Structural and functional studies on two molybdopterin and iron-sulfur containing enzymes : transhydroxylase from pelobacter acidigallici and acetylene hydratase from pelobacter acetylenicus

(1)

S TRUCTURAL AND F UNCTIONAL STUDIES ON TWO M OLYBDOPTERIN AND I RON -S ULFUR CONTAINING E NZYMES :

T RANSHYDROXYLASE FROM P ELOBACTER ACIDIGALLICI AND A CETYLENE H YDRATASE FROM P ELOBACTER ACETYLENICUS

Z

UR

E

RLANGUNG DES AKADEMISCHEN

G

RADES EINES

D

OKTORS DER

N

ATURWISSENSCHAFTEN

DEM

F

ACHBEREICH

B

IOLOGIE

, U

NIVERSITÄT

K

ONSTANZ VORGELEGTE

D

ISSERTATION VON

M. S

C

. C

HEM

. G

RAŻYNA

B

ERNADETA

S

EIFFERT

KONSTANZ,MAI 2007

REFERENT:PROF.DR.P.M.H.KRONECK

KOREFERENT:PROF.DR.O.EINSLE Konstanzer Online-Publikations-System (KOPS)

URL: http://www.ub.uni-konstanz.de/kops/volltexte/2007/3935/

URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-39357

(2)

(3)

Z podziękowaniem dla mojej kochanej mamy oraz Piotrkowi

(4)

(5)

I. Zusammenfassung

(8)

Zusammenfassung

In dieser Doktorarbeit wurden zwei neuartige Molybdopterin-abhängige Enzyme aus strikt anaeroben Bakterien untersucht. Sowohl die Pyrogallol-Phloroglucin Transhydroxylase (TH) als auch die Acetylen Hydratase (AH) katalysieren chemisch ungewöhnliche Reaktionen unter dem strikten Ausschluss von Luftsauerstoff. Obwohl es sich bei beiden Reaktionen formal um keine Redoxreaktionen handelt, setzen beide Enzyme komplexe Metallzentren ein für die Katalyse.

I.a. Transhydroxylase aus Pelobacter acidigallici

Das strikt anaerobe Bakterium Pelobacter acidigallici vergärt Gallussäure (3,4,5- Trihydroxybenzoesäure), Pyrogallol (1,2,3-Trihydroxybenzol), Phloroglucin (1,3,5-Tri- hydroxybenzol) oder 2,4,6-Trihydroxybenzoesäure zu drei Molekülen Acetat (plus CO2) [Brune et al., 1990; Schink et al., 1982]. Ein Schlüsselenzym dieses Abbauweges ist die Pyrogallol-Phloroglucinol Transhydroxylase (TH), welche in Abwesenheit von O2

Pyrogallol zu Phloroglucinol umsetzt – in Anwesenheit von 1,2,3,5-tetrahydroxybenzene als Cosubstrat. Ein möglicher Reaktionsmechanismus wurde von Messerschmidt et al., (2004) auf der Grundlage einer hochaufgelösten Röntgenstruktur vorgeschlagen.

Die funktionelle Einheit der Transhydroxylase von P. acidigallici ist ein Heterodimer bestehend aus einer großen (100.4 kDa) und einer kleinen Untereinheit (31.3 kDa).

Dieses Enzym ist nahe verwandt mit Enzymen aus der Familie der DMSO-Reduktasen.

Obwohl die von der TH katalysierte Reaktion formal keine echte Redox-Reaktion ist, enthält das Enzym drei [4Fe-4S] -Zentren und ein Mo(MGD)2 als Redox-Kofaktoren.

Die EPR Spektren (Proben eingefroren bei verschiedenen Redoxpotentialen) zeigten einen Satz komplizierter Resonanzen in Abhängigkeit vom eingebauten Molybdän- Isotop, vom pH der Lösung, und von der Anwesenheit von Anionen wie zB. Chlorid. Die EPR Daten der Mo-TH deuteten auf ein Gleichgewicht hin zwischen zwei Mo(V)- Spezies: Mo(V)-OH2 und Mo(V)-OH. Die Aminosäuren Aspartat und Histidin am aktiven Zentrum scheinen essentiell zu sein für die Katalyse. Inkubationsexperimente mit

(9)

Zusammenfassung

VII dem Substrat Pyrogallol und dem möglicher Intermediat 2,4,6,3’,4’,5’- hexahydroxydiphenylether ergaben eine geringe Abnahme des Mo(V) EPR-Signals. Im Falle reduzierter Mo-TH führen diese beiden Verbindungen zu einer partiellen Oxidation der Metallzentren, was durch EPR Spektroskopie gezeigt wurde. Dagegen wurde unter gleichen Bedingungen mit dem Kosubstrat 1,2,3,5-tetrahydroxybenzol eine Abnahme des Mo(V) EPR-Signals und die Reduktion der [4Fe-4S]-Zentren erreicht. Ein möglicher Elektronen-Transfer-Weg ausgehend von Mo(MGD)2 in der α-Untereinheit via eines Arginin-Restes zum benachbarten [4Fe-4S]-Cluster in der β-Untereinheit wurde vorgeschlagen.

I.b. Acetylen Hydratase aus Pelobacter acetylenicus

Das strikt anaerobe, mesophile δ-Bakterium Pelobacter acetylenicus wandelt den ungesättigten Kohlenwasserstoff Ethin (Trivialname: Acetylen) zu Acetat und Ethanol um, mit Acetaldehyd als Zwischenprodukt [Abt, 2001; Kisker et al., 1998; Schink, 1985]. Der erste Schritt dieses Gärungsweges ist die Anlagerung von Wasser an die Dreifachbindung von Acetylen, die von der Acetylen Hydratase (AH) katalysiert wird [Rosner et al., 1995].

AH aus P. acetylenicus wurde bis zur Homogenität gereinigt. Die funktionelle Einheit ist das Monomer mit einer molekularen Masse der Aminosäurenkette von 81.9 kDa.

Sequenz- und Strukturvergleiche ordnen das Enzym der Familie der DMSO-Reduktasen zu. Es enthält ein W(MGD)2-Zentrum und ein [4Fe-4S]-Cluster vom Cubane-Typ.

Wolfram kann durch Molybdän ersetzt werden, die entsprechende Mo-AH zeigt eine signifikant geringere Aktivität. ICP/MS, EPR und UV/VIS-Spektroskopie zeigen, dass P. acetylenicus sowohl Wolfram als auch Molybdän in das Aktivzentrum der AH einbauen kann. Die W-AH sticht unter den andere Mitgliedern der DMSO Familie hervor, da sie mit der Hydratisierung von Acetylen zu Acetaldehyd keine Redox- Reaktion katalysiert. Dabei ist zu beachten, dass das Enzym für die Katalyse durch Natriumdithionit oder Ti(III)-Zitrat reduktiv aktiviert werden muss.

(10)

Zusammenfassung

EPR-spektroskopische Untersuchungen Dithionit-reduzierter W-AH und Mo-AH zeigen die charakteristischen Resonanzen eines [4Fe-4S]-Zentrums, mit gav = 1.966 ± 0.001.

Die mit K3[Fe(CN)6] oxidierte W-AH zeigt das EPR Signal eines W(V)-Zentrums, mit gav = 2.02. Im gegensatz dazu zeigt die Mo-AH („wie isoliert“) Resonanzen eines Mo(V)-Zentrums, mit gav = 1.99.

