Biosynthesis of vitamin B 2 An essential zinc ion at the catalytic site of GTP cyclohydrolase II Johannes Kaiser 1 , Nicholas Schramek 1 , Sabine Eberhardt 1 , Stefanie Pu¨ ttmer 2 , Michael Schuster 2 and Adelbert Bacher 1 1 Lehrstuhl fu ¨ r Organische Chemie und Biochemie, Technische Universita ¨ tMu ¨ nchen, Garching, Germany; 2 Lehrstuhl fu ¨ r Anorganische und Analytische Chemie, Technische Universita ¨ tMu ¨ nchen, Garching, Germany GTP cyclohydrolase II catalyzes the hydrolytic release of formate and pyrophosphate from GTP producing 2,5-diamino-6-ribosylamino-4(3H)-pyrimidinone 5¢-phos- phate, the first committed intermediate in the biosynthesis of riboflavin. The enzyme was shown to contain one zinc ion per subunit. Replacement of cysteine residue 54, 65 or 67 with serine resulted in proteins devoid of bound zinc and unable to release formate from the imidazole ring of GTP or from the intermediate analog, 2-amino-5-formylamino- 6-ribosylamino-4(3H)-pyrimidinone 5¢-triphosphate. How- ever, the mutant proteins retained the capacity to release pyrophosphate from GTP and from the formamide-type intermediate analog. The data suggest that the enzyme catalyzes an ordered reaction in which the hydrolytic release of pyrophosphate precedes the hydrolytic attack of the imidazole ring. Ring opening and formate release are both dependent on a zinc ion acting as a Lewis acid, which acti- vates the two water molecules involved in the sequential hydrolysis of two carbon–nitrogen bonds. Keywords: formate; GTP cyclohydrolase; imidazole ring; pyrophosphate; zinc ion. GTP cyclohydrolases catalyze the first steps in the biosyn- thetic pathways of riboflavin, tetrahydrofolate and tetra- hydrobiopterin. More specifically, GTP cyclohydrolase I catalyzes the release of C8 of GTP (Compound 1, Fig. 1) followed by the formation of a novel pyrazine ring with inclusion of carbon atoms 1¢ and 2¢ of the ribose side chain [1,2]. The reaction product, dihydroneopterin triphosphate (Compound 3, Fig. 1), is the first precursor in the biosyn- thetic pathways of tetrahydrofolate and tetrahydrobiopterin [3,4]. GTP cyclohydrolase II catalyzes the hydrolytic release of C8 of GTP accompanied by the release of pyrophosphate from the carbohydrate side chain of GTP. The enzyme product, 2,5-diamino-6-ribosylamino-4(3H)-pyrimidinone 5¢-phosphate (Compound 4, Fig. 1), is the first committed precursor in the biosynthesis of riboflavin (vitamin B 2 )[5]. Recently, GTP cyclohydrolase II was also shown to catalyze the formation of GMP from GTP at  10% the rate of formation of the main product, Compound 4 [6]. It was also shown that the 5¢-triphosphates of 8-oxo-7,8-dihydro-2¢- deoxyguanosine and 8-oxo-7,8-dihydroguanosine can be converted into the respective monophosphates, although the enzyme is unable to open the imidazole ring of the structurally modified guanine residues of these nucleotides [7]. Despite certain similarities in their reaction mechanisms, GTP cyclohydrolases I and II have no detectable sequence similarity. Recently, we found that GTP cyclohydrolase I contains an essential zinc ion at each active site [8]. Mutant proteins unable to bind zinc are totally unable to catalyze the opening of the imidazole ring of GTP. In this paper, we show that a zinc ion complexed to three cysteine residues is absolutely required for the release of formate from GTP by GTP cyclohydrolase II, whereas the metal ion is not required for the enzyme-catalyzed release of pyrophosphate from the substrate. EXPERIMENTAL PROCEDURES Materials Oligonucleotides were custom-synthesized by MWG Bio- tech, Ebersberg, Germany. Nucleotide triphosphates were purchased from Sigma-Aldrich Fine Chemicals, Munich, Germany. 