Health Articles

 

x-ray structure of tmp kinase from mycobacterium tuberculosis complexed with tmp at 1.95 Å resolution

doi:10.1006/jmbi.2001.4843 available online at on J. Mol. Biol. (2001) 311, 87±100 X-ray Structure of TMP Kinase from Mycobacterium tuberculosis Complexed with TMP at 1.95 AÊ ResolutionI. Li de la SierraH. , A. M. , O. Baà 1Unite de Biochimie Structurale The X-ray structure of Mycobacterium tuberculosis TMP kinase at 1.95 AÊ resolution is described as a binary complex with its natural substrate 2Laboratoire de Chimie TMP. Its main features involve: (i) a clear magnesium-binding site; (ii) an alpha-helical conformation for the so-called LID region; and (iii) a high MacromoleÂcules, URA 2185 du density of positive charges in the active site. There is a network of inter- C.N.R.S., Institut Pasteur, 28 actions involving highly conserved side-chains of the protein, the mag- rue du Dr. Roux, 75724 Paris nesium ion, a sulphate ion mimicking the b phosphate group of ATP and the TMP molecule itself. All these interactions conspire in stabilizing what appears to be the closed form of the enzyme. A complete multia- lignement of all (32) known sequences of TMP kinases is presented.
Subtle differences in the TMP binding site were noted, as compared to the Escherichia coli, yeast and human enzyme structures, which have been reported recently. These differences could be used to design speci®c inhibitors of this essential enzyme of nucleotide metabolism. Two cases of compensatory mutations were detected in the TMP binding site of eukarotic and prokaryotic enzymes. In addition, an intriguing high value of the electric ®eld is reported in the vicinity of the phosphate group of TMP and the putative binding site of the g phosphate group of ATP.
# 2001 Academic Press Keywords: crystal structure; rational drug design; AZTMP; thymidylate *Corresponding author kinase; Mycobacterium tuberculosis principal microbial agent involved for humans.
Tuberculosis is primarily transmitted via airborne The incidence of tuberculosis (TB) has been aerosoled secretions. A peculiar aspect of its patho- increasing during the last 20 years and it is now genicity comes from the fact that it can remain the ®rst cause of mortality among infectious dis- quiescent and become active decades later. One of eases in the world, killing more than two million the most signi®cant risk factor for developing people a Mycobacterium tuberculosis is the tuberculosis is human immunode®ciency virus (HIV) infection. The current treatment of active TB includes four drugs (isoniazid, rifampicin, pyrazi- Present address: I. Li de la Sierra, Laboratoire namide and ethambutol) for at least six months. A d'Enzymologie et de Biochimie Structurales, C.N.R.S.
signi®cant proportion of patients do not complete Bat. 34, Avenue de la Terrasse, 91198 Gif/Yvette Cedex, the therapy, especially in developing countries, and this has led to the appearance of resistant Abbreviations used: TMPK, thymidylate kinase; strains of M. tuberculosis. Therefore, there is cur- TMPKMtub, thymidylate kinase from Mycobacterium tuberculosis; TMPK rently a large effort to identify new potential tar- Ecoli, thymidylate kinase from Escherichia coli; TMPKYeast, thymidylate kinase from gets for inhibitors and to develop new antibiotics.
yeast; CMPK, cytidylate kinase; UMPK, uridylate In this work, one essential enzyme of nucleotide kinase; AMPK, adenylate kinase; NMPK, nucleoside metabolism, namely thymidine monophosphate monophosphate kinase; TMP, thymidine kinase (TMPK), is taken as a potential target for monophosphate; TDP, thymidine diphosphate; TTP, developing rationally designed inhibitors.
thymidine triphosphate; TB, tuberculosis; HIV, human TMPK (E.C.2.7.4.9, ATP:TMP phosphotransfer- immunode®ciency virus; MIR, multiple isomorphous replacement; HSV, herpes simplex virus.
ase) belongs to a large superfamily of nucleoside E-mail address of the corresponding author: monophosphate kinases (NMPK). It catalyses the phosphorylation of thymidine monophosphate # 2001 Academic Press X-ray Structure of M. tuberculosis Thymidylate Kinase (TMP) to thymidine diphosphate (TDP) utilizing crystal structure was determined by multiple iso- ATP as its preferred phosphoryl donorIt lies at morphous replacement (MIR) using ®ve heavy- the junction of the de novo and salvage pathways atom derivatives, with one of them giving reliable of thymidine triphosphate (TTP) metabolism and is phase information up to 2.7 AÊ resolution the last speci®c enzyme for its synthesis. These The initial 2.7 AÊ MIR map was improved by den- characteristics make the TMPK a good target for sity modi®cation techniques and allowed the con- the design of new antibiotics drugs.
struction of the model without ambiguity. The The high-resolution structure of TMPK should model was re®ned to a crystallographic R-factor of be a good starting point to devising novel inhibi- tory compounds using a structure-based drug- free 25 %The quality of the struc- ture was assessed using PROCHECd gives an design approach. It may be worth mentioning here overall G-factor of 0.24. TMPK that one of the most successful antiviral drugs Mtub has 214 residues and a molecular mass of 24 kDa. The re®ned against herpes simplex virus (aciclovir) is directed structure consists of 208 amino acid residues (the against thymidine kinase, which is responsible for six C-terminal residues were not observed in the the synthesis of both TMP and TDP in cells electron density map), one TMP molecule , one sulphate group and one magnesium ion bound to Here, we report the structure of the TMPK from the catalytic domain M. tuberculosis (TMPKMtub) bound to its natural substrate, TMP, at 1.95 AÊ resolution. It is the ®rst The strictly conserved residue Arg95 (located in NMPK reported structure from M. tuberculosis and the consensus sequence DR(Y/F/H), residues 94- from a pathogen in general. The structure, com- 96 in the M. tuberculosis amino acid sequence; bined with a careful analysis of the alignment of is the only non-glycine residue lying out- all known sequences annotated as TMPKs, allows side the allowed regions of the Ramanchandran us to identify the residues involved in the TMP- plot. This particular conformation results from its binding site and those probably important for cata- location in the catalytic site, Arg95 being in direct lysis. The structure of the TMPK contact with the phosphate moiety of the TMP Mtub catalytic site is quite different from other bacterial or eukaryotic molecule in our complex structure (see below). The enzymes: arginine residues from both LID and P- M. tuberculosis enzyme also has one cis-residue loop regions are probably implicated in the phos- conformation at the strictly conserved proline resi- phoryl transfer, one magnesium ion is readily vis- due of motif (F/E)P at position 37, which is also ible in the active site and the LID region is in a observed in the cis conformation in all TMPK struc- helical conformation characteristic of the closed tures reported so far form of the molecules of this family, even though The global folding of the M. tuberculosis enzyme the second substrate is not present in the crystal.
