Báo cáo khoa học: Effect of valine 106 on structure–function relation of cytosolic human thymidine kinase Kinetic properties and oligomerization pattern of nine substitution mutants of V106 ppt

Effect of valine 106 on structure–function relation of cytosolic human thymidine kinase Kinetic properties and oligomerization pattern of nine substitution mutants of V106 Hanne Frederiksen*†, Dvora Berenstein† and Birgitte Munch-Petersen Department of Life Sciences and Chemistry, Roskilde University, Denmark Information on the regulation and structure–function rela- tion of enzymes involved in DNA precursor synthesis is pivotal, as defects in several of these enzymes have been found to cause depletion or deletion of mitochondrial DNA resulting in severe diseases. Here, the effect of amino acid 106 on the enzymatic properties of the cell-cycle-regulated human cytosolic thymidine kinase 1 (TK1) is investigated. On the basis of the previously observed profound differences between recombinant TK1 with Val106 (V106WT) and Met106 (V106M) in catalytic activity and oligomerization pattern, we designed and characterized nine mutants of amino acid 106 differing in size, conformation and polarity. According to their oligomerization pattern and thymidine kinetics, the TK1 mutants can be divided into two groups. Group I (V106A, V106I and V106T) behaves like V106WT, in that pre-assay exposure to ATP induces reversible transition from a dimer with low catalytic activity to a tetr- amer with high catalytic activity. Group II (V106G, V106H, V106K, V106L and V106Q) behaves like V106M in that they are permanently high activity tetramers, irrespective of ATP exposure. We conclude that size and conformation of amino acid 106 are more important than polarity for the catalytic activity and oligomerization of TK1. The role of amino acid 106 and the sequence surrounding it for dimer–tetramer transition was confirmed by cloning the putative interface fragment of human TK1 and investigating its oligomeri- zation pattern. Keywords: dimer–tetramer formation; enzyme kinetics; enzyme mutants; structure–function relation; thymidine kinase. Enzymes involved in salvage and metabolism of deoxy- nucleosides have an important role in the regulation of DNA precursors for DNA synthesis and repair. Recently, severe syndromes, such as mitochondrial neurogastrointes- tinal encephalomyopathy and mitochondrial DNA deple- tion syndrome which lead to multiple mitochondrial DNA abnormalities, were found to be caused by defects in the cytoplasmic thymidine phosphorylase [1,2] or the two mitochondrial deoxynucleoside kinases: deoxyguanosine kinase (dGK) and thymidine kinase 2 (TK2) respectively [3,4]. In contrast with earlier work suggesting spatial and metabolic separation of thymidine phosphate pools between the cytosol and mitochondria [5,6], recent evidence suggests the two compartments are connected by a rapid and dynamic exchange [7]. These findings may explain why defects in deoxynucleotide metabolic enzymes, mitochond- rial as well as cytoplasmic, lead to severe mitochondrial DNA abnormality syndromes. Therefore, it is of great importance to acquire detailed knowledge about the prop- erties of the enzymes involved in balancing the cellular and mitochondrial dNTP pools. Human cytosolic thymidine kinase (TK1; EC 2.7.1.21) is a salvage pathway enzyme in the synthesis of the DNA precursor dTTP. It catalyzes the first step of this pathway, in which thymidine is phosphorylated to dTMP [8]. In turn, intracellular dTMP is rapidly phosphorylated to dTTP, an allosteric effector of ribonucleotide reductase [9]. Imbal- ances in the dTTP pool are thus followed by an imbalanced supply of the four deoxyribonucleoside triphosphates for DNA synthesis and repair, and result in increased rates of mutation and the probability of carcinogenesis [10]. TK1 is cell-cycle regulated and its activity fluctuates with DNA synthesis [11,12]. The subunit size of TK1 is 24 kDa [13,14], and the native enzymes purified from human lymphocytes [14] and HeLa cells [15] were found to be tetramers in the presence of ATP. In the presence of thymidine instead of Correspondence to B. Munch-Petersen, Department of Life Sciences and Chemistry, Roskilde University, PO Box 260, DK-4000 Roskilde, Denmark. Fax: + 45 46743011, Tel.: + 45 46742418, E-mail: bmp@ruc.dk Abbreviations: dCK, deoxycytidine kinase; dGK, deoxyguanosine kinase; dNK, multisubstrate nucleoside kinase from Drosophila mel- anogaster; GST, glutathione S-transferase; HSV1-TK, Herpes simplex type-1 thymidine kinase; TK1, human cytosolic thymidine kinase; rLy-TK1 Val106 , recombinant TK1 expressed from cDNA derived from human lymphocytes, the same as rLy-TK1 (V106WT); rLy-TK1 66)136 , the putative interface fragment of TK1 corresponding to residues 66–136; TK1+ATP, rLy-TK1 incubated and stored with 2.5 m M ATP/MgCl 2 ;TK1)ATP, rLy-TK1 incubated and stored without ATP/MgCl 2 ; TK2, human mitochondrial thymidine kinase. Enzyme: Human cytosolic thymidine kinase (TK1; EC, 2.7.1.21). *Present address: Institute of Food and Veterinary Research, Department of Toxicology and Risk Assessment, Mørkhøj Bygade 19, DK-2860 Søborg, Denmark. These authors contributed equally to this publication. Note: A website is available at http://www.ruc.dk (Received 1 March 2004, revised 6 April 2004, accepted 16 April 2004) Eur. J. Biochem. 