Structure and function of N -acetylglucosamine kinase Identification of two active site cysteines Markus Berger, Hao Chen, Werner Reutter and Stephan Hinderlich Institut fu ¨ r Molekularbiologie und Biochemie, Freie Universita ¨ t Berlin, Berlin-Dahlem, Germany N-Acetylglucosamine is a major component of com- plex carbohydrates. The mammalian salvage pathway of N-acetylglucosamine recruitment from glycoconjugate deg- radation or nutritional sources starts with phosphorylation by N-acetylglucosamine kinase. In this study we describe the identification of two active site cysteines of the sugar kinase by site-directed mutagenesis and computer-based structure prediction. Murine N-acetylglucosamine kinase contains six cysteine residues, all of which were mutated to serine residues. The strongest reduction of enzyme activity was found for the mutant C131S, followed by C143S. Deter- mination of the kinetic properties of the cysteine mutants showed that the decreased enzyme activities were due to a strongly decreased affinity to either N-acetylglucosamine for C131S, or ATP for C143S. A secondary structure prediction of N-acetylglucosamine kinase showed a high homology to glucokinase. A model of the three-dimensional structure of N-acetylglucosamine kinase based on the known struc- ture of glucokinase was therefore generated. This model confirmed that both cysteines are located in the active site of N-acetylglucosamine kinase with a potential role in the binding of the transfered c-phosphate group of ATP within the catalytic mechanism. Keywords: N-acetylglucosamine kinase; N-acetylglucos- amine; aminosugar metabolism; ATP binding domain; cysteine residue. N-Acetylglucosamine kinase (GlcNAc kinase; EC 2.7.1.59) catalyzes the phosphorylation of N-acetylglucosamine (GlcNAc) to GlcNAc 6-phosphate. GlcNAc, from lysos- omal degradation of oligosaccharides or nutritional sourc- es, is a main substrate for the synthesis of UDP-GlcNAc. This activated nucleotide sugar is then used in the biosynthesis of N- and O-glycans [1]. UDP-GlcNAc can be derived either from the salvage pathway involving GlcNAc kinase, or by de novo synthesis from the fructose- 6-phosphate produced in glycolysis (Fig. 1). UDP-GlcNAc is the main substrate of the glycoconjugate biosynthesis. It is used not only in N-/O-glycan biosynthesis, but also as a substrate of O-GlcNAc transferase, which modifies cyto- solic and nuclear proteins at serine or threonine residues by addition of a single GlcNAc. The latter possibly plays a role in signal transduction as an antagonist of protein phosphorylation [2]. Finally, it is the key substrate for the biosynthesis of sialic acids [3]. The function of the salvage pathway of UDP-GlcNAc biosynthesis is not completely clarified, but there is evidence that it compensates the de novo pathway. Thus, tissues with high energy requirements, for example neuro- nal cells, sperms or the apical zone of transporting epithelia, convert glucosamine 6-phosphate (GlcN 6-P) to fructose 6-phosphate by the action of GlcN-6-phosphate deaminase [4], suggesting that in these tissues UDP- GlcNAc is provided by the salvage pathway. Furthermore, GlcN-6-phosphate deaminase is activated by GlcNAc-6-P, the product of the GlcNAc kinase reaction [5]. In mice, elimination of GlcN-6-phosphate-N-acetyltransfer- ase, another enzyme of the de novo pathway (Fig. 1) resulted in embryonal lethality at embryonic day 7.5 [6]. Fibroblasts derived from these mice displayed a reduced UDP-GlcNAc pool and consequently a reduced prolifer- ation. An external supplement of GlcNAc, metabolized by the salvage pathway, completely normalized the pheno- type. Finally, it was recently found that the subcellular localization of GlcNAc kinase in fibroblasts is different from the localization of the enzymes of the