2-Amino-5-formylamino-6-ribosylamino-4(3H)- pyrimidinone 5¢-triphosphate was prepared as described previously using the H179A mutant of GTP cyclohydrolase I [9]. DNA sequencing was performed by MWG Biotech. Micro-organisms and plasmids Bacterial strains and plasmids used in this study are summarized in Table 1. Site-directed mutagenesis Site-directed mutagenesis was performed by PCR using the overlap extension technique [10]. PCR was performed with Pfx polymerase (Gibco BRL, Karlsruhe, Germany) to minimize the error rate. The internal mismatch primers are showninTable1. The general scheme of mutagenetic PCR involved three rounds of amplification cycles using two mismatch and two flanking primers (primers MF and BamH1rev, Table 1). During the first round, 20 amplification cycles were carried Correspondence to A. Bacher, Lehrstuhl fu ¨ r Organische Chemie und Biochemie, Technische Universita ¨ tMu ¨ nchen, Lichtenbergstr. 4, D-85747 Garching, Germany. Fax: + 49 89 289 13363, Tel.: + 49 89 289 13360, E-mail: adelbert.bacher@ch.tum.de (Received 12 June 2002, revised 5 September 2002, accepted 9 September 2002) Eur. J. Biochem. 269, 5264–5270 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03239.x out with one of the flanking primers and the corresponding mismatch primer. The plasmid pECH2 [11] was used as template. During the second amplification cycle, 20 ampli- fication cycles were carried out using the second flanking primer and the corresponding mismatch primer. The plasmid pECH2 was used as template. Compounds from both amplification rounds were purified by agarose gel electrophoresis. During the third round, the products of both round one and two were used as templates, and 20 amplification cycles were carried out using the two flanking primers. The resulting compound was subjected to agarose gel electrophoresis, digested with BamHI and EcoRI, purified using the QIAquick PCR purification kit, and ligated into plasmid pNCO-113 which had been digested with the same restriction enzymes. The ligation mixture was transformed into Escherichia coli XL1-blue cells (Strata- gene, Heidelberg, Germany). All gene constructs were verified by DNA sequencing. Protein expression was performed in E. coli M15 (Table 2). Enzyme purification Recombinant GTP cyclohydrolase II of E. coli was purified by published procedures [12]. Enzyme assays Enzyme-catalyzed formation of Compound 4 by GTP cyclohydrolase II was monitored by a published procedure [6]. Assay of pyrophosphate release Reaction mixtures containing 100 m M Tris/HCl, pH 8.0, 5m M MgCl 2 ,2m M substrate (GTP or Compound 2) and protein were incubated at 37 °C. Aliquots of 100 lLwere retrieved at intervals. The reaction was stopped by removal of the enzyme by ultrafiltration (Nanosep 10K Omega; Pall Life Sciences, Ann Arbor, MI, USA), and 30 lL aliquots of the filtrates were applied to an HPLC anion-exchange column (Gromsil SAX; 200 · 4 mm; 5 lm; Grom, Herren- berg, Germany). The column was washed with 30 mL 5 m M ammonium phosphate, pH 2.7, and was developed with a linear gradient of 5–530 m M ammonium phosphate, pH 3.8. The flow rate was 1 mLÆmin )1 . The effluent was monitored photometrically (Knauer Wellchrom K-2600; Knauer, Berlin, Germany) at 254 nm, 272 nm, 293 nm and 330 nm. Zinc determination A solution containing 2 M HCl and 400 lgÆmL )1 protein was incubated at 90 °C for 5 h and then analyzed using a Unicam 919 flame atomic absorption spectrometer (Unicam, Cambridge, UK). RESULTS Earlier studies on GTP cyclohydrolase II suggested the complex reaction pathway shown in Fig. 1 [6]. Briefly, it was proposed that the reaction is initiated by the hydrolytic release of pyrophosphate, possibly involving the formation of a covalent phosphoguanosyl derivative of the enzyme. Carbon 8 of the guanine moiety is then assumed to be released as formate by two consecutive hydrolytic reactions. Thereactionisbelievedtobeterminatedbyhydrolysisof the presumed phosphodiester or phosphoamide bond between enzyme and substrate. This reaction sequence can explain the formation of GMP as a side product accounting for about 10% of the enzyme activity [6]. Fig. 1. Reactions catalyzed by GTP cyclohydrolases. (A) GTP cyclo- hydrolase I; (B) GTP cyclohydrolase II [6]. Ó FEBS 2002 Biosynthesis of riboflavin (Eur. J. Biochem. 