Báo cáo khoa học: Active site mutants of Drosophila melanogaster multisubstrate deoxyribonucleoside kinase pot

Active site mutants of Drosophila melanogaster multisubstrate deoxyribonucleoside kinase Nicola Solaroli 1,2 , Mia Bjerke 1 , Marjan H. Amiri 1 , Magnus Johansson 1 and Anna Karlsson 1 1 Division of Clinical Virology F68, Karolinska Institute, Huddinge University Hospital, Stockholm, Sweden; 2 Dipartimento di Scienze Farmaceutiche, Universita ` di Ferrara, Ferrara, Italy The multisubstrate deoxyribonucleoside kinase of Droso- phila melanogaster (Dm-dNK) is sequence-related to three human deoxyribonucleoside kinases and to herpes simplex virus type-1 thymidine kinase. Dm-dNK phos- phorylates both purine and pyrimidine deoxyribonucleo- sides and nucleoside analogues although it has a preference for pyrimidine nucleosides. We performed site-directed mutagenesis on residues that, based on structural data, are involved in substrate recognition. The aim was to increase the phosphorylation efficiency of purine nucleoside sub- stratestocreateanimprovedenzymetobeusedinsuicide gene therapy. A Q81N mutation showed a relative increase in deoxyguanosine phosphorylation compared with the wild-type enzyme although the efficiency of deoxythymidine phosphorylation was 10-fold lower for the mutant. In addition to residue Q81 the function of amino acids N28, I29 and F114 was investigated by different substitutions. All of the mutated enzymes showed decreased efficiency of thymidine phosphorylation in comparison with the wild- type enzyme supporting their importance for substrate binding and/or catalysis as proposed by the recently solved structure of Dm-dNK. Keywords: gene therapy; nucleoside analog; suicide gene. The deoxyribonucleoside kinase of the fruit fly Drosophila melanogaster (Dm-dNK) is a multisubstrate enzyme that phosphorylates pyrimidine and purine deoxyribonucleo- sides as well as several anticancer and antiviral nucleoside analogues [1–3]. Dm-dNK is sequence-related to the human deoxycytidine kinase (dCK), deoxyguanosine kinase (dGK) and thymidine kinase (TK2), as well as to the herpes simplex virus type-1 thymidine kinase (HSV-1 TK) [2]. Although the human and viral deoxyribonucleoside kinases can phos- phorylate multiple deoxyribonucleosides, Dm-dNK is the only enzyme in the nucleoside kinase enzyme family that has the ability to phosphorylate all naturally occurring deoxy- ribonucleosides required for DNA replication. In addition to its broad substrate specificity, Dm-dNK also exhibits higher catalytic rates for nucleoside and nucleoside analogue phosphorylation compared with other nucleoside kinases. Nucleoside kinases are presently being investigated for possible use as suicide genes in combined gene/chemo- therapy of cancer [4]. The most commonly studied nucleoside kinase suicide gene is the HSV-1 TK gene used in combination with the guanosine nucleoside analogue ganciclovir (GCV) [4–6]. The suicide nucleoside kinase is rate-limiting in the pharmacological activation of the cytotoxic nucleoside analogues, and mutants of HSV-1 TK with improved biochemical properties for nucleoside analogue phosphorylation are more efficient suicide genes [7–9]. The broad substrate specificity of Dm-dNK and its high catalytic rate makes it an interesting candidate gene for suicide gene therapy. We have recently evaluated the possible use of Dm-dNK as a suicide gene and shown that over-expression of Dm- dNK enhances the sensitivity of cancer cells to several cytotoxic nucleoside analogues [10]. Although Dm-dNK phosphorylates both purine and pyrimidine nucleosides, the enzyme has a preference for pyrimidine nucleosides [1–3]. The maximal catalytic rate of purine and pyrimidine nucleoside phosphorylation is simi- lar, but the enzyme exhibits higher affinity for pyrimidine nucleosides and nucleoside analogues. For suicide gene therapy application, purine nucleoside analogues may be preferred