is similar to that of the other TMPKs, despite the low degree of similarity of their amino acid Results and Discussion sequences (26 %, 25 % and 22 % sequence identity over about 200 aligned residues with TMPKEcoli, Structure determination and overall description TMPKMtub has nine a-helices, surrounding a ®ve- The recombinant TMPKMtub in the presence of stranded b-sheet core conserved in all TMPKs.
TMP yielded crystals suitable for X-ray study Comparing the TMPKMtub-TMP structure with Table 1. Heavy-atom parameters, data-collection and phasing statistics Soaking time (hrs) ESRF, ID14-3 LURE, DW21b ESRF, ID14-3 (wavelength (AÊ)) Unit-cell parameters 40-1.95 (2.0-1.95) 12-3.0 (4.0-3.0) 40-2.4 (2.46-2.4) 12-3.0 (3.11-3) Total number of refl.
Total of unique refl.
Phasing power (Res. AÊ) a Rmerge ˆ hijIhi ÿ hIhij/hiIhi, where Ihi is the ith observation of the re¯ection h, while hIhi is the mean intensity of re¯ection h.
b Riso ˆ jFPH ÿ FPj/FP, where FPH and FP are the derivative and the native structure-factor amplitudes, respectively.

X-ray Structure of M. tuberculosis Thymidylate Kinase Table 2. Re®nement statisticsResolution range (AÊ) No. of reflections Used for refinement Rfree calculation No. of non-hydrogen atoms rmsd from ideality Bond lengths (AÊ) Bond angle distances (AÊ) Average temperature factors Ramanchandran plot Residues in most favoured regions (%) Residues in additional allowed regions (%) a Rfactor ˆ jjFoj ÿ jFcjj/jFoj.
b Rfree was calculated with a small fraction (5 %) of randomly c G-factor is the overall measure of structure quality from both the similar structure from yeast (1tmk) or the TMPKYeast-TP5A (3tmk) and TMPKEcoli-TP5A (4tmk) structures results in good overlap of the b-sheet core (rmsd for ®ve strands, 23 Ca atoms, 0.74 AÊ, 0.71 and 0.81 AÊ for 1tmk, 3tmk and 4tmk, respectively) but the surrounding helices and loops have moved signi®cantly relative to each other (for 163 Ca superposing atoms, the rmsd is 10.89 AÊ, 10.93 AÊ and 8.44 AÊ for 1tmk, 3tmk and 4tmk, respectively; see This, together with the low level of sequence identity conservation, could explain the failure of the molecular replace- ment attempts to solve TMPMtub structure using both yeast and Escherichia coli models and the need for an MIR structure determination.
The TMP kinase family All known TMPK sequences have been aligned Figure 1. TMPKMtub-TMP binary complex structure.
together in This includes 17 bacterial (a) Ribbon diagram of the protein complexed with TMP, enzymes, seven archaebacterial enzymes and eight the sulphate and the magnesium ions bound to the eukaryotic enzymes. The alignment has been active site, showing in red the LID and the P-loop adjusted manually, especially in the LID region, to regions. (b) Superposition of the Ca traces of TMPKMtub- take into account all the available structural infor- TMP (red), TMPKYeast-TMP (green) and TMPKEcoli-TP5A mation; it becomes less certain after residue 180 or (blue) structures. The LID region adopts a helical confor- mation in the TMPK so, rendering the identi®cation of any residue Mtub-TMP complex, in contrast to a disordered region in the TMPK essential for adenine recognition dif®cult. The Yeast-TMP complex.
multialignment provides enlightening functional information by looking at strictly conserved resi- dues and at those that are strictly conserved inside subfamilies and systematically changed from one

X-ray Structure of M. tuberculosis Thymidylate Kinase Figure 2 (legend shown on page 92) subfamily to the other; they are coloured yellow in The three-dimensional structures of TMPK from It then becomes necessary to look at all yeastE. coland more recently humahave available three-dimensional structures to under- been solved and show a similar fold to other stand the origin of this phenomenon.
NMPKs; namely, a core of ®ve-stranded parallel

X-ray Structure of M. tuberculosis Thymidylate Kinase Figure 2 (legend shown on page 92) b-sheets surrounded by nine a-helices. Three of the phosphate donor; second, the loop contain- regions contain the essential residues for the func- ing the strictly conserved arginine residue that tion of this enzyme ®rst, the P-loop brings the donor and the acceptor nucleotides motif (consensus sequence GxxxxGKS/T)which together with consensus sequence DR(Y/H/F) controls the positioning of the phosphoryl groups and third the LID region, a ¯exible stretch that X-ray Structure of M. tuberculosis Thymidylate Kinase closes on the phosphoryl donor when it binds (see peculiar: it does not have the positive charge of the also Via et afor a recent review of these eukaryotes in position 10 (it has glycine instead, as sequence motifs).
in prokaryotes), but it has another extra arginine In addition to these functionally essential regions residue at position 14. Therefore, the M. tuberculosis just mentioned, the most striking features revealed enzyme appears, together with its closely related by the multialignment are the (F/E)P sequence cousin Mycobacterium leprae, as unique among all motif at the junction between strand b2 and helix known TMPK sequenceh a possible horizon- a2, the KPD motif just before the b4 strand and the tal gene transfer event with eukaryotes. Whatever strictly conserved serine residue at position 99 (see the reality of this gene transfer, the same event which, to our knowledge, has remained happened in M. leprae.
unnoticed so far.