W-AH wurde unter striktem Ausschluß von Luftsauerstoff in seiner aktiven, reduzierten Form kristallisiert. Die Kristallstruktur wurde mit einer Auflösung von 1.26 Å erarbeitet.

Das W(MGD)2-Zentrum bindet neben einem Cystein-Schwefel einen Sauerstoff- Liganden – Wasser oder eine Hydroxygruppe – der mit dem benachbarten Aspartat 13 wechselwirkt. Aufbauend auf diesen strukturellen und spektroskopischen Ergebnissen wurde ein erster möglicher Reaktionsmechanismus für die Anlagerung von Wasser an die Acetylen-Dreifachbindung unter Ausschluß von Luftsauerstoff formuliert. Acetylen wird in eine hydrophobe Tasche des Enzyms modelliert, direkt über einem Wasssermolekül, welches an das W(MGD)2-Zentrum gebunden ist. Damit das Substrat Zugang zu diesem neuartigen Aktivzentrum erhalten kann, entwickelt das Protein einen speziellen Substratkanal, weit entfernt von der Stelle, an der dieser normalerweise in anderen Molybdän- und Wolfram-haltigen Enzymen der DMSO Reduktase Familie gefunden wird.

(11)

II. Summary

(12)

Summary

In this doctoral thesis, two novel molybdopterin-dependent enzymes from anaerobic bacteria have been investigated. Both pyrogallol–phloroglucinol transhydroxylase (TH) and acetylene hydratase (AH) catalyze rather unusual reactions in the absence of dioxygen. Despite the presence of highly sophisticated metal centers in their active sites, these reactions are formally reactions without a net transfer of electrons.

II.a. Transhydroxylase from Pelobacter acidigallici

The strictly anaerobic bacterium Pelobacter acidigallici ferments gallic acid (3,4,5- trihydroxybenzoic acid), pyrogallol (1,2,3-trihydroxybenzene), phloroglucinol (1,3,5- trihydroxy-benzene), or 2,4,6-trihydroxybenzoic acid to three molecules of acetate (plus CO2) [Brune et al., 1990; Schink et al., 1982]. A key enzyme in the fermentation pathway is pyrogallol–phloroglucinol transhydroxylase (TH), which converts pyrogallol to phloroglucinol in the absence of O2, in the presence of 1,2,3,5-tetrahydroxybenzene as a cosubstrate. The proposed reaction scheme was described by Messerschmidt et al., (2004) based on a high resolution structure of TH.

Transhydroxylase from P. acidigallici is a heterodimer consisting of a large subunit (100.4 kDa) and a small subunit (31.3 kDa). This enzyme is closely related to enzymes of the DMSO-reductase family. Although the overall reaction of transhydroxylase is no redox reaction it contains three [4Fe-4S] centers and one Mo(MGD)2 as redox-cofactors.

The EPR spectra of the Mo(V) site (samples frozen at different redox potentials) revealed a very complex pattern of resonances dependent on the molybdenum isotope inserted, the pH of the solution, and the presence of anions such as chloride. The EPR study of the Mo(V) center suggested the presence of two Mo(V) species in equilibrium, Mo(V)-OH2

and Mo(V)-OH, with at 1.996, 1.982, 1.963. The amino acids aspartate and histidine near the active site seem to be essentiell for catalysis. Incubation experiments of Mo-TH (as isolated and reduced by dithionite) with pyrogallol, or the putative intermediate 2,4,6,3’,4’,5’-hexahydroxydiphenyl ether, showed a rather small decrease in intensity of

(13)

Summary

XI

the Mo(V) EPR signal; in the case of reduced Mo-TH, these compounds caused partial oxidation of the metal sites. Surprisingly, the co-substrate 1,2,3,5-tetrahydroxybenzene led to a decrease of the Mo(V) EPR signal, and to a reduction of the Fe-S centers as shown by EPR spectroscopy. A putative electron transfer pathway from the Mo(MGD)2

unit on the α-subunit of TH via an arginine residue to the close by [4Fe-4S] on the β- subunit has been assigned.

II.b. Acetylene hydratase from Pelobacter acetylenicus

The strictly anaerobic, mesophilic, δ-bacterium Pelobacter acetylenicus converts the unsaturated hydrocarbon ethine (trivial name, acetylene) to acetate and ethanol via acetaldehyde as an intermediate [Abt, 2001; Kisker et al., 1998; Schink, 1985]. The first step of the fermentation pathway, the hydration of acetylene to acetaldehyde, is catalyzed by the enzyme acetylene hydratase [Rosner et al., 1995].

Acetylene hydratase (AH) from P. acetylenicus was purified to homogeneity. It is a monomer with a molecular mass of the amino acid chain of 81.9 kDa. Sequence and structure comparisonsgroup the protein into the DMSO-reductase family. It contains a W(MGD)2 center and a cubane-type [4Fe-4S] cluster. Tungsten can be replaced by molybdenum, the corresponding Mo-AH is quite less active. ICP/MS, EPR, and UV/Vis- spectroscopy revealed that P. acetylenicus is able to insert tungsten as well as molybdenum into the bisMGD cofactor of acetylene hydratase. W-AHstands out from its family class because it catalyzes a non-redox reaction, the hydration of acetylene to acetaldehyde. Note that the enzyme requires reduction by either Na-dithionite or Ti(III) citrate.

EPR spectroscopic investigation of dithionite-reduced W-AH and Mo-AH showed signals of a [4Fe-4S] center, with gav = 1.966 ± 0.001. W-AH oxidized by K3[Fe(CN)6] exhibited resonances of a W(V) center, with gav = 2.02. The “as isolated” Mo-AH showed resonances of a Mo(V) center, with gav = 1.99.

(14)

Summary

W-AH has been crystallized under the strict exclusionof dioxygen in its active, reduced state. The crystal structure of W-AH was determined at 1.26 Å resolution. The structure showed thatthe tungsten center binds a water molecule (or OH-group) which can closely interact with a nearby aspartate residue. These structural and spectroscopic results led to the proposal of a reaction mechanism. Hereby, the acetylene molecule is placed into hydrophobic pocket, directly above a water/OH molecule coordinated to tungsten. Two different reactions mmight then take place: electrophilic addition leading to vinyl cation as intermediate or nucleophilic attack yielded vinyl anion with acetylene. Access to this novel W–Asp active site has been made possible by a substrate channel at a position distantfrom those found in other molybdenum and tungsten enzymes.

(15)

1. Introduction

(16)

Introduction

1.1. Molybdenum and tungsten in biological systems

1.1.1. Chemistry of molybdenum and tungsten

Molybdenum and tungsten belong to the transition metals in Group 6 of the Periodic Table. The atomic and ionic radii of W and Mo, as well as their electron affinity, are virtually identical, and both occur naturally as a mixture of isotopes. Radioactive isotopes are available for both elements (¹⁸⁵W and ⁹⁹Mo), as well as there are stable ¹⁸³W and

95Mo isotopes with a nuclear spin suitable for the study of hyperfine interactions by magnetic resonance techniques. The high solubility of Mo(VI) and W(VI) oxyanions, 10^-

7 and 10^-9 M, respectively, in sea water, ensures that both metals are bioavailable [Young, 1997].The difference between both metals can be clearly seen under oxic vs. anoxic conditions. Generally, in the presence of dioxygen, molybdate (MoO42-) and tungstate (WO42-) are the principal forms, and these compete with each other for binding sites in proteins and uptake systems. In the absence of dioxygen, in sulfur rich environments, molybdenum can exist in the Mo(V) or Mo(IV) state, usually coordinated to sulfide, such as in MoS2, or to organosulfur ligands, such as the dithiolene moiety of molybdopterin [Sigel et al., 2002]. The organometallic chemistry of Mo and W, respectively, will be described in a separate paragraph when discussing its coordination in metalloenzymes.

