Biochemical characterization of recombinant dihydroorotate dehydrogenase from the opportunistic pathogenic yeast Candida albicans Elke Zameitat 1 , Zoran Gojkovic ´ 2, *, Wolfgang Knecht 2,† , Jure Pis ˇ kur 2,‡ and Monika Lo ¨ ffler 1 1 Institute for Physiological Chemistry, Philipps-University, Marburg, Germany 2 BioCentrum-DTU, Technical University of Denmark, Lyngby, Denmark Many fungi, including certain yeasts, have been known for decades as human pathogens. Candida albicans rep- resents the major group of yeast species identified in clinical isolates. This opportunistic pathogen causes both trivial infections in normal people and serious infections in immuno-compromised patients, especially HIV-infected individuals [1]. Yeast infections represent a severe problem for clinicians, as a limited number of antifungal agents are available. In addition, these organisms are becoming resistant to current classes of antifungal agents, particularly the azoles [2]. Expres- sion of efflux pumps that reduce drug accumulation, and mutation or overexpression of antifungal target proteins are strategies that may be used by the patho- gens [3]. The clinical consequences of antifungal resistance can be seen in treatment failures in patients and in changes in the prevalences of Candida spe- cies [4]. Keywords antimycotics; Candida albicans; dihydroorotate dehydrogenase; pyrimidines; redoxal Correspondence E. Zameitat or M. Lo ¨ ffler, Institute for Physiological Chemistry, Philipps-University, Karl-von-Frisch-Str. 1, D-35033 Marburg, Germany Fax: +49 6421 2865116 Tel: +49 6421 2865022 E-mail: zameitat@gmx.de; loeffler@staff.uni-marburg.de Present address *AstraZeneca R&D Mo ¨ lndal, SE-431 83 Mo ¨ lndal, Sweden †ZGene A ⁄ S, Anker Engelundsvej 1, Build- ing 301, 2800 Lyngby, Denmark ‡Department of Cell and Organism Biology, Lund University, So ¨ lvegatan 35, SE-223 62 Lund, Sweden (Received 17 March 2006, revised 16 May 2006, accepted 18 May 2006) doi:10.1111/j.1742-4658.2006.05327.x Candida albicans is the most prevalent yeast pathogen in humans, and recently it has become increasingly resistant to the current antifungal agents. In this study we investigated C. albicans dihydroorotate dehydroge- nase (DHODH, EC 1.3.99.11), which catalyzes the fourth step of de novo pyrimidine synthesis, as a new target for controlling infection. We propose that the enzyme is a member of the DHODH family 2, which comprises mitochondrially bound enzymes, with quinone as the direct electron accep- tor and oxygen as the final electron acceptor. Full-length DHODH and N-terminally truncated DHODH, which lacks the targeting sequence and the transmembrane domain, were subcloned from C. albicans, recombinant- ly expressed in Escherichia coli, purified, and characterized for their kinetics and substrate specificity. An inhibitor screening with 28 selected com- pounds was performed. Only the dianisidine derivative, redoxal, and the biphenyl quinoline-carboxylic acid derivative, brequinar sodium, which are known to be potent inhibitors of mammalian DHODH, markedly reduced C. albicans DHODH activity. This study provides a background for the development of antipyrimidines with high efficacy for decreasing in situ pyrimidine nucleotide pools in C. albicans. Abbreviations DHO, L-dihydroorotate; DHODH, dihydroorotate dehydrogenase; FeCy, potassium hexacyanoferrate(III); Q 0 , 2,3-dimethoxy-5-methyl-1,4- benzoquinone; Q 6 , ubiquinone 30; Q 10 , ubiquinone 50; Q D , decylubiquinone. FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS 3183 Whereas the pyrimidine metabolism of Saccharomy- ces cerevisiae has received considerable attention, that of C. albicans has been addressed only indirectly. For example, 5-fluorocytosine possesses antifungal activity in C. albicans but no antineoplastic activity, as does 5-fluorouracil in humans [5]. Expression of the salvage enzymes cytosine deaminase and uracil phosphoribo- syltransferase in C. albicans makes pyrimidine salvage different from that in mammals, because mammals can only take up pyrimidine nucleosides for recycling [6]. As a prerequisite for development of antipyrimidine agents that can enter cells through the salvage path- way, permeases for pyrimidines and purines have been well studied in yeast species [7]. The enzymes of the de novo pyrimidine synthesis pathway have been shown to be drug targets, or potential drug targets, in eukaryotes [8,9]. This biosyn- thetic pathway results in the formation of UMP and consists of six enzymatic activities found in all organ- isms [10,11]. In most eukaryotes studied so far, five of the corresponding enzymes are located in the cytosol, whereas the fourth enzymatic reaction catalyzed by dihydroorotate dehydrogenase takes place at the inner mitochondrial membrane [10,12]. The reaction mech- anism of dihydroorotate dehydrogenase (DHODH, EC 1.3.99.11) (Fig. 1) includes the stereospecific oxidation of (S)-5,6-dihydroorotate to orotate with reduction of flavin [13,14], and the transfer of electrons to ubiqui- none, which is part of the respiratory chain. Because of this connection, pyrimidine formation requires a sufficient concentration of oxygen in the cells. Whereas Schizosaccharomyces pombe possesses a mitochondrial