On the basis of the location of the active-site arginine residues in either the P-loop or LID sequences, TMPKs were categorized into two type- Dimerization mode e I TMPKs (e.g. yeast and human) have, in addition to the invariant lysine (residue 13 in the Mtub, as well as TMPKYeast or TMPKEcoli, is a homodimer in solution. The TMPK M. tuberculosis amino acid sequence), an invariant plexed with TMP crystallizes with one molecule basic residue at position 10 in their P-loop per asymmetric unit; the functional dimer can be sequence that can interact with the g phosphate recovered by using the symmetry operator x ÿ y, group of ATP and lack such a positively charged ÿy, ÿz. The interface of the dimer consists of three residue in the LID region. In contrast, type II pairs of helices (a2, a3 and a6) as observed in TMPKs (e.g. E. coli), have a glycine residue in the E. coli enzyme complexes, which also crystallizes P-loop at position 10 and one additional basic resi- with one molecule per asymmetric The due in the LID region that interacts with ATP same helices are observed in the interfacial zone of (Arg153); however, this last residue is not strictly TMPKYeast complexes, which crystallize either with conserved in prokaryotes or in archaebacteria one dimer four dimerr asymmetric unit; in fact, the E. coli and the yeast dimers are remark- It came as a surprise, however, that the multia- ably superimposable. However, having best super- imposed the ®rst monomers of all three enzymes, sequence is actually closer to the eukaryotic the second monomer of M. tuberculosis appears to enzymes (including the viral sequences) and poss- be oriented upside-down as compared to either its ibly the archeal enzymes than to the bacterial yeast or E. coli counterparts. In other words, enzymes. This is especially true in the region just whereas the 2-fold axis of TMPKEcoli or TMPKYeast upstream from motif KPD (a6-b4) and in the loop dimer runs orthogonal to helix a3, across the dimer between helix a2 and strand b2 (motif FP in eukar- interface, relating a2, a3 and a7 to their antiparallel yotes and EP in prokaryotes), as well as in the equivalents, the 2-fold axis of TMPKMtub runs par- region just downstream from the LID region (helix allel to a3, relating a2, a3 and a6 to a60, a30 and a8, especially residue Q172). However, the LID a20, respectively. The dimer interface contains a region itself clearly resembles more closely the bac- closely packed hydrophobic core but also an ion terial ones, albeit with a characteristic arginine and pair Glu50-Arg127, which could explain the origin aspartate-rich large insertion. The P-loop is quite of this different dimerization mode, since it is not Figure 2. Multialignment of all known TMPK sequences. The alignment was done using the program PILEUP of package GCG (version h opening and extension gap penalties of 6 and 2, respectively. The alignment was adjusted manually in the LID region and in the helix a8 region, taking into account the structural elements of all other known structures (E. coli, yeast, human TMPKs). The Figure was generated with the program ESPriptIn addition to the generic name kthy (which stands for thymidylate kinase), the following abbreviations have been used: myctu M. tuberculosis; mycle M. leprae; helpj, Helicobacter pylori J99; helpy, H. pylori; mycge, Mycoplama genitalis; mycpn, Mycoplasma pneumoniae; ecoli, E. coli; yerpe, Y. pestis; haein, H. in¯uenzae; bucai, Buchnera aphidicola; bacsu, Bacillus subtilis; ricpr, Rickettsia prowazekii; chlpn, Chlamydia pneumoniae; chltr, Chlamydia trachomatis; caucr, Caulobacter crescentus; aerpr, Aeropyrum pernix; syny3, Synechocystis sp.; aquae, Aquifex aeolicus; thema, Thermotoga maritama; arcfu, Archeoglobus fulgidus; metja, Methanococcus janaschii; pyrho, Pyrococcus horikoshii; sulso, Sulfolobus solfataricus; metth, Methanobacterium thermoautotrophicum; schpo, Schizosaccharomyces pombe; yeast, Saccharomyces cerevisiae; vaccv, Vaccinia virus; variv, Variola virus; human, Homo sapiens; mouse, Mus musculis; caee, Caernorhabditis elegans; asfb7, african swine fever virus. Residues are boxed if more than 50 % identity was reached at this position. Strictly conserved pos- itions are in bold red. Residues that are more than 80 % identical in each group de®ned below, but different from one group to the other are indicated in yellow. Four groups of sequences have been de®ned in the program two sequences of M. tuberculosis and M. leprae come ®rst, then the seven archaebacterial sequences, then a group of 15 bacterial sequences, then the last eukaryotic eight sequences as the fourth group. The numbering and the second- ary structures of the M. tuberculosis enzyme are indicated at the ®rst line. Special symbols were added in the last line: green upper triangles for residues involved in TMP binding, red ellipses for residues putatively involved in ATP binding and blue crosses for the ones involved in the Mg2‡ binding.
X-ray Structure of M. tuberculosis Thymidylate Kinase conserved in enzymes other than M. tuberculosis or Stabilization of the structure of the LID region One of the main characteristics of the TMPKMtub- TMP complex structure is that the LID region is observed in an a-helix conformation, even though the ATP binding site is unoccupied.
The LID segment is described in the other known NMPK structures as highly ¯exible changes when the ATP molecule is ®xed to the enzyme.Similarily, there is a transition between a coil to an helical conformation in the LID region of TMPKYeast when ATP is bound.In all known ternary complexes of TMPKEcoli, it is also structured as an a-helix because ATP ana- logues were present in the active site (no structure of the E. coli enzyme with an empty ATP-binding site has been reported so far).