269) 5265 The recent finding that GTP cyclohydrolase I requires zinc for the hydrolytic opening of the imidazole ring of GTP [8] prompted us to analyze GTP cyclohydrolase II from E. coli for the presence of zinc ions by atomic absorption spectrometry. As shown in Table 3, the wild-type protein from E. coli prepared by a published procedure [12] was found to contain 0.71 zinc ions per subunit. As structural information on GTP cyclohydrolase II is not available, we decided to screen for amino acids involved in zinc chelation by site-directed mutagenesis. Sequence comparison revealed a CX 2 GX 7 CXC motif which occurs in all cyclohydrolase II sequences (Fig. 2). Each of the three cysteine residues was replaced with serine by PCR-assisted mutagenesis. The mutations were confirmed by DNA sequencing. The mutant genes could be expressed to high levels, and the proteins could be purified by the protocol reported for the wild-type enzyme. Replacement of any of the three cysteine residues in GTP cyclohydrolase II resulted in proteins that were devoid of zinc within the limit of experimental accuracy (Table 3). The conversion of GTP into the product, Compound 4, can be monitored photometrically. The series of spectra shown in Fig. 3 shows isosbestic points at 231 and 271 nm. Product formation can best be monitored photometrically at 300 nm at which the substrate, GTP, shows only very low absorbance. Each of the mutant proteins failed to convert GTP into 2,5-diamino-6-ribosylamino-4(3H)-pyrimidinone 5¢-phos- phate (Compound 4, Table 4). The catalytic rates were less than 1 nmolÆmg )1 Æmin )1 . This translates into less than one product molecule formed per subunit and per hour. We have previously shown that 2-amino-5-formylamino- 6-ribosylamino-4(3H)-pyrimidinone 5¢-triphosphate (Com- pound 2, Table 4) can serve as substrate for GTP cyclohydrolase II, although it does not qualify as a kinetically competent intermediate [13]. The compound is Fig. 2. Sequence comparison of GTP cyclohydrolase II. Aae, Aquifex aeolicus;Aac,Actinobacillus actinomycetemcomitans;Apl,Actinoba- cillus pleuropneumoniae;Ath,Arabidopsis thaliana;Bsu,Bacillus sub- tilis;Cmu,Chlamydia muridarum;Cpn,Chlamydophila pneumoniae; Ctr, Chlamydia trachomatis;Eco,Escherichia coli;Hin,Haemophilus influenzae;Hpy,Helicobacter pylori;Mtu,Mycobacterium tuberculosis; Nme, Neisseria meningitides;Pgu,Pichia guilliermondii;Sce,Sac- charomyces cerevisiae;Ssp,Synechocystis species;Tma,Thermotoga maritima. Table 1. Oligonucleotides used in this study. Oligonucleotides were designed to hybridize to the sense (–) and antisense (+) strand of the ribA. Designation Primer orientation Sequence MF + ACACAGAATTCATTAAAGAGGAGAAATTAACCATG BamH1rev – GCAAATGGGATCCACAATGCAAGAGG P-C54S-f + CATTCCGAATCTCTGACTGGTGAC P-C54S-r – GTCACCAGTCAGAGATTCGGAATG P-C65S-f + GCTTGCTGTCTGATTGTGGCTTC P-C65S-r – GAAGCCACAATCAGAACGCAAGC P-C67S-f + GCGCTGCGATTCCGGCTTCCAGC P-C67S-r – GCTGGAAGCCGGAATCGCAGCGC Table 3. Zinc content of GTP cyclohydrolase II of E. co li. Mutant Zn 2+ per subunit (mol/mol) Wild-type 0.71 C54S < 0.1 C65S < 0.1 C67S < 0.1 Table 2. Micro-organisms and plasmids used in this study. Strain or plasmid Genotype or relevant characteristic Ref. or source E. coli XL1-Blue recA1, endA1, gyrA96, thi – 1, hsdR17, supE44, relA1, lac [F¢, proAB, lacl q ZDM15, Tn10(tet r )] Stratagene [28] E. coli M15 [pREP4] lac, ara, gal, mtl, recA + , uvr + , [pREP4: lacl, kana r ] [29] pNCO113 High-copy expression vector [29] 5266 J. Kaiser et al.