de novo pathway [7], indicating a spatial separation of the two pathways, presumably with different regulation. GlcNAc kinase has been cloned from man [8] and mouse [9]. Like N-acetylgalactosamine kinase [10] and N-acetyl- mannosamine kinase, as a part of the bifunctional enzyme UDP-GlcNAc 2-epimerase/N-acetylmannosamine kinase [11,12], it is an N-acetylhexosamine kinase. This group is part of the sugar kinase/heat shock protein 70/actin super- family. Common to all these proteins is an ATPase domain of known three-dimensional structure [13]. Sequence align- ments of GlcNAc kinase with the two most prominent sugar kinases, hexokinase and glucokinase, revealed a strong homology of the three enzymes [9]. In the present study, comparison of GlcNAc kinase with a model of the three-dimensional structure of glucokinase provides the first picture of the possible structure of the enzyme. Furthermore, two active site cysteines were identified by site-directed mutagenesis, agreeing with the predicted structureofGlcNAckinase. Correspondence to M. Berger, Institut fu ¨ r Molekularbiologie und Biochemie, Freie Universita ¨ t Berlin, Arnimallee 22, D-14195, Berlin-Dahlem, Germany. Fax: + 493084451541, Tel.: + 493084451547, E-mail: mberger@zedat.fu-berlin.de Abbreviations: GlcN 6-phosphate, glucosamine 6-phosphate; GlcNAc, N-acetylglucosamine. Enzyme: N-acetylglucosamine kinase (EC 2.7.1.59). (Received 13 March 2002, revised 29 May 2002, accepted 10 July 2002) Eur. J. Biochem. 269, 4212–4218 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03117.x EXPERIMENTAL PROCEDURES Materials [1- 14 C]GlcNAc was from ICN (Eschwege, Germany). Restriction enzymes were obtained from Gibco BRL (Gaithersburg, MD, USA). Nitrocellulose membrane was from Schleicher & Schuell (Dassel, Germany). PCR primers were from MWG Biotech (Ebersberg, Germany). All other chemicals were from Sigma (Deisenhofen, Germany). Enzyme assays GlcNAc kinase activity was determined as described [14]. In brief, the final volume of incubation mixtures was 225 lL, containing 60 m M Tris/HCl, pH 7.5, 20 m M MgCl 2 ,5m M GlcNAc, 50 nCi [1- 14 C]-GlcNAc, 10 m M ATP (disodium salt), 10 m M phosphoenolpyruvate, 2.5 U pyruvate kinase and variable amounts of protein extracts. Incubations were carried out at 37 °C for 30 min, and reactions were stopped by addition of 350 lL of ethanol. Radiolabeled compounds were separated by paper chromatography. Radioactivity was determined in the presence of Ultima Gold XR (Packard, Groningen, Netherlands) in a Tri-Carb 1900 CA liquid scintillation analyser (Packard). Protein concen- tration was measured by the method of Bradford (1976), using bovine serum albumin as a standard. K m values were determined by Lineweaver–Burk plots with Ni-nitrilotriacetic acid purified enzymes (see below). For determination of the K m of GlcNAc, different concentrations of GlcNAc were used in the presence of 10 m M ATP. For determination of the K m of ATP, different concentrations of ATP were used in the presence of 5 m M GlcNAc. Site directed mutagenesis Site directed mutagenesis was performed using the Quick- Change TM site directed mutagenesis kit (Stratagene, Hei- delberg, Germany). In brief, the expression vector pRSETC containing mouse GlcNAc kinase cDNA [9] was used as a template in a PCR-like amplification using Pfu-polymerase and primers containing the desired mutation. The primers used to generate the mutated cDNAs are shown in Table 1. Fig. 1. De novo and salvage pathway of UDP-GlcNAc biosynthesis. Table 1. Oligonucleotides used for generation of GlcNAc kinase cysteine mutants by site-directed mutagenesis. Mismatches with the template are underlined. Name Sequence C45S 5¢- GGCACAGACCAGAGTGTGGAGAGGATCA ATGAG C131S 5¢-GGAACAGGCTCCAACAGTAGGCTTATCAACCC TGATGG C143S 5¢-GATGGCTCCGAGAGTGGCAGTGGAGGCTGGGG C211S 5¢-CCCATTTGTATAGGGACTTTGATAAAAGTAAG TTTGCTGGATTTTGCCAGAAAATTGC C217S 5¢-GCTGGATTTAGTCAGAAAATTGCAGAAGGTG CACATCAGGG C268S 5¢-CCCATTCTGAGTGTGGGCTCAGTGTGG Ó FEBS 2002 Structure and function of GlcNAc kinase (Eur. J. Biochem. 