To explain the stability of the LID region in our structure, which lacks the phosphate donor, the role of a sulphate ion in the active site should be emphasized. Indeed, this sulphate ion partly explains the structural ordering of the LID region, because it provides a direct link between Arg153 of the LID and Lys13 of the P-loop (see and 4(a)); in a second shell of interactions (5-6 AÊ), it is at the center of a galaxy of positive charges (Arg14, Arg95, Arg149 and Arg160). This sulphate ion is in the position expected to be occupied by the b phosphate group of ATP and this is a recur- rent situation that has been observed in many Figure 3. (a) The sulphate ion-binding site, showing NMPKs. In thymidine kinase, the presence of this the bridge between the LID region (residues 149 and sulphate ion has been proved to be linked directly 153) and the P-loop (residues 13 and 14). (b) The Mg2 ‡- to the a-helix structuring of the LID region through binding site, showing that the Mg2 ‡ ligands (residues molecular dynamics simulations (M. Orozco et al., 166 and 9, the 50 phosphate oxygen and water mol- personal communication). There is yet another link ecules) are arranged in an octahedral con®guration.
between the LID region and the P-loop through the magnesium ion, of which two of the ligands are Glu166 and Asp9 (from the LID region and the P-loop, respectively); this is described in more detail below and located further away along the ATP-binding site, between the b and g phosphate groups Three water molecules (W1009, W1018 and W1050) and an oxygen atom from the phosphoryl The second main structural characteristic of the group are the other ligands that coordinate the M. tuberculosis structure is that it contains a mag- Mg2 ‡ in an octahedral con®guration nesium site, even though no ADP or non-hydroly- and 4(b)), in addition to the carboxylate oxygen sable analogue of ATP is present in the crystal. The atoms of Asp9 and of Glu166 already mentioned.
chemical nature of this positive peak in the initial One of these water molecules is in direct contact electron density maps was inferred from its octa- with a carboxylate oxygen atom of Asp163.
hedral coordination (all the distances between the magnesium ion and its oxygen ligands are in the M. tuberculosis TMP kinase is in the fully range 2.2-2.25 AÊ) and the fact that it is the only closed conformation divalent cation present in millimolar amounts in the mother liquor.
NMPKs have been described in several states, The magnesium ion observed here is a unique especially the AMPKan open conformation is feature of the TMPKMtub structure. No magnesium observed without substrate, a partially closed con- atom was reported for the yeast or the E. coli formation with a single substrate and a fully closed enzFor the human enzyme, one mag- conformation in the presence of both substrates.
nesium binding site has been reported but it is Both open and partially closed conformations were X-ray Structure of M. tuberculosis Thymidylate Kinase away from the main-chain nitrogen atom of Asp9; Arg95 is itself maintained in place through the strictly conserved Ser99, thus completing this intri- cate network (see All these inter- actions conspire to make the TMPKMtub structure presented here as most closely related to the fully closed TMPK structure.
The TMP-binding site As mentioned earlier, TMP is essential for the crystallization of TMPKMtub; the experimental elec- tron density of the TMP molecule could be seen in the initial MIR-DM maps. shows the electron density in a 2Fobs ÿ Fcalc map using a model where the TMP molecule has been omitted.
There are three main interactions that character- ize the TMP binding and (i) a stack- ing interaction involving the pyrimidine ring and the Phe70 side-chain; (ii) the interaction with Tyr103, which helps select deoxy-ribonucleotides versus ribonucleotides; (iii) and the hydrogen bond between the O4 atom in the base moiety and the Figure 4. (a) A schematic drawing of the TMP-binding site in the Mg2‡ and sulphate ion region (CHEM- DRAW), displaying all the residues in direct contact with the substrate. (b) As in (a), with the base moiety of the TMP-binding site.
observed for the CMPK from E. collly closed conformations were observed for the UMPK- CMPK from yeast and Dictyostelium discoideu Finally, partially closed fully close formations were observed for the TMPK from E. coli, yeast and human, respectively.
The presence of the magnesium ion and of the sulphate ion already pointed to an explanation for the structuring and the closing of the LID region.
Here, we argue that the enzyme is in its fully closed conformation, even though the second sub- strate is not bound; indeed, the 30OH group of the sugar moiety of TMP is at the center of an exten- sive hydrogen bond network, which can be described as follows. As mentioned earlier, Asp9 is essential in holding this 30OH in place, but the Asp9 carboxylate group is in turn hydrogen bonded to Tyr103 and Glu172 which are all strictly conserved The hydroxyl Figure 5. (a) A drawing of the thymidine moiety- group of the Tyr103 side-chain is in turn hydrogen binding site in TMPK bonded to Arg95, as in the human enz Mtub. (b) Network of (hydrogen- bond) interactions involved in the stabilization of 30OH the Gln172 side-chain OE1 atom is located 3.1 AÊ group of the TMP molecule.
X-ray Structure of M. tuberculosis Thymidylate Kinase interaction of its hydrophobic part with Phe36, a residue strictly conserved in M. tuberculosis, M. leprae and eukaryotes and located just before the structurally important Pro37, which is strictly conserved in all TMPKs and whose cis confor- mation forms the very bottom of the cavity respon- sible for the thymidine binding It is rewarding to ®nd that this hydrophobic interaction is replaced in prokaryotes by a (compensating) direct hydrogen bond between the NE atom of Arg74 and the carboxylate group of the strictly conserved glutamate residue that replaces Phe36 in all prokaryotes, while the interaction with Glu124 is replaced by an interaction with Thr105, Asp102 and Tyr75 (E. coli numbering). Therefore, the situ- ations of Arg74 in prokarotes in eukaryotes are almost mirror images of one another, with TMPKMtub behaving as a eukaryotic TMPK.
All direct contacts between TMP and TMPKMtub are shown in There are 11 residues located less than 3.9 AÊ away from the TMP mol- ecule, and six of these make hydrogen bonds with the nucleotide: Asp9, Phe36, Tyr39, Arg74, Arg95 and Asn100. Also Ser99, a strictly conserved resi- due that had hitherto escaped notice, is crucial in the positioning of Arg95 Several subtle differences are observed when comparing with other TMPKs. This concerns: (i) the hydrogen bond between the N3 atom of the pyrimidine ring with the Asn100 side-chain (changed to a glycine residue in both yeast and human, and to a threo- nine residue in the E. coli enzyme); (ii) the Tyr39 residue makes two polar contacts with the TMP, one with a phosphate oxygen atom (3.1 AÊ) and the Figure 6. (a) Model of bound TMP in 2Fo ÿ Fc map; other with the oxygen atom at the 50 position the TMP molecule was omitted from the model used to (3.4 AÊ). Arginine and glycine residues replace this calculate the phases. Contours are drawn at the 2s level above the mean electron density. (b) Fourier-difference tyrosine in human and yeast enzymes respectively.