(Eur. J. Biochem. 269) Ó FEBS 2002 deformylated by the wild-type enzyme at a rate of 122 nmolÆmg )1 Æmin )1 and has been interpreted as an intermediate analog that can be converted into the enzyme product, Compound 4, but does not occur as an interme- diate in the physiological reaction starting with GTP as substrate. All mutants shown in Table 4 have lost the ability to catalyze the release of formate from the formamide-type compound. It follows that a zinc ion is absolutely required for the opening of the imidazole ring of GTP as well as for the subsequent hydrolysis of the resulting formamide motif of Compound 4. Studies with the H179A mutant of GTP cyclohydrolase I had shown the formation of the formamide-type Com- pound 2 (Fig. 1) from GTP to be a reversible reaction with an equilibrium constant of  0.1 at 30 °C and pH 7.0 [9]. It was therefore in order to check whether the proteins under study can catalyze ring closure of Compound 2 with formation of a guanosine nucleotide. Attempts to detect GMP or GTP in reaction mixtures containing one of the proteins in Table 3 and Compound 2 as substrate were unsuccessful. We previously showed that GTP cyclohydrolase II catalyzes the formation of GMP as a minor product by the release of pyrophosphate from GTP [6]. Specifically, GMP was formed at  10% of the rate of formation of the enzyme product, Compound 4. All mutants shown in Table 3 can catalyze the formation of GMP from GTP, albeit at a reduced velocity. Specifically, the rate for the C54S mutant was  60% of that of the wild-type, and the relative rates of the C65S and C67S mutants were in the range 10–20%. It follows that the zinc ion is not required for the release of pyrophosphate from GTP. However, the release of phosphate from GTP by the wild-type and mutant proteins requires magnesium ions. The wild-type enzyme has been shown to catalyze the release of pyrophosphate from the intermediate analog, Compound 2. We have now found that the mutants in Table 3 retain the ability to catalyze that reaction with for- mation of 2-amino-5-formylamino-6-ribosylamino-4(3H)- pyrimidinone 5¢-monophosphate, which was identified by ion-exchange HPLC (Table 4). GMP and GDP do not serve as substrates for formation of Compound 2 by wild-type GTP cyclohydrolase II, as shown already by Foor & Brown [5]. These authors also reported that GTP cyclohydrolase II is unable to use nucleotide triphosphates other than GTP as substrate. A reinvestigation using the recombinant E. coli enzyme showed that pyrophosphate can be catalytically released from deoxyGTP by the wild-type enzyme as well as by the zinc-deficient mutants obtained by replacement of cysteine residues. Moreover, the wild-type enzyme was able to catalyze ring-opening reactions with deoxyGTP as substrate as shown by photometric analysis (Fig. 4). UV absorbance changes observed with GTP and deoxyGTP were similar (data not shown). The rate of the ring-opening reaction catalyzed by wild-type GTP cyclohydrolase II was 182 nmolÆmin )1 Æmg )1 for GTP and 38 nmolÆmin )1 Æmg )1 fordeoxyGTPassubstrate. DISCUSSION Flavin coenzymes are indispensable in all cellular organisms because of their involvement in redox processes of central metabolic pathways that are crucial for energy transduction. The precursor of flavocoenzymes, riboflavin (vitamin B 2 ), is Fig. 3. Ultraviolet spectra. A reaction mixture containing 100 m M Tris/HCl,pH8.0,10m M MgCl 2 ,90l M GTP, and 0.25 mg protein was incubated at 30 °C. Spectra were recorded at intervals of 75 s. Fig. 4. Formation of products from deoxyGTP by GTP cyclohydrolase II. Table 4. Catalytic activity of GTP cyclohydrolase II mutants. The activity with different substrates (first row) and products (second row) is shown. Protein Activity (nmolÆmg )1 Æmin )1 ) GTP 4 dGTP 12 2 4 GTP GMP dGTP dGMP 2 9 Wild-type 182 38 122 16 25 15 C54S <1 <1 <1 10 1.1 5 C65S <1 <1 <1 2 1.2 <1 C67S <1 <1 <1 3 1.1 <1 Ó FEBS 2002 Biosynthesis of riboflavin (Eur. J. Biochem. 