) ourexperimental strategy was to take �1 week from breaking the E. barkeri cells

used to purify the protein until freezing the protein crystals used for structuredetermination. For all steps, including crystallization of NDH, buffer condi-tions were used for which NDH was reported to be most stable and active (34).Frozen E. barkeri cells were resuspended in 50 mM Tris�HCl (pH 7.8), 10 mMNaCl, 1 mM EDTA, and 2 mM DTT (buffer A) and broken by 5 cycles ofsonication. After centrifugation for 30 min the cell-free extract was loaded ona Source 30Q column. NDH was eluted from the anionic exchange column bya linear gradient of buffer A containing 0.5 M NaCl. Pooled fractions wereloaded onto a hydroxyapatite column without preceding buffer exchange.Active NDH eluted with 120 mM KPO4, pH 7.3. Active fractions were collected,rebuffered in 50 mM Tris�HCl (pH 7.8), 0.2 M KCl, 1 mM EDTA, and 2 mM DTTand concentrated to 20 mg/mL. The protein was stored at �4 °C and usedwithin 5 days. Approximately 1.5 mg NDH could be obtained from 10 g of cells.To obtain protein with higher purity, an additional gelfiltration step (Super-dex 200) was performed in 50 mM Tris�HCl (pH 7.8), 0.2 M KCl, 1 mM EDTA, and2 mM DTT. All purification steps were carried out under anoxic conditions ina glove box containing 95% N2 and 5% H2.

Activity Measurement. Enzyme assays were conducted according to Gladyshevet al. (34) with some modifications. Hydroxylase activity was measured underanoxic conditions in 100 mM KPO4 (pH 7.0), 50 mM nicotinate (pH 7.5), 1 mMNADP�, and 5 mM DTT. The reaction was started by addition of NADP� andfollowed by absorption increase at 340 nm.

Crystallographic Methods. NDH was crystallized by the vapor diffusion methodin a hanging drop. The enzyme preparations used for crystallization hadspecific activities of 11–20 units�mg�1. The drop contained a 1:1 mixture ofprotein solution (10 mg/mL�1) with reservoir solution containing 18–20% PEG3350, 0.1 M Tris�HCl (pH 7.5), 75 mM NaNO3, 5% glycerol, or 1% 2,4-methylpentandiol. Crystals were harvested in the corresponding soakingsolutions containing 15% (vol/vol) (2R,3R)-butanediol (Sigma), shock-frozen,and stored in liquid nitrogen. Diffraction data were collected at �180 °C on arotating anode X-ray generator (Nonius FR591; Bruker) equipped with animage plate detector (mar345dtb; Marresearch) (native, Table 1) and at theBM-14, European Synchrotron Radiation Facility, Grenoble (above and belowSe-edge, Table 1). The protein structure model was build with Coot (41) andMAIN 2000 (42). Positional and temperature refinement was carried out withCNS (43). Anomalous difference Fourier maps were calculated by using CNS(43). The final refinement statistics and stereochemical analyses using PRO-CHECK (44) are shown in Table 1. Ramachandran statistics revealed 86% in themost favored, 13% in the additionally allowed, 1% in the generously allowed,and 0% in the disallowed regions. Simulated annealing omit maps were usedto validate the active-site structure reported.

Model Calculations. Calculations were carried out by using hybrid densityfunctional theory (B3LYP) as implemented in Gaussian 03. The structures werefully optimized and confirmed as minima or transition state by calculating thevibrational frequency. The structures were optimized with B3LYP usingLANL2DZ ECP (45, 46) augmented with f polarization functions (47) for Mo and6–31G(d) for the rest of atoms. For the hydrogen that performs the hydridetransfer 6–31��G(d,p) was used. The zero point energies were calculatedwith same basis set and level of theory. All quoted energies are corrected tozero point energy and are without temperature correction.

ACKNOWLEDGMENTS. H.D. was supported by Deutsche Forschungsgemein-schaft Grant DO 785/2–2 and the Fonds der Chemischen Industrie. R.H. andH.D. received travel funds from the Bavaria California Technology Center toinitiate the collaboration.

1. Bock A, et al. (1991) Selenocysteine: The 21st amino acid. Mol Microbiol 5:515–520.2. Stadtman TC (1996) Selenocysteine. Annu Rev Biochem 65:83–100.3. Stadtman TC (1990) Selenium biochemistry. Annu Rev Biochem 59:111–127.4. Wessjohann LA, Schneider A, Abbas M, Brandt W (2007) Selenium in chemistry and

biochemistry in comparison to sulfur. Biol Chem 388:997–1006.5. GladyshevVN(2002)Comparisonofselenium-containingmolybdoenzymes.Molybdenum

and tungsten: Their roles in biological processes. Metal Ions Biol Syst 39:655–672.6. Zhang Y, Gladyshev VN (2008) Molybdoproteomes and evolution of molybdenum

utilization. J Mol Biol 379:881–899.7. Gladyshev VN, Khangulov SV, Axley MJ, Stadtman TC (1994) Coordination of selenium

to molybdenum in formate dehydrogenase H from Escherichia coli. Proc Natl Acad SciUSA 91:7708–7711.