because these compounds appear to induce a higher bystander cell killing [11], i.e. killing of untransduced neighbouring cells by transfer of phosphorylated nucleoside analogues via gap junctions. The recently solved structures of Dm-dNK and dGK [12], and the already available structure of HSV-1 TK, reveal a common folding of these enzymes and in particular the amino acid residues involved in substrate interactions are highly conserved. However, the substrate binding site also exhibits some major differences between Dm-dNK and HSV-1 TK (Fig. 1). In particular the HSV-1 TK configuration with P57, H58 and Y172, that corres- ponds to Dm-dNK N28, I29 and P114, is unique in comparison with the other kinases (Fig. 2). HSV-1 TK Correspondence to N. Solaroli, Division of Clinical Virology F68, Karolinska Institute, Huddinge University Hospital, S-14186 Stock- holm, Sweden. Fax: + 46 8 58587933, Tel.: + 46 8 58587932, E-mail: nicola.solaroli@labmed.ki.se Abbreviations: Dm-dNK, Drosophila melanogaster deoxyribonucleo- side kinase; dCK, deoxycytidine kinase; dGK, deoxyguanosine kinase; TK2, thymidine kinase 2; HSV-1 TK, herpes simplex virus type-1 thymidine kinase; dAdo, deoxyadenosine; dCyd, deoxycytidine; dGuo, deoxyguanosine; dThd, deoxythymidine; dTMP, deoxythy- midine monophosphate; CdA, 2-chloro-2¢-deoxyadenosine; dFdC, 2¢,2¢-difluorodeoxycytidine; AraC, 1-b- D -arabinofuranosylcytosine; BVDU, (E)-5-(2-bromovinyl)-2¢-deoxyuridine; AZT, azidothymidine. (Received 12 March 2003, revised 16 April 2003, accepted 13 May 2003) Eur. J. Biochem. 270, 2879–2884 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03666.x residue H58, coordinated by P57 and by a hydrogen bond with Y172, has the ability to make a hydrogen bond with ganciclovir indicating that this particular structure may be important for the ganciclovir binding that is unique for HSV-1 TK. We have used the structural information to perform site-directed mutagenesis of the residues N28, I29, F114 and Q81 in the Dm-dNK active site. The aim of the study was in part to understand the determinants of the substrate specificity of the enzyme and in part to find Dm-dNK mutants with improved kinetic properties for application in suicide gene therapy. Materials and methods Sequence and structural analysis The amino acid sequences were aligned and analysed with GENEDOC software [13]. SWISS - PDB VIEWER v3.6 (Glaxo Wellcome) [14], MACROMODEL v7.0 (Schroedinger) [15] were used for structural analysis. HSV-1 TK three- dimensional structure was 1VTK, and Dm-dNK three- dimensional structure was 1J90, both downloaded from Protein Data Bank [16]. ICM - LITE v2.8 was used for the creation of Fig. 2. Site-directed mutagenesis Site-directed mutagenesis of the Dm-dNK cDNA was performed using the QuikChange site-directed mutagenesis kit (Stratagene). Primers used were: for N28 to P28, 5¢-CTCATCGAGGGC CCAATCGGCAGCGGGAAGA CCACG-3¢ and 5¢-CGTGGTCTTCCCGCTGCCGAT TGGGCCCTCGATGAG-3¢;forI29toH29,5¢-CTCAT CGAGGGCAAC CATGGCAGCGGGAAGACC-3¢ and 5¢-GGTCTTCCCGCTGCC ATGGTTGCCCTCG ATGAG-3¢;forN28+I29toP28+H29,5¢-CTC ATCGAGGGC CCACATGGCAGCGGGAAGACCA CG-3¢ and 5¢-CGTGGTCTTCCCGCTGCC ATGTGG GCCCTCGATGAG-3¢; for Q81 to N81, 5¢-GGG CCATGCCCTTT AACAGTTATGTCACGCTGACC-3¢ and 5¢-GGTCAGCGTGACATAACT GTTAAAGGGCA TGGCCC-3¢; for Q81 to D81, 5¢-GGGCCATGCCCT Fig. 1. Alignment of Dm-dNK with HSV-1 TK,dGK,dCK,andTK2.Black boxes indicate completely conserved amino acid residues (upper case) and grey boxes indicate semi- conserved residues (lower case). Arrows indicate the location of the studied mutants. Numbering is based on the amino acid sequence of Dm-dNK. Fig. 2. Structure overlay of Dm-dNK (black) and HSV-1 TK (grey) active sites. 2880 N. Solaroli et al. (Eur. J. Biochem. 