membrane-bound enzyme (classified as family 2 DHODH), S. cerevisiae has a cytosolic DHODH (clas- sified as family 1 DHODH), the activity of which is independent of ubiquinone and the presence of oxygen [15–17]. This feature promotes growth of this yeast under anaerobic conditions. Saccharomyces kluyveri,a species relatively closely related to S. cerevisiae, is the only yeast known to date that contains both enzyme forms [16,17]. Even though S. cerevisiae is a close rel- ative of Candida species, and is often used as a model pathogen, its DHODH of family 1 is unsuitable as a prototype for the search for enzyme inhibitors in other yeasts. We subcloned a gene coding for C. albicans DHODH (accession number AY230865), overexpressed and purified the recombinant enzyme, and compared it with the DHODH from humans and other yeasts [17,18]. This work evaluates C. albicans DHODH as a target for the development of highly specific antimycot- ic drugs against this widespread pathogen. Results Genetic code and overexpression In C. albicans the standard leucine CUG codon is translated as serine [19]. We found two CUG codons in the DHODH ORF and changed them to UCG (L11S and L78S) by site-directed PCR mutagenesis for gene expression in a bacterial system. By sequence alignment, we identified C. albicans DHODH as a family 2 enzyme (Fig. 2). In this class of enzyme, a catalytic serine residue corresponds to the active-site cysteine in family 1. A typical bipartite N-terminal sequence was identified in the sequence consisting of a targeting sequence that, analogously to the rat and human enzyme [12], promotes import into mitochon- dria and a hydrophobic transmembrane domain neces- sary for the correct insertion into the inner mitochondrial membrane. Expression vectors were constructed to produce full-length CaDHODH and an N-terminally truncated mutant (DNCaDHODH), lack- ing the putative bipartite mitochondrial targeting motif and transmembrane domain. After purification of the full-length and truncated CaDHODH by affinity chromatography, SDS ⁄ PAGE (Fig. 3) showed that the purified enzymes were of the expected molecular mass of 48 kDa for the full-length enzyme and 42 kDa for the truncated enzyme. The yield of recombinant proteins purified from 1 L E. coli BL21 cultures was different for the full-length and truncated enzyme when cultured under similar condi- tions: 0.5 mg CaDHODH and 1.2 mg DNCaDHODH. Compared with other mitochondrial yeast DHODH, the protein abundancies were in the same range: Sch. pombe DHODH, 0.4 and 1.8 mg; S. kluyveri DHODH, 1.4 and 1.8 mg (unpublished data). For the truncated and full-length human DHODH, the yields Fig. 1. Scheme of dihydroorotate dehydrogenase (DHODH) reaction with chemical formulae. Electron transfer from dihydroorotate to FMN and further on to quinone. Candida albicans dihydroorotate dehydrogenase E. Zameitat et al. 3184 FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS of purified proteins were approximately 10 times higher [20]. Obviously, truncated forms of the yeast DHODH were expressed more efficiently than the full- size enzymes. It was not possible to increase the yield of full-length DHODH by changing expression condi- tions (temperature, oxygen supply, induction point and period of expression) or using more or less Triton X-100 as nonionic detergent through the purification protocol (data not shown). The Western blot in Fig. 3, performed with human DHODH antibodies, showed cross-reactivity with the DHODH protein from C. albicans. Cross-reactivity was also observed with recombinant DHODH from Arabidopsis thaliana (unpublished data). The flavin ⁄ protein ratio (mol ⁄ mol) as estimated from fluorimetric cofactor analyses of the two recom- binant enzymes was in the range 0.2–0.3 mol flavin per mol protein. Kinetic characterization Activity measurements of CaDHODH and DNCaD- HODH in various buffers revealed maximum activity at pH 8.0–8.5. From the characteristic bell-shaped activity profile, two pK a values could be calculated: pK a1 , 6.7 ± 0.05 for both enzymes; pK a2 , 9.5 ± 0.1 for CaDHODH and 9.9 ± 0.15 for DNCaDHODH. We compared the activity of CaDHODH and DNCaDHODH with a variety of native and two artifi- cial electron acceptors. CaDHODH and DNCaD- HODH could use the artificial acceptors potassium hexacyanoferrate(III) (FeCy) and 2,6-dichloroindophe- nol. FeCy was the best electron acceptor. Studies with different quinone acceptors [2,3-dimethoxy-5-methyl- 1,4-benzoquinone (Q 0 ), ubiquinone 30 (Q 6 ), ubiquinone 50 (Q 10 ), decylubiquinone (Q D )] indicated a better acceptance of the ubiquinone derivative Q 6 than Q 10 , which is the ubiquinone of most higher eukaryotes (Table 1). Fumarate and NAD were inadequate elect- ron acceptors for CaDHODH and DNCaDHODH. A B Fig. 2. Dihydroorotate dehydrogenase (DHODH) amino-acid sequences. (A) Alignment of the N-termini of the recombinantly expressed C. albicans DHODH and human DHODH ( CLUSTAL W version 1.8). Amino-acid residues that are identical in the human and C. albicans enzyme are highlighted in black. L11S and L78S mutations are shown in red. Approximate positions of the domain that direct mitochondrial import and the hydrophobic, putative transmembrane domain are indicated. In addition, the membrane-association motif forming a hydrophobic tun- nel for the electron acceptor in DHODH is indicated. Numbers refer to amino-acid residues of the C. albicans protein. (B) Alignment of the catalytic centre of the recombinantly expressed C. albicans DHODH and amino-acid sequences of different DHODH family 2 enzymes ( CLUSTAL W version 1.8). The highly conserved serine residue is marked in green. AB Fig. 3. Recombinant C. albicans dihydroorotate dehydrogenase (DHODH). (A) SDS ⁄ PAGE. Lanes: M, molecular mass marker; 1, CaDHODH; 2, DNCaDHODH. 