271, 2248–2256 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04166.x ATP or without substrates present, TK1 appears as a dimer [14]. Human TK1 has 234 amino acids, and, in the originally published primary sequence, amino acid 106 was methio- nine [16,17]. Our group has recently analysed TK1 cDNA and genomic DNA from 22 normal or transformed cell lines, and in all cases we found a valine at amino acid position 106 [18]. Also, alignment of mammalian TK1 and TK from vaccinia virus (Fig. 1) demonstrates the presence of valine at the site corresponding to amino acid 106 in human TK1, which is located in a highly conserved area thought to encompass the magnesium-binding and thymi- dine-binding sites [19,20]. We have found a remarkable difference in catalytic activity between recombinant TK1 expressed from human lymphocyte cDNA, rLy-TK1 Val106 (V106WT) and its mutant rLy-TK1 Met106 (V106M) [18]. V106WT was a dimer with low catalytic activity (K 0.5 for thymidine about 15 l M ), but pre-assay exposure to ATP induced an enzyme concentration-dependent reversible transition from a dimer to a tetramer with an % 30-fold higher catalytic activity (K 0.5 for thymidine % 0.5 l M ) [14,18,21]. The maximal velocities for the ATP exposed and unexposed forms were the same. In contrast, irrespect- ive of pre-assay exposure to ATP, V106M was a permanent tetramer with low K 0.5 for thymidine (% 0.5 l M ) and similar maximal velocities, which were % 2–3-fold lower than that of V106WT [18,21]. Until recently, the only deoxyribonucleoside kinase with a known 3D structure solved by X-ray crystallography was the Herpes simplex virus type-1 thymidine kinase (HSV1- TK) [22–26]. In 2001, the X-ray crystallographic structure was reported for two cellular deoxynucleoside kinases – the Drosophila melanogaster multisubstrate deoxynucleoside kinase (dNK) and the human deoxyguanosine kinase (dGK [27]) – and in 2003 the X-ray crystallographic structure of the human deoxycytidine kinase (dCK) was solved [28]. The amino acid sequence identity is 34% between dNK and dGK [29], and 47% between dGK and dCK [28], and the structures of dNK, dGK and dCK appeared to be very similar [27,28]. Despite the very low sequence identity of the cellular kinases with the Herpes virus TK (% 10%),thecorestructureshaveasimilarfold and there is also a close resemblance to the human and yeast thymidylate kinases [8,27]. Therefore, although the sequence identity of TK1 with HSV1-TK and the other cellular kinases belonging to the dNK group is too low for a reliable homology model (% 10%), TK1 may have the same overall structure as the other nucleoside kinases. Furthermore, a prediction of the secondary structure of TK1 [30–32] places Val106 in the middle of an a-helix which aligns in CLUSTAL W [33] with one of the interface helices (a-helix 4) of HSV1- TK. This may indicate that the area surrounding Val106 is integrated into the oligomerization interface. To obtain more information about this putative interface region of TK1, we sought to clarify the importance of amino acid 106 for the structure and function of the enzyme by mutating Val106 to amino acids differing in polarity, size and conformation, and subsequently investigating their effect on the quaternary structure and kinetics. Further- more, we confirmed that amino acid 106 and the neigh- bouring residues are involved in dimer–tetramer transition by cloning the putative interface fragment of human TK1, rLy-TK1 66)136 , and investigating the influence of the V106M mutation on the oligomerization properties of this fragment. Materials and methods Bacterial strains and plasmids The thymidine kinase-deficient strain of Escherichia coli, KY895 [34], and E.colistrain BL21 were used to propagate bacterial plasmids. BL21 was used for expression of recombinant TK1 enzymes. We have previously cloned the entire TK1 coding sequence into the BamHI–EcoRI restriction sites of the glutathione S-transferase (GST) fusion vector pGEX-2T, as described in [18]. This vector encodes a thrombin cleavage site between the GST gene and the multiple cloning site. Construction of pGEX-2T-LyTK1 Val106X mutants The plasmid pGEX-2T-LyTK1 Val106 [18] was used as template DNA for PCR, and mutations in the GTG codon coding for Val106 were introduced with the Quick Change TM site-directed mutagenesis kit from Stratagene (according to the instructions of the manufacturer). The sense [5¢-TTTTTCCCTGACATCGTGGAGTTCTGCGA GGCC(358–390)-3¢] and antisense [5¢-GGCCTCGCAGA ACTCCACGATGTCAGGGAAAAA(390–358)-3¢]muta- genic primers were substituted as follows in the target codon for Val106 (bold): G CG/CGC for Ala106, CAG/CTG for Gln106, G GT/ACC for Gly106, CAC/GTG for His106, ATC/GAT for Ile106, CTG/CAG for Leu106, AAA/TTT for Lys106, ATG/CAT for Met106 and ACC/GGT for Thr106. The altered bases are underlined and the codons are given in the sense/antisense primer, respectively. The base Fig. 1. Amino-acid sequence alignment of the putative interface region of human TK1 with related enzymes. Val106 is in bold and italics, the putative Mg 2+ -binding motif VIGID 97 [19,20] and the putative thymidine-binding motif FQRK 131 [20] are in italics in all sequences. In the human sequence, the b-branching amino acids and the a-helix breaking glycines and prolines are in bold and underlined. The sequences have the following GenBank identifier numbers: gi/23503074, human; gi/6678357, mouse; gi/125428, Chinese hamster; gi/125427, chicken; gi/9791018, vaccinia virus. Identical amino acids are indicated by asterisks. Ó FEBS 2004 Effect of amino acid 106 on human TK1 (Eur. J. Biochem. 