(iii) The 30-hydroxyl group of the ribose moiety 5I-dUMP ÿ Fc map. Fc denotes structure factors ampli- tudes of the TMPKMtub model without TMP. The peak makes three polar contacts one with in the electron density map corresponding to an iodine a water molecule (2.72 AÊ) involved in the Mg2‡ atom in position 5 of the pyrimidine ring is observed coordination (see above) and two others (2.72 AÊ with a cut-off of 5s above the mean electron density.
and 3.3 AÊ) with Asp9 of M. tuberculosis sequence.
The Figure was drawn with BOBSCRIPT The direct interaction of the P-loop with the sugar moiety of the monophosphate substrate is a unique feature of TMPKst is interesting to note that residues Asn100 and Tyr39 are precisely those Arg74 side-chain, which favours thymidine or ura- painted in yellow in the multialignement drawn cil over cytosine and These with EPScripusing the option of highlighting three positions are almost universally conserved in residues conserved in subfamilies but changed all known TMPK sequences (see The role from one subfamily to the other of the last interaction in the kinase activity was con®rmed recently in the case of the thymidine Interaction with other kinase from herpes simplex virus (HSV) type 1: steady-state kinetic studies showed that mutating Gln125 (Arg74 equivalent residue in the TMPK Mtub phosphorylates dUMP analogues was observed for other TMPKsion of sequence) into Glu, Asp or Asn in the thymidine the methyl group of position 5 of the pyrimidine kinase from HSV-type 1 has a devastating effect on ring by iodine (5I-dUMP) affects both the K the phosphorylation of TMP, the thymidylate kinase activity of all three mutants being decreased m parameters of the enzyme: the Km value is 3.5-fold higher than that of TMP and reaction rate by over 90 %.
is 70 % of that with TMP (see Moreover, Apart from forming an ion pair with Glu124, the both 5I-dUMP and TMP bind to the phosphate Arg74 side-chain is stabilized through stacking acceptor binding site in a very similar fashion, as X-ray Structure of M. tuberculosis Thymidylate Kinase Table 3. Steady-state kinetic parameters Km for 5I-dUMP (mM) Vm (ATP, 5I-dUMP) (mM/min mg) Vm (ATP, TMP) (mM/min mg) kcat (sÿ1) with TMP and ATP Km for AZTMP (mM) kcat for TMP/kcat for AZTMP Ratio kcat/Km for TMP ‡ AZTMP KI for AZTMP (mM) Results for E. coli, Y. pestis and Yeast enzymes were extracted from the following publications: b Chenal-Francisque et c Munier-Lehmann et a shown in the experimental X-ray structure of the E. coli enzymes (when the ATP is ®xed) are also observed in our structure. The side-chain of resi- obtained by soaking the original crystals in a due Arg149 in the LID helix is located as observed 2 mM solution of 5I-dUMP.
in the E. coli enzyme, namely opposite both the AZTMP was observed to be a competitive inhibi- loop between b5 and a9 (comprising Leu 193 in tor of TMPKMtub with a KI of 10 mM: the presence E. coli) and the Thr15 residue (a highly conserved of an azido group totally abolishes the catalysis residue in TMPKs), ready to interact by stacking without changing the af®nity. It is the ®rst interaction with the adenine ring of ATP reported TMPK that does not phosphorylate the These three regions together form the AZTMP molecule, in contrast to other TMPKs apparent binding site of ATP in the complex with from prokaryotes or eukaryotesAZTMP is a AZTP5A in E. coli Following both the substrate for the E. coli, Yersinia pestis and yeast structural work on yeast and E. coli TMPK as well enzymes (with a reduction of kcat of only 2.5-fold, as a recent article on sequence determinants map- 96-fold and 200-fold for the E. coli, Y. pestis and ping the binding sites of both substrates in yeast enzymes, respectively; see , and for NMPis possible to identify a semi-invariant Salmonella typhi and Haemophilus in¯uenzae small residue (alanine, serine, glycine, threonine) at S. typhi and E. coli TMPKs phosphorylate AZTMP position 196 in the loop between helix a9 and at comparable rates, whereas H. in¯uenzae enzyme strand b5, common to all TMPKs (see was more similar to the Y. pestis TMPK.
Except for the fact that the loop between b5 and a9 Even though we have not been able to exchange is much shorter in M. tuberculosis than in E. coli, the TMP for AZTMP in our original crystals in con- everything is in place to accommodate the adenine ditions similar to those leading to the exchange of moiety of ATP in our structure.
TMP with 5I-dUMP, nor to grow large enough co- crystals, we can still postulate a plausible expla- nation for the inhibitory effect of AZTMP. By look- Electrostatic potential in the active site ing at the structure of the active site, it is possible A number of positive or negative charges are to imagine a direct interaction between the mag- pointing towards the active site of TMPK nesium ion and a modelled azido group of helical conformation of the LID region allows side- AZTMP, as inspired by the superimposed structure chains of Arg153, Arg156 and Arg160 to be located of the AZTP5A-TMPKEcoli complex, because the around the P-loop segment, which itself places two distance between the last nitrogen atom of the positive charges in this region, thereby contribut- azido group and the cation can be reduced to only ing to create a highly positive electrostatic potential 2.0 AÊ, after rotation around the C30-N30 bond. This and mapping directly the binding of the tripho- would displace one of the ligands of the mag- sphate moiety of the ATP molecule.
nesium ion and deeply perturb the geometry of the In addition, the Lys13 and Asp163 residues active site, since this magnesium ion is involved in interact with the phosphoryl group of TMP via the positioning of several key chemical groups for water molecules W1014 and W1009, respectively.
the reaction, namely one of the phosphate oxygen On the other hand, two acidic residues, Asp9 atoms and the essential Asp9.
and Glu166, create a negative electrostatic potential near the TMP phosphate binding site, which con- Inferring the ATP binding site tributes to the binding of the magnesium ion. This creates a high gradient of the electric potential (i.e.