2 lg protein per lane. (B) Western blot. Lanes: M, molecular mass marker; 1, CaDHODH; 2, DNCaD- HODH. 1 lg protein per lane. E. Zameitat et al. Candida albicans dihydroorotate dehydrogenase FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS 3185 However, the presence of atmospheric oxygen seemed to promote very low DHODH activity, suggesting that molecular oxygen may be used as a poor electron accep- tor. The specific activity of the enzymes using Q D and 2,6-dichloroindophenol as acceptor was % 6UÆmg )1 . K m values for 2,6-dichloroindophenol, dihydroorotate (DHO), and Q D for CaDHODH and DNCaDHODH, respectively, were similar, as were k cat values for both enzyme forms (Table 2). Inhibition of the recombinant DHODH Specific inhibitors for yeast DHODH have not yet been described. We studied the recombinant enzymes from C. albicans for their susceptibility to various compounds, which have already been proven to be inhibitors of human DHODH or DHODH from other species or compounds implicated in interfering with electron transport in mitochondria or pyrimidine meta- bolism [18,21–23]. Only the dianisidine derivative red- oxal (0.5 mm) exhibited an inhibitory effect of more than 50% on CaDHODH and DNCaDHODH com- pared with the noninhibited reaction. As compounds such as redoxal may have redox activity, we tested it as a putative direct electron acceptor for the C. albi- cans DHODH. Redoxal (up to 100 lm) did not pro- mote oxidation of dihydroorotate to orotate (data not shown). Also, 1 mm brequinar reduced the activity by more than 50% of the full-length enzyme (Table 3). IC 50 values as a practical reflection of the relative effects of different substances on enzyme activity under comparable assay and laboratory conditions were obtained from dose–response curves. The IC 50 for redoxal was 106 ± 12 lm (CaDHODH) and 102 ± 12 lm (DNCaDHODH), respectively; that for brequinar was 439 ± 83 lm (CaDHODH). Discussion The availability and characterization of recombinant DHODH from C. albicans in this work permitted the first screening of compounds as putative enzyme inhib- itors, with the rationale to interfere with the pyrimid- ine nucleotide pools of this pathogen. All DHODH proteins of family 2 must be translo- cated from the cytosol to the inner membrane of mito- chondria. The proteins are directed by targeting sequences, which usually consist of various numbers of amino acids at the N-terminus [24]. Although no con- sensus sequence has been identified, the pre-sequences have a high content of basic, hydrophobic and hydrox- ylated amino acids, and a length of about 10–80 amino acids. Generally, the pre-sequence is cleaved on import, as it is not necessary for protein function [25]. The length of the targeting sequence in the C. albicans DHODH (37 amino acids) would suggest that there should be a cleavable site. However, in silico studies (‘PeptideCutter’, http://www.expasy.org/tools) could not identify cleavage sites of known proteases. Mam- malian DHODHs have a shorter targeting sequence of Table 1. Alternative electron-accepting substrates for recombinant C. albicans dihydroorotate dehydrogenase (DHODH). Activities are expressed relative to that with FeCy as the electron acceptor, and mean ± SEM from three determinations is given as a percentage. All reaction mixtures contained molecular oxygen at atmospheric pressure (equivalent to about 230 l M)and1mM DHO. DCIP, 2,6- dichloroindophenol. Electron acceptor Relative velocity (%) CaDHODH DNCaDHODH FeCy (1 m M) 100 100 DCIP (1 m M) 15.2 ± 3.1 31.7 ± 7.9 DCIP + QD (1 m M +0.1 mM) 40.8 ± 1.7 50.6 ± 3.1 Q D (0.1 mM) 28.8 ± 2.3 36.4 ± 2.4 Q 10 (0.1 mM) 8.8 ± 2.5 3.5 ± 0.3 Q 6 (0.1 mM) 83.9 ± 3.8 73.1 ± 2.5 Q 0 (0.1 mM) 16.8 ± 1.8 28.3 ± 3.9 Fumarate (1 m M) 2.4 ± 0.4 2.0 ± 1.2 NAD (0.1 m M) 0.2 ± 0.1 0.3 ± 0.1 None 1.5 ± 0.8 1.9 ± 0.2 Table 2. Kinetic constants of the purified full-length and truncated C. albicans dihydroorotate dehydrogenase (DHODH). All measurements were performed in triplicate. For K m and V max , the best fit (± asymptotic SEM) of the Michaelis–Menten equation to all data is given. The k cat values were calculated using the equation V max ¼ k cat [E], where [E] is the total enzyme concentration and is based on one active site ⁄ monomer. U is the enzyme activity as lmol substrateÆmin )1 . DHODH V max (UÆmg )1 ) K m a (lM DHO) K m b (lM Q D ) K m c (lM DCIP) k cat (s )1 ) k cat ⁄ K m a (M DHO )1 Æs )1 ) k cat ⁄ K m b (M Q À 1 D Æs )1 ) k cat ⁄ K m c (M DCIP )1 Æs )1 ) Ca 6.0 ± 1 108 ± 12 42 ± 12 122 ± 13 2.1 1.9 · 10 3 5.0 · 10 4 1.7 · 10 3 DNCa 5.6 ± 0.6 111 ± 18 65 ± 22 40 ± 4 2.2 2.0 · 10 3 3.4 · 10 4 5.5 · 10 4 a The concentration of DHO was varied (0–1.0 mM) at fixed concentrations of 100 lM Q D and 60 lM 2,6-dichloroindophenol (DCIP) as elec- tron acceptors. b The concentration of Q D was varied (0–0.2 mM) at a fixed DHO concentration of 1 mM. c The concentration of DCIP was varied (0–0.2 m M) at a fixed DHO concentration of 1 mM. Candida albicans dihydroorotate dehydrogenase E. Zameitat et al. 3186 FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS only 11–13 amino acids, which was found not to be cleaved off after import into the inner mitochondrial membrane [12]. The targeting sequences of yeast DHODH possess up to 5 times more amino acids [17]; therefore, proteolytic processing may be possible. The length of the targeting sequence seems to influence the recombinant expression rate. All yeast DHODHs have % 40 amino acids in contrast with 10 amino acids in the human DHODH [17]. A similar observation was made with the A. thaliana DHODH, which has a tar- geting sequence of 57 amino acids. The expression rate of the truncated plant DHODH was higher than that of the full-size protein [26]. At the N-terminus, the adjoining hydrophobic region, which was identified as a transmembrane domain of 17 amino acids in rat DHODH [12], can be presumed to be a membrane-spanning a-helix. Here, we were able to predict a transmembrane domain of 16 amino acids with ‘ProtScale’ (http://www.expasy. org/tools) in the C. albicans DHODH amino-acid sequence (Fig. 2). As the recombinant CaDHODH and DNCaD- HODH had the same kinetic parameters, the trunca- tion seemed not to influence the enzyme activity of yeast DHODH. However, the specific activity of C. albicans DHODH was considerably lower than that obtained with recombinant mammalian DHODH preparations, which were determined using the same assay (e.g. human enzyme, 100–150 UÆmg )1 ) [20]. In comparison with human species (K m ¼ 6–15 lm for DHO and K m ¼ 9–14 lm for Q D ), the K m values for C. albicans DHODH were 10-fold higher. On the other hand, C. albicans DHODH was very similar to other yeast DHODHs with regard to its kinetic properties. Higher K m values for Q D were described for full-length S. kluyveri and Sch. pombe DHODH [17] compared with the truncated forms. Two a-helices after the hydrophobic domain at the N-terminus were predicted by structural alignment using ‘Swiss-PdbViewer 3.7’ comparing the structures of human (RCSB PDB-ID, 1D3G) and E. coli (RCSB PDB-ID, 1F78) with the amino-acid sequence of C. albicans DHODH (data not shown). They are sim- ilar to those of human DHODH and are thought to be essential for membrane association and for facilita- ting the contact between the ubiquinone from the inner membrane and the active site of DHODH [23,27]. Although there are some differences in processing and association in the organization of the fungal and mam- malian respiratory chain complexes, the assembly ensures the transfer of electrons from different sources to oxygen by the respiratory chain complexes and the coupling of proton uptake from the matrix compart- ment [28]. The nature of the quinone in C. albicans is not known. In this study, recombinant C. albicans DHODH was shown to use several native and two artificial electron acceptors, FeCy and 2,6-dichloroin- dophenol. Q 6 , which has been described as a physiolo- gical electron acceptor in the respiratory chain of S. cerevisiae [29], was found to be superior to all the other quinones studied here. Fumarate and NAD + , the physiological acceptors for DHODH of family 1A and 1B, respectively, were not acceptors for C. albicans DHODH. This provides functional evidence, addi- tional to its sequence similarity and catalytic-site Table 3. Activity of recombinant C. albicans dihydroorotate dehy- drogenase (DHODH) in the presence of putative inhibitors. Relative velocities determined in chromogen reduction assays with 1 m M DHO, 0.1 mM Q D and 0.1 mM 2,6-dichloroindophenol (DCIP) as final electron acceptor are given. Values are mean ± SEM from three determinations. The activity of each enzyme without inhibitor was set as 100%. If not otherwise stated the concentration of the com- pound was 1 m M. TTFA, 2-thenoyltrifluoroacetone. Compound Relative velocity (%) CaDHODH DNCaDHODH Control with dimethyl sulfoxide 100 100 Control with buffer 100 100 3,4-Dihydroxybenzoic acid 98 ± 3 95 ± 2 3,5-Dihydroxybenzoic acid 98 ± 2 98 ± 1 5-Fluorocytosine 108 ± 5 99 ± 10 5-Fluoroorotate 93 ± 6 101 ± 5 5-Fluorouracil 105 ± 7 103 ± 5 A77-1726 98 ± 17 84 ± 8 Acetylsalicylic acid 67 ± 12 104 ± 11 Alloxan (10 m M)89±9103±6 Amytal 57 ± 14 103 ± 22 Antimycin A (0.5 m M)53±356±3(1mM) Atovaquone (0.5 m M) 106 ± 13 93 ± 5 Brequinar 44 ± 1 80 ± 3 Carboxin 87 ± 2 102 ± 5 Ciprofloxacin 99 ± 4 101 ± 12 Clindamycin 98 ± 4 105 ± 4 Dichloroallyllawson 50 ± 12 58 ± 1 Ectosin 90 ± 12 93 ± 9 Lawson 69 ± 11 95 ± 26 Licochalcone A (0.5 m M) 100 ± 25 100 ± 5 Menadione 86 ± 17 122 ± 6 Polyporic acid 87 ± 11 87 ± 8 Redoxal 5 ± 3 6 ± 5 Redoxal (0.5 m M) 25 ± 15 25 ± 21 Salicylhydroxamic acid 97 ± 14 99 ± 5 Salicylic acid 94 ± 17 90 ± 10 4-Trifluoromethylaniline 91 ± 15 100 ± 20 Toltrazuril 96 ± 6 103 ± 9 Tournaire acid 3 (2.5 m M)98±2107±5 TTFA (2 m M) 101 ± 10 115 ± 1 E. Zameitat et al. Candida albicans dihydroorotate dehydrogenase FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS 3187 features, that C.albicans DHODH belongs to the DHODH family 2 enzymes (Fig. 2). The respiratory chain complexes of fungi have been shown to be inhibited by standard agents, e.g. rote- none, myxothiazol, antimycin A, and CN – , extensively used to assay animal mitochondria [30]. In contrast with this high conservation of sensitivity, drugs that have been shown to suppress DHODH activity both in vitro and in vivo were found here not to interfere with the DHODH of C. albicans, e.g. A77-1726, ato- vaquone, and licochalcone A. Some of these drugs (Table 3) are in clinical use today: the antirheumatic drug leflunomide ⁄ A771726 (Arava TM ) [31], the antima- larial drug atovaquone (Malarone TM ) [22], and the anticoccidial toltrazuril (Baycox TM ) [21]. The develop- ment of effective compounds against Plasmodium falci- parum and Pneumocystis carinii took advantage of species-specific differences between DHODH from family 2. By structure–activity relationship studies, some of these drugs have been shown to interfere with the ubiquinone-binding site of mammal DHODH [23,32] but not with that of E. coli [27]. Detailed kin- etic investigations of the bisubstrate reaction catalyzed by full-length rat DHODH revealed a noncompetitive type of inhibition by brequinar with respect to the co- substrate Q D [33]. LicA was described as a potent inhibitor of E. coli DHODH, but it affected neither DHODH-1A and 1B from Lactococcus lactis (M. Han- sen, University of Copenhagen, personal communica- tion) nor human DHODH (unpublished data). Structural alignment using Swiss-PdbViewer 3.7 to compare the structures of human and E. coli DHODH [23,27] and the amino-acid sequence of C. albicans DHODH showed considerable differences between the inhibitor-binding sites (data not shown). Mainly, hydrophobic interactions, which are important for the binding of A771726 and brequinar, were reduced. In the structural alignment, we found hydrophobic amino acids, which are important for inhibitor binding, replaced with smaller or larger residues. This may explain the difference in binding of these drugs by the fungal and animal DHODH and again highlights DHODH as a very species-specific target for potential intervention and drug discovery. In this study, considerable interference was observed in the oxidation of DHO with Q D by redoxal. The IC 50 value of % 100 lm is higher for the fungal enzyme than for the human (IC 50 ¼ 368 nm) and rat (IC 50 ¼ 214 nm) enzyme [34]. Interestingly, the distinct species- related efficacy of inhibition of the human and rodent enzyme observed with isoxazol, cinchoninic acid and naphthoquinone derivatives seemed to be less obvious with redoxal. It was concluded that the binding of o-dianisidines may be divergent from that of the other classes [34]. As redoxal was superior to all the other compounds tested here in inhibiting fungi DHODH, it can be considered an attractive lead for the synthesis of molecules with higher activity. The high-resolution X-ray crystallographic structure of C. albicans DHODH in complex with an o-dianisidine derivative will be necessary to understand the mode of binding and interference with enzyme catalysis. As the inactivation of any enzyme involved in a metabolic chain will render the whole chain inoperat- ive, the inactivation of any of the six proteins involved in pyrimidine de novo synthesis should result in the same profound effect on the pyrimidine nuc- leotide pools in C. albicans. However, in mammalian cell lines, the development of drug resistance was observed with other agents and other enzymes of the de novo pathway to a much greater degree than with DHODH [35]. Therefore, it is reasonable to assume that the overexpression and proper location of an integral membrane protein would happen to a limited extent only. Thus DHODH rather than a cytosolic enzyme of pyrimidine biosynthesis should be the preferential target for drug development. The availab- ility of recombinant DHODH should expedite discov- ery of more potent agents for growth control strategies in C. albicans, and permit the screening of a large number of compounds, the examination of structure–activity relationships of inhibitors, and determination of the 3D structure of enzyme–inhib- itor complexes. Experimental procedures Reagents Unless otherwise stated, the following chemicals were from Roche Diagnostics (Mannheim, Germany), Serva (Heidel- berg, Germany), Merck (Darmstadt, Germany) or Sigma (Sigma-Aldrich, Taufkirchen, Germany) at the purest grade available: anhydrotetracycline (Acros Organics, Geel, Bel- gium), DHO, dimethyl sulfoxide, Q D ,Q 0 ,Q 6 ,Q 10 , FeCy, fumarate, NAD, 2,6-dichloroindophenol. The inhibitors studied were: 2-hydroxyethylidene-cyano- acetic acid 4-trifluoromethyl anilide (A77-1726; Sanoif- Aventis, Frankfurt, Germany); trans-2-[4-(chlorophenyl)- cyclohexyl]-3-hydroxy-1,4-naphthoquinone (atovaquone, 566C80; Wellcome Foundation, Dartford, UK); 6-fluoro-2- (2¢-fluoro-1,1¢-biphenyl-4yl)-3-me thyl-4-quinoline carboxy- lic acid (brequinar sodium, NSC 368390; DuPont Pharma GmbH, Bad Homburg, Germany); licochalcone A [36]; acetylsalicylic acid; alloxan; antimycin A; 3,4 dihydroxy- benzoic acid; 3,5 dihydroxybenzoic acid; 