1.1.2. Beneficial and toxic effects of molybdenum and tungsten in higher organisms

Molybdenum is an essential trace element for all higher organisms where molybdenum enzymes have a number of important roles. Plants for example require Mo in nitrate reductase for proper nitrogen assimilation. In humans Mo is coordinated in the active site of enzymes like xanthine oxidase, sulfite oxidase or aldehyde oxidase, which are involved in diseases including gout, combined oxidase deficiency, and radical damage following cardiac failure. Combined oxidase deficiency is a rare genetic disease responsible for severe neurological disorders observed in infant children. There also exists a link between one of the many proteins required to synthesize the molybdopterin cofactor of these enzymes, and a protein used in neuronal synapses [Burgmayer, ;

(17)

Introduction

3

Schwarz et al., 2004]. This illustrates how the biological impact of the molybdenum enzymes may reach beyond the limited context of their unique catalytic function.

The impact of tungsten on humans has been studied in the case of workers chronically exposed to hard-metal dust. Tungsten had been shown to accelerate the development of mammary cancer in rats. This was attributed to a decrease in hepatic molybdenum, which severely disturbs the function of the liver, the site of estrogen metabolism [Sigel et al., 2002].

1.1.3. Molybdenum and tungsten in the microbial world

According to one possible theory, early life on this planet arose in a hot and reducing environment, with dioxygen being more or less absent [Hille, 2002]. Most likely, transitions metals, such as nickel and tungsten, and sulfur compounds played an important role in building efficient early life catalysts [Kroneck, 2005]. Later, when the concentration of dioxygen increased in the atmosphere, tungsten may have been replaced by molybdenum in enzymes. This hypothesis is supported by the finding that molybdenum enzymes are found in all forms of life in opposite to tungsten enzymes which have been only discovered in anaerobic mesophiles or thermophiles [Hille, 2002].

Depending on the growth conditions, some bacteria can exchange molybdenum by tungsten, which has been described for several enzymes, especially dimethyl sulfoxide reductase [Sigel et al., 2002]. In the case of the tungsten enzyme acetylene hydratase, molybdenum could be inserted, and the corresponding Mo acetylene hydratase was active [Rosner et al., 1995].

The question why some enzymes can function with either metal ion while others seem to be exclusive for one is far from being elucidated. Possible explanations are linked to (i) the redox properties of the enzyme under investigation, and (ii) to the bioavailability of tungsten vs molybdenum.

(i) Which metal will be chosen by a microorganism can be determined by the relationship of temperature and redox potential of the metal-sulfur site. The molybdopterin-dependent enzymes show a remarkably small difference of 30 mV at 25 °C for the redox potential E1/2 M(IV)/M(V). The oxidation process M(V)/M(VI) indicated that redox potentials for the tungsten compounds are

(18)

Introduction

(at 25°C) more positive than the potentials for the molybdenum compound but the difference is extraordinarily small. At 70^oC, is it appears that E1/2(Mo) >

E1/2(W), consequently tungsten is prefer, entially inserted [Schulzke, 2005].

(ii) Concerning bioavailability and insertion of molybdenum or tungsten, several interesting observations have been made. In DMSO reductase from Rhodobacter capsulatus, and TMAO reductase from Escherichia coli, the metal which was of higher concentration in the in the growth medium became incorporated. Similarly, Desulfovibrio alaskensis could insert both molybdenum and tungsten into formate dehydrogenase (Fdh) depending on the concentration in the medium. In the medium where Desulfovibrio alaskensis produced either Mo- or W-FDH, existed other enzymes contained only Mo (aldehyde oxidoreductase) [Andrade et al., 2000]. In contrast, Fdh from both Desulfovibrio gigas and Syntrophobacter fumaroxidans incorporated only tungsten, despite the presence of a suitable concentration of molybdenum in the medium [Moura et al., 2004]. The same relation was obtained for Pyrococcus furiosus containing typical tungsten enzymes (AOR family). The bacterium was grown on medium containing also Mo in an attempt to replace the tungsten, however, in the enzymes only tungsten ion was determined [Sigel et al., 2002].

These experimental findings suggest that bioavailability is not the only parameter which governs the choice of the metal to be incorporated into the active site. Obviously, some enzymes are very selective for either tungsten or molybdenum, whereas other enzymes can function only with a specific ion [Moura et al., 2004].

1.1.4. Molybdenum and tungsten enzymes: classification and general properties

Despite the high similarity between the chemical properties of Mo and W, the enzymes containing tungsten represent only a small percentage compared to the molybdenum enzymes [Johnson et al., 1996].

(19)

Introduction

5

Mononuclear centers of molybdenum and tungsten are found in the active sites of a diverse group of enzymes that are involved in important reactions of the biogeochemical cycles of carbon, nitrogen, and sulfur. The metals themselves are catalytically inactive unless they are complexed by a complex cofactor.

Figure 1: The structure of Molybdopterin = Moco cofactor. (A) In prokaryotes, molybdopterin cofactors normally contain a nucleoside bonded via two phosphates. A nucleoside consists of a ribose and one of the four nitrogenous bases shown here. A nucleotide is a nucleoside which is bonded to one or more phosphates; (B) Two molybdopterin guanine dinucleotide (bis-MGD) cofactors coordinated to a molybdenum or tungsten atom, as found in enzymes that belong to DMSO reductase family.

With the exception of nitrogenase, where Mo is a constituent of the so-called FeMo- cofactor, molybdenum or tungsten, is bound to a pterin, thus forming the molybdenum cofactor (Moco, Fig. 1) which is the active compound at the catalytic site of all molybdenum- and tungsten-enzymes [Brondino et al., 2006].

Molybdopterin cofactors (termed either Moco or pyranopterin) are found in several enzymes of both prokaryotes and eukaryotes [Hille, 1996]. Molybdopterin cofactors from

(20)

Introduction

eukaryotes normally contain only one pterin ring, whereas prokaryotic molybdopterin cofactors may contain either one or two pterin rings and a nucleotide (Fig. 1 A) covalently linked to the pterin moiety. The nucleotides can contain one of four different bases: guanine, adenine, cytosine and uracil. In same bacterial enzymes was found the molybdopterin-guanine-dinucleotide (MGD) as a cofactor, in which each pterin molecule has a guanine nucleotide. One or two MGD molecules can complex the molybdenum or tungsten atom (Fig. 1 B) [Hille, 2002; Moura et al., 2004]. The effect of the nucleotide on the chemical properties of the molybdopterin cofactor is not known. However, it is assumed that molybdopterin cofactors of similar structure have similar chemical properties.

In addition, molybdenum and tungsten proteins may also carry other redox cofactors, such as iron-sulfur centers, hemes, or flavins, which are involved in intra- and intermolecular electron - transfer processes [Mendel et al., 2006]. Based on structural and genomic data, tungsten and molybdenum enzymes have been divided into four families:

sulfite oxidase, xanthine oxidase, dimethyl sulfoxide (DMSO) reductase and aldehyde / ferredoxin oxidoreductase [Hille, 2002; Moura et al., 2004] (Fig. 2).