269) 4213 The parental template is then digested by the restriction enzyme DpnI, which specifically cuts methylated DNA. The nicked vector DNA with the desired mutations was transformed into Escherichia coli InvaF¢ cells. All mutant constructs were controlled by sequencing with the Sanger didesoxychain termination reaction for double stranded DNA. Expression of GlcNAc kinase and mutants in E. coli The vectors were transformed into E. coli BL21 cells (Invitrogen) following the manufacturer’s instructions. Posi- tive clones were selected with chloramphenicol and ampi- cillin. Cells were grown to an absorbance of 0.5, induced with 1m M isopropylthio-a-galactoside for 2 h, harvested and resuspended in 20 m M Na 2 HPO 4 , pH 7.5, and then lysed by freezing/thawing. The lysate was centrifuged at 10 000 g for 15 min and the supernatant was analyzed for GlcNAc kinase activity and by Western blot analysis as described below. The supernatant was purified using a Ni-nitrilotriacetic acid column (Qiagen) as described earlier [9]. Western blot analysis Supernatants of E. coli lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis as des- cribed by Laemmli [15] using 10% acrylamide gels. Separ- ated proteins were electroblotted onto nitrocellulose membranes. The membranes were blocked for 3 h in 5% skim milk in buffer A (0.1% Tween 20, 150 m M NaCl, 3m M KCl in 9 m M NaH 2 PO 4 , pH 7.2) and then incubated overnight in a 1 : 5000 dilution of Anti-Xpress antibody (Invitrogen) in buffer A. Detection was performed using a peroxidase-conjugated goat anti-mouse second Ig (Dianova; Hamburg, Germany) and an enhanced chemiluminescence detection kit (Amersham). To normalize different expres- sion rates the scanned Western blots were analyzed by using the IPLabGel software. Multiple sequence alignment Overall sequence similarities were investigated using the PSI - BLAST software [16] and the NR - DATABASE .Protein sequences were aligned by using the MULTALIN software [17]. The algorithms used for secondary structure prediction were PHDSEC [18], PSIPRED [19] and PROFPREDICTION [20]. The different algorithms were used to find the lowest common denominator with regard to a-helices and the b-sheets. Three-dimensional modelling was performed using the RASMOL software. RESULTS Construction of cysteine mutants and functional expression in E. coli BL21 In an earlier study, the use of specific thiol-modifying chemical reagents revealed the presence of cysteine residues in or near the active site of GlcNAc kinase [14]. The number of functionally relevant cysteines was quantified to two, and dithiol-modifying reagents predicted a structural vicinity of these cysteines. The amino acid sequences of murine and human GlcNAc kinase showed six conserved cysteine residues [8,9]. For functional characterization therefore all six cysteines were mutated to serines by site-directed mutagenesis. Wild-type and mutated GlcNAc kinase cDNAs were expressed in E. coli BL21 cells. The proteins were fused to a His-tag, which allowed purification of the proteins by Ni-nitrilotriacetic acid chromatography and detection by a specific antibody using Western-blot analysis. GlcNAc kinase cDNA encodes for a protein with 39 kDa, the His-tag for a 3 kDa protein part. The analysis of wild-type and mutated GlcNAc kinases in BL21 cytosols therefore detected 42 kDa polypeptides in all cases (Fig. 2). The expression of functionally intact proteins was checked by determination of GlcNAc kinase activities. BL21 cells are well suited for the expression of GlcNAc