270) Ó FEBS 2003 TTGACAGTTATGTCACGCTGACC-3¢ and 5¢-GGTC AGCGTGACATAACT GTCAAAGGGCATGGCCC-3¢; for F114 to Y114, 5¢-GCGCTCGCTATTGCT ACGT GGAGAACATGCG-3¢ and 5¢-CGCATGTTCTCCACG TAGCAATAGCGAGCGC-3¢. The Dm-dNK cDNA mutations were verified by sequence determinations of both strands using an ABI 310 sequencer (Applied Biosystems) and the BigDye cycle sequencing kit. Expression and purification We expressed the Dm-dNK cDNA in Escherichia coli as a fusion protein to glutathione S-transferase. The plasmids were transformed into HMS174(DE3) (Novagen) and single colonies were inoculated into Luria–Bertani medium supplemented with 100 lgÆmL )1 ampicillin. The bacteria were grown at 37 °C and protein expression was induced at D 600  0.8 with 1 m M isopropyl-1-thio-b- D -galacto- pyranoside for  12 h at 27 °C. The expressed protein was purified using glutathione-sepharose 4B (Amersham Pharmacia Biotech) as described [17]. The purity of the enzymes are verified by SDS/PAGE (Phast system, Amersham Pharmacia Biotech) and the protein concen- tration was determined with Bradford Protein Assay (Bio- Rad) using BSA as the concentration standard. The protein was aliquoted, frozen in liquid nitrogen, and stored at )80 °C. Enzyme assays The activity of the purified recombinant enzymes was assayed in a 50-lL reaction mixture containing: 50 m M Tris/HCl pH 7.6, 0.1 mgÆmL )1 BSA, 2.5 m M ATP, 5 m M MgCl 2 ,5m M dithiothreitol, and 0.15 l M [methyl- 3 H]dThd (Amersham Pharmacia Biotech). The samples were incu- bated for 30 min at 37 °C and every 10 min 10 lL aliquots were spotted on Whatman DE-81 filter paper disks. The filters were dried 1 h, washed three times for 5 min in 5 m M ammonium formate and once in sterile water. The filter bound nucleoside monophosphates were eluted in 500 lL 0.1 M HCl and 0.1 M KCl and the radioactivity quantified by scintillation counting. The Michaelis–Menten constants were calculated using the GRAPHPAD PRISM software. The substrate specificity of the purified enzymes was assayed by TLC as described [18]. Briefly, the assay was performed in 50 m M Tris/HCl pH 7.6, 0.5 mgÆmL )1 BSA, 5m M MgCl 2 ,5m M dithiothreitol, 10 m M ATP, 15 lCi [c- 32 P]ATP (Amersham Pharmacia Biotech) and recombin- ant Dm-dNK. The samples were incubated for 30 min at 37 °C. Two lL of the reaction mixtures were spotted on polyethyleneimine-cellulose F TLC sheets (Merck), and the nucleosides and nucleotides were separated in a buffer containing NH 4 OH : isobutyric acid : dH 2 O (1 : 66 : 33). The sheets were autoradiographed using phosphorimaging plates (BAS-1000, Fujix). Results We have performed site-directed mutagenesis of four active site amino acid residues of Dm-dNK. The residues are N28, I29, Q81 and F114 that according to primary and tertiary structural alignments correspond to residues P57, H58, Q125 and Y172 of HSV-1 TK, respectively (Figs 1 and 2). Dm-dNKQ81andHSV-1TKQ125areinvolvedin substrate binding, and this residue forms two hydrogen bonds to the nucleoside base [12,19]. Mutagenesis of HSV-1 TK Q125 to asparagine has been shown to decrease the enzyme affinity for dThd and delete the dTMP kinase activity while the affinity for the purine nucleoside analogue ganciclovir is relatively less affected [20,21]. We decided to mutate Q81 in Dm-dNK, which corresponds to HSV-1 TK Q125, and determine the kinetic properties of the recom- binant enzyme. The Q81N mutant showed, compared with the wild-type enzyme, a marked relative increase in dGuo phosphorylation, and a minor increase in dAdo phosphory- lation (Table 1). The Q81N enzyme also phosphorylated 2¢,2¢-difluorodeoxyguanine (dFdG) with similar efficiency as dGuo, but not 9-b- D -arabinofuranosylguanosine (AraG) or acyclovir (data not shown). However, similar to the Q125N HSV-1 TK mutant, the absolute V max and the affinity for dThd were decreased (Table 2). The efficiency of substrates phosphorylation determined as V max /K m and compared with wild-type enzyme results was  10-fold lower for dThd,  fourfold lower for dAdo, but  fourfold higher for dGuo. Mutation of Dm-dNK Q81 to aspartate, corresponding to HSV-1 TK Q125D mutant, showed almost a complete loss