1.1.5. Catalytic center of molybdenum and tungsten enzymes

In the dimethyl sulfoxide (DMSO) reductase family, both Mo-containing and W- containing enzymes are found (Fig. 2). In general, the metal is coordinated by the dithiolene sulfur of Moco, and by a variable number of oxygen (oxo, hydroxo, water, serine, aspartic acid), sulfur (cysteine), and selenium ligands (seleno-cysteine).

The proteins of the DMSO reductase family exhibit a wide diversity of properties among its members. The crystal structure of DMSO reductase from Rhodobacter capsulatus was the first structure reported for a member of this family [Bailey et al., 1996; McAlpine et al., 1998]. Within the family nitrate reductase (NAR) and formate dehydrogenase (FDH) have been studied most, so far. In this work I will focus on acetylene hydratase (AH) and pyrogallol-phloroglucinol transhydroxylase (TH) which also belong to the DMSO reductase family.

(21)

Introduction

7

Figure 2: Active site structure of the four families of Mo- and W-containing enzymes based on three- dimensional structure. Active site of enzymes belonging to DMSO reductase family are exemplified by (A) molybdenum-containing enzymes, where different types of ligands identified as X: O-Ser by DMSO reductase, transhydroxylase; Se-Cys by formate dehydrogenase; S-Cys by periplasmic nitrate reductase; O- Asp respiratory nitrate reductase; (B) tungsten-containing enzymes including acetylene hydratase (S-Cys ligand) and formate dehydrogenase (Se-Cys ligand) [Hille, 2002; Moura et al., 2004].

1.1.6. Properties of molybdenum and tungsten dithiolene complexes In dithiolene complexes of both tungsten and molybdenum, the interaction between metal and sulfur is very similar. The majority of the oxo-metal-bis(dithiolene) complexes that have been investigated by vibrational spectroscopy, exhibit octahedral geometry in the Mo(VI)/W(VI) state, and square-pyramidal geometry in the Mo(V, IV)/W(V, IV) state.

The X-ray structure of FDH from E. coli reveals two MGD coordinated to molybdenum.

In the reduced Mo(IV) form of the enzyme the metal carries in the fifth position ligand SH or OH in a distorted square pyramidal geometry, in the oxidized Mo(VI) form the metal binds a seleno-cystein as the sixth ligand, and the geometry changes to distorted

Tungsten Enzymes Molybdenum

Enzymes

(22)

Introduction

trigonal bipyramidal [Boyington et al., 1997; Raaijmakers et al., 2006]. Such geometrical differences between the reduced and oxidized states have been also observed in biomimetic complexes for the active site of DMSO reductase which showed a tetragonal pyramidal geometry (coordination number 5) or a trigonal prismatic geometry (coordination number 6) for the Mo(IV) state, compared to a distorted octahedral geometry (coordination number 6) for the Mo(VI) state [Hofmann, in press].

In the case of tungsten enzymes, the active site geometry does not vary significantly, usually a distorted pseudo-trigonal prismatic geometry is found [Stiefel, 2002].

1.2. EPR properties of metal centers in biological systems

1.2.1 Molybdenum

The Mo(V) state is paramagnetic, with one unpaired electron. The observed EPR signals are not as sensitive towards changes in temperature compared to the spectra of [4Fe-4S]

centers. Nevertheless, the line shapes of Mo(V) EPR signals are complex, usually they are highly anisotropic and show hyperfine structure. Both isotopes ⁹⁵Mo and ⁹⁷Mo (natural abundance 25%, nuclear spin I=5/2) give sextets of lines of relatively low amplitude compared to the intense center line resulting from isotopes (92,94,96,98,100Mo), with I = 0. Analysis and assignment of the individual lines in the EPR spectrum will be facilitated by substituting the enzyme with either pure ⁹⁵Mo,⁹⁷Mo, or ⁹⁸Mo. In addition, further lines can be assigned to superhyperfine interactions resulting from coordinated nitrogens and protons at the active site, shown for example in Figure 3. In the latter case, an exchange of the protons by deuterons might help the assignment. Pioneering studies have been performed by R.C. Bray on the enzyme xanthine oxidase [Bray et al., 1966].

(23)

Introduction

9

Figure 3: EPR spectra of formate-treated FDH from E. coli obtained at 130 K, (a) The spectrum is a superposition of Mo(V) signal with g-factors of 2.094 and 2.0 (b) The same spectrum as in (a) at higher gain. The gz, gy, and gx components of Mo(V) species with I = 0 are marked by dashed lines. The hyperfine features from Mo(V) isotopes with I = ⁵/2 are marked by short sticks. The sample has been frozen after incubation of FDH with 20 mM formate for 10 s [Gladyshev et al., 1996; Khangulov et al., 1998].

1.2.2. [4Fe-4S] centers

The first application of electron paramagnetic resonance spectroscopy (EPR) to iron- sulfur proteins was reported by Orme-Johnson et al., (1968). The geometric structure of a generic [4Fe-4S] cluster in a protein environment is a distorted cube, which can be

(24)

Introduction

depicted as a tetrahedron of four iron atoms interpenetrating a larger tetrahedron of four sulfur atoms (Fig. 4).

Figure 4: The oxidized and reduced state of a [4Fe-4S] cluster presenting the mixed and equal-valence iron atom pairs. The iron and sulfide atoms of the [4Fe-4S] clusters as well as sulfur and Cβ atoms of the cysteines liganting the [4Fe-4S] cluster, are shown with iron atoms as large black spheres, sulfur atoms as light-gray spheres and Cβ as small black spheres [Vassiliev et al., 2001].

The atom-to-atom distances range from 2.67 to 2.81 Å for the iron tetrahedron, and 3.49 to 3.61 Å for the sulfur tetrahedron in different [4Fe-4S] proteins [Vassiliev et al., 2001].

Most commonly, the iron atoms are coordinated to cysteine SH groups which can be detected in the amino acid sequence by the characteristic motif Cys-X2-Cys-X3-Cys [Bruschi et al., 1988].

[4Fe-4S] clusters are diamagnetic in the oxidized state, and paramagnetic in the reduced state. In the oxidized state, the two Fe(III) and two Fe(II) atoms are magnetically coupled, giving rise to an effective total spin of S=0. Any paramagnetism of oxidized iron–sulfur centers at room temperature is due to a population of excited paramagnetic states. In the reduced state, one Fe(III) and three Fe(II) atoms are magnetically coupled, giving rise to an effective total spin of S=1/2. Although in the reduced state a [4Fe-4S]⁺ cluster may be considered to contain the formal valances of three Fe(II) atoms and one Fe(III) atom, two localized Fe pairs exist: one equal-valence pair (Fe²⁺- Fe²⁺), and one mixed-valence pair (Fe^2.5+- Fe^2.5+) in which the electron is delocalized over the two iron sites. Since EPR signals of [4Fe-4S] clusters have very short relaxation times, they usually can only be detected at temperatures below 30 K. The samples must be frozen and, except in single crystals, the molecules will be randomly oriented. As a result, the

(25)

Introduction

11

spectra represent a summation of signals from redox centers oriented in all directions relative to the magnetic field, and are termed ‘polycrystalline powder spectra’. The EPR spectrum of a [4Fe-4S] cluster typically shows three distinct g-values reflecting

‘rhombic’ symmetry. For example the typical rhombic [4Fe–4S]⁺ signal was obtained by FDH from E. coli (Fig. 5).