269) 5267 biosynthesized by plants and many micro-organisms, whereas animals depend on nutritional sources. For numerous pathogenic micro-organisms, the enzymes of the riboflavin biosynthetic pathway are essential proteins. Specifically, Enterobacteriaceae are virtually unable to absorb flavins from the environment and are therefore absolutely dependent on their endogenous production [14]. The same has been shown for several yeast species including Candida guilliermondii [15,16]. Mycobacterium tuberculosis and Mycobacterium leprae both have complete sets of riboflavin biosynthesis genes. As these genes have apparently survived the extensive frag- mentation of genes in M. leprae [17], they are likely to be essential for the intracellular lifestyle of Mycobacteria. The genes of riboflavin biosynthesis are therefore putative targets for the treatment of infections caused by Gram- negative bacteria and possibly by Mycobacteria and pathogenic yeasts. The exploration of novel anti-infective targets is of supreme importance in the light of the rapid progression of resistance development in all microbial pathogens. In contrast with the riboflavin biosynthetic pathway, the dihydrofolate pathway was already validated as an anti- infective target in the first half of the last century (for reviews see references [18,19]). In fact, sulfonamides inhibiting dihydropteroate synthase were the first chemotherapeutic agents with a broad antimicrobial and antiprotozoal spectrum of activity. Later, trimethoprim, an inhibitor of dihydrofolate reductase, was introduced for the treatment of bacterial infections, often in combination with sulfona- mides. Both the riboflavin and tetrahydrofolate pathway start from GTP (Fig. 1). The first step of each pathway involves the hydrolytic opening of the imidazole ring of the substrate with formation of formate as a byproduct. However, the enzyme products are different in structure. In the case of GTP cyclohydrolase I, the ring-opening step is followed by a complex series of reactions leading to formation of a dihydropterin [1,2,20–24]. Another difference is the release of pyrophosphate by GTP cyclohydrolase II but not by GTP cyclohydrolase I. Although GTP cyclohydrolase I has been known for more than three decades, it was only recently shown that the hydrolytic opening of the imidazole ring and the subsequent release of formate requires a zinc ion acting as a Lewis acid, which sequentially activates the two water molecules that serve as nucleophiles in the two consecutive hydrolytic reactions [8]. Because of the mechanistic similarities of the two different GTP cyclohydrolases, we investigated the type II enzyme for the presence of zinc. The data are consistent with the presence of one zinc ion per subunit of the homodimeric enzyme of E. coli. The comparison of numer- ous putative GTP cyclohydrolase II sequences indicated a pattern of three absolutely conserved thiols (Fig. 2). The sequence motif fits well with the typical short spacer/long spacer motif found in a number of catalytic zinc-binding sites [25]. The replacement of any of the three conserved cysteine residues produced mutants with zinc levels below the level of detection. This suggests that the catalytic zinc ion is chelated by cysteine residues 54, 65 and 67 of the E. coli enzyme, and that loss of any one of the three thiol groups is sufficient to abolish the zinc-binding capacity of the protein. For comparison, the catalytic zinc of GTP cyclohydrolase I is chelated by two cysteine and one histidine residues, whereas a second histidine residue contacts the metal via an interpolated water