kinase, because they have negligible background activities [9]. Figure 3 shows the relative GlcNAc kinase activites of wild-type and mutated GlcNAc kinases, normalized to the varying expression levels by analyzing protein expression with Western blots. All cysteine mutants showed detect- Fig. 2. Western blot analysis of overexpressed GlcNAc kinase and cysteine mutants in E. coli BL21. Supernatants after cell lysis were separated by SDS-polyacrylamide gel electrophoresis and blotted onto nitrocellulose as described in Experimental procedures. The blot was stained with the Anti-Xpress Ig recognizing the His-tag fusion part of the recombinant proteins. Mock, BL21 cells transformed with pRSETC without GlcNAc kinase cDNA. Fig. 3. Relative specific activities of GlcNAc kinase and cysteine mutants. GlcNAc kinase activities were determined in the cytosolic supernatants of E. coli BL21 cells as described in Experimental proce- dures. All values are means ± SD of five independent expressions. 4214 M. Berger et al.(Eur. J. Biochem. 269) Ó FEBS 2002 able GlcNAc kinase activities, indicating a successful functional expression. However, these activities differ widely among the mutants, in comparison with the wild- type. The highest reduction of enzyme activity was found for the mutant C131S, whereas the activity of C45S was almost unchanged. This was a first hint that distinct cysteines may have significant roles in substrate binding or the catalytic mechanism of GlcNAc kinase. Kinetic characterization of GlcNAc kinase and cysteine mutants In order to get a more detailed insight into the functions of specific cysteine residues of GlcNAc kinase, the K m values for both substrates, GlcNAc and ATP were determined for all mutants and for the wild-type enzyme (Table 2). In general, the decrease in enzyme activity of the mutants correlated with increased K m values, suggesting a decreased affinity to GlcNAc or ATP or both substrates. C131S showed a 17-fold increased K m for GlcNAc, whereas the K m for ATP was increased only 2.5-fold. The opposite was found for C143S, where the respective increases in K m were fivefold for GlcNAc and 10-fold for ATP. Both mutants therefore displayed a decreased substrate affinity, but the K m values indicate that C131 probably plays a role in GlcNAc binding, whereas C143 participates in binding of both substrates. The mutant C217S showed a sixfold increase in the K m for GlcNAc, whereas the K m for ATP was unchanged. This suggests that C217 may have a role in GlcNAc binding. For C211 and C268 the K m values for both substrates were increased, suggesting that these mutations result in change that influences the binding of GlcNAc as well as ATP. The K m values of C45S are the same as for the wild-type enzyme, so that C45 seems to have no role in substrate binding of GlcNAc kinase. Structure prediction of GlcNAc kinase The results from the cysteine mutants suggest that C131 and C143 may be active site cysteines of GlcNAc kinase. To support this a prediction of the secondary structure based on the amino acid sequence was performed using three different algorithms: PHDsec [18], PsiPRED [19] and ProfPrediction [20]. Figure 4 shows the secondary structures, which were predicted by all three methods. The Table 2. Kinetic data of wild-type and mutated GlcNAc. All values are means of at least three independent experiments. Mutants K m GlcNAc (l M ) K m ATP (l M ) Wild-type 230 ± 70 520 ± 100 C45S 240 ± 110 600 ± 270 C131S 3900 ± 600 1300 ± 300 C143S 1150 ± 270 5300 ± 1900 C211S 530 ± 150 1600 ± 250 C217S 1350 ± 310 600 ± 180 C268S 1100 ± 300 1000 ± 350 Fig. 4. Sequence alignment of glucokinase and GlcNAc kinase. The amino acid sequences are numbered beginning with the first amino acid of the N-terminus. Both amino acid sequences share the ATP binding subdomains (Phosphate1, Connect1, Phosphate2, Adenosine, Connect2), and secondary structures (I, II, and III). a-helices are indicated in bold, b-sheets are underlined. Mutated cysteine residues of GlcNAc kinase are labeled with a star. Ó FEBS 2002 Structure and function of GlcNAc kinase (Eur. J. Biochem. 