of activity for all substrates tested. Table 1. Nucleoside and nucleoside analogue phosphorylation by recombinant Dm-dNK mutant enzymes. Relative levels of phosphorylation expressed in relation to percentage dThd phosphorylation of the corresponding mutant. n.d., Not detectable. Dm-dNK mutations Protein amount Substrates (100 l M ) dThd dCyd dAdo dGuo CdA dFdC AraC GCV BVDU Wild-type 2 ng 100 142 55 <1 138 24 69 <1 59 N28P 2 lg n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. I29H 2 lg 100 159 52 <1 134 76 <1 <1 193 Q81N 200 ng 100 97 98 92 80 103 109 <1 102 Q81D 2 lg n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. F114A 200 ng 100 575 <1 <1 59 <1 156 <1 113 F114Y 200 ng 100 95 10 6 17 50 58 <1 59 N28P + I29H 2 lg n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. I29H + F114Y 2 lg n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. N28P + I29H + F114Y 2 lg n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Ó FEBS 2003 Dm-dNK active site mutants (Eur. J. Biochem. 270) 2881 The Dm-dNK amino acid F114 is conserved in dCK, dGK and TK2, whereas HSV1-TK has a tyrosine residue at this position. In HSV-1 TK this residue, Y172, stacks against the nucleoside base [19]. Random mutagenesis of the herpes enzyme has shown that this residue can be functionally replaced only by a phenylalanine [22]. We decided to investigate whether replacing the Dm-dNK phenylalanine at this position with a tyrosine, would make the enzyme more similar to HSV-1 TK. The F114Y Dm- dNK mutant showed an eightfold decreased dThd affinity and a 50-fold decrease in catalytic rate compared with the wild-type enzyme (Table 2). No major difference in phos- phorylation relative to thymidine was detected for the tested nucleosides and nucleoside analogues (Table 1). Replacing F114 with an alanine residue resulted in a K m for dThd similar to that of the wild-type enzyme, but the V max was at least 4000-fold lower than for the wild-type enzyme. Although a shift in the nucleoside and nucleoside analogue phosphorylation relative to dThd was detected for this mutant enzyme, with a predominantly increased dCyd phosphorylation, the overall low catalytic rate of the enzyme makes these results difficult to interpret. The glycine-rich loop, which corresponds to the ATP- phosphate binding site is highly conserved in all members of the nucleoside kinase enzyme family (Fig. 2). In particular in the middle of the HSV-1 TK glycine-rich loop there is H58 which forms a stabilizing interaction with Y172 [19] and is hydrogen bonded with the sugar moiety, or with the antiviral drug ganciclovir. These interactions have been suggested to be important for the ability of HSV-1 TK to bind and phosphorylate the acyclic guanosine analogues. The corresponding amino acid residue in Dm-dNK is an isoleucine (I29). To investigate whether replacement of this residue with a histidine would alter the substrate specificity of the enzyme, we created the I29H mutant. However, this enzyme showed very low activity and no major alteration in substrate specificity (Tables 1 and 2). Adjacent to amino acid H58 of HSV-1 TK is P57, which is not conserved in any of the other nucleoside kinases. The chemical properties of the proline residue may cause it to affect the angle and position of the glycine-rich loop. We therefore decided to insert a proline at the same position in Dm-dNK. However, the N28P mutant showed a complete loss of activity. We also decided to investigate the effect of combining different mutations in Dm-dNK. As stated above, H58 may interact with Y172 in HSV-1 TK and this interaction may be important for the binding and phosphorylation of ganciclovir [19]. We therefore created a Dm-dNK enzyme containing both I29H and F114Y mutations. However, this enzyme did not show any activity for any of the investigated compounds. Neither the combination of N28P with I29H nor the triple combination of N28P with I29H and F114Y showed any