molecule (J. Rebelo, G. Auerbach, A. Bracher, G. Bader, H. Nar, M. Fischer, C. Ho ¨ sl, N. Schramek, J. Kaiser, R. Huber, and A. Bacher, unpublished work). The replacement of any of the four amino-acid residues involved in zinc chelation is sufficient to abolish the zinc-binding capacity as well as the catalytic activity of the enzyme. The mutant proteins resulting from the replacement of any of the conserved cysteine residues in GTP cyclohydro- lase II with serine failed to catalyze the formation of the enzyme product, Compound 2, from GTP at a detectable rate. Moreover, these mutants failed to release formate from the formamide-type Compound 4, which is a substrate of wild-type GTP cyclohydrolase II, although it lacks the characteristics of a kinetically competent intermediate [13]. We conclude that the catalytic action of zinc is required for the hydrolytic opening of the imidazole ring as well as for the subsequent hydrolysis of the formamide-type product 7 with formation of formate. These findings suggest the hypothetical mechanism shown in Fig. 5. The formation of a covalent linkage between the substrate and the enzyme with formation of pyrophosphate is the first and rate-determining step [13]. A nucleophilic attack by a zinc-activated water molecule leads to the formation of a GTP hydrate, Compound 14. Cleavage of the C8–N9 bond leads to the formamide intermediate, Compound 15. In analogy with the mechan- ism of zinc proteases (Fig. 6) [26], the co-ordination number of zinc could then be increased to five through complexation of an additional water molecule, which attacks the zinc- complexed formyl group of the intermediate 16. The resulting tetrahedral intermediate could lose formate, and product 4 could be released by hydrolysis of the covalent bond between enzyme and substrate. In a final hydrolytic step, the product is released from the enzyme. It was recently shown that the 5¢-triphosphates of 8-oxo-7,8-dihydro-2¢-deoxyguanosine and 8-oxo-7,8-di- hydroguanosine can be converted into the respective monophosphates by GTP cyclohydrolase II, although the enzyme is unable to open the imidazole ring of the structurally modified guanine residues of these nucleotides [7]. GTP cyclohydrolase II has also been shown to catalyze the conversion of GTP into GMP. This side reaction occurs at a rate of about 10% compared with the formation of the product, Compound 4, in the case of the wild-type enzyme of E. coli. Mutants obtained by replacement of cysteine 54, 65 or 67 retain the capacity to produce GMP from GTP by hydrolytic release of pyrophosphate, although at a reduced rate. It follows that zinc is not required for the hydrolytic release of pyrophosphate. On the other hand, magnesium ions are required for pyrophosphate release. In fact, none of the partial reactions specified in Fig. 1 can be observed in the absence of magnesium ions. These observations are all consistent with the hypothesis of an ordered mechanism in which the release of pyrophos- phate depending on the co-operation of magnesium ions must precede all other reaction steps (Fig. 7). In parallel to many other reactions involving nucleoside triphosphates, magnesium may be required for complexation of the triphosphate motif before substrate binding. The formation 5268 J. Kaiser et al.(Eur. J. Biochem. 