5-fluorouracil; Candida albicans dihydroorotate dehydrogenase E. Zameitat et al. 3188 FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS 5-fluoroorotate; 5-fluorocytosine; 2-methyl-1,4-naphthoqui- none; menadione; salicylic acid; salicylhydroxamic acid (Sigma); amytal (Serva); ciprofloxacin; toltrazuril (Bayer AG, Leverkusen, Germany); clindamycin; carboxin; ectosin (Fluka, Buchs, Switzerland); redoxal (NCI 73735) [35]; di- chloroallyllawson (NIH Drug Synthesis and Chemistry Branch, Development Therapeutics Program, Division of Cancer Treatment, Bethseda, MD, USA); lawson (Aldrich); polyporic acid (Langner, University of Halle, Germany); 4-trifluoromethylaniline (Chemos GmbH, Regenstrauf, Germany); tournaire acid 3 [37]. Oligonucleotides ZGCaURA1–5¢, ATGTTTCGTCCAAGTATCAAAT TC ZGCaURA1–3¢, TCACTTATCATCAGAGCC Ca-forlong2, ATGTTTCGTCCAAGTATCAAATTC AAACAGTCGACTTTGTCC CaKDHODH-mutfor1, CACAGATGCAGAGTCGG GACATAAGTTGGGGGTT CaKDHODH-mutrev1, CCAACTTATGTCCCGACT CTGCATCTGTGAAAGT CaDHODH-rev, CCGGAATTCCTTATCATCAGAG CCAATTAT Ca-BamHI-for, GCGGATCCCGAATGTTTCGTCC AAGTATCAAATTCAAACAGTCG Cak-BamHI-for, GCGGATCCCGAATGTCAAGAT CAGCAATCCATGAATATGTTTTGTGC CaDHODH-rev3, CCGGAATTCTCACTTATCATC AGAGCCAATTATTTGCTCCCATG Expression plasmids The C. albicans URA1 gene (accession number AY230865) was subcloned with the oligonucleotides ZGCaURA1–5¢ and ZGCaURA1–3¢. The 1335-bp ORF was then amplified from the URA1 PCR fragment with primers Ca-forlong2 and CaDHODH-rev. Mutations were inserted by PCR with primers Ca-forlong2 and CaKDHODH-mutrev1 for a first fragment and with primers CaKDHODH-mutfor1 and CaDHODH-rev for a second fragment. The overlapping PCR fragments were then used as templates for PCR with primers Ca-forlong2 and CaDHODH-rev. For subcloning of the DHODH ORF the restriction sides for BamHI ⁄ EcoRI were created with primers Ca-BamHI-for and CaDHODH-rev3 for full-length C. albicans DHODH. The resulting PCR fragment was cut by BamHI ⁄ EcoRI and sub- sequently ligated into pGEX-6P-3, cut by BamHI ⁄ EcoRI. The resulting plasmid was named pGEX-6P-3-CaDHODH, and the recombinant expressed enzyme is referred to as CaDHODH. A 55-amino acid N-terminal truncated form of C. albicans DHODH was constructed using CaK-Bam- HI-for and CaDHODH-rev, cut by BamHI ⁄ EcoRI and subsequently ligated into corresponding sites of pGEX-6P- 3. The resulting plasmid was named pGEX-6P-3-DNCaD- HODH, and the recombinant expressed enzyme is referred to as DNCaDHODH . Protein expression and purification All recombinant DHODHs were expressed as fusion pro- teins containing an N-terminal glutathione S-transferase (GST) tag. The proteins were expressed in the E. coli strain BL21 for 24 h at 18 °C after induction (A 600 ¼ 0.5–0.6) with 1 mm isopropyl b-d-thiogalactoside in Luria–Bertani broth ⁄ ampicillin (100 lgÆmL )1 ) medium plus 0.1 mm FMN. For purification of the recombinant proteins, the cells were harvested at 4000 g for 15 min, resuspended in binding buf- fer (140 mm NaCl, 2.7 mm KCl, 0.1 mm FMN, 10 mm Na 2 HPO 4 , 1.8 mm KH 2 PO 4 , 1% Triton X-100, pH 7.3), and disrupted by sonification. After centrifugation for 60 min at 15 000 g, the supernatant was applied to a 1-mL GSTrapTM FF column (Amersham Biosciences Europe, Freiburg, Germany). The column was washed with 10 vol. binding buffer and 10 vol. pre-scission buffer (50 mm Tris ⁄ HCl, 150 mm NaCl, 1 mm EDTA, 1 mm dithiothrei- tol, 1% Triton X-100, pH 7). The recombinant proteins were cut by pre-scission protease (Amersham Biosciences) at 4 °C overnight. The recombinant proteins without GST tag were eluted with 5 vol. pre-scission buffer. The exchange to buffer C [50 mm Tris ⁄ HCl, 150 mm KCl, 10% (v ⁄ v) glycerol, 0.1% (v ⁄ v) Triton X-100, pH 8] was per- formed using a PD-10 column (Amersham Biosciences). Protein determination and SDS ⁄ PAGE were performed as described previously [17]. For fluorimetric determination of flavin, 0.5–0.7 lgÆmL )1 protein was denatured by heating up to 100 °C for 10 min. After being allowed to cool, the solution was centrifuged and protected from light until measurement using a spectrofluorimeter (SFM 25, Bio-Tek Instruments, Bad Friedrichshall, Germany) at excita- tion ⁄ emission wavelengths of 465 ⁄ 518 nm, with FMN as standard marker (0–100 lm). Immunological methods Before immunodetection, recombinant C. albicans DHODH from SDS ⁄ PAGE was transferred on to Immobilon P (Milli- pore, Schwalbach, Germany) by semidry blotting (1.5 h at 0.8 mAÆcm )2 ; SDS ⁄ polyacrylamide gel). After being blocked with 5% nonfat dried milk in 10 mm sodium phosphate buf- fer, pH 7.5, containing 150 mm NaCl, the membrane was exposed to affinity-purified rabbit antibodies to human DHODH (diluted 1 : 15000) [38]. As secondary antibodies, goat anti-rabbit horseradish peroxidase-conjugated IgG (Sigma), diluted 1 : 10 000, were used. Bound antibodies were detected with an ECL detection kit (Amersham Bio- sciences). E. Zameitat et al. Candida albicans dihydroorotate dehydrogenase FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS 3189 Biochemical analysis of DHODH The assay to determine enzyme activity and kinetic parame- ters was performed in 50 mm Tris ⁄ HCl, 150 mm KCl, 10% (v ⁄ v) glycerol, 0.1% (v ⁄ v) Triton X-100, pH 8 [18]. At 30 °C, the