A B

Figure 5: EPR spectrum of formate-treated FDH from E. coli obtained at 42 K, (A) The spectrum is a superposition of [4Fe-4S]⁺ signal with g-factors of 1.840, 1.957, and 2.045, and the Mo(V) signal with g- factors of 2.094 and 2.0; (B) EPR signal of [4Fe-4S] clusters obtained after subtraction of the Mo(V)

"2.094" signal from spectrum A; FDH was incubated with 10 mM formate for 1 min [Gladyshev et al., 1996; Khangulov et al., 1998].

Each g-value, gx, gy, and gz, corresponds to one of the value obtained when the magnetic field is parallel to one of the three special directions of the paramagnetic molecule. The principal g-values obtained from powder spectra are therefore approximately equal to the single crystal’s principal g-values. The relationship between the crystal axis and the g- tensor axes is not well understood [Vassiliev et al., 2001].

(26)

Introduction

1.3. Acetylene, substrate for microbial growth and metalloenzyme inhibitor

Acetylene (C2H2) is a highly flammable gas that forms an explosive mixture with air, and polymerizes exothermically. The carbon-carbon triple bond involves one σ-bond and two orthogonal π-bonds (Fig. 6).

1^stπ bond in x, y plane 2^stπ bond in x, z plane Cylindrically symmetrical set of π electrons Figure 6: Molecular orbitals of acetylene

The hydrogen atoms of alkynes are relatively acidic compared to hydrogen atoms of ethylene or ethane. Acetylene itself has a pKa of about 25. Acetylene has a rather rich chemistry. There exist a number of different reactions, such as reductions as well as electrophilic and nucleophilic additions [Yurkanis-Bruice, 2004]. The role of acetylene as an inhibitor of important microbial metabolic processes is well known [Hyman et al., 1988]. The close association between acetylene and the nitrogen cycle can be traced back to the original description of C2H2 gas as an inhibitor and an alternative substrate of nitrogenase.

In the list of prebiotic molecules observed in the interstellar gas, we also find acetylene and higher alkynes [Thaddeus, 2006]. Some interesting information came recently from research on the moon Titan which appears to have environmental conditions similar to those on Earth some 4 billion years ago. This is the time when life most likely originated on Earth. Life may have originated on Titan during its warmer early history and then developed adaptation strategies to cope with the increasingly cold conditions. If organisms originated and persisted, metabolic strategies could exist that would provide sufficient energy for life to persist, even today. Metabolic reactions might include the

(27)

Introduction

13

catalytic hydrogenation of photochemically produced acetylene, or involve the recombination of radicals created in the atmosphere by ultraviolet radiation (Fig. 7) [Schulze-Makuch et al., 2005].

Figure 7: Proposed environmental conditions at Titan, schematic. Acetylene and radicals are produced by photochemical reactions in the atmosphere. Due to its high specific gravity acetylene will sink to Titan’s surface and to the bottom of a hydrocarbon reservoir, where it can be used by putative organisms for metabolic reactions (insert). The metabolic end-product methane rises to the atmosphere and is detected to be isotopically lighter than predicted by Titan formation theories [Schulze-Makuch et al., 2005].

The first report on the utilization of acetylene gas (C2H2) by a bacterium was published 75 years ago [Birch-Hirschfeld, 1932]. Later, Norcadia rhodochrous was described to use C2H2 as its sole source of carbon and energy [Kanner et al., 1979]. To date, the major source of atmospheric C2H2 appears to result from automobile exhaust, although, natural sources can not be completely excluded [Whitby et al., 1978]. The solubility of C2H2 gas in water is rather high (1.03 g/l) when compared to other substrate gases, consequently, lowest atmospheric concentrations may allow microbial growth. De Bont and Peck (1980) reported on the metabolism of C2H2 and methylacetylene by Rhodococcus A1, and mentioned the occurance of the enzyme acetylene hydratase at high levels in cell-free extracts. The enzyme was inhibited by dioxygen but not the product acetaldehyde.

(28)

Introduction

Yamada and Jakoby (1958, 1959) described the enzymatic utilization of acetylenedi- and monocarboxylic acid, with addition of H2O to the C≡C bond. Later, pure cultures of acetylene-fermenting anaerobes were obtained by enrichment with acetylene from freshwater and marine sources [Schink, 1985].

1.4. Transhydroxylase from Pelobacter acidigallici

Pelobacter acidigallici strain MaGal2 (DSM 2377) is a strictly anaerobic, chemoorganothroph, and gram-negative bacterium that ferments gallic acid, pyrogallol, phloroglucinol, and 2,4,6,-trihydroxybenzoic acid to three molecules of acetate [Hille, 1999; Schink et al., 1982]. It was isolated from black, anaerobic marine mud of Rio Marin, a channel about 2.5 m wide and 70 cm deep, located in the city of Venice, Italy.

The cells are rod-shaped with 0.5 – 0.8 x 1.5 – 3.5 µm in size. The DNA base ratio is 51.8% ± 0.4 mol% G + C [Schink et al., 1982]. Pelobacter acidigallici belongs to the group of fermenting bacteria like Eubacterium oxidoreducens, and Pelobacter massiliensis. These bacteria anaerobically degrade trihydroxybenzenes and their carboxylated derivatives, via the intermediate phloroglucinol, which is subsequently reduced and cleaved hydrolytically [Reichenbecher et al., 1999].

A crucial step in the fermentation of pyrogallol is the transhydroxylation of pyrogallol to phloroglucinol [Brune et al., 1992]. This reaction is catalyzed by the Mo/Fe-S dependent enzyme pyrogallol:phloroglucinol hydroxyltransferase (transhydroxylase, TH).

Conversion of pyrogallol to phloroglucinol was studied with the molybdenum enzyme transhydroxylase isolated from Pelobacter acidigallici [Baas et al., 1999; Reichenbecher et al., 1994; Reichenbecher et al., 1996].

Transhydroxylation experiments in H218O revealed that none of the hydroxyl groups of phloroglucinol was derived from water, confirming the concept that this enzyme transfers a hydroxyl group from the cosubstrate 1,2,3,5-tetrahydroxybenzene (tetrahydroxybenzene) to the acceptor pyrogallol, and simultaneously regenerates the cosubstrate [Reichenbecher et al., 1999].

(29)

Introduction

15

OH

O

H OH

OH

O

H OH

OH O

H OH

OH

O

H OH

OH

*

TRANSHYDROXYLASE Pyrogallol

1,2,3,5-Tetrahydroxybenzene Phloroglucinol

Figure 8: Proposed reaction mechanism of transhydroxylase from P. acidigallici [Brune et al., 1992;

Reichenbecher et al., 1999].

This concept requires a reaction which synthesizes the cofactor de novo to maintain a sufficiently high intracellular pool during growth. Some sulfoxides and aromatic N- oxides were found to act as hydroxyl donors to convert pyrogallol to tetrahydroxybenzene. Again, water was not the source of the added hydroxyl groups; the oxides reacted as cosubstrates in a transhydroxylation reaction rather than as true oxidants in a net hydroxylation reaction. No oxidizing agent was found that supported a formation of tetrahydroxybenzene via a net hydroxylation of pyrogallol. However, conversion of pyrogallol to phloroglucinol in the absence of tetrahydroxybenzene was achieved if little pyrogallol and a high amount of enzyme preparation was used which had been pre-exposed to air. Obviously, the enzyme was oxidized by air to form sufficient amounts of tetrahydroxybenzene from pyrogallol to start the reaction. A reaction mechanism has been proposed which combines an oxidative hydroxylation with a reductive dehydroxylation via the molybdenum cofactor, and allows the transfer of a hydroxyl group between tetrahydroxybenzene and pyrogallol without involvement of water. With this, the transhydroxylase differs basically from all other hydroxylating molybdenum enzymes which all use water as hydroxyl source [Reichenbecher et al., 1999].