269) Ó FEBS 2002 of a covalent linkage between the substrate and GTP cyclohydrolase via a phosphodiester or phosphoamide motif is likely to be the rate-limiting step [27]. After the hydrolytic release of formate, the covalent linkage between enzyme and reaction intermediates can be cleaved hydro- lytically. Cleavage of the phosphodiester or phosphoamide bond can also occur without preliminary ring opening, thus affording GMP from GTP. However, it should be noted that the covalent binding of the intermediate to the enzyme has not yet been documented directly. ACKNOWLEDGEMENTS This work was supported by the Deutsche Forschungsgemeinschaft,by European Community Grant ERB FMRX CT98-0204, the Fonds der Chemischen Industrie and the Hans Fischer-Gesellschaft. We thank Angelika Werner for expert help with the preparation of the manuscript. REFERENCES 1. Burg, A.W. & Brown, G.M. (1968) The biosynthesis of folic acid. VIII. Purification and properties of the enzyme that catalyzes the production of formate from carbon atom 8 of guanosine tripho- sphate. J. Biol. Chem. 243, 2349–2358. 2. Shiota, T. & Palumbo, M.P. (1965) Enzymatic synthesis of the pteridine moiety of dihydrofolate from guanine nucleotides. J. Biol. Chem. 240, 4449–4453. 3. Brown, G.M. & Williamson, J.M. (1987) Biosynthesis of folic acid, riboflavin, thiamine, and pantothenic acid. In Escherichia coli and Salmonella typhimurium (Neidhardt, F.C., Ingraham, J.L., Low, K.B., Magasanik, B., Schaechter, M. & Umbarger, H.E., eds), pp. 521–538. American Society for Microbiology, Wash- ington, DC. 4. Nichol, C.A., Smith, G.K. & Duch, D.S. (1985) Biosynthesis and metabolism of tetrahydrobiopterin and molybdopterin. Annu. Rev. Biochem. 54, 729–764. 5. Foor, F. & Brown, G.M. (1975) Purification and properties of guanosine triphosphate cyclohydrolase II from Escherichia coli. J. Biol. Chem. 250, 3545–3551. 6. Ritz, H., Schramek, N., Bracher, A., Herz, S., Eisenreich, W., Richter, G. & Bacher, A. (2001) Biosynthesis of riboflavin. Studies on the mechanism of GTP cyclohydrolase II. J. Biol. Chem. 276, 22273–22277. 7. Kobayashi, M., Ohara-Nemoto, Y., Kaneko, M., Hayakawa, H., Sekiguchi, M. & Yamamoto, K. (1998) Potential of Escherichia coli GTP cyclohydrolase II for hydrolyzing 8-oxo-dGTP, a mutagenic substrate for DNA synthesis. J. Biol. Chem. 273, 26394–26399. 8. Auerbach, G., Herrmann, A., Bracher, A., Bader, G., Gu ¨ tlich, M., Fischer,M.,Neukamm,M.,Garrido-Franco,M.,Richardson,J., Nar,H.,Huber,R.&Bacher,A.(2000)Zincplaysakeyrolein human and bacterial GTP cyclohydrolase I. Proc. Natl Acad. Sci. USA 97, 13567–13572. 9. Bracher, A., Fischer, M., Eisenreich, W., Ritz, H., Schramek, N., Boyle,P.,Gentili,P.,Huber,R.,Nar,H.,Auerbach,G.&Bacher, A. (1999) Histidine 179 mutants of GTP cyclohydrolase I catalyze the formation of 2-amino-5-formylamino-6-ribofuranosylamino- 4(3H)-pyrimidinone triphosphate. J. Biol. Chem. 274, 16727– 16735. 10. Horton, R.M. & Pease, L.R. (1991) Recombination and muta- genesis of DNA sequences using PCR. In Directed Mutagenesis (McPherson, M.J., ed.), pp. 217–246. Oxford University Press, New York. 11. Eberhardt, S., Richter, G., Ritz, H., Brandt, J. & Bacher, A. (1994) Biosynthesis of riboflavin. Cloning, sequencing, mapping, and Fig. 5. Hypothetical mechanism for release of formate by GTP cyclo- hydrolase II. Fig. 6. Reaction mechanism of zinc proteases [26]. Fig. 7. Cleland notation for the GTP cyclohydrolase II mechanism. Ó FEBS 2002 Biosynthesis of riboflavin (Eur. J. Biochem. 269) 5269 hyperexpression of the genes ribA coding for GTP cyclohydrolase II and ribC coding for riboflavin synthase of Escherichia coli.In Flavins Flavoproteins 1993: Proceedings of the 11th International Symposium. pp.63–66.DeGruyter,Berlin,Germany. 12. Bacher, A., Richter, G., Ritz, H., Eberhardt, S., Fischer, M. & Krieger, C. (1997) Biosynthesis of riboflavin: GTP cyclohydrolase II, deaminase, and reductase. Methods Enzymol. 280, 382–389. 13. Schramek, N., Bracher, A. & Bacher, A. (2001) Biosynthesis of riboflavin. Single turnover kinetic analysis of GTP cyclohydrolase II. J. Biol. Chem. 276, 44157–44162. 14. Shavlovskii, G.M., Tesliar, G.E. & Strugovshchikova, L.P. (1982) Flavinogenesis regulation in riboflavin-dependent Escherichia coli mutants. Mikrobiologiia 51, 986–992. 