271) 2249 numbering is as described in [16] where the translation initiation is at position 58, and therefore the codon for amino acid 106 starts at base number 373. Construction of pGEX-2T-LyTK1 66)136(Val106) Plasmid pGEX-2T-LyTK1 Val106 [18]wasusedastemplate for PCR with a sense primer: 5¢- GGG GGATCCTGCA CACATGACCGGAACACC(247–273)-3¢ designed to contain a GGG overhang and an antisense primer: 5¢-CGGCACC GAATTCTAGATGGCCCCAAATGGC TTCCT(480–445)-3¢. The numbering is as described in [16]. The underlined bases are changed in comparison with the original sequence to introduce a BamHI site (in bold) and the coding sequence for thrombin cleavage in the sense primer, and an EcoRI site (in bold) in the antisense primer. Thus, the N-terminal amino acids of the expressed fragment will be GS 66 CTHD instead of 66 CTHD. The PCR condi- tions were: 4 lgÆmL )1 template DNA, 3 m M MgCl 2 , 0.2 m M each dNTP and 0.36 l M each primer in 10 m M Tris/HCl buffer (pH 8.3) and 1 unit of Thermus aquaticus DNA polymerase (Stratagene) in a total volume of 25 lL; 30 cycles; 1 min at 94 °C, 1 min at 55 °C,and1minat 72 °C. The purified PCR product was ligated into the BamH1–EcoR1 restriction sites of the pGEX-2T vector and transformed into competent E.colicells. Codon CTG(466– 468) was mutated to a UAG stop signal by site-directed mutagenesis performed with the QuickChange TM site- directed mutagenesis kit according to the manufacturer’s instructions. The following mutagenic primers were used: sense, 5¢-CCATTTGGGGCCATC TAGAACCTGGTGC CGCTG(451–483)-3¢;antisense,5¢-CAGCGGCACCAG GTTC TAGATGGCCCCAAATGG(483–451)-3¢. The underlined bases were changed in comparison with the original sequence; the stop codon is in bold. Construction of pGEX-2T-LyTK1 66)136(Met106) GTG at positions 373–375 (bold), corresponding to amino acid 106 (numbers as described in [16]), was mutated to ATG with the QuickChange TM site-directed mutagenesis kit with the following primers: sense primer, 5¢-CAGTTTT TCCCTGACATC ATGGAGTTCTGCGAGGCCATG (355–393)-3¢; antisense primer, 5¢-CATGGCCTCGCAGA ACTCCATGATGTCAGGGAAAAACTG(393–355)-3¢. The changed bases are underlined. DNA sequencing pGEX-2T-LyTK1 Val106 , pGEX-2T-LyTK1 Val106X mutant plasmids and pGEX-2T-LyTK1 66)136(Met106) plasmid were sequenced on both strands using the Thermo Sequenase sequencing kit (Amersham Biosciences), and pGEX-2T- LyTK1 66)136(Val106) plasmid was sequenced on both strands with Sequenase TM version 2.0 DNA Sequencing Kit (Amersham Biosciences). Expression and purification of rLy-TK1 recombinant enzymes and rLy-TK1 66)136 proteins Expression and purification of the GST-TK1 fusion proteins have been described in detail previously [18]. Briefly, induction was performed at 25 °C by the addition of 0.1 m M isopropyl thio-b- D -galactopyranoside, the bacterial lysate was filtered and applied to a glutathione–Sepharose 4B column (Amersham Biosciences), and TK1 was cleaved from the GST part with thrombin (Amersham Biosciences). After addition of glycerol, dithiothreitol, MgCl 2 and Triton X-100 to final concentrations of 10%, 5 m M ,5 m M and 1%, respectively, the thrombin cleavage fractions were stored at )80 °C. The yield of enzyme protein from 300 mL bacterial culture was 1–3 mg in the thrombin cleavage fractions, and the purification fold, calculated as the ratio between the specific activity in the pooled cleavage fractions and in the crude bacterial extract, was % 20. The yield of rLy- TK1 66)136 proteins in the cleavage fractions was 3–6 mg per litre bacterial culture. The purity of the preparations was estimated to be over 90% by SDS/PAGE (not shown). ATP incubation and storage of the rLy-TK1 recombinant enzymes for kinetic experiments and gel filtration The thrombin cleavage fractions were diluted to 5 lgÆmL )1 in dilution buffer A (50 m M Tris/HCl, pH 7.5, 5 m M MgCl 2 ,0.1 M KCl, 2 m M Chaps, 10% glycerol and 5 m M dithiothreitol) with and without 2.5 m M ATP, and incuba- tedonicefor2hbeforestorageat)80 °C. The enzymes incubated and stored with and without ATP are referred to as the +ATP and –ATP forms, respectively. Estimation of subunit molecular size of rLy-TK1 66)136 by tricine/ethylene glycol/SDS/PAGE Because the standard SDS/PAGE methods resulted in diffuse protein bands and insufficient resolution of the relatively small rLy-TK1 66)136 peptide (< 8 kDa), a method of Scha ¨ gger & von Jagow [35] modified according to Separation Technique File no. 112 from Pharmacia (now Amersham Biosciences) was used. The upper and lower gel was made 4.5% and 13% with polyacrylamide, respectively, and the gel buffer was 30% ethylene glycol/0.112 M acetate/ 0.112 M Tris/HCl, pH 6.5. The electrode buffer consisted of 0.2 M Tris, 0.2 M tricine (instead of glycine) and 0.55% SDS, pH 8.1. The Peptide marker kit, molecular mass 2512–16 949 Da, from Amersham Biosciences was used as the molecular mass standard. Native molecular size The apparent molecular size of recombinant enzymes was determined by gel filtration on a Superdex 200 column (10 · 300 mm) connected to a Gradifrac automatic sampler (Amersham Biosciences) as described previously [14,18]. A 200-lL portion of thrombin cleavage fraction stored at )80 °C at a protein concentration of 5 lgÆmL )1 was mixed with 100 lL of the equilibration and elution