Transhydroxylase is a heterodimer consisting of a large subunit (100.4 kDa) and a small subunit (31.3 kDa). This enzyme is closely related to enzymes of the DMSO-reductase family. Although the overall reaction of transhydroxylase is no redox reaction it contains three [4Fe-4S] (β-subunit) centers and one Mo-bisMGD as redox-cofactors. It contains

(30)

Introduction

11.56 ± 1.72 Fe, 0.96 ± 0.21 Mo (atomic absorption spectroscopy), and 13.13 ± 1.68 acid labile sulfur per heterodimer [Baas et al., 1999; Reichenbecher et al., 1994;

Reichenbecher et al., 1996]. The isoelectric point is 4.1, the specific activity of transhydroxylase is highest at pH 7.0, and the temperature optimum is between 53 and 58°C [Reichenbecher et al., 1994]. The crystal structure of the reduced enzyme was published by Messerschmidt et al., 2004. It represented the largest structure (1149 amino acid residues per molecule, 12 independent molecules per unit cell), which had been solved so far by the single anomalous diffraction technique (SAD). The crystals were analyzed with synchrotron radiation and the three-dimensional structure of TH could be solved at 2.5 Å resolution.

The role of the 3 [4Fe-4S] clusters in the ß-subunit remains unclear at this point. The distance between the closest [4Fe-4S] cluster and the molybdenum with 23.4 Å appears too far for an efficient electron transfer.

Figure 9: Metal centers in transhydroxylase from P. acidigallici [Messerschmidt et al., 2004].

β-β-ssuubbuunniitt α

α--ssuubbuunniitt

[4[4FFee--44SS]]²²⁺^+/^/+⁺ f

feerrrreeddooxxiinn clcluusstteerr

MoMo--bbiissMMGGDD

(31)

Introduction

17

In addition, crystal structures of TH in complex with acetate, pyrogallol, and the inhibitor 1,2,4-trihydroxybenzene were successfully solved at a resolution of 2.35 Å, 2.20 Å, and 2.00 Å. In the structure of the complex of TH with its substrate, pyrogallol bound with the O1 oxygen to the Mo centre (Fig. 10). Additionally, pyrogallol was ligated by Asp174 and Arg153. The side chain of Tyr560 can adopt two different conformations and is in the open conformation for cosubstrate, with the substrate pyrogallol bound to the Mo active site. The OH group of Tyr404 and the SH group of Cys557 are in hydrogen bonding distances to pyrogallol and probably play important roles during catalysis. The 3D structure of transhydroxylase supports the participation of a cosubstrate and led to a new possible reaction mechanism [Messerschmidt et al., 2004; Niessen, 2004].

Figure 10: The binding site of pyrogallol in the active site of TH with open conformation of Tyr 560 in channel (pdb VLE).

(32)

Introduction

1.5. Acetylene hydratase from Pelobacter acetylenicus

Pelobacter acetylenicus is a strictly anaerobic and mesophilic bacterium that is able to grow on acetylene as single energy and carbon source. The first step in the metabolization of acetylene is the transformation of acetylene to acetaldehyde [Schink, 1985]. Growth of P. acetylenicus with acetylene and specific acetylene hydratase activity depended on tungstate or, to a lower degree, on the supply of molybdate in the medium.

The specific enzyme activity in the cell extracts was highest after growth in the presence of tungstate [Abt, 2001; ten Brink, 2006].

To date, acetylene is the only hydrocarbon known to be metabolized in the absence and presence of dioxygen in the same manner [Schink, 1985]. The novel tungsten-[4Fe-4S]- enzyme acetylene hydratase (W-AH) is the first enzyme involved in fermentative conversion of C2H2 to acetate and ethanol by the strict anaerobe Pelobacter acetylenicus.

W-AH converts C2H2 to acetaldehyde, a reaction distinct from the reduction of acetylene to ethylene by nitrogenase:

HC≡CH + H2O → [H2C=C(OH)H] → CH3CHO

W-AH belongs to the superfamily of molybdopterin-dependent enzymes. The addition of one H2O molecule to the C≡C bond - formally a non-redox reaction - requires a strong reductant and the presence of chemically complex metal centers. When P. acetylenicus grew at elevated levels of molybdate, the less active variant Mo-AH could be isolated [Abt, 2001; ten Brink, 2006]. AH is highly specific for acetylene.

AH from P. acetylenicus has been purified to homogeneity [Meckenstock et al., 1999;

Rosner et al., 1995]. It is a monomer with a molecular mass of 81.9 kDa (amino acid sequence) compared to 73 kDa by SDS-PAGE. BLASTP searches revealed that the enzyme is highly similar to enzymes of the DMSO-reductase family.

W-AH had an isoelectric point at pH 4.2. Per mol of enzyme, 4.8 mol of iron, 3.9 mol of acid-labile sulfur, and 0.4 mol of tungsten but no molybdenum, were detected. The Km

for acetylene was 14 µM as assayed in a coupled photometric test with yeast alcohol dehydrogenase and NADH, and the Vmax was 69 µmol ⋅ min^-1 ⋅ mg of protein^-1 [Rosner et al., 1995]. The optimum temperature for specific activity of the AH(W) was 50°C , and the apparent pH optimum was 6.0 to 6.5 [Meckenstock et al., 1999]. The N-terminus of AH shows a typical binding motif for a iron-sulfur cluster of the type Cys-X2-Cys-X3- Cys [Rosner et al., 1995]. According to spectroscopic data the midpoint redox potential

(33)

Introduction

19

of the [4Fe-4S] cluster was determined at -410 ± 20mV. Redox titrations gave a midpoint redox potential of – 410mV for the [4Fe-4S] cluster, and –340 mV for 50% maximum activity. Setting the potential to ≤ –410 mV brought the iron sulfur center to the [4Fe- 4S]⁺ state but did not change the activity of the enzyme [Meckenstock et al., 1999].

Model studies demonstrated the likely participation of a W(IV) site in the catalysis of the hydration of acetylene, whereas the corresponding W(VI) remained inactive (Fig. 11), [Yadav et al., 1997].

NC

NC S

S

CN

CN S

S W

O

NC

NC S

S

CN

CN S

S W

O 2-

IV

2-

VI

oxidized

Na₂S₂O₄

reduced

Figure 11: Reduction of [Et4N]2[W^VIO2(mnt)2], where mnt = 1,2-dicyanoethylenedithiolate [Yadav et al., 1997]

Acetylene hydratase is rather oxygen-sensitive; when purified under air the [4Fe-4S]

cluster becomes degraded to a [3Fe-4S] cluster as demonstrated by EPR spectroscopy [Meckenstock et al., 1999].