Despite the fact that TMPKMtub was co-crystal- electric ®eld) in the vicinity of the a phosphate lized as a binary complex with TMP, but in the group of TMP and the g phosphate group of ATP.
absence of ATP, two structural characteristics In fact, it could be calculated that this gradient is observed in the ternary complexes from yeast and such that one goes from the ‡30 kT/e to the ÿ30

X-ray Structure of M. tuberculosis Thymidylate Kinase TMPKMtub and (b) the observed AZTP5A-binding site in TMPKEcoliOnly the adenine moiety of ATP is rep- resented. The side-chains of the a-helical LID region Figure 8. (a) Electrostatic potential surfaces in the have been omitted for the sake of clarity, except for M. tuberculosis enzyme in the vicinity of the TMP-ATP- R149. A small adjustment of the most extreme part of binding sites, calculated with program this arginine side-chain is necessary in the M. tuberculo- ing the magnesium and sulphate ions, at zero ionic sis enzyme to accommodate (and stack under) the ade- force. The blue surface represents the ‡30 kT/e potential nine ring. The Figures were drawn with surface and the red one the ÿ30 kT/e surface. The TMP molecule and the magnesium ion are in ball and stick representation (b) As in (a) for the E. coli enzyme, except that the isopotential surfaces are contoured at ‡15 kT/e and ÿ20 kT/e, respectively. The AZTP5A molecule is also included.
kT/e potential surface in less than 8 AÊ in TMPKMtub, creating an electric ®eld as high as 10‡7 V cmÿ1 This is calculated in the absence of the magnesium and sulphate ion, at suf®ciently close to the phosphate group of TMP to zero ionic strength. Interestingly, this effect is make a new covalent bond.
qualitatively maintained in the E. coli enzyme, albeit with an amplitude divided by a factor of 2 Possible catalytic residues This may be the reason why a mag- nesium ion is observed in the M. tuberculosis In AMPK and UMPK enzymes, the arginine resi- enzyme and not the E. coli enzyme, because bind- dues located in the LID region have been shown to ing of this cation by and large supresses this very play a role in catalyzing phosphoryl transf high electric ®eld. Nevertheless, should the mag- TMPKEcoli, the arginine residues located in the LID nesium ion move during catalysis, the electric ®eld region could also play a similar role, i.e. stabilize could develop again and be used to break down the transition but in the eukaryotic the covalent bond between the b and g phosphate enzymes one of the key basic residues is located in groups of ATP, and attract the displaced electrons the P-loop and not in the LID region. In the human X-ray Structure of M. tuberculosis Thymidylate Kinase enzyme, there seems to be a different mechanism described by the electrostatic potential (and its associated electric ®eld) will require more detailed Apart from the side-chains of Arg153, Arg156 theoretical studies. The role of the magnesium ion and Arg160 from the LID region, the active site of deserves special care, because any movement of TMPKMtub also contains the Arg14 residue (in the this cation during catalysis would develop an enor- P-loop segment). This last residue is not conserved mous electric ®eld in the vicinity of the chemical and is replaced by a threonine residue in E. coli bond to be broken.
and yeast or a serine residue in human TMPK; it is located at the N terminus of the a-helix 1. Its side- chain is located in such a position that it could be Materials and Methods engaged in the binding of the g phosphoryl group of the ATP molecule, as inferred from the Crystallization and data collection complexes of E. coli and yeast enzymes with the Crystals of M. tuberculosis TMPK in complex with bisubstrate inhibitor TP5A.
TMP were obtained as , a 6 ml drop of In addition to these residues, the TMP phosphor- a 1:1 mixture of the protein solution at 5-8 mg mlÿ1 and yl group of the TMPK the reservoir solution was equilibrated with a 35 % Mtub structure makes direct polar contacts with three residues (Asp9, Tyr39 (w/v) ammonium sulphate solution, 0.1 Mes (pH 6.0), and Arg95) and solvent-mediated contacts with six containing 2 % (w/v) PEG 600. Crystals appear as hexa- residues (Gly12, Lys13, Phe36, Arg153, Asp163 and gonal bipyramids within three weeks. The space group Glu166) of the enzyme molecule (see is P6522 with cell dimensions 76.6 AÊ, 76.6 AÊ and 134.4 AÊ.
Numerous heavy-atom compounds were screened at These tyrosine and arginine residues are good can- different concentrations and varying incubation times.
didates in assisting the transfer of a phosphoryl The heavy-atom derivatives were prepared by transfer- group to TMP: indeed, an arginine residue in the ring the TMPKMtub-TMP crystals to a stabilisation sol- position of Arg95 role in the yeast, E. coli ution of 60 % (w/v) ammonium sulphate, 10 % (w/v) and human It follows that Tyr39, PEG 2000 and 100 mM Mes (pH 6.0) containing the which is unique to TMPK heavy-atom reagents. The concentration and soaking sequence, stands out as the second possible target time for each of the isomorphous derivatives obtained for the design of inhibitors speci®c to TMPK with mercury, samarium, platinum, and uranium salts, addition to Asn100 as already mentioned. It is and the TMPKMtub binary complex with the nucleotide replaced by an arginine residue at the same analogue 5I-dUMP are indicated i sequence position (39) in eukaryotes or at position X-ray data were collected from cryo-cooled crystals using 25 % (w/v) glycerol as cryoprotectant. The X-ray 47 (M. tuberculosis numbering) in prokaryotes, sources and detectors used for data collection are listed where it is strictly conserved among all but one in Table 1. Diffraction data were processed using MARX- Dor DENZO/SCALEPACpackages. The CCP4 packags used to calculate structure factors from the observed intensities (TRUNCATE) and scale native to derivative data (FHSCAL).
This study of TMPKMtub complexed to TMP pro- Structure determination vides the ®rst structure of a pathogen TMPK and the ®rst example of a TMPK with an LID domain Resolution of the crystal structure of the TMPMtub- structured in a-helix in the absence of a bound TMP complex was performed by the multiple isomor- ATP molecule. The spatial con®guration described phous replacement (MIR) method. The Patterson maps as selective for the adenine ring is also in place in were interpreted with the automated procedure devel- our X-ray structure. The interaction of the oped in the program HEAVd checked by cross- Fourier differences. The heavy-atom positions were then TMPKMtub with the TMP shows three differences re®ned with MLPHARE (CCP4, 1994) and in Solo- in the contacts when compared to the yeast, mon s used for solvent ¯attening calculations.
human or E. coli enzymes: Arg14 and Tyr39, which Details of the crystallographic data sets used for struc- interact with the phosphate moiety and Asn100, ture solution and re®nement are given in with the base moiety. The side-chains from four overall ®gure-of-merit of MLPHARE phases increased arginine residues, 14, 153, 156, 160, are observed from 0.43 to 0.75 after solvent ¯attening at 2.7 AÊ resol- around the phosphate-binding site and could be ution, assuming 45 % solvent. The resulting map was of excellent quality.