buffer B (50 m M imidazole/HCl, pH 7.5, 5 m M MgCl 2 ,0.1 M KCl, 2m M Chaps and 5 m M dithiothreitol), containing 0.17 mgÆmL )1 Blue Dextran 2000 as internal marker for determination of column void volume. Then 200 lLofthis mixture (protein concentration 3.25 lgÆmL )1 ) was applied to the column. The +ATP enzyme samples contained 2.5 m M ATP,andwereelutedinbufferBwith2.5 m M ATP. Fractions of 200 lL were collected and mixed with 100 lL 2250 H. Frederiksen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 buffer B containing 30% glycerol and 2 m M ATP for enzyme stabilization, and assayed for thymidine kinase activity at standard assay conditions with 100 l M thymidine. The native molecular size of the rLy-TK1 66)136 proteins was estimated on a Superose 12 column (10 · 300 mm) connected to a Gradifrac automatic sampler (Amersham Biosciences) as described previously [14,18]. Protein from Fig. 2. Gel filtration of rLy-TK1(V106WT) and rLy-TK1(V106X) enzymes. Approxi- mately 0.65 lgproteinin200lLwasinjected into a Superdex 200 column. (A) Dimeric enzymes: V106WT, V106A, V106I, and V106T; (B) tetrameric enzymes: V106G, V106H, V106K, V106L, V106M, and V106Q. The molecular mass markers (|) are (from left to right): b-amylase (200 kDa), BSA (66 kDa), ovalbumin (45 kDa), carbonic anhydrase (29 kDa) and cytochrome c (12.4 kDa). V e is the elution volume, and V 0 is the void volume estimated with blue dextran 2000. The horizontal bars indicate the range of duplicate determinations. Ó FEBS 2004 Effect of amino acid 106 on human TK1 (Eur. J. Biochem. 271) 2251 the thrombin cleavage fraction (200 lL; protein concentra- tion 0.5 mgÆmL )1 ), containing 0.17 mgÆmL )1 Blue Dextran 2000 as internal marker for determination of the column void volume, was applied. The column was equilibrated and eluted with buffer B without Chaps. Five hundred microliter fractions were collected for estimation of protein concen- tration by the method of Bradford [36]. Thymidine kinase assay TK1 activity was assayed by measuring the initial velocities using the DE-81 filter paper method as described previously [14,18]. Standard assay conditions were 5 ngÆmL )1 enzyme, 50 m M Tris/HCl, pH 7.5, 2.5 m M MgCl 2 ,10m M dithio- threitol, 2.5 m M ATP, 0.5 m M Chaps, 3 mgÆmL )1 BSA, 3m M NaF and the indicated concentrations of [methyl- 3 H]thymidine (Amersham Biosciences) in a final volume of 50 lL. For each velocity, four time samples were taken. The enzymes, stored without ATP at 5 lgÆmL )1 , were diluted immediately before the start of the reaction with ice-cold enzyme dilution buffer C (50 m M Tris/HCl, pH 7.5, 1 m M Chaps and 3 mgÆmL )1 BSA). For dilution of the TK1 +ATP form, 2.5 m M ATP and 2.5 m M MgCl 2 were included in the dilution buffer. Enzyme kinetics The kinetic parameters and the degree of co-operativity were determined as previously described [18]. The experi- mental data were fitted to the Hill equation v ¼ VÁs n K n 0:5 þ s n and the kinetic parameters determined with the nonlinear regression software from Graphpad PrismÒ. V is the maximal velocity, n is the Hill constant, and K 0.5 , like K m in the Michaelis–Menten equation, defines the substrate concentration S where v ¼ 0.5 V max [37]. Results Subunit and native molecular size In a previous study we have shown that replacement of Val106 with methionine affected the dimer–tetramer ratio and kinetic properties of recombinant TK1 from human lymphocytes [18]. To identify the functional group of amino acid 106 responsible for this dimer–tetramer transition and change in thymidine K 0.5 , we introduced the following nine mutations: V106A, V106G, V106H, V106I, V106K, V106L, V106M, V106Q and V106T. We then characterized the enzymatic properties of the mutant enzymes. Theapparentnativesizesofthe–ATPformsofV106WT and of the mutant recombinant enzymes (subunit size 24 kDa, in agreement with previous results [13,14]) were determined by gel filtration, and the profiles are shown in Fig. 2. The applied volume was 200 lLwithanenzyme concentration of 3.25 lgÆmL )1 ,becausewewishedto operate at the supposed physiological concentration of TK1 protein calculated to be % 4 lgÆmL )1 in S-phase cells [21]. V106A, V106I and V106T were eluted essentially as V106WT: a substantial part of each of these enzymes was eluted from the Superdex column with an approximate size of 50 kDa, i.e. as dimers (Fig. 2A). In contrast, when the same concentrations of V106G, V106H, V106K, V106L and V106Q were applied, they were eluted similarly to V106M with an approximate size of 100 kDa, i.e. as tetramers (Fig. 2B). Accordingly, we named the first group of enzymes Ôthe dimeric enzymesÕ, and the second group Ôthe tetrameric enzymesÕ. The oligomerization pattern of the peptides rLy- TK1 66)136(Val106) and rLy-TK1 66)136(Met106) is shown in Fig. 3. rLy-TK1 66)136(Val106) was eluted as two separate peaks with molecular sizes of % 29 kDa and 12 kDa, whereas rLy-TK1 66)136(Met106) was eluted as a single sharp peak of % 29 kDa. According to the calculated (and verified by SDS/PAGE) subunit size of 7.7 kDa, rLy- TK1 66)136(Val106) appeared to be eluted as a mixture of a tetramer and a dimer, whereas rLy-TK1 66)136(Met106) was eluted as a tetramer only. This oligomerization pattern strongly supports our assumption that the peptide rLy- TK1 66)136 is an integral part of the TK1 oligomerization interface, and that amino acid 106 is indeed of significance for