oxidation of the substrate DHO with the quinone cosub- strate was coupled to the reduction of the chromogen 2,6-di- chloroindophenol. The K m of DHO was determined by varying the concentration of DHO (1–1000 lm) at a fixed concentration (200 lm)ofQ D . The K m of Q D was determined by varying the concentration of Q D (0.1–200 lm) at a fixed concentration (1 m m) of DHO. The additional K m value for 2,6-dichloroindophenol (0.01–200 lm) was determined using the same assay but without Q D . Kinetic data were evaluated under initial velocity conditions [33]; the Michaelis–Menten equation v ¼ V[S] ⁄ (K m + [S]) was fitted to all data. The pH-dependence of initial velocities was measured at saturating substrate concentrations (1 mm DHO, 0.1 mm Q D ) in different buffer systems (Mes ⁄ HCl, Hepes ⁄ HCl, Tris ⁄ HCl) covering the pH range 5–9, using the chromogen reduction assay with 2,6-dichloroindophenol as final elec- tron acceptor. Overlapping pH ranges were measured in two buffer systems to exclude salt effects. The equation v ¼ V ⁄ [(10 –pH ⁄ 10 –pKa1 ) + (10 –pKa2 ⁄ 10 –pH ) +1] was fitted to the data. Various natural and artificial electron acceptors were compared in the optimal Tris ⁄ HCl buffer system at pH 8.0. Reduction of the electron acceptors was measured at the indicated wavelength: 2,6-dichloroindophenol (600 nm, e ¼ 18800 m )1 Æcm )1 ), FeCy (420 nm, e ¼ 1020 m )1 Æcm )1 ), NAD + (340 nm, 6200 m )1 Æcm )1 ). In an alternative assay, UV absorption of the product orotate was monit- ored in the presence of the electron acceptor fumarate (280 nm, e ¼ 7500 m )1 Æcm )1 ) or oxygen only (280 nm, e ¼ 7500 m )1 Æcm )1 ), and at the appropriate isosbestic wave- length, with Q D (300 nm, e ¼ 2950 m )1 Æcm- 1 ), Q 10 (300 nm, e ¼ 2950 m )1 Æcm )1 ), Q 6 (293 nm, e ¼ 4700 m )1 Æcm )1 ), Q 0 (287 nm, e ¼ 5680 m )1 Æcm )1 ), respectively. To determine the inhibitory potency of 28 different com- pounds, the chromogen reduction assay was used with the putative inhibitor up to a concentration of 1 mm as des- cribed above. Stock solutions of all inhibitors were prepared freshly in Tris ⁄ HCl buffer, pH 8.0, or in dimethyl sulfoxide. The appropriate controls were run in buffer or in the pres- ence of dimethyl sulfoxide; 2% dimethyl sulfoxide in the assays was found not to interfere with the DHODH activity. All measurements were performed in triplicate. Percentage of inhibition was related to controls (100% activity). To determine the IC 50 values for redoxal and breqinar, the initial velocity of the DHODH reaction was measured at saturating substrate concentrations of DHO (1 mm) and Q D (0.1 mm) with various concentrations of the putative inhibitors (redoxal, 1 lm)1mm; brequinar, 1 lm)8mm). The equation v ¼ V ⁄ {1 + [I] ⁄ IC 50 }, where [I] is the inhib- itor concentration, was fitted to the initial velocities to find the drug concentration causing 50% inhibition of the enzyme activity (IC 50 value). Acknowledgements This study was supported by the Deutsche Fors- chungsgemeinschaft, Marburger Graduiertenkolleg ‘Protein Function at the Atomic Level’ to ML and by the Danish Research Council to JP. We thank Maria- Bettina Kowalski and Ute Beck for technical assist- ance. References 1 Whiteway M & Oberholzer U (2003) Candida morpho- genesis and host–pathogen interactions. Curr Microbiol 7, 350–357. 2Lo ¨ ffler J & Stevens DA (2003) Antifungal drug resis- tance. Clin Infect Dis 36, 31–41. 3 Sanglar D & Odds FC (2002) Resistance of Candida species to antifungal agents: molecular mechanisms and clinical consequences. Lancet Infect Dis 2, 73–85. 4 Prasad R & Kapoor K (2005) Multidrug resistance in yeast Candida. Int Rev Cytol 242, 215–248. 5 Hope WW, Taberno L, Denning DW & Anderson MJ (2004) Molecular mechanisms of primary resistance to flucytosine in Candida albicans. Antimicrob Agents Chemother 48, 4377–4386. 6Lo ¨ ffler M, Fairbanks L, Zameitat E, Marinaki T & Simmonds HA (2005) Pyrimidine pathways in health and disease. Trends Mol Med 11, 430–437. 7 Rao TVG, Verme RS & Prasad R (1983) Transport of purine, pyrimidine bases and nucleosides in Candida albicans, a pathogenic yeast. Biochem Int 6, 409–417. 8 Knecht W & Lo ¨ ffler M (1998) Species-related inhibition of human and rat dihydroorotate dehydrogenase by immunosuppressive isoxazol and cinchoninic acid deri- vatives. Biochem Pharmacol 56, 1259–1264. 9 Christopherson RI, Lyons SD & Wilson PK (2002) Inhibitors of de novo nucleotide biosynthesis as drugs. Acc Chem Res 35, 961–971. 10 Jones ME (1980) Pyrimidine nucleotide biosynthesis in animals: gene, enzymes and regulation of UMP biosyn- thesis. Annu Rev Biochem 49, 253–279. 11 Nara T, Hshimoto T & Aoki T (2000) Evolutionary implications of the mosaic pyrimidine-biosynthetic path- way in eurkayotes. Gene 257, 209–222. 12 Rawls J, Knecht W, Diekert K, Lill R & Lo ¨ ffler M (2000) Requirements for the mitochondrial import and localization of dihydroorotate dehydrogenase. Eur J Biochem 267, 2079–2087. 13 Blattmann P & Retey J (1972) Stereospecificity of the dihydroorotate-dehydrogenase reaction. Eur J Biochem 30, 130–137. Candida albicans dihydroorotate dehydrogenase E. Zameitat et al. 3190 FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS 14 Mohsen AA, Rigby SEJ, Jensen KF, Munro AW & Scrutton NS (2004) Thermodynamic basis of electron transfer in dihydroorotate dehydrogenase B from Lactococcus lactis: analysis by potentiometry, EPR spectroscopy and ENDOR spectroscopy. Biochemistry 43, 6498–6510. 