Proc. Natl. Acad. Sci. USA Vol. 95, pp. 14045–14050, November 1998 Biochemistry Structural basis for efficient phosphorylation of 3*-azidothymidine monophosphate by Escherichia coli thymidylate kinase ARNON LAVIE*, NILS OSTERMANN*†,RALF BRUNDIERS†‡,ROGER S. GOODY*, JOCHEN REINSTEIN*, MANFRED KONRAD‡, AND ILME SCHLICHTING*§ *Department of Physical Biochemistry, Max Planck Institute for Molecular Physiology, Rheinlanddamm 201, 44139 Dortmund, Germany; and ‡Department of Molecular Genetics, Max Planck Institute for Biophysical Chemistry, 37018 Go¨ttingen, Germany Edited by Perry A. Frey, University of Wisconsin, Madison, WI, and approved September 11, 1998 (received for review June 29, 1998) ABSTRACT The crystal structures of Escherichia coli that efficiently phosphorylate the antiviral drugs acyclovir or thymidylate kinase (TmpK) in complex with P1-(5*-adenosyl)- ganciclovir but not AZT or AZT-MP (6, 7). P5-(5*-thymidyl)pentaphosphate and P1-(5*-adenosyl)P5-[5*- Therefore, we set out to understand the structural basis for (3*-azido-3*-deoxythymidine)] pentaphosphate have been the slow activation of AZT-MP. The structures of the yeast solved to 2.0-Å and 2.2-Å resolution, respectively. The overall TmpK (TmpKyeast) complexed with either the physiological structure of the bacterial TmpK is very similar to that of yeast substrate dTMP or the partially activated prodrug AZT-MP TmpK. In contrast to the human and yeast TmpKs, which indicate that it is the interaction of the 39-deoxyribose sub- phosphorylate 3*-azido-3*-deoxythymidine 5*-monophosphate stituent with a carboxylic acid side chain located in the P loop (AZT-MP) at a 200-fold reduced turnover number (kcat)in sequence that is responsible for the reduced rate with comparison to the physiological substrate dTMP, reduction of AZT-MP (8, 9). This interaction causes a P loop movement to kcat is only 2-fold for the bacterial enzyme. The different accommodate the bulky and rigid azido group of AZT-MP. kinetic properties toward AZT-MP between the eukaryotic The residue after this carboxylic acid is Arg-15 and it was TmpKs and E. coli TmpK can be rationalized by the different shown to play a catalytic role (10). Thus, we postulated that it ways in which these enzymes stabilize the presumed transition is the resulting arginine displacement that is responsible for the state and the different manner in which a carboxylic acid side reduced phosphorylation rate. Our interest in TmpK from chain in the P loop interacts with the deoxyribose of the Escherichia coli (TmpKcoli) arose when we realized that it lacks monophosphate. Yeast TmpK interacts with the 3*-hydroxyl of this catalytic arginine in the P loop (having a glycine instead) dTMP through Asp-14 of the P loop in a bidentate manner: although the P loop carboxylic acid is retained, albeit as binding of AZT-MP results in a shift of the P loop to glutamic acid in contrast to the aspartic acid in the yeast accommodate the larger substituent. In E. coli TmpK, the enzyme. This led to the hypothesis that even if a similar P loop corresponding residue is Glu-12, and it interacts in a side-on shift were to occur upon AZT-MP binding to TmpKcoli,it fashion with the 3*-hydroxyl of dTMP. This different mode of should not have such a drastic effect on catalysis (10). Kinetic interaction between the P loop carboxylic acid with the 3* measurements support our hypothesis, showing that the rate of AZT-MP phosphorylation at saturating substrate concentra- substituent of the monophosphate deoxyribose allows the tions is reduced by only a factor of 2 in comparison to dTMP, accommodation of an azido group in the case of the E. coli whereas this factor is 200 for the TmpK (10). The crystal enzyme without significant P loop movement. In addition, yeast structures of TmpK complexed with P1-(59-adenosyl)-P5- although the yeast enzyme uses Arg-15 (a glycine in E. coli)to coli (59-thymidyl)pentaphosphate (TP A) or P1-(59-adenosyl)-P5- stabilize the transition state, E. coli seems to use Arg-153 from 5 [59-(39-azido-39-deoxythymidine)] pentaphosphate (AZT- a region termed Lid instead. Thus, the binding of AZT-MP to P5A) reported herein provide a structural explanation for our the yeast TmpK results in the shift of a catalytic residue, which kinetic observations and supply important insights as to how is not the case for the bacterial kinase. TmpKs achieve catalysis. In addition, this is a structure of what we term a type II TmpK, which differs from type I TmpKs (e.g., Phosphorylation of AZT-MP by thymidylate kinase (TmpK; yeast and human TmpKs) in their P loop and Lid sequences. EC 2.7.4.9, ATP:dTMP phosphotransferase) has been impli- cated as the rate-limiting step in the activation pathway of the anti-HIV prodrug 39-azido-39-deoxythymidine (AZT). AZT MATERIALS AND METHODS must be phosphorylated by cellular enzymes to azidothymidine Synthesis and Kd