the subunit arrangement of the enzyme molecule. Kinetic properties We have previously shown that replacement of Val106 with methionine results in a rLy-TK1 with high catalytic activity (K 0.5 ¼ 0.5 l M ), irrespective of pre-assay exposure to ATP, and associated with the tetrameric state of the enzyme [18,21]. Figure 4 shows the relation between the initial velocity and the thymidine concentration for both the –ATP and +ATP form of the V106 mutant enzymes at saturating concentration of ATP. The calculated kinetic parameters are given in Table 1. The substrate kinetics of the dimeric enzymes V106A, V106I and V106T (Fig. 4A) is essentially the same as that previously observed for V106WT [18] and for the endogenous TK1 purified from human lymphocytes Fig. 3. Gel filtration of rLy-TK1 66)136 proteins. About 100 lgofrLy- TK1 66)136(Val106) (d)andrLy-TK1 66)136(Met106) (r) were injected into a Superose 12 column. The molecular mass markers (|) are (from left to right): b-amylase (200 kDa), BSA (66 kDa), ovalbumin (45 kDa), carbonic anhydrase (29 kDa) and cytochrome c (12.4 kDa). V e is the elution volume, and V 0 is the void volume estimated with blue dextran 2000. 2252 H. Frederiksen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 [14]: The –ATP form of these enzymes displays nonhyper- bolic, ÔcreepingÕ binding curves, with high K 0.5 values and n (Hill coefficient) values < 1 (Table 1), whereas their corresponding +ATP forms have low K 0.5 and n values slightly above 1. The substrate kinetics of the permanently tetrameric mutants V106G, V106H, V106K, V106L and V106Q (Fig. 4B) is essentially the same as that previously described for V106M [18], as both the –ATP and +ATP form of these enzymes have low K 0.5 and n values above 1 (Table 1). Although the –ATP form of V106K does not gain the V max value of its +ATP form, both the +ATP and –ATP forms have low K 0.5 values of 0.7 and 1.2 l M and n values of 1.4, similar to the other enzymes in the tetrameric group. The ratios between the K 0.5 values for the –ATP and +ATP forms clearly justify the above proposed grouping as dimeric and tetrameric enzymes. The K 0.5 (–ATP) to K 0.5 (+ATP) ratios for V106WT, V106A, V106I, and V106T are 30 or higher (Table 1), in agreement with the previous observations for V106WT [14,18]. In contrast, both the –ATP form and +ATP form of the tetrameric enzymes, V106G, V106H, V106K, V106L, V106M, and V106Q have the same low K 0.5 values (0.3–1.2 l M ), and their K 0.5 (–ATP) to K 0.5 (+ATP) ratios are %1. Despite the low K 0.5 values, the phosphorylating capacity of the tetrameric enzymes seems to be compromised, as the V max values of both the +ATP and –ATP forms are 2–3-fold lower than those of the dimeric enzymes (Table 1). Discussion There is no known 3D structure for the group of enzymes to which TK1 belongs. The only available 3D structures of the deoxynucleoside kinases are for the thymidine kinase from Herpes virus [22–26] and for the TK2-like enzymes, i.e. the multisubstrate deoxynucleoside kinase from Drosophila melanogaster, dNK, the human deoxyguanosine kinase, dGK [27], and the human deoxycytidine kinase, dCK [28]. Despite the very low overall amino-acid sequence homology (% 10%), the region of mammalian TK1 enzymes with amino acid 106 aligns with the dimerization region of HSV1-TK.Aminoacid106ispositionedinanareaofTK1 that is (a) highly conserved among vertebrates and viruses of the pox family and may be important for the regulation and substrate affinity of the enzyme and (b) predicted to form an amphipathic helix facilitating subunit interaction [8]. Con- sequently, we cloned, expressed and purified the putative interface domain of rLy-TK1, rLy-TK1 66)136 , and investi- gated the oligomerization properties of rLy-TK1 66)136 with valine or methionine as amino acid 106. Our results confirmed the importance of amino acid 106 for the subunit arrangement of the enzyme molecule, because in gel- filtration experiments, the Met106 rLy-TK1 interface frag- ment was eluted as a tetramer, whereas the Val106 rLy-TK1 fragment was eluted as a mixture of a dimer and a tetramer. For further investigation of the role of size, conformation and polarity of amino acid 106 for the function and structure of human TK1, we created nine mutant enzymes at amino acid site 106 by site-directed mutagenesis of the recombinant human lymphocyte TK1, rLy-TK1 Val106 (V106WT). After expression and purification, the effect of the mutated amino acids on the oligomerization pattern and kinetic properties was examined. Our results suggested that the recombinant enzymes could be divided into two groups. Group I, the dimeric enzymes, containing V106A, V106I and V106T, shared their oligomerization and kinetic properties with V106WT, i.e. their –ATP form had high K 0.5 values for thymidine, % 27– 43 l M for valine, isoleucine and threonine, and 13 l M for alanine. The thymidine substrate kinetic pattern was nonhyperbolic, with ÔcreepingÕ velocity vs. substrate curves, and the Hill coefficient was determined to be 0.8, indica- ting a negative co-operative reaction mechanism. At the Fig. 4. Relation between the initial velocity of dTMP formation and thymidine concentration. Open symbols, +ATP forms; closed sym- bols, –ATP forms. (A) Dimeric enzymes: V106WT, V106A, V106I and V106T; (B) tetrameric enzymes: V106G, V106H, V106K, V106L, V106M and V106Q. v is the initial velocity. Ó FEBS 2004 Effect of amino acid 106 on human TK1 (Eur. J. Biochem. 