Both the location along the sequence (P-loop and LID region) of those residues supposed to play a Model building and refinement catalytic role and the kinetic results suggest that Model building and adjustments were done with the Mtub is not similar to the other TMPKs reported until now and that its ®ne structure could program t into the solvent-¯attened MIR map, not have been predicted accurately using state-of- then into the SIGMAA-weighted mapsRe®nement was performed up to 1.95 AÊ resolution with REFMAC the-art homology modelling methods.
(CCP4, Standards protocols, including maximum The question of ascertaining whether one of likelihood target, bulk solvent correction and isotropic B- these residues plays a role more important than factors were used. The model was inspected manually the others or whether it is a collective effect best with SIGMAA-weighted 2Fo ÿ Fc and Fo ÿ Fc maps, and X-ray Structure of M. tuberculosis Thymidylate Kinase progress in the model re®nement was evaluated by the type-1 thymidine kinase by X-ray crystallography of decrease in the free R-factor. Re®nement statistics can be complexes with aciclovir and other ligands. Proteins: Struct. Funct. Genet. 32, 350-361.
The current model includes 208 residues (1-208), one 5. Li de la Sierra, I., Munier-Lehmann, H., Gilles, A. M., molecule of TMP, two sulphate ions, one metal ion and Barzu, O. & Delarue, M. (2000). Crystallization and 150 water molecules. One of the sulphate ions is located preliminary X-ray analysis of the thymidylate kinase at the interface between two symmetry-related mol- from M. tuberculosis. Acta Crystallog. sect. D, 56, 226- ecules, while the other one is in the active site. There is one cis-proline (Pro37). The following residues were 6. Laskowski, R. A., McArthur, M. W., Moss, D. S. & modeled as alanine residues, because their side-chain Thornton, J. M. (1993). PROCHECK, a program to density was either poorly de®ned or non-existent: Arg86, assess the validity of crystallographic models. J. Appl.
Glu144, Leu145. The residues 209-214 of the C terminus Crystallog. 26, 283-291.
were not observed.
7. Ostermann, N., Schlichting, I., Brundiers, R., Konrad, M., Reinstein, J. & Veit, T. et al. (2000).
Insights into the phosphoryltransfer mechanism of human thymidylate kinase gained from crystal All calculations were using the program structures of enzyme complexes along the reaction Delphi run on an SGI machinePartial charges were coordinate. Struct. Fold. Des. 8, 629-642.
assigned according to the dictionary of AMBER. The 8. Lavie, A., Vetter, I. R., Konrad, M., Goody, R. S., electrostatic potential (in units of kT/e) was mapped Reinstein, J. & Schlichting, I. (1997). Structure of onto the molecular surface using the program Grasp thymidylate kinase reveals the cause behind the The ionic strength of the buffer was set to zero, with the limiting step in AZT activation. Nature Struct. Biol.
interior dielectric constant set to 2-4 and that of the sol- vent to 80. The TMP molecule and the magnesium and 9. Lavie, A., Konrad, M., Brundiers, R., Goody, R. S., sulphate ions were omitted from the calculation. To Schlichting, I. & Reinstein, J. (1998). Crystal structure achieve better accuracy, and especially to remove arte- of yeast thymidylate kinase complexed with the facts at the border of the grid, a focussing technique was bisubstrate inhibitor P1-(50-adenosyl) P5-(50-thymi- used and performed in three steps, with the molecule dyl) pentaphosphate (TP5A) at 2.0 AÊ resolution: occupying gradually more and more of the grid (25 %, implications for catalysis and AZT activation.
50 % and 75 %). The ®nal grid spacing was 1.5 grid unit/ Biochemistry, 37, 3677-3686.
AÊ. The ®nal map was interpolated into a 65  65  65 10. Lavie, A., Ostermann, N., Brundiers, R., Goody, R., grid to allow for visualization with Grasp, or, alterna- Reinstein, J., Konrad, M. & Schlichting, I. (1998).
tively, was converted into an O-style map using the Structural basis for ef®cient phosphorylation of 30- description of the map ®le format provided in the Delphi azidothymidine monophosphate by Escherichia coli thymidylate kinase. Proc. Natl Acad. Sci. USA, 95, Protein Data Bank accession numbers 11. Saraste, M., Sibbald, P. R. & Wittinghofer, A. (1990).
The P-loop-a common motif in ATP- and GTP-bind- Coordinates of the TMPKMtub binary complex have ing proteins. Trends Biochem. Sci. 15, 430-434.
been deposited in the RSCB Protein Databank with 12. Reynes, J. P., Tiraby, M., Baron, M., Drocourt, D. & accession number 1G3U.
Tiraby, G. (1996). Escherichia coli thymidylate kinase: molecular cloning, nucleotide sequence, and genetic organization of the corresponding tmk locus.
J. Bacteriol. 178, 2804-2812.
13. Via, A., FerreÁ, F., Branetti, B., Valencia, A. & Helmer-Citerich, M. (2000). Three-dimensional view We thank the staff of ID14 (ESRF, Grenoble) and of of the surface motif associated with the P-loop struc- LURE (Orsay) for excellent facility with X-ray data col- ture: cis and trans cases of convergent evolution.
lection. This work was supported by grants from the J. Mol. Biol. 303, 455-465.
EEC (BIO98 CT-0354), Institut Pasteur, INSERM and 14. Munier-Lehmann, H., Chaffotte, A., Pochet, S. & CNRS (URA 2185). We thank M. Orozco for communi- Labesse, G. (2001). Thymidylate kinase of M. tuber- cating results prior to publication.
culosis: a chimera sharing properties common to eukaryotic and bacterial enzymes. Protein Sci. 10, 15. Berry, M. B., Meador, B., Bilderback, T., Liang, P., 1. Stokstad, E. (2000). Drug-resistant TB on the rise.