Determination of TP5A and AZT-P5A. triphosphate before it can exert its antiviral effect by inhibiting AZT-MP was synthesized by incubating AZT with two equiv- HIV reverse transcriptase and acting as a chain terminator of alents of phosphorus oxychloride under conditions of basic nascent DNA strands. The bottleneck of AZT activation lies catalysis (2,6-dimethylpyridine). The bisubstrate inhibitors in the second phosphorylation step, the step that adds a TP5A and AZT-P5A were prepared by condensation of aden- phosphate group to azidothymidine monophosphate (AZT- osine tetraphosphate and dTMP or AZT-MP (activated by MP) to yield azidothymidine diphosphate and is catalyzed by TmpK. Cells treated with AZT accumulate the toxic AZT-MP This paper was submitted directly (Track II) to the Proceedings office. 1 9 5 9 (1–3) in millimolar concentration (4, 5), and thus the active Abbreviations: TP5A, P -(5 -adenosyl)-P -(5 -thymidyl)pentaphos- 1 9 5 9 9 9 AZT triphosphorylated metabolite is at very low concentra- phate; AZT-P5A, P -(5 -adenosyl)-P -[5 -(3 -azido-3 -deoxythymi- dine)]pentaphosphate; AZT, 39-azido-39-deoxythymidine; AZT-MP, tion. This situation is only slightly improved upon coinfection AZT 59-monophosphate; TmpK, thymidylate kinase; GAP, GTPase- with herpes simplex virus I thymidine kinase-carrying vectors activating protein. Data deposition: The atomic coordinates have been deposited in the The publication costs of this article were defrayed in part by page charge Protein Data Bank, Biology Department, Brookhaven National Lab- oratory, Upton, NY 11973 (PDB ID codes 4TMK for TP5A1 and payment. This article must therefore be hereby marked ‘‘advertisement’’ in 5TMP for AZTP5A). accordance with 18 U.S.C. §1734 solely to indicate this fact. †N.O. and R.B. contributed equally to this work. © 1998 by The National Academy of Sciences 0027-8424y98y9514045-6$2.00y0 §To whom reprint requests should be addressed. e-mail: ilme. PNAS is available online at www.pnas.org. [email protected]. 14045 Downloaded by guest on October 1, 2021 14046 Biochemistry: Lavie et al. Proc. Natl. Acad. Sci. USA 95 (1998) diphenyl phosphorochloridate) as described (11) for the syn- that surround a five-stranded b-sheet core and is likewise a thesis of AP5A. homodimer. Overlaying the yeast TmpK model with the E. coli The fluorescent TP5A analog TP5A-MANT [the fluorescent TmpK model results in good overlap of the b-sheet core (rms N-methylanthraniloyl (MANT) group is on the ribose of the deviation for five strands, 27 Ca atoms, 0.36 Å) but the adenosine moiety] was prepared as described (12) for the surrounding helices have moved relative to each other (rms preparation of MANT-AP5A. Fluorescence measurements deviation for nine helices, 113 Ca atoms, 2.88 Å). This explains were performed with a SLM 8100 spectrofluorimeter as the failure of the molecular replacement attempt and the need described (12) with an excitation wavelength of 360 nm and an for the de novo structure determination. emission wavelength of 440 nm. The experiments were carried NMP kinases and many other nucleotide binding proteins z y out at 25°C in a buffer containing 100 mM Tris HCl (pH 7.5), contain a sequence motif (GX4GKS T), called the P loop, 5 mM MgCl2, 2 mM EDTA, and 100 mM KCl. whose function is to bind nucleoside di- or triphosphates (18). Protein Preparation and Crystallographic Analysis. E. coli NMP kinases bind an additional nucleotide substrate, a NMP, TmpK was cloned from genomic DNA into a pJC20 vector, and catalyze the reversible phosphoryl transfer from the expressed in E. coli BL21(DE3), and purified to homogeneity triphosphate [preferably ATP for TmpKs (19)] to the mono- as described (10). Crystals of the complex of E. coli TmpK with phosphate (dTMP). In all previously determined NMP kinase either TP5A or AZT-P5A were grown by the vapor diffusion structures, the monophosphate is bound in a deep cleft with method using the hanging drop geometry. An enzyme solution the nucleobase surrounded by protein residues. The triphos- y (15 mg ml) was premixed with the nucleotide solution to phate is bound (mainly by the P loop) with its phosphates y contain (final concentrations) enzyme (12 mg ml) and 2 mM facing that of the monophosphate and is observed to interact nucleotide. Typically, 4 ml of the enzymeynucleotide mixture m with a region called Lid. The closure of the Lid region upon was mixed with 2 l of a solution containing 1.25 M ammonium triphosphate binding acts in most NMP kinases to bring phosphate (pH 8.0) and left to equilibrate at 20°C against a catalytic arginines to the reaction center. TmpKyeast was the reservoir of 2.5 M ammonium phosphate. Crystals with typical 3 3 m first NMP kinase observed to deviate from this principle; dimensions of 500 400 100 m grew after a few days in although there is a Lid region near the triphosphate, it lacks space group R32.