271) 2253 investigated concentrations in gel-filtration experiments, they appeared as both dimers and tetramers with native molecular sizes of about 50 and 100 kDa, respectively. The +ATP form of the group I enzymes had low K 0.5 values for thymidine, % 0.3–0.9 l M , and Hill coefficients slightly above 1. Except for alanine, the amino acids at site 106 in the dimeric group are of similar size and conforma- tion, but different polarities, and the hydroxyl moiety of threonine does not seem to cause any disturbances. Hence, the hydrophobicity of the residue at site 106 is not critical for the function and conformation of rLy-TK1. Group II, the tetrameric enzymes, containing V106G, V106H, V106K, V106L and V106Q, have properties similar to V106M, i.e. in both the absence and presence of ATP they have low K 0.5 values for thymidine, % 0.3–1.2 l M ,and Hill coefficients between 1.4 and 2, indicating positive co-operativity, and they are eluted as tetramers in gel- filtration experiments. Dilution experiments have shown that the stability of the –ATP form of group II enzymes is strikingly low compared with the group I enzymes. If diluted to 25 ngÆmL )1 and incubated at 4 °C for 30 min, the dimeric group I enzymes retained 80–100% of their enzymatic activity, while the tetrameric group II enzymes retained only 0.5–3%. As the amino acids at site 106 in group II differ in polarity as well as in size and conforma- tion, these properties do not appear to explain the decreased K 0.5 values, the decreased stability, or the conformational changes. Although it is clear that group I comprises enzymes that have a substantial portion eluted as a dimer and biphasic kinetics with a high K 0.5 value, the correlation between the K 0.5 value and the dimer–tetramer ratio is less clear. This may rely on the fact that the mutation not only interferes with the dimer–tetramer transition but also the interaction with the substrate. Because valine, isoleucine and leucine are nonpolar amino acids with similar hydrophobicity, size and side chain conformation, grouping of V106I with the V106WT in the dimeric group I was expected, but the absence of V106L was not. The presence of the polar V106T in group I was also unexpected. However, the side chains of valine, isoleucine and threonine have one property in common, i.e. branching at the b-carbon atom classically considered to destabilize a-helices because of steric clashes. The side chain of leucine differs from valine, isoleucine and threonine by having a branch at its c-carbon atom, and, although the a-helical propensity of leucine is nearly as high as that of alanine [38], the long hydrophobic side chain of leucine resembles the side chain of methionine in its length and the absence of the branched b-carbon. This may explain why the oligomerization and kinetic properties of V106L are the same as those of V106M [18] and other enzymes of the tetrameric group II (Table 1). Dimer–tetramer transition, which is dependent on enzyme concentration and pre-exposure to ATP [14,18,21], would require an enzyme with substantial conformational flexibility. b-Branching amino acids may have a regulatory role in such conformation-dependent transitions, as they are known to increase the strain within an a-helix, and so to destabilize helix–helix interaction [39–41]. In fact, Val106 in human TK1 is preceded by another b-branched residue, Ile105, and among the 71 conserved residues in the segment 66–136, there are six a-helix-breaking glycines, three prolines and 15 b-branching amino acids (Fig. 1). The activity of TK1 correlates with the DNA synthesis [11,12], and we have previously proposed a model in which fluctuation of TK1 activity during the cell cycle is due to a shift from a low activity dimer dominating at low TK1 concentrations in G1 to a high activity tetramer dominating in the S phase with high TK1 concentrations [21]. Perturbation in transition pattern from a low thymidine affinity dimer to a high thymidine affinity tetramer has recently been reported for a recombinant TK1 (V106WT) enzyme, in which Ser13 was substituted with aspartate [42]. The S13D substitution mimics phosphorylation of Ser13, shown to be the site of heavy mitotic phosphorylation in HeLa cells [43–45]. Thymidine kinetics and gel-filtration experiments show that the S13D mutation causes an equilibrium shift from a tetramer to a dimer paralleled by an % 10-fold increase in K m [42]. These results explain the previously observed downregulated activity of phosphoryl- ated TK1 at G2/M phases in proliferating cells [43–45]. Table 1. Kinetic parameters of rLy-TK1(V106WT) and the mutant enzymes. V max , K 0.5 and the Hill constant n were determined as described in Materials and Methods. The best fit ± SE to all data is given. Enzyme V max (lmolÆmin )1 Æmg )1 ) K 0.5 (l M ) n K À ATP 0:5 /K þ ATP 0:5 )ATP +ATP )ATP +ATP –ATP +ATP Group I – dimeric enzymes V106WT 11.0 ± 1.3 9.4 ± 0.2 27.1 ± 8.5 0.6 ± 0.05 0.8 ± 0.07 1.2 ± 0.05 45 V106A 8.0 ± 0.8 6.6 ± 0.2 12.7 ± 3.2 0.3 ± 0.03 0.8 ± 0.06 1.3 ± 0.1 42 V106I 10.3 ± 1.8 8.3 ± 0.3 29.4 ± 16.4 0.9 ± 0.1 0.8 ± 0.1 1.2 ± 0.1 33 V106T 8.4 ± 1.9 7.1 ± 0.2 43 ± 28 0.4 ± 0.04 0.8 ± 0.1 1.2 ± 0.1 108 Group II – tetrameric enzymes V106G 3.6 ± 0.2 3.0 ± 0.1 0.4 ± 0.1 0.3 ± 0.1 1.9 ± 0.4 1.7 ± 0.4 1.3 V106H 3.7 ± 0.2 3.1 ± 0.1 0.3 ± 0.2 0.4 ± 0.1 1.6 ± 0.7 2.0 ± 0.4 0.8 V106K 2.2 ± 