15. Shavlovskii, G.M., Logvinenko, E.M., Kashchenko, V.E., Kol- tun, L.V. & Zakal’skii, A.E. (1976) Detection in Pichia guillier- mondii of GTP-cyclohydrolase, an enzyme involved in the 1st stage of flavinogenesis. Dokl. Akad. Nauk SSSR 230, 1485–1487. 16. Oltmanns, O., Bacher, A. & Lingens, F. (1968) Growth of Candida guilliermondii deficiency mutants with ethyl methanesulfonate. Z. Naturforsch. B 23, 1556. 17. Cole, S.T., Eiglmeier, K., Parkhill, J., James, K.D., Thomson, N.R., Wheeler, P.R., Honore, N., Garnier, T., Churcher, C., Harris, D., Mungall, K., Basham, D., Brown, D., Chillingworth, T., Connor, R., Davies, R.M., Devlin, K., Duthoy, S., Feltwell, T., Fraser, A., Hamlin, N., Holroyd, S., Hornsby, T., Jagels, K., Lacroix, C., Maclean, J., Moule, S., Murphy, L., Oliver, K., Quail, M.A., Rajandream, M.A., Rutherford, K.M., Rutter, S., Seeger, K., Simon, S., Simmonds, M., Skelton, J., Squares, R., Squares, S., Stevens, K., Taylor, K., Whitehead, S., Woodward, J.R. & Barrell, B.G. (2001) Massive gene decay in the leprosy bacillus. Nature (London) 409, 1007–1011. 18. Schweitzer, B.I., Dicker, A.P. & Bertino, J.R. (1990) Dihydrofolate reductase as a therapeutic target. FASEB J. 4, 2441–2452. 19. Hitchings, G.H. (1971) Folate antagonists as antibacterial and antiprotozoal agents. Ann. NY Acad. Sci. 186, 444–451. 20. Wolf, W.A. & Brown, G.M. (1969) The biosynthesis of folic acid. X. Evidence for an Amadori rearrangement in the enzymatic formation of dihydroneopterin triphosphate from GTP. Biochim. Biophys. Acta 192, 468–478. 21. Shiota, T., Baugh, C.M. & Myrick, J. (1969) The assignment of structure to the formamidopyrimidine nucleoside triphosphate precursor of pteridines. Biochim. Biophys. Acta 192, 205–210. 22. Bracher, A., Eisenreich, W., Schramek, N., Ritz, H., Go ¨ tze, E., Herrmann, A., Gu ¨ tlich, M. & Bacher, A. (1998) Biosynthesis of pteridines. NMR studies on the reaction mechanisms of GTP cyclohydrolase I, pyruvoyltetrahydropterin synthase, and sepiapterin reductase. J. Biol. Chem. 273, 28132–28141. 23. Schramek, N., Bracher, A. & Bacher, A. (2001) Ring opening is not rate-limiting in the GTP cyclohydrolase I reaction. J. Biol. Chem. 276, 2622–2626. 24. Schramek,N.,Bracher,A.,Fischer,M.,Auerbach,G.,Nar,H., Huber, R. & Bacher, A. (2002) Reaction mechanisms of GTP cyclohydrolase I. Single turnover experiments using a kinetically competent reaction intermediate. J. Mol. Biol. 316, 829–837. 25. Vallee, B.L. & Auld, D.S. (1989) Short and long spacer sequences and other structural features of zinc binding sites in zinc enzymes. FEBS Lett. 257, 138–140. 26. Christianson, D.W., David, P.R. & Lipscomb, W.N. (1987) Mechanism of carboxypeptidase A: hydration of a ketonic sub- strate analogue. Proc.NatlAcad.Sci.USA84, 1512–1515. 27. Bacher, A., Eisenreich, W., Kis, K., Ladenstein, R., Richter, G., Scheuring, J. & Weinkauf, S. (1993) Biosynthesis of flavins. In Bioorganic Chemistry Frontiers (Dugas, H. & Schmidtchen, F.P., eds), pp. 147–192. Springer, Berlin. 28. Bullock, W.O., Fernandez, J.M. & Short, J.M. (1987) XL1-Blue: a high efficiency plasmid transforming recA Escherichia coli with beta galactosidase selection. Bio Techniques 5, 376–380. 29. Stu ¨ ber,D.,Matile,H.&Garotta,G.(1990)Systemforhigh-level production in Escherichia coli and rapid purification of recombinant proteins: application to epitope mapping, prepara- tion of antibodies, and structure-function analysis. In Immuno- logical Methods (Lefkovits, I. & Pernis, P., eds), pp. 121–152. Academic Press, Orlando, FL. 5270 J. Kaiser et al.(Eur. J. Biochem. 269) Ó FEBS 2002 . Biosynthesis of vitamin B 2 An essential zinc ion at the catalytic site of GTP cyclohydrolase II Johannes Kaiser 1 , Nicholas. tetrahydrobiopterin [3,4]. GTP cyclohydrolase II catalyzes the hydrolytic release of C8 of GTP accompanied by the release of pyrophosphate from the carbohydrate