269) 4215 secondary structure of GlcNAc kinase was compared with that of glucokinase, because glucokinase seems to be most closely related, at least among the well-characterized sugar kinases. Thus, glucokinase and GlcNAc kinase share all five subdomains of the ATP-binding domain of sugar kinases with high sequence similarity (Fig. 4; [9]). Both proteins have a similar molecular mass (glucokinase 50 kDa, GlcNAc kinase 39 kDa) without potential regu- latory domains. Finally, GlcNAc kinase can also act as a glucokinase when glucose is present at millimolar concen- trations [21], which are necessary also for the activity of glucokinase [22]. The distribution of a-helices and b-sheets in GlcNAc kinase and glucokinase showed a very similar pattern. Highest similarities were found within the ATP-binding subdomains, indicating a common structure of the ATP- binding sites of the two enzymes. Interestingly, five of the six cysteines of GlcNAc kinase had an equivalent cysteine in glucokinase, not always in a direct sequence homology, but always within the common secondary structure: C45 GlcNAc kinase –C129 Glucokinase (a-helix); C131 GlcNAc kinase – C233 Glucokinase (b-sheet); C143 GlcNAc kinase – C252 Glucokinase (b-sheet); C211 GlcNAc kinase –C371 Glucokinase (a-helix); C217 GlcNAc kinase – C382 Glucokinase (a-helix). This suggests that these homologous cysteines may have the same function in GlcNAc kinase as in glucokinase. In order to visualize a potential three-dimensional structure of GlcNAc kinase with respect to the localization of the cysteine residues, a three-dimensional model of glucokinase was used [23]. Based on the sequence align- ments of glucokinase and GlcNAc kinase and the common ATP-binding domain, the GlcNAc kinase was fitted into the glucokinase model using RasMol software. Figure 5 shows the predicted structure of GlcNAc kinase with the localiza- tion of the five ATP-binding subdomains and the secondary structures common to glucokinase. Typical v-like lobes are present, comparable to the hexokinase subunits [24]. One lobe consists of the phosphate1 subdomain and the predicted a-helix I. The other lobe consists of the phos- phate2 and the adenosine subdomains, the predicted b-sheet II and the predicted a-helix III. The two lobes are linked with the Connect1 and Connect2 subdomains. Localization of the substrate binding pockets is analogous to that of hexokinase, with the ATP pocket in the adenosine sub- domain and the sugar pocket on the opposite side within the b-sheet II [25]. The positions of the cysteines of GlcNAc kinase are displayed in Fig. 5B. Cysteines C211 and C45, whose Fig. 5. Predicted three-dimensional structure of GlcNAc kinase. (A) Front view of the three- dimensional structure of GlcNAc kinase based on a model of glucokinase. ATP binding subdomains are highlighted in different gray scales. Secondary structures in common with glucokinase are indicated with: I, a-helix; II, b-sheet; III, a-helix. (B) Top view of the three- dimensional structure of GlcNAc kinase. The relative positions of cysteines are labeled with astar. 4216 M. Berger et al.(Eur. J. Biochem. 