0.06 4.1 ± 0.1 0.7 ± 0.1 1.2 ± 0.1 1.4 ± 0.1 1.4 ± 0.1 0.6 V106L 4.1 ± 0.1 3.9 ± 0.1 0.4 ± 0.1 0.7 ± 0.1 1.9 ± 0.2 1.8 ± 0.1 0.6 V106M 2.9 ± 0.06 2.9 ± 0.06 0.6 ± 0.08 0.6 ± 0.1 1.5 ± 0.1 1.5 ± 0.1 1.0 V106Q 4.9 ± 0.2 4.2 ± 0.1 0.8 ± 0.1 0.8 ± 0.1 1.5 ± 0.2 1.6 ± 0.1 1.0 2254 H. Frederiksen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 The observations described above [42] extend our model [21] by showing that phosphorylation of TK1 is involved in the dimer–tetramer transition, as well. Taken together, these observations imply that the shift in TK1 between a low activity dimer with apparently negative co-operativity and a high activity tetramer with apparently hyperbolic reaction mechanism plays a significant physiological role in the regulation of TK1 activity and hence the biosynthesis of dTTP. The vital importance of enzyme regulation by co- operative mechanisms has recently been underlined by the H121N mutant of the mitochondrial TK2, found in some patients with the mitochondrial DNA depletion syndrome, combined with severe myopathy and early death [46]. It is therefore of great importance to obtain as much informa- tion as possible about regulation and enzymatic properties of enzymes in DNA precursor metabolism. Acknowledgements We are indebted to Marianne Lauridsen for excellent technical assistance. This work was supported by the Danish Research Council and the NOVO research foundation. References 1. Nishino, I., Spinazzola, A. & Hirano, M. (1999) Thymidine phosphorylase gene mutations in MNGIE, a human mitochon- drial disorder. Science 283, 689–692. 2. Spinazzola, A., Marti, R., Nishino, I., Andreu, A.L., Naini, A., Tadesse, S., Pela, I., Zammarchi, E., Donati, M.A., Oliver, J.A. & Hirano, M. (2002) Altered thymidine metabolism due to defects of thymidine phosphorylase. J. Biol. Chem. 277, 4128–4133. 3. Mandel, H., Szargel, R., Labay, V., Elpeleg, O., Saada, A., Sha- lata,A.,Anbinder,Y.,Berkowitz,D.,Hartman,C.,Barak,M., Eriksson, S. & Cohen, N. (2001) The deoxyguanosine kinase gene is mutated in individuals with depleted hepatocerebral mito- chondrial DNA. Nat. Genet. 29, 337–341. 4. Saada, A., Shaag, A., Mandel, H., Nevo, Y., Eriksson, S. &Elpeleg, O. (2001) Mutant mitochondrial thymidine kinase in mitochon- drial DNA depletion myopathy. Nat. Genet. 29, 342–344. 5. Berk, A.J. & Clayton, D.A. (1973) A genetically distinct thymidine kinase in mammalian mitochondria. Exclusive labeling of mito- chondrial deoxyribonucleic acid. J. Biol. Chem. 248, 2722–2729. 6. Bogenhagen, D. & Clayton, D.A. (1976) Thymidylate nucleotide supply for mitochondrial DNA synthesis in mouse 1-cells. Effect of 5-fluorodeoxyuridine and methotrexate in thymidine kinase plus and thymidine kinase minus cells. J. Biol. Chem. 251, 2938– 2944. 7. Pontarin, G., Gallinaro, L., Ferraro, P., Reichard, P. & Bianchi, V. (2003) Origins of mitochondrial thymidine triphosphate: dynamic relations to cytosolic pools. Proc. Natl Acad. Sci. USA 100, 12159–12164. 8. Eriksson, S., Munch-Petersen, B., Johansson, K. & Eklund, H. (2002) Structure and function of cellular deoxyribonucleoside kinases. Cell. Mol. Life Sci. 59, 1327–1346. 9. Thelander, L. & Reichard, P. (1979) Reduction of ribonucleotides. Annu. Rev. Biochem. 48, 133–158. 10. Kunz, B.A., Kohalmi, S.E., Kunkel, T.A., Mathews, C.K., McIntosh, E.M. & Reidy, J.A. (1994) International Commission for Protection Against Environmental Mutagens and Carcinogens. Deoxyribonucleoside triphosphate levels: a critical factor in the maintenance of genetic stability. Mutat. Res. 318,1– 64. 11. Sherley, J.L. & Kelly, T.J. (1988) Regulation of human thymidine kinase during the cell cycle. J. Biol. Chem. 263, 8350–8358. 12. Kristensen, T., Jensen, H.K. & Munch-Petersen, B. (1994) Over- expression of human thymidine kinase mRNA without corre- sponding enzymatic activity in patients with chronic lymphatic leukemia. Leuk. Res. 18, 861–866. 13. Munch-Petersen, B., Cloos, L., Tyrsted, G. & Eriksson, S. (1991) Diverging substrate specificity of pure human thymidine kinases 1 and 2 against antiviral dideoxynucleosides. J. Biol. Chem. 266, 9032–9038. 14. Munch-Petersen, B., Tyrsted, G. & Cloos, L. (1993) Reversible ATP-dependent transition between two forms of human cytosolic thymidine kinase with different enzymatic properties. J. Biol. Chem. 268, 15621–15625. 15. Sherley, J.L. & Kelly, T.J. (1988) Human cytosolic thymidine kinase. Purification and physical characterization of the enzyme from HeLa cells. J. Biol. Chem. 263, 375–382. 16. Bradshaw, H.D. Jr & Deininger, P.L. (1984) Human thymidine kinase gene: molecular cloning and nucleotide sequence of a cDNA expressible in mammalian cells. Mol. Cell. Biol. 4, 2316– 2320. 17. Flemington, E., Bradshaw, H.D. Jr, Traina-Dorge, V., Slagel, V. & Deininger, P.L. (1987) Sequence, structure and promoter characterization of the human thymidine kinase gene. Gene 52, 267–277. 18. Berenstein, D., Christensen, J.F., Kristensen, T., Hofbauer, R. & Munch-Petersen, B. (2000) Valine, not methionine, is amino acid 106 in human cytosolic thymidine kinase (TK1). Impact on oligomerization, stability, and kinetic properties. J. Biol. Chem. 275, 32187–32192. 19. Black, M.E. & Hruby, D.E. (1992) Site-directed mutagenesis of a conserved domain in vaccinia virus thymidine kinase. Evidence for a potential role in magnesium binding. J. Biol. Chem. 267, 6801– 6806. 