1.6. Scope of the study

The work described in this thesis has been designed to obtain detailed biochemical, spectroscopic and structural information on the two novel molybdopterin-dependent enzymes acetylene hydratase and transhydroxylase. The data obtained will help to get a deeper insight into the unusual chemistry of both enzymes and their mode of catalytic action. Acetylene hydratase and transhydroxylase catalyze two unusual non-redox reactions, although both enzymes contain sophisticated metal centers.

By combining biological, biochemical and biophysical methods we planed to elucidate the 3D-structure of the proteins at high resolution as well as the electronic design and

(34)

Introduction

function of their metal centers at the atomic level. The knowledge of the architecture and the electronic features of the active sites is essential for understanding the mode of catalysis and the substrate specificity. While the 3D-structure provides the proximity and spatial relationship of amino acid residues and coordinated ligands, the nature of their bonding will be provided by spectroscopic and computational methods. Spectroscopies will yield information on the binding of the redox cofactors in their different oxidation states, and the interaction with the substrate.

(35)

2. Materials and Methods

(36)

Materials and Methods

2.1. Chemicals

If not specified chemicals were obtained in p.a. quality and used without further purification.

(i) Buffers

Fluka: MES (2-(N-morpholino)ethane sulphonic acid), KH2PO4; Merck: K2HPO4, NaH2PO4; Riedel-de-Haën: Na-citrate dihydrate; Roth: HEPES (N-[2- hydroxyethyl]piperazine-N’-[ethane sulfonic acid]), TRIS (tris-(hydroxymethyl)- aminomethane);

(ii) Chromatographic resins

GE Healthcare (Amersham Biosciences): Resource Q15, Resource Q30; Pharmacia Biotech: Superdex^TM 200, HiLoad^TM26/60, Q-Sepharose Fast Flow

(iii) Crystallization factorials

Fluka: MPD (2-methyl-2,4-pentanediol, Ultra)

Crystal screen solutions were obtained from Hampton Research Corporation (USA), Jena Biosciences GmbH (Germany) and NeXtal Biotechnologies (Qiagen, Germany)

(iv) Dyes

Serva: bromphenol blue (sodium salt), Coomassie-brilliant blue G-250

Gas

Messer Griesheim: Argon 5.0, Helium 4.6, Acetylene 2.6; Sauerstoffwerk Friedrichshafen: N2 5.0, N2/CO2 (80:20 v/v), N2/H2 (96:4 v/v). Liquid helium was delivered by the Department of Physics, Universität Konstanz.

(37)

23

(v) General chemicals

Merck: NaOH, 37.5 % HCl, MgCl2⋅6H2O, MnCl2⋅4H2O; Riedel-de-Haën: Na-acetate, CuSO4⋅5H2O, Na2S2O4, KH2PO4, Na2CO3, KCl; Roth: NaCl.

(vi) Proteins and enzymes

BioRad: low range SDS/PAGE molecular weight standards; Fluka: DNase I (deoxyribonuclease I); Serva: BSA (bovine serum albumin); Sigma: gelfiltration molecular weight marker kit. Boehringer Mannheim: Yeast Alkohol-Dehydrogenase (400 U/mg)

(vii) Reagents

Merck: K3[Fe(CN)6]; Sigma-Aldrich: BCA (bicinchonic acid solution); Boehringer, Mannheim: NADH (nicotinamide adenine dinucleotide); Cayman Chemical: PAPA NONOate

(viii) Non-commercially available compounds

1,2,3,5-Tetrahydroxybenzene was synthesized and kindly provided by Dr. T. Huhn, Universität Konstanz. Purity was checked by NMR. Titanium(III) citrate was synthesized and kindly provided by D. Abt, Universität Konstanz [Zehnder et al., 1976]. 2,4,6,3’,4’,5’-Hexahydroxydiphenyl ether was a generous gift by Prof. J. Retey, Universität Karlsruhe [Paizs et al., 2007].

(38)

2.2. Organisms and cultivation

2.2.1 Pelobacter acidigallici

Ma Gal 2 strain DSMZ 2377 was grown in batch cultures (0.1, 1, 50 l) at 30°C in bicarbonate-buffered, sulfide-reduced saltwater mineral medium [Brune et al., 1990]

The medium was sterilized and cooled under a N2/CO2 (80 : 20, v/v) atmosphere.

After addition of modified trace element solution SL 10 (contains molybdate) and vitamin solutions, the pH was adjusted to 7.2–7.4 with 1 M HCl [Niessen, ; Widdel, 1980; Widdel et al., 1981]. To replace molybdenum (from natural abundant MoO42-) by stable molybdenum isotopes ⁹⁵Mo or ⁹⁸Mo, the culture was transferred at least four times in medium containing molybdate ⁹⁵MoO42- or ⁹⁸MoO42-, respectively (Table 1).

Cultures were inoculated with 10% (v/v) of a stock culture. The substrate gallic acid was dissolved in water under exclusion of air, neutralized to pH 7.0 with concentrated NaOH, sterilized by filtration (0.2 µm), and fed at the start (7 mM) and twice (7 mM) during cultivation. Growth was monitored at 578 nm, and the pH of the medium was maintained at pH 7.2 with 2 M Na2CO3. Cells of a 50 l batch culture were harvested at the end of the exponential growth phase after 1-2 days (A578 = 0.8) with a Pellicon ultrafiltration unit (cutoff 100 kDa, Millipore). The concentrate was centrifuged at 10.000 g (30 min, 4°C) and the resulting cell pellet was stored at –70°C prior to use.

2.2.2 Pelobacter acetylenicus

WoAcy1 strain DSMZ 3246 was grown in batch cultures (0.1, 1, 20 l) at 30°C in freshwater medium [Schink, 1985]. The medium was sterilized and cooled under a N2/CO2 (80: 20, v/v) atmosphere, buffered with 30 mM NaHCO3, and reduced with Na2S. After addition of modified trace element solution SL 10 (contains tungstate) and vitamin solution, the pH was adjusted to 7.0 – 7.4 with 1 M HCl [Niessen, 2004;

Widdel, 1980; Widdel et al., 1981]. To replace tungsten by ¹⁸³W isotope, the culture was transferred at least four times in ¹⁸³WO42- -containing medium (Table 1). The substrate acetylene was added to the gas phase to provide a concentration of 7-10% in the headspace of the glass bottle. The acetylene consumption during the growth of P.

acetylenicus was controlled with a gasometer. Cultures were inoculated with 10% (by vol.) of a stock culture. The growth was monitored at 578 nm, and the pH of the

(39)

25

medium was maintained at pH 7.0 with 2 M Na2CO3. Cells of a 20 l batch culture were harvested after 2-3 days (A578 = 0.75)

Molybdate cultivation of P. acetylenicus was carried out in freshwater medium [Abt, 2001; Schink, 1985] as described above. To replace tungsten by molybdenum, the culture was transferred at least eight times in medium containing 2 µM MoO42- and 2 nM WO42- (Table 1) [Abt, 2001]. Cells were harvested after six days (A578 = 0.9) with a Pellicon ultrafiltration unit (cutoff 100kDa, Millipore). The concentrate was centrifuged at 10.000 g (30 min, 4°C) and the resulting cell pellets were frozen in liquid nitrogen and stored at -70°C.