8. Boyington JC, et al. (1997) Crystal structure of formate dehydrogenase H: Catalysis involv-ing Mo, molybdopterin, selenocysteine, and an Fe4S4 cluster. Science 275:1305–1308.

9. Raaijmakers HC, Romao MJ (2006) Formate-reduced Escherichia coli formate dehydro-genase H: The reinterpretation of the crystal structure suggests a new reaction mech-anism. J Biol Inorg Chem 11:849–854.

10. Hille R (1996) The mononuclear molybdenum enzymes. Chem Rev 96:2757–2816.11. Hille R (2005) Molybdenum-containing hydroxylases. Arch Biochem Biophys 433:107–116.12. Holcenberg JS, Stadtman ER (1969) Nicotinic acid metabolism. 3. Purification and

properties of a nicotinic acid hydroxylase. J Biol Chem 244:1194–1203.13. Imhoff D, Andreesen JR (1979) Nicotinic acid hydroxylase from Clostridium barkeri:

Selenium-dependent formation of active enzyme. FEMS Microbiol Lett 5:155–158.14. Dilworth GL (1982) Properties of the selenium-containing moiety of nicotinic acid

hydroxylase from Clostridium barkeri. Arch Biochem Biophys 219:30–38.15. Self WT, Stadtman TC (2000) Selenium-dependent metabolism of purines: A selenium-

dependent purine hydroxylase and xanthine dehydrogenase were purified fromClostridium purinolyticum and characterized. Proc Natl Acad Sci USA 97:7208 –7213.

16. Wagner R, Cammack R, Andreesen JR (1984) Purification and characterization ofxanthine dehydrogenase from Clostridium acidiurici grown in the presence of sele-nium. Biochim Biophys Acta 791:63–74.

17. Schrader T, Rienhofer A, Andreesen JR (1999) Selenium-containing xanthine dehydro-genase from Eubacterium barkeri. Eur J Biochem 264:862–871.

Wagener et al. PNAS � July 7, 2009 � vol. 106 � no. 27 � 11059

BIO

CHEM

ISTR

Y

Dow

nloa

ded

by g

uest

on

Mar

ch 2

8, 2

021

18. Gladyshev VN, Khangulov SV, Stadtman TC (1994) Nicotinic acid hydroxylase fromClostridium barkeri: Electron paramagnetic resonance studies show that selenium iscoordinated with molybdenum in the catalytically active selenium-dependent enzyme.Proc Natl Acad Sci USA 91:232–236.

19. Romao MJ, et al. (1995) Crystal structure of the xanthine oxidase-related aldehydeoxido-reductase from Desulfovibrio gigas. Science 270:1170–1176.

20. Rebelo J, et al. (2000) Gene sequence and crystal structure of the aldehyde oxidoreduc-tase from Desulfovibrio desulfuricans ATCC 27774. J Mol Biol 297:135–146.

21. Enroth C, et al. (2000) Crystal structures of bovine milk xanthine dehydrogenase andxanthine oxidase: Structure-based mechanism of conversion. Proc Natl Acad Sci USA97:10723–10728.

22. Truglio JJ, et al. (2002) Crystal structures of the active and alloxanthine-inhibited forms ofxanthine dehydrogenase from Rhodobacter capsulatus. Structure (London) 10:115–125.

23. Unciuleac M, et al. (2004) Structure of a xanthine oxidase-related 4-hydroxybenzoyl-CoA reductase with an additional [4Fe-4S] cluster and an inverted electron flow.Structure (London) 12:2249–2256.

24. Bonin I, et al. (2004) Active site geometry and substrate recognition of the molybde-num hydroxylase quinoline 2-oxidoreductase. Structure (London) 12:1425–1435.

25. Dobbek H, Gremer L, Meyer O, Huber R (1999) Crystal structure and mechanism of COdehydrogenase, a molybdo iron-sulfur flavoprotein containing S-selanylcysteine. ProcNatl Acad Sci USA 96:8884–8889.

26. Hanzelmann P, et al. (2000) The effect of intracellular molybdenum in Hydrog-enophaga pseudoflava on the crystallographic structure of the seleno-molybdo-iron-sulfur flavoenzyme carbon monoxide dehydrogenase. J Mol Biol 301:1221–1235.