20. Folkers, G., Trumpp-Kallmeyer, S., Gutbrod, O., Krickl, S., Fetzer, J. & Keil, G.M. (1991) Computer-aided active-site-directed modeling of the herpes simplex virus 1 and human thymidine kinase. J. Comput. Aided Mol. Des. 5, 385–404. 21. Munch-Petersen, B., Cloos, L., Jensen, H.K. & Tyrsted, G. (1995) Human thymidine kinase 1. Regulation in normal and malignant cells. Adv. Enzyme Regul. 35, 69–89. 22.Wild,K.,Bohner,T.,Aubry,A.,Folkers,G.&Schulz,G.E. (1995) The three-dimensional structure of thymidine kinase from Herpes simplex virus type 1. FEBS Lett. 368, 289–292. 23. 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 type-I in complex with deoxythymidine and Ganciclovir. Nat. Struct. Biol. 2, 876–881. 24. Wild, K., Bohner, T., Folkers, G. & Schulz, G.E. (1997) The structures of thymidine kinase from herpes simplex virus type 1 in complex with substrates and a substrate analogue. Protein Sci. 6, 2097–2106. 25. Champness, J.N., Bennett, M.S., Wien, F., Visse, R., Summers, W.C., Herdewijn, P., Declercq, E., Ostrowski, T., Jarvest, R.L. & Sanderson, M.R. (1998) Exploring the active site of herpes simplex virus type-1 thymidine kinase by X-ray crystallography of com- plexes with acyclovir and other ligands. Protein Struct. Funct. Genet. 32, 350–361. 26. Bennett, M.S., Wien, F., Champness, J.N., Batuwangala, T., Rutherford, T., Summers, W.C., Sun, H., Wright, G. & Sander- son, M.R. (1999) Structure to 1.9 A ˚ resolution of a complex with herpes simplex virus type-1 thymidine kinase of a novel, non- substrate inhibitor: X-ray crystallographic comparison with binding of acyclovir. FEBS Lett. 443, 121–125. Ó FEBS 2004 Effect of amino acid 106 on human TK1 (Eur. J. Biochem. 271) 2255 27. 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 deoxyribonucleoside kinases. Nat. Struct. Biol. 8, 616–620. 28. Sabini, E., Ort, S., Monnerjahn, C., Konrad, M. & Lavie, A. (2003) Structure of human dCK suggests strategies to improve anticancer and antiviral therapy. Nat. Struct. Biol. 10, 513–519. 29. Munch-Petersen, B., Knecht, W., Lenz, C., Sondergaard, L. & Piskur, J. (2000) Functional expression of a multisubstrate deoxyribonucleoside kinase from Drosophila melanogaster and its C-terminal deletion mutants. J. Biol. Chem. 275, 6673–6679. 30. Cuff, J.A., Clamp, M.E., Siddiqui, A.S., Finlay, M. & Barton, G.J. (1998) JPred: a consensus secondary structure prediction server. Bioinformatics 14, 892–893. 31. Cuff, J.A. & Barton, G.J. (1999) Evaluation and improvement of multiple sequence methods for protein secondary structure pre- diction. Proteins 34, 508–519. 32. Cuff, J.A. & Barton, G.J. (2000) Application of multiple sequence alignment profiles to improve protein secondary structure pre- diction. Proteins 40, 502–511. 33. Higgins, D.G. & Thompson, J.D. (1996) Using CLUSTAL for multiple sequence alignments. Methods Enzymol. 266, 383–402. 34. Igarashi, K., Hiraga, S. & Yura, T. (1967) A deoxythymidine kinase deficient mutant of Escherichia coli. Mapping and trans- duction studies with phage phi-80. Genetics 57, 643–654. 35. Scha ¨ gger, H. & von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166,368– 379. 36. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. 37. Cornish-Bowden, A. (ed.) (1995) Control of enzyme activity. In Fundamentals of Enzyme Kinetics, pp. 203–237. Portland Press Ltd, London. 38. Creighton, T.E. (1993) Proteins. Structures and Molecular Prop- erties. 2nd edn. Freeman and Company, New York. 39. Dao-pin, S., Baase, W.A. & Matthews, B.W. (1990) A mutant T4 lysozyme (Val131–Ala) designed to increase thermostability by the reduction of strain within an alpha-helix. Proteins 7,198– 204. 40. Deber,C.M.,Li,Z.,Joensson,C.,Glibowicka,M.&Xu,G.Y. (1992) Transmembrane region of wild-type and mutant M13 coat proteins. Conformational role of beta-branched residues. J. Biol. Chem. 267, 5296–5300. 41. Deber, C.M., Khan, A.R., Li, Z., Joensson, C., Glibowicka, M. & Wang, J. (1993) Val fi Ala mutations selectively alter helix–helix packing in the transmembrane segment of phage M13 coat pro- tein. Proc. Natl Acad. Sci. USA 90, 11648–11652. 42. Li, C.L., Lu, C.Y., Ke, P.Y. & Chang, Z.F. (2004) Perturbation of ATP-induced tetramerization of human cytosolic thymidine kinase by substitution of serine-13 with aspartic acid at the mitotic phosphorylation site. Biochem. Biophys. Res. Commun. 313, 587– 593. 43. Chang, Z.F. & Huang, D.Y. (1993) The regulation of thymidine kinase in HL-60 human promyeloleukemia cells. J. Biol. Chem. 268, 1266–1271. 44. Chang, Z.F., Huang, D.Y. & Hsue, N.C. (1994) Differential phosphorylation of human thymidine kinase in proliferating and M phase-arrested human cells. J. Biol. Chem. 269, 21249– 21254. 45. Chang, Z.F., Huang, D.Y. & Chi, L.M. (1998) Serine 13 is the site of mitotic phosphorylation of human thymidine kinase. J. Biol. Chem. 273, 12095–12100. 46. Wang, L., Saada, A. & Eriksson, S. (2003) Kinetic properties of mutant human thymidine kinase 2 suggest a mechanism for mitochondrial DNA depletion myopathy. J. Biol. Chem. 278, 6963–6968. 2256 H. Frederiksen et al.(Eur. J. Biochem. 271) Ó FEBS 2004 . Effect of valine 106 on structure–function relation of cytosolic human thymidine kinase Kinetic properties and oligomerization pattern of nine substitution. following nine mutations: V106A, V106G, V106H, V106I, V106K, V106L, V106M, V106Q and V106T. We then characterized the enzymatic properties of the mutant enzymes. Theapparentnativesizesofthe–ATPformsofV106WT and