activity. Discussion We have performed site-directed mutagenesis of Dm-dNK in order to gain further understanding of the catalytic mechanism of the enzyme and the properties determining the broad substrate specificity of the enzyme. Although the maximal rate of purine and pyrimidine nucleoside phos- phorylation is similar, the enzyme exhibits higher affinity for pyrimidine nucleosides and nucleoside analogues [1–3]. We aimed to generate mutant enzymes with improved kinetic properties for possible use in suicide gene therapy. In particular we wanted to investigate the possibility of creating an enzyme with increased affinity for the purine nucleoside analogue ganciclovir. Ganciclovir is a parti- cularly interesting nucleoside analogue as it is known to have efficient bystander cell killing. Bystander cell killing is likely to be an important factor for the success of suicide gene therapy because presently available vectors allow only a minor portion of the cells in a tumour to be transduced. We have studied the bystander effect of the pyrimidine nucleoside analogues (E)-5-(2-bromovinyl)-2¢-deoxyuridine and 1-b- D -arabinofuranosylthymine in cells expressing Dm- dNK [23,24]. Although a bystander effect is detected, it is less efficient than the bystander effect observed for purine nucleoside analogues such as ganciclovir in combination with HSV-1 TK. Accordingly, a modified Dm-dNK with enhanced purine nucleoside analogue phosphorylation may be more useful. The four active site amino acid residue mutants charac- terized in the present study showed partial or complete loss of activity compared with the wild-type enzyme in terms of dThd phosphorylation. The Dm-dNK residue Q81, corres- ponding to HSV-1 TK Q125, forms hydrogen bonds with thenucleosidebase[12].TheQ81NDm-dNK mutation, similar to the HSV-1 TK Q125N mutant, exhibited decreased dThd phosphorylation but a relative increase in purine nucleoside phosphorylation. This finding supports a role of this residue in binding the nucleoside base, but this amino acid residue alone does not explain the differences observed in substrate specificity between the different nucleoside kinases as this residue is conserved in all members of the enzyme family. The tyrosine residue 172 can functionally replace phenylalanine in HSV-1 TK [22], however, the Dm-dNK F114Y mutant, corresponding to HSV-1 TK Y172, exhibited a marked loss of activity. The Dm-dNK F114A mutant showed very low catalytic rate, but interestingly exhibited a similar K m fordThdasthe wild-type enzyme. Table 2. Kinetics of dThd, dGuo and dAdo phosphorylation by recom- binant Dm -dNK mutants. Dm-dNK mutation K m (l M ) V max (nmolÆmg )1 Æmin )1 ) V max /K m dThd Wild-type 2.5 4350 1740 I29H >100 <1 <0.01 Q81N 4.7 835 180 Q81D >100 <1 <0.01 F114A 3.1 <1 <0.3 F114Y 22 93 4.2 dGuo Wild-type a 2000 3666 1.8 Q81N 985 6700 6.8 dAdo Wild-type a 373 15166 40 Q81N 625 7260 11.6 a Data from Johansson 1999. 2882 N. Solaroli et al. (Eur. J. Biochem. 270) Ó FEBS 2003 We also hypothesized that the amino acids adjacent to the ATP-binding phosphate loop may be important in deter- mining the specificity of the enzymes. However, the single amino acid mutants of N28P and I29H or the combination of the two mutants with or without the F114Y showed complete or almost complete loss of activity for all the investigated substrates. There are principally two ways to improve the activation of nucleoside analogue phosphorylation in vivo. Either the affinity and catalytic rate for the cytotoxic nucleoside analogue is increased, or the affinity to competing nucleo- sides is decreased. As all nucleoside kinases studied for possible use in gene therapy also phosphorylate naturally occurring deoxyribonucleosides, there will be competition between the deoxyribonucleosides and the analogue for phosphorylation by the nucleoside kinase. Accordingly, by decreasing the enzyme affinity to natural deoxyribonucleo- sides an