15 Nagy M, Lacroute F & Thomas D (1992) Divergent evolution of pyrimidine biosynthesis between anaerobic and aerobic yeast. Proc Natl Acad Sci USA 89, 8966– 8970. 16 Gojkovic ´ Z, Knecht W, Zameitat E, Warneboldt J, Coutelis JB, Pynyaha Y, Neuveglise C, Møller K, Lo ¨ f- fler M & Pis ˇ kur J (2004) Horizontal gene transfer pro- moted evolution of the ability to propagate under anaerobic conditions in yeasts. Mol Genet Genomics 271, 387–393. 17 Zameitat E, Knecht W, Pis ˇ kur J & Lo ¨ ffler M (2004) Two different dihydroorotate dehydrogenases from yeast Saccharomyces kluyveri. FEBS Lett 568, 129–134. 18 Knecht W, Bergjohann U, Gonski S, Kirschbaum B & Lo ¨ ffler M (1996) Functional expression of a fragment of human dihydroorotate dehydrogenase by means of the baculovirus expression system, and kinetic investigation of the purified recombinant enzyme. Eur J Biochem 240, 292–301. 19 Santos MAS, Keith G & Tuite MF (1993) Non-stan- dard translational events in Candida abicans mediated by an unusual seryl-tRNA with a 5¢-CAG-3¢ (leucine) anticodon. EMBO J 12, 607–617. 20 Bader B, Knecht W, Fries M & Lo ¨ ffler M (1998) Expression, purification, and characterization of histi- dine-tagged rat and human flavoenzyme dihydroorotate dehydrogenase. Protein Expr Purif 13, 414–422. 21 Haberkorn A (1996) Chemotherapy of human and ani- mal coccidiosis: state and perspectives. Parasitol Res 82, 193–199. 22 Olliaro P & Wirth D (1997) New targets for antimalar- ial drug discovery. J Pharm Pharmacol 49, 29–33. 23 Liu S, Neidhardt EA, Grossman TH, Ocain T & Clardy J (2000) Structures of human dihydroorotate dehydro- genase in complex with antiproliferative agents. Struc- ture 8, 25–33. 24 Schatz G & Dobberstein B (1996) Common principles of protein translocation across membranes. Science 271, 1519–1526. 25 Truscott KN, Brandner K & Pfanner N (2003) Mechan- isms of protein import into mitochondria. Curr Biol 13, R326–R337. 26 Ullrich A, Knecht W, Fries M & Lo ¨ ffler M (2001) Recombinant expression of N-terminal truncated mutants of the membrane bound mouse, rat and human flavoenzyme dihydroorotate dehydrogenase. A versatile tool to rate inhibitor effects? Eur J Biochem 268, 1861– 1868. 27 Nørager S, Jensen KF, Bjo ¨ rnberg O & Larsen S (2002) E. coli dihydroorotate dehydrogenase reveals structural and functional distinctions between different classes of dihydroorotate dehydrogenase. Structure 10, 1211–1223. 28 Joseph-Horne T, Hollomon DW & Wood PM (2001) Fungal respiration: a fusion of standard and alternative components. Biochim Biophys Acta 1504, 179–195. 29 Tsai AL, Olsen JS & Palmer G (1987) The kinetics of reoxidation of yeast complex III. J Biol Chem 262, 8677–8684. 30 Helmerhorst EJ, Murphy MP, Troxler RF & Oppen- heim FG (2002) Characterization of the mitochondrial respiratory pathways in Candida albicans. Biochim Bio- phys Acta 1556, 73–80. 31 Herrmann M, Schleyerbach R & Kirschbaum BJ (2000) Leflunomide: an immunomodulatory drug for the treat- ment of rheumatoid arthritis and other autoimmune dis- eases. Immunopharmacology 47, 273–289. 32 Hansen M, LeNours J, Johansson E, Antal T, Ullrich A, Lo ¨ ffler M & Larsen S (2004) Inhibitior binding in a class 2 dihydroorotate dehydrogenase causes variations in the membrane-associated N-terminal domain. Protein Sci 13, 1031–1042. 33 Knecht W, Henseling J & Lo ¨ ffler M (2000) Kinetics of inhibition of human and rat dihydroorotate dehydro- genase by atovaquone, lawsone derivatives, brequinar sodium and polyporic acid. Chem Biol Interact 124, 61–76. 34 Knecht W & Lo ¨ ffler M (2000) Redoxal as a new lead structure for dihydroorotate dehydrogenase inhibitors: a kinetic study of the inhibition mechanism. FEBS Lett 467, 27–30. 35 Lo ¨ ffler M, Klein A, Hayek-Ouassini M, Knecht W & Konrad L (2004) Dihydroorotate dehydrogenase mRNA and protein expression analysis in normal and drug-resistant cells. Nucleosides, Nucleotides, Nucleic Acids 23, 1281–1285. 36 Chen M, Christensen SB, Blom J, Lemmich E, Nadel- mann L, Fich K, Theander TG & Kharazmi A (1993) Licochalcone A, a novel antiparasitic agent with potent activity against human pathogenic protozoan species of Leishmania. Antimicrob Agents Chemother 37, 2550– 2556. 37 Tournaire C, Caujolle R, Payard M, Commenges G, Bessie ` res MH, Bories C, Loiseau P & Gayral P (1996) Synthesis and protozoocidal activities of quinones. Eur J Med Chem 31, 507–511. 38 Dietz C, Hinsch E & Lo ¨ ffler M (2000) Immunocytochem- ical detection of mitochondrial dihydroorotate dehydro- genase in human spermatozoa. Int J Androl 23, 294–299. E. Zameitat et al. Candida albicans dihydroorotate dehydrogenase FEBS Journal 273 (2006) 3183–3191 ª 2006 The Authors Journal compilation ª 2006 FEBS 3191 . Biochemical characterization of recombinant dihydroorotate dehydrogenase from the opportunistic pathogenic yeast Candida albicans Elke Zameitat 1 ,. residues of the C. albicans protein. (B) Alignment of the catalytic centre of the recombinantly expressed C. albicans DHODH and amino-acid sequences of different