269) Ó FEBS 2002 mutation to serine had little or no effect on specific activity, are localized outside the catalytic center. The cysteines, whose mutation had a more marked effect on specific activity, are localized near (C217) or within (C268) the ATP binding domain. C131 and C143, which were suggested to be active site residues, were found to be very close to the area of phosphate transition. Therefore it can be concluded that C131 and C143 are directly involved in binding the two relevant functional groups of the kinase reaction, the c-phosphate of ATP and the hydroxyl group at C6 of GlcNAc, or they may even be involved in the catalytic mechanism. DISCUSSION The most reactive functional group in a protein is, in general, the sulfhydryl group of cysteine. Chemical reagents are used to investigate the role of cysteine residues in substrate binding and catalytic activity of enzymes [26]. Earlier investigations by our group, using chemical reagents, identified two adjacent cysteine residues in GlcNAc kinase [14]. In the present work all six cysteine residues of the GlcNAc kinase were mutated to serine residues. With the added help of a three-dimensional model, each of the cysteine residues was checked for its role in the catalytic mechanism of GlcNAc kinase. C131 and C143 seem to be directly involved in the transfer of the c-phosphate from ATP to the hydroxyl group at C6 of GlcNAc. This is confirmed by the measured K m values for GlcNAc and ATP, whereby mutations of C131 and C143 resulted in strongly decreased affinities to both substrates. It has been suggested that two cysteine residues in vicinal positions are a general feature of the active site of enzymes with phosphate binding sites [27]. In GlcNAc kinase these cysteines seem to be C131 and C143. Their counterparts in glucokinase are C233 and C252. Mutations of these cysteines to serines have very similar effects on enzyme activity and substrate binding in glucokinase and GlcNAc kinase [28]. It can therefore be concluded that these cysteines have an important and identical role in the enzymatic mechanism of both kinases, presumably catalyzing the transfer of the c-phosphate from ATP to GlcNAc/glucose. C131 appears to be the cysteine of the phosphate donor site during the catalytic mechanism, whereas C143 accepts the phosphate and transfers it to C6 of GlcNAc. The closeness of C131 to C143 can be seen in a three-dimensional model of glucokinase in complex with glucose [29], where the corresponding C233 is in the direct neighborhood of Asn231, which binds the C4-hydroxyl group of glucose. Kinetic data for C217S clearly display a role in the binding of GlcNAc. C217 is located on a-helix III (Fig. 5B). It can therefore be concluded that this a-helix influences the sugar binding. C211 is also located on a-helix III, and its presence results in an increase in the K m value for GlcNAc. But C211S also showed an increased K m value for ATP. This can be explained by the closeness of this part of a-helix III to the adenosine subdomain (Fig. 5). Structural changes due to the C211S and C217S mutation may also have an allosteric effect on other parts of the enzyme. The same explanation can be given for the increased K m values of C268S; although located on the adenosine subdomain, the mutation obviously changes the structure of the sugar binding domain. The results for C211S, C217S and C268S can also be explained by an ineffective induced fit mechanism of the mutated GlcNAc kinases. The induced fit mechan- ism for sugar kinases was first described for hexokinase [30]. Binding of the sugar results in a strong conforma- tional change allowing the binding of ATP. Mutations of amino acids involved in this mechanism therefore result in lower affinities to both substrates. Datta [31] reported for hog spleen GlcNAc kinase that its activity is feed-back inhibited by UDP-GlcNAc. This postulates an allosteric binding site for UDP-GlcNAc within the GlcNAc kinase protein. But the structural model of GlcNAc kinase does not reveal an obvious UDP- GlcNAc binding site. Furthermore, purified rat liver GlcNAc kinase [14] and recombinant mouse GlcNAc kinase did not show