increased nucleoside analogue phosphorylation may be achieved. In a recent study, random mutagenesis of Dm-dNK was used in order to identify mutant enzymes with improved properties for phosphorylation of nucleoside analogues [25]. In this study a double mutant (N45D, N64D) was selected based on improved sensitivity towards azidothymidine (AZT) and dideoxycytidine (ddC) when expressed in a thymidine kinase-deficient E. coli strain. The characterization of the mutant enzyme showed decreased efficiency for thymidine phosphorylation resulting in a more favourable phosphorylation of AZT and ddC compared with wild-type Dm-dNK. In addition, this mutant enzyme exhibited a decreased feedback regulation by dTTP and this was suggested to be an additional mechanism resulting in an increased accumulation of the triphosphates of AZT and ddC. Similar strategies have previously been used for generating improved HSV-1 TK mutants [9]. In the HSV-1 TK study, mutants with enhanced ability to phosphorylate AZT were identified. The characterization of these mutants showed a reduced K m for AZT and decreased specificity for thymidine indicating that the mutant enzymes did better accommodate the azido group of AZT. However, randomly selected mutant enzymes may contain additional substitu- tions and thus prediction of the importance of specific residues is speculative. In conclusion, site-directed muta- genesis contributes with important information regarding the role of specific amino acids of an enzyme. However, additional methods, such as random mutagenesis in com- bination with efficient screening systems, may be more useful to obtain Dm-dNK mutants with specific properties for suicide gene therapy. Acknowledgements This work was supported by grants from the Swedish Research Council, the Swedish Cancer Foundation, the European Commission, the Medical Faculty of the Karolinska Institute, and the Hedlund Foundation. References 1. Munch-Petersen, B., Piskur, J. & Sondergaard, L. (1998) Four deoxynucleoside kinase activities from Drosophila melanogaster are contained within a single monomeric enzyme, a new multi- functional deoxynucleoside kinase. J. Biol. Chem. 273, 3926–3931. 2. Johansson, M., van Rompay, A.R., Degreve, B., Balzarini, J. & Karlsson, A. (1999) Cloning and characterization of the multisubstrate deoxyribonucleoside kinase of Drosophila melano- gaster. J. Biol. Chem. 274, 23814–23819. 3. Munch-Petersen, B., Knecht, W., Lenz, C., Sondergaard, L. & Piskur, J. (2000) Functional expression of a multisubstrate deoxy- ribonucleoside kinase from Drosophila melanogaster and its C-terminal deletion mutants. J. Biol. Chem. 275, 6673–6679. 4. Springer, C.J. & Niculescu-Duvaz, I. (2000) Prodrug- activating systems in suicide gene therapy. J. Clin. Invest. 105, 1161–1167. 5. Moolten, F.L. (1986) Tumor chemosensitivity conferred by inserted herpes thymidine kinase genes: paradigm for a prospect- ive cancer control strategy. Cancer Res. 46, 5276–5281. 6. Culver, K.W., Ram, Z., Wallbridge, S., Ishii, H., Oldfield, E.H. & Blaese, R.M. (1992) In vivo gene transfer with retroviral vector- producer cells for treatment of experimental brain tumors. Science 256, 1550–1552. 7. Kokoris, M.S., Sabo, P., Adman, E.T. & Black, M.E. (1999) Enhancement of tumor ablation by a selected HSV-1 thymidine kinase mutant. Gene Ther. 6, 1415–1426. 8. Black, M.E., Newcomb, T.G., Wilson, H.M. & Loeb, L.A. (1996) Creation of drug-specific herpes simplex virus type 1 thymidine kinase mutants for gene therapy. Proc. Natl Acad. Sci. USA 93, 3525–3529. 9. Christians, F.C., Scapozza, L., Crameri, A., Folkers, G. & Stemmer, W.P. (1999) Directed evolution of thymidine kinase for AZT phosphorylation using DNA family shuffling. Nat. Biotechnol. 17, 259–264. 10. Zheng, X., Johansson, M. & Karlsson, A. (2000) Retroviral transduction of cancer cell lines with the gene encoding Drosophila melanogaster multisubstrate deoxyribonucleoside kinase. J. Biol. Chem. 275, 39125–39129. 