Glaser, M. & Phillips, G. N., Jr (1994). The closed Science, 287, 2391.
conformation of a highly ¯exible protein: the struc- 2. Anderson, E. P. (1973). Nucleoside and nucleotide ture of E. coli adenylate kinase with bound AMP kinases. In The Enzymes (Boyer, P. D., ed.), 3rd edit., and AMPPNP. Proteins: Struct. Funct. Genet. 19, 183- vol. 8, pp. 49-96, Academic Press, New York.
3. Wild, K., Bohner, T., Folkers, G. & Schulz, G. E.
16. Vonrhein, C., Schlauderer, G. J. & Schulz, G. E.
(1997). The structures of thymidine kinase from (1995). Movie of the structural changes during a cat- Herpes simplex virus type I in complex with alytic cycle of nucleoside monophosphate kinases.
substrates and a substrate analogue. Protein Sci. 6, Structure, 3, 483-490.
17. Muller, C. W., Schlauderer, G. J., Reinstein, J. & 4. Champness, J. N., Bennett, M. S., Wien, F., Visse, R., Schulz, G. E. (1996). Adenylate kinase motions Summers, W. C. & Herdewijn, P. et al. (1998).
during catalysis: an energetic counterweight balan- Exploring the active site of herpes simplex virus cing substrate binding. Structure, 4, 147-156.
X-ray Structure of M. tuberculosis Thymidylate Kinase 18. Briozzo, P., Golinelli-Pimpaneau, B., Gilles, A. M., 28. Collaborative Computing Project No. 4 (1994). The Gaucher, J. F., Burlacu-Miron, S. & Sakamoto, H.
CCP4 suite: programs for protein crystallography.
et al. (1998). Structures of Escherichia coli CMP kinase Acta Crystallog. sect. D, 50, 760-763.
alone and in complex with CDP: a new fold of the 29. Terwilliger, T. C., Kim, S. H. & Eisenberg, D. (1987).
nucleoside monophosphate binding domain and Generalized method for determining heavy-atoms insights into cytosine nucleotide speci®city. Struc- parameters using the difference Patterson function.
ture, 6, 1517-1527.
Acta Crystallog. sect. A, 43, 1-5.
19. MuÈller-Dieckmann, H. J. & Schulz, G. E. (1994). The 30. Cowtan, K. & Main, P. (1998). Miscellaneous algor- structure of uridylate kinase with its substrates, ithms for density modi®cation. Acta Crystallog. sect.
showing the transition state geometry. J. Mol. Biol.
D, 54, 487-93.
236, 361-367.
31. Abrahams, J. P. & Leslie, A. G. W. (1996). Methods 20. Schlichting, I. & Reinstein, J. (1997). Structures of used in the structure determination of bovine active conformations of UMP kinase from Dictyoste- mitochondrial F1-ATPase. Acta Crystallog. sect. D, 52, lium discoideum suggest phosphoryl transfer is associative. Biochemistry, 36, 9290-9296.
32. Jones, T. A., Zou, J. Y., Cowan, S. W. & Kjeldgaard, 21. Hinds, T. A., Compadre, C., Hurlburt, B. K. & M. (1991). Improved methods for the building of Drake, R. R. (2000). Conservative mutations of Gln- protein models in electron density maps and the 125 in HSV-1 thymidine kinase result in a ganciclo- location of errors in these models. Acta Crystallog.
vir with minimal deoxypyrimidine kinase activities.
sect. A, 47, 110-119.
Biochemistry, 39, 4105-4111.
33. Read, R. J. (1986). Improved Fourier coef®cients for 22. Gouet, P., Courcelle, E., Stuart, D. & Metoz, F.
maps using phases from partial structures with (1999). ESPript: multiple sequence alignments in errors. Acta Crystallog. sect. A, 42, 140-149.
Postscript. BioInformatics, 15, 305-308.
34. Murshudov, G., Vagin, A. & Dodson, E. (1997).
23. Chenal-Francisque, V., Tourneux, L., Carniel, E., Re®nement of macromolecular structures by the Christova, P., Li de la Sierra, I., Barzu, O. & Gilles, maximum-likelihood method. Acta Crystallog. sect.
A. M. (1999). The highly similar TMP kinases of D, 53, 240-255.
Yersinia pestis and Escherichia coli differ markedly 35. Honig, B. & Nicholls, A. (1995). Classical electro- in their AZTMP phosphorylating activity. Eur. J. Bio- statics in biology and chemistry. Science, 268, 1144- chem. 265, 112-119.
24. Abele, U. & Schulz, G. E. (1995). High resolution 36. Nicholls, A., Sharp, K. & Honig, B. (1991). Protein structures of adenylate kinase from yeast ligated folding and association: insights from the interfacial with inhibitor AP5A showing the pathway of phos- and thermodynamics properties of hydrocarbons.
phoryl transfer. Protein Sci. 4, 1262-1271.
Proteins: Struct. Funct. Genet. 11, 281-291.
25. Kabsch, W. (1988). Automatic indexing of rotation 37. GCG (1997). Wisconsin Package Version 9.1, Gen- diffraction patterns. J. Appl. Crystallog. 21, 67-71.
etics Computer Group, Madison, USA, WI.
26. Kabsch, W. (1988). Evaluation of single-crystal X-ray 38. Esnouf, R. (1999). Extension of the program diffraction data from a position-sensitive detector.
Molscript to display electron density maps. Acta J. Appl. Crystallog. 21, 916-924.
Crystallog. sect. D, 55, 938-940.
27. Otwinowski, Z. & Minor, W. (1997). Processing of 39. Kraulis, P. J. (1991). MOLSCRIPT: a program to pro- X-ray diffraction data collected in the oscillation duce both detailed and schematic plots of protein mode. Methods Enzymol. 276, 307-326.
structures. J. Appl. Crystallog. 24, 946-950.
Edited by R. Huber (Received 7 February 2001; received in revised form 23 May 2001; accepted 25 May 2001)

Source: http://lorentz.dynstr.pasteur.fr/website/publi/jmb_tmpk.pdf