any inhibition of their activities when assayed in the presence of UDP-GlcNAc (data not shown). This suggests that, at least for the rodent enzymes, no regulation by UDP-GlcNAc takes place. The present study combines results from site-directed mutagenesis of distinct amino acids of an enzyme with theoretical structural predictions based on known struc- tures of related enzymes. Although the results from both methods show good correlations, more detailed investiga- tions are required. In order to obtain a detailed three- dimensional structure of the enzyme by X-ray diffraction analysis the crystallization of recombinant GlcNAc kinase is in progress. ACKNOWLEDGEMENTS This work was supported by the Bundesministerium fu ¨ r Bildung und Forschung, Bonn, the Fonds der Chemischen Industrie, Franfurt/ Main, and the Sonnenfeld-Stiftung, Berlin. We thank Dr T. A. Scott for improving the English style of the manuscript. REFERENCES 1. Schachter, H. (1991) The Ôyellow brick roadÕ to branched complex N-glycans. Glycobiology 1, 453–461. 2. Hart, G.W., Kreppel, L.K., Comer, F.I., Arnold, C.S., Snow, D.M.YeZ., Cheng, X., DellaManna, D., Caine, D.S., Earles, B.J., Akimoto, Y., Cole, R.N. & Hayes, B.K. (1996) O-GlcNAcylation of key nuclear and cytoskeletal proteins: reciprocity with O-phosphorylation and putative roles in protein multimerization. Glycobiology 6, 711–716. 3. Reutter, W., Sta ¨ sche, R., Stehling, P. & Baum, O. (1997) The biology of sialic acids. Insights into their structure, metabolism and function in particular during viral infection. In Glycoscience, Status and Perspectives (Gabius, H J. & Gabius, S., eds), pp. 245– 259. Chapman & Hall, Weinheim, Germany. 4. Wolosker, H., Kline, D., Bian, Y., Blackshaw, S., Cameron, A.M., Fralich, T.J., Schnaar, R.L. & Snyder, S.H. (1998) Molecularly cloned mammalian glucosamine-6-phosphate deaminase localizes to transporting epithelium and lacks oscillin activity. Faseb J. 12, 91–99. 5. Comb, D.G.R.S. (1958) Glucosamine metabolism. IV. Glucos- amine 6-phosphate deaminase. J. Biol. Chem. 232, 807–827. 6. Boehmelt, G., Wakeham, A., Elia, A., Sasaki, T., Plyte, S., Potter, J., Yang, Y., Tsang, E., Ruland, J., Iscove, N.N., Dennis, J.W. & Mak, T.W. (2000) Decreased UDP-GlcNAc levels abrogate proliferation control in EMeg32-deficient cells. EMBO J. 19, 5092–5104. 7. Ligos, J.M., Lain De Lera, T., Hinderlich, S., Guinea, B., Sanchez, L., Roca, R., Valencia, A. & Bernad, A. (2002) Functional Ó FEBS 2002 Structure and function of GlcNAc kinase (Eur. J. Biochem. 269) 4217 interaction between the ser/thr kinase PKL12 and GlcNAc kinase, a prominent enzyme implicated in the salvage pathway for Glc- NAc recycling. J. Biol. Chem. 277, 6333–6343. 8. Yamada-Okabe, T., Sakamori, Y., Mio, T. & Yamada-Okabe, H. (2001) Identification and characterization of the genes for N-acetylglucosamine kinase and N-acetylglucosamine-phosphate deacetylase in the pathogenic fungus Candida albicans. Eur. J. Biochem. 268, 2498–2505. 9. Hinderlich, S., Berger, M., Schwarzkopf, M., Effertz, K. & Reutter, W. (2000) Molecular cloning and characterization of murine and human N-acetylglucosamine kinase. Eur. J. Biochem. 267, 3301–3308. 10. Pastuszak, I., Drake, R. & Elbein, A.D. (1996) Kidney N-acet- ylgalactosamine (GalNAc)-1-phosphate kinase, a new pathway of GalNAc activation. J. Biol. Chem. 271, 20776–20782. 11. Hinderlich, S., Sta ¨ sche, R., Zeitler, R. & Reutter, W. (1997) A bifunctional enzyme catalyzes the first two steps in N-acetylneur- aminic acid biosynthesis of rat liver. Purification and characterization of UDP-N-acetylglucosamine 2-epimerase/ N-acetylmannosamine kinase. J. Biol. Chem. 272, 24313–24318. 