27. Pauff JM, Zhang J, Bell CE, Hille R (2008) Substrate orientation in xanthine oxidase:Crystal structure of enzyme in reaction with 2-hydroxy-6-methylpurine. J Biol Chem283:4818–4824.

28. Dobbek H, et al. (2002) Catalysis at a dinuclear [CuSMo(AAO)OH] cluster in a COdehydrogenase resolved at 1.1-Å resolution. Proc Natl Acad Sci USA 99:15971–15976.

29. Gladyshev VN, Lecchi P (1995) Identification of molybdopterins in molybdenum- andselenium-containing enzymes. Biofactors 5:93–97.

30. Alhapel A, et al. (2006) Molecular and functional analysis of nicotinate catabolism inEubacterium barkeri. Proc Natl Acad Sci USA 103:12341–12346.

31. Caldeira J, et al. (2000) Analysis of the electron paramagnetic resonance properties ofthe [2Fe-2S]1� centers in molybdenum enzymes of the xanthine oxidase family: As-signment of signals I and II. Biochemistry 39:2700–2707.

32. Canne C, et al. (1997) Comparative EPR and redox studies of three prokaryotic enzymesof the xanthine oxidase family: Quinoline 2-oxidoreductase, quinaldine 4-oxidase, andisoquinoline 1-oxidoreductase. Biochemistry 36:9780–9790.

33. Gremer L, et al. (2000) Binding of flavin adenine dinucleotide to molybdenum-containing carbon monoxide dehydrogenase from Oligotropha carboxidovorans.J Biol Chem 275:1864–1872.

34. Gladyshev VN, Khangulov SV, Stadtman TC (1996) Properties of the selenium- andmolybdenum-containing nicotinic acid hydroxylase from Clostridium barkeri. Bio-chemistry 35:212–223.

35. Leimkuhler S, et al. (2004) The role of active site glutamate residues in catalysis ofRhodobacter capsulatus xanthine dehydrogenase. J Biol Chem 279:40437–40444.

36. Pauff JM, et al. (2007) The role of arginine 310 in catalysis and substrate specificity inxanthine dehydrogenase from Rhodobacter capsulatus. J Biol Chem 282:12785–12790.

37. Ma X, Schulzke C, Schmidt HG, Noltemeyer M (2007) Structural, electrochemical, andoxygen atom transfer properties of a molybdenum selenoether complex[Mo2O4(OC3H6SeC3H6O)2] and its thioether analogue [Mo2O4(OC3H6SC3H6O)2]. DaltonTrans 2007:1773–1780.

38. Self WT, Wolfe MD, Stadtman TC (2003) Cofactor determination and spectroscopiccharacterization of the selenium-dependent purine hydroxylase from Clostridiumpurinolyticum. Biochemistry 42:11382–11390.

39. Pai EF, Nishino T (2002) The molybdenum-containing xanthine oxidoreductases andpicolinate dehydrogenases. Molybdenum and tungsten: Their roles in biological pro-cesses. Metal Ions Biol Syst 39:431–454.

40. Kuwabara Y, et al. (2003) Unique amino acids cluster for switching from the dehydro-genase to oxidase form of xanthine oxidoreductase. Proc Natl Acad Sci USA 100:8170–8175.

41. Emsley P, Cowtan K (2004) COOT: Model-building tools for molecular graphics. ActaCrystallogr D 60:2126–2132.

42. Turk D (1992) PhD thesis (Technische Universitat, Munich).43. Brunger AT, et al. (1998) Crystallography and NMR system: A new software suite for

macromolecular structure determination. Acta Crystallogr D 54:905–921.44. Laskowski RA, MacArthur MW, Moss DS, Thornton JM (1993) PROCHECK: A program to

check the stereochemical quality of protein structures. J Appl Crystallogr 26:283–291.45. Hay PJ, Wadt WR (1985) Ab initio effective core potentials for molecular calculations:

Potentials for the transition-metal atoms Sc to Hg. J Chem Phys 82:270–283.46. Hay PJ, Wadt WR (1985) Ab initio effective core potentials for molecular calculations:

Potentials for K to Au including the outermost core orbitals. J Chem Phys 82:299–310.47. Ehlers AW, et al. (1993) A set of F-polarization functions for pseudo-potential basis sets

of the transition metals Sc-Cu, Y-Ag and La-Au. Chem Phys Lett 208:111–114.48. DeLano WL (2002) PyMol (DeLano Scientific, San Carlos, CA).

11060 � www.pnas.org�cgi�doi�10.1073�pnas.0902210106 Wagener et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

8, 2

021