Table 1: Isotope composition of molybdenum and tungsten compounds used in cultivation experiments

95Mo (I=5/2) ⁹⁷Mo (I=5/2) Mo (I=0) Supplier

MoO42- 15.92 % 9.55 % 74.53 % Merck

95 MoO3 96.8 % - 3.2 % ORNL

98 MoO3 - - 98.0% Russia/Kroneck

183W (I=1/2) W (I=0)

WO42- 14.31% 85.69% Fluka

183WO3 96.5 % 3.5% ORNL

95MoO3, ⁹⁸MoO3 and ¹⁸³WO3 were diluted in concentrated NaOH giving ⁹⁵MoO42-,

98MoO42- and ¹⁸³WO42- [Greenwood et al., 1990].

2.3. Purification protocols

2.3.1. Transhydroxylase from Pelobacter acidigallici

All purification steps were performed in the presence of air at 4°C on a FPLC system (GE Healthcare). Frozen cells were thawed and suspended in 50 mM triethanolamine

(40)

(TEA) pH 7.5 containing a few crystals of desoxyribonuclease I / 10 mM MgCl2·6H2O.

Cells were broken by three passages in a French press (137 MPa; Amicon). The crude extract was centrifuged at 100000 g (60 min, 4°C) giving the soluble fraction as supernatant. The solution was applied to a Q-Sepharose column previously equilibrated with 50 mM TEA pH 7.5. After loading, the column was flushed with three bed volumes of buffer containing 175 mM NaCl. A linear gradient of buffer with increasing NaCl concentration from 175 mM to 500 mM eluted transhydroxylase. Fractions showing transhydroxylase on a SDS-PAGE gel were pooled and concentrated to a final volume of 2 ml by ultrafiltration (cut-off 30 kDa, 30-YM membrane, Millipore;

Amicon).The concentrate was loaded onto Superdex 200 HiLoad^® 26/60 column (GE Healthcare), equilibrated with 200 mM NaCl in 50 mM TEA pH 7.5 and eluted with the same buffer. Active fractions were pooled, concentrated by ultrafiltration (cut-off 30 kDa, 30-YM membrane, Millipore; Amicon), frozen in liquid nitrogen and stored in aliquots of 50-250 µl (10-15 mg/ml) at -70°C.

2.3.2. Acetylene hydratase from Pelobacter acetylenicus

All purification steps were performed in the absence of dioxygen. Freshly prepared or frozen cells were brought into the anaerobic chamber, suspended in 50 mM Tris-HCl pH 7.5, to a cell density A578 = 135 containing 1mM PMSF and a few crystals of desoxyribonuclease I / 10 mM MgCl2·6H2O. Cells were broken by three passages in a French press (137 MPa; Amicon) under Ar or He gas and the lysate collected in N2/H2

containing glass bottles were sealed with a rubber septum. The cell lysate was brought again into the anaerobic chamber and transferred into centrifuge tubes. Cell debris was removed by centrifugation at 10.000 g (30 min, 4°C) giving the crude extract (supernatant).

All chromatographic steps were performed at 18°C in an anaerobic chamber (94% N2, 6% H2; Coy, Pd-Cat) on a FPLC system (GE Healthcare) equipped with a SPD- M10Avp Diode Array Detector.

The crude extract was subjected to precipitation by (NH4)2SO4. In the first precipitation step, 4 M (NH4)2SO4 in water was added slowly to a final concentration of 2.3 M. The solution was stirred on ice for 30 min. After centrifugation (30 min, 10000 g) acetylene

(41)

27

hydratase was precipitated in a second step by adding 4 M (NH4)2SO4 (final concentration of 3.2 M) and stirring on ice for 30 min. After centrifugation (next day) the pellet was resuspended in 50 mM Tris/HCl pH 7.5 and desalted by ultrafiltration (cut-off 30 kDa, 30-YM membrane, Millipore; Amicon) with the same buffer. The desalted protein was centrifuged at 10000 g (5 min, 4°C) and the supernantant was loaded onto a Resource 30Q column (HR 10/16, GE Healthcare) equilibrated with 50 mM Tris/HCl pH 7.5. Fractions containing AH were pooled, again desalted by ultrafiltration (cut-off 30 kDa, 30-YM membrane, Millipore; Amicon) with the same buffer and loaded onto a high resolution Resource 15Q column (HR 10/16, GE Healthcare) equilibrated with 50 mM Tris/HCl pH 7.5. A linear gradient (0.1-0.3 M NaCl) led to the elution of AH at 0.2 M NaCl. Active fractions were pooled and concentrated by ultrafiltration (cut-off 30 kDa, 30-YM membrane, Millipore; Amicon) to 1.5 ml. The concentrate was loaded on a HiLoad^® 26/60 Superdex 200 column (GE Healthcare), equilibrated with 200 mM NaCl in 50 mM Tris/HCl pH 7.5 and eluted with the same buffer. The pure protein was concentrated with a Vivaspin ultrafiltration spin column (30 kDa PES, Vivascience), frozen in liquid nitrogen and stored in aliquots of 50-250 µl (10 mg/ml) at -70°C.

2.4. Analytical methods

2.4.1. Protein

Protein was determined by the bicinchoninic acid (BCA) method [Smith et al., 1985].

100 µl of unknown or standard proteins (5 − 20 µg) Bovine serum albumin (BSA) were mixed with 1 ml of a solution containing 50 : 1 (v/v) BCA and CuSO4 · 5H2O (4%, w/v). The reaction mixture was incubated for 20 min at 60°C. The calibration curve was recorded on a Cary 50 spectrometer (Varian, Darmstadt) and the absorbance was monitored at 562 nm.

(42)

2.4.2. ICP-MS analysis of metals

Metals were determinated by Inductively Coupled Plasma Mass Spectrometry (ICP- MS) at the Spurenanalytisches Laboratorium Dr. Baumann (Maxhütte-Haidhof). Iron, molybdenum and tungsten were determined in samples of acetylene hydratase purified from different cultivations (200 µl; approx. 2 mg/ml).

2.5. Enzymatic activities

2.5.1. Determination of Transhydroxylase activity with HPLC

Transhydroxylase activity was measured by a discontinuous assay [Brune et al., 1990;

Reichenbecher et al., 1999]. All steps were performed under exclusion of dioxygen in an anaerobic chamber. In a typical assay 775 µl of 100 mM potassium phosphate buffer (KPi), pH 7.0 were mixed with 100 µl pyrogallol (100 µM in KPi), 100 µl 1,2,3,5- tetrahydroxybenzene (100 µM in KPi), and incubated for one minute. The reaction was started by addition of 25 µl transhydroxylase solution. Aliquots (100 µl) were taken after 3 and 5 min incubation and immediately added to 100 mM H3PO4 (400 µl).

Samples were transported from the anaerobic chamber in a gas-tight syringe and analyzed quantitatively for pyrogallol and phloroglucinol on a HPLC system (Sykam) equipped with a Gromsil C-18 Reverse Phase column (Grom Analytik + HPLC GmbH);

temperature 40°C, solvent methanol/phosphate (12.5% methanol, by vol.; 100 mM KPi pH 2.6). Samples (20 µl) were injected and eluted at a flow rate of 1.5 ml min^-1. Aromatic compounds were detected at 205 nm. Peak identification was performed by comparison of retention times and UV spectra with those of standard samples and quantified by comparison with standards of known concentration. The same procedure was applied to analyze the potential intermediate 2,4,6,3’,4’,5’-hexahydroxydiphenyl ether (syntheses by Prof. J. Retey).

Structural and functional studies on two molybdopterin and iron-sulfur containing enzymes : transhydroxylase from pelobacter acidigallici and acetylene hydratase from pelobacter acetylenicus