12. Sta ¨ sche, R., Hinderlich, S., Weise, C., Effertz, K., Lucka, L., Moormann, P. & Reutter, W. (1997) A bifunctional enzyme cat- alyzes the first two steps in N-acetylneuraminic acid biosynthesis of rat liver. Molecular cloning and functional expression of UDP- N-acetyl-glucosamine 2-epimerase/N- acetylmannosamine kinase. J. Biol. Chem. 272, 24319–24324. 13. Bork, P., Sander, C. & Valencia, A. (1992) An ATPase domain common to prokaryotic cell cycle proteins, sugar kinases, actin, and hsp70 heat shock proteins. Proc. Natl Acad. Sci. USA 89, 7290–7294. 14. Hinderlich, S., Nohring, S., Weise, C., Franke, P., Stasche, R. & Reutter, W. (1998) Purification and characterization of N-acetyl- glucosamine kinase from rat liver – comparison with UDP-N- acetylglucosamine 2-epimerase/N-acetylmannosamine kinase. Eur. J. Biochem. 252, 133–139. 15. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680– 685. 16. Altschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J. (1990) Basic local alignment search tool. J. Mol. Biol. 215, 403– 410. 17. Corpet, F. (1988) Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res. 16, 10881–10890. 18. Rost, B. & Sander, C. (1993) Prediction of protein secondary structure at better than 70% accuracy. J. Mol. Biol. 232, 584–599. 19. Jones, D.T. (1999) Protein secondary structure prediction based on position-specific scoring matrices. J. Mol. Biol. 292, 195– 202. 20. Ouali, M. & King, R.D. (2000) Cascaded multiple classifiers for secondary structure prediction. Protein Sci. 9, 1162–1176. 21. Allen, M.B. & Walker, D.G. (1980) The isolation and preliminary characterization of N-acetyl- D -glucosamine kinase from rat kidney and liver. Biochem. J. 185, 565–575. 22. Meglasson, M.D. & Matschinsky, F.M. (1986) Pancreatic islet glucose metabolism and regulation of insulin secretion. Diabetes Metab. Rev. 2, 163–214. 23. St Charles, R., Harrison, R.W., Bell, G.I., Pilkis, S.J. & Weber, I.T. (1994) Molecular model of human beta-cell glucokinase built by analogy to the crystal structure of yeast hexokinase B. Diabetes 43, 784–791. 24. Fletterick, R.J., Bates, D.J. & Steitz, T.A. (1975) The structure of a yeast hexokinase monomer and its complexes with substrates at 2.7-A resolution. Proc. Natl Acad. Sci. USA 72, 38–42. 25. Aleshin, A.E., Kirby, C., Liu, X., Bourenkov, G.P., Bartunik, H.D., Fromm, H.J. & Honzatko, R.B. (2000) Crystal structures of mutant monomeric hexokinase I reveal multiple ADP binding sites and conformational changes relevant to allosteric regulation. J. Mol. Biol. 296, 1001–1015. 26. Lundblad, R.L.N. & C.M. (1984) Chemical Reagents for Protein Modification. CRC Press, New York. 27. Rippa, M., Bellini, T., Signorini, M. & Dallocchio, F. (1981) Evidence for multiple pairs of vicinal thiols in some proteins. J. Biol. Chem. 256, 451–455. 28. Tiedge, M., Richter, T. & Lenzen, S. (2000) Importance of cysteine residues for the stability and catalytic activity of human pancreatic beta cell glucokinase. Arch. Biochem. Biophys. 375, 251–260. 29. Mahalingam, B., Cuesta-Munoz, A., Davis, E.A., Matschinsky, F.M., Harrison, R.W. & Weber, I.T. (1999) Structural model of human glucokinase in complex with glucose and ATP: implica- tions for the mutants that cause hypo- and hyperglycemia. Diabetes 48, 1698–1705. 30. DelaFuente, G., Lagunas, R. & Sols, A. (1970) Induced fit in yeast hexokinase. Eur. J. Biochem. 16, 226–233. 31. Datta, A. (1971) Studies on hog spleen N-acetylglucosamine kin- ase. II. Allosteric regulation of the activity of N-acetylglucosamine kinase. Arch. Biochem. Biophys. 142, 645–650. 4218 M. Berger et al.(Eur. J. Biochem. 269) Ó FEBS 2002 . Structure and function of N -acetylglucosamine kinase Identification of two active site cysteines Markus Berger, Hao Chen, Werner Reutter and Stephan. N-acetylglucosamine kinase. In this study we describe the identification of two active site cysteines of the sugar kinase by site- directed mutagenesis and