11. Degreve, B., De Clercq, E. & Balzarini, J. (1999) Bystander effect of purine nucleoside analogues in HSV-1 tk suicide gene therapy is superior to that of pyrimidine nucleoside analogues. Gene Ther. 6, 162–170. 12. Johansson, K., Ramaswamy, S., Ljungcrantz, C., Knecht, W., Piskur, J., Munch-Petersen, B., Eriksson, S. & Eklund, H. (2001) Structural basis for substrate specificities of cellular deoxy- ribonucleoside kinases. Nat. Struct. Biol. 8, 616–620. 13. Nicholas, K.B., Nicholas, H.B.J. & Deerfield, D.W.I. (1997) GeneDoc: analysis and visualization of genetic variation. EMB- NEW.NEWS 4, 14. 14. Guex, N. & Peitsch, M.C. (1997) SWISS-MODEL and the Swiss- PdbViewer: an environment for comparative protein modeling. Electrophoresis 18, 2714–2723. 15. Mohamadi, F., Richards, N.G.J., Guida, W.C., Liskamp, R., Lipton, M., Caufield, C., Chang, G., Hendrickson, T. & Still, W.C. (1990) MacroModel – an integrated software system for modeling organic and bioorganic molecules using molecular mechanics. J. Comput. Chem. 11, 440–467. 16. Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig,H.,Shindyalov,I.N.&Bourne,P.E.(2000)TheProtein Data Bank. Nucleic Acids Res. 28, 235–242. 17. Johansson, M. & Karlsson, A. (1997) Cloning of the cDNA and chromosome localization of the gene for human thymidine kinase 2. J. Biol. Chem. 272, 8454–8458. 18. Eriksson, S., Kierdaszuk, B., Munch-Petersen, B., Oberg, B. & Johansson, N.G. (1991) Comparison of the substrate specificities of human thymidine kinase 1 and 2 and deoxycytidine kinase toward antiviral and cytostatic nucleoside analogs. Biochem. Biophys. Res. Commun. 176, 586–592. 19. Brown, D.G., Visse, R., Sandhu, G., Davies, A., Rizkallah, P.J., Melitz, C., Summers, W.C. & Sanderson, M.R. (1995) Crystal structures of the thymidine kinase from herpes simplex virus Ó FEBS 2003 Dm-dNK active site mutants (Eur. J. Biochem. 270) 2883 type-1 in complex with deoxythymidine and ganciclovir. Nat. Struct. Biol. 2, 876–881. 20. Drake, R.R., Wilbert, T.N., Hinds, T.A. & Gilbert, K.M. (1999) Differential ganciclovir-mediated cell killing by glutamine 125 mutants of herpes simplex virus type 1 thymidine kinase. J. Biol. Chem. 274, 37186–37192. 21. Hinds, T.A., Compadre, C., Hurlburt, B.K. & Drake, R.R. (2000) Conservative mutations of glutamine-125 in herpes simplex virus type 1 thymidine kinase result in a ganciclovir kinase with minimal deoxypyrimidine kinase activities. Biochemistry 39, 4105–4111. 22. Munir,K.M.,French,D.C.,Dube,D.K.&Loeb,L.A.(1992) Permissible amino acid substitutions within the putative nucleo- side binding site of herpes simplex virus type 1 encoded thymidine kinase established by random sequence mutagenesis [corrected] [published erratum appears in J Biol Chem 1992 July 25; 267 (21): 15258]. J. Biol. Chem. 267, 6584–6589. 23. Zheng, X., Johansson, M. & Karlsson, A. (2001) Nucleoside analog cytotoxicity and bystander cell killing of cancer cells expressing Drosophila melanogaster deoxyribonucleoside kinase in the nucleus or cytosol. Biochem. Biophys. Res. Commun. 289, 229–233. 24. Zheng, X., Johansson, M. & Karlsson, A. (2001) Bystander effects of cancer cell lines transduced with the multisubstrate deoxyribonucleoside kinase of Drosophila melanogaster and synergistic enhancement by hydroxyurea. Mol. Pharmacol. 60, 262–266. 25. Knecht, W., Munch-Petersen, B. & Piskur, J. (2000) Identification of residues involved in the specificity and regulation of the highly efficient multisubstrate deoxyribonucleoside kinase from Droso- phila melanogaster. J. Mol. Biol. 301, 827–837. 2884 N. Solaroli et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . Active site mutants of Drosophila melanogaster multisubstrate deoxyribonucleoside kinase Nicola Solaroli 1,2 , Mia Bjerke 1 ,. Italy The multisubstrate deoxyribonucleoside kinase of Droso- phila melanogaster (Dm-dNK) is sequence-related to three human deoxyribonucleoside kinases