The enantioselectivities of the active and allosteric sites of mammalian ribonucleotide reductase Jian He 1 ,Be ´ atrice Roy 2 , Christian Pe ´ rigaud 2 , Ossama B. Kashlan 3 and Barry S. Cooperman 1 1 Department of Chemistry, University of Pennsylvania, PA, USA 2 Laboratoire de Chimie Organique Biomole ´ culaire de Synthe ` se, Universite ´ Montpellier II, France 3 Department of Medicine, University of Pittsburgh, PA, USA Ribonucleotide reductases (RRs, EC 1.17.4.1) form a family of allosterically regulated enzymes that cata- lyze the conversion of ribonucleotides to 2¢-deoxy- ribonucleotides and are essential for de novo DNA biosynthesis and repair, regulating other enzymes in the DNA synthesis pathway via control of the nuc- leotide pool [1]. Of the four known classes of RR (Ia, Ib, II and III) class Ia, which requires two dif- ferent subunits R1 and R2 for activity and catalyzes the reduction of all four common NDPs, is the most widespread, comprising all eukaryotic RRs as well as some from eubacteria, bacteriophages and viruses. The R1 subunit contains the active site as well as allosteric sites. We have recently demonstrated that there are three such sites in murine R1 (mR1) (the specificity or s-site, the adenine or a-site, and the hexamerization or h-site) [2,3], leading to a complex pattern of regulation of enzymatic activity, the major features of which are summarized in Scheme 1, as follows: (a) ATP, dATP, dGTP, or dTTP binding to the s-site drives formation of R1 2 ; (b) ATP or dATP binding to the a-site drives formation of R1 4 , which exists in two conformations, R1 4a and R1 4b , with the latter predominating at equilibrium; (c) ATP binding to the rather low affinity (K d 1–4 mm) h-site, which occurs at physiologically significant concentrations, drives formation of R1 6 – dATP does not bind to this site at physiologically significant concentrations; (d) the R2 2 complexes of R1 2 ,R1 4a , and R1 6 are enzymatically active, whereas the R2 2 complex of mR1 4b has little, if any, activity; and (e) the sub- strate specificity of RR is determined by the ligand occupying the s-site: ATP and dATP stimulate the reduction of CDP and UDP, dTTP stimulates the Keywords allosteric sites; enantioselectivity; L-ADP; L-ATP; mammalian ribonucleotide reductase Correspondence B. S. Cooperman, Department of Chemistry, University of Pennsylvania, Philadelphia, PA 19104–6323, USA Fax: 215 8982037 Tel: 215 8986330 E-mail: cooprman@pobox.upenn.edu (Received 25 October 2004, revised 20 December 2004, accepted 7 January 2005) doi:10.1111/j.1742-4658.2005.04557.x Here we examine the enantioselectivity of the allosteric and substrate bind- ing sites of murine ribonucleotide reductase (mRR). l-ADP binds to the active site and l-ATP binds to both the s- and a-allosteric sites of mR1 with affinities that are only three- to 10-fold weaker than the values for the corresponding d-enantiomers. These results demonstrate the potential of l-nucleotides for interacting with and modulating the activity of mRR, a cancer chemotherapeutic and antiviral target. On the other hand, we detect no substrate activity for l-ADP and no inhibitory activity for N 3 -l-dUDP, demonstrating the greater stereochemical stringency at the active site with respect to catalytic activity. Abbreviations mRR, mammalian ribonucleotide reductase; mR1, large subunit of mammalian ribonucleotide reductase; mR2, small subunit of mammalian ribonucleotide reductase; N 3 -D-dUDP, 2¢-azido-2¢-deoxy-b-D-uridine 5¢-diphosphate; N 3 -L-dUDP, 2¢-azido-2¢-deoxy-b-L-uridine 5¢-diphosphate; RR, ribonucleotide reductase. 1236 FEBS Journal 272 (2005) 1236–1242 ª 2005 FEBS reduction of GDP and dGTP stimulates the reduc- tion of ADP. dATP is a universal inhibitor of RR activity due to its induction of R1 4b formation as a result of a-site binding, whereas ATP is a universal activator because it induces R1 6 formation as a result of h-site binding. RR is a well-recognized target for cancer chemo- therapeutic and antiviral agents [4–7], as illustrated by the anticancer drugs hydroxyurea [8] and gemcitabine [9,10]. In recent years, l-nucleoside analogues have been examined as novel therapeutic agents, which have been shown to sometimes have comparable or higher antiviral activity than their d-counterparts, as well as more favorable toxicological profiles and superior metabolic stability [11–15]. Studies at the individual enzyme level have shown that l-nucleosides and l-nucleotides can bind to human or other mammalian enzyme active sites (reviewed in [16,17]) and sometimes act as substrates, in particular as phosphate donors and acceptors, as is the case for l -ATP and deoxycyti- dine kinase [18,19], l-nucleosides and deoxycytidine kinase, thymidine kinase 2 and deoxyguanosine kinase [20], l-nucleoside-5¢-monophosphates and UMP-CMP kinase [21], and, most recently l-nucleoside-5¢-diphos- phates and phosphoglycerate kinase [22,23]. In contrast, almost no investigations of interactions of l-nucleosides or l-nucleotides with allosteric sites have been reported (although see [13]). In the present work, we examine the abilities of l-ATP and of l-ADP (Fig. 1) to interact with the allosteric sites and active site of mR1, respectively. We also demonstrate that mRR displays enantiospecificity with respect to the b-d-configuration of the sugar moiety of 2¢-azido-2¢- deoxyuridine-5¢-diphosphate, paralleling our recent results with Escherichia coli RR [24]. Results and Discussion L-ATP is an allosteric effector of CDP reductase CDP is the only one of the four ribonucleoside diphos- phate substrates that is reduced by mRR with a signi- ficant activity in the absence of allosteric effectors, albeit with a high K m and a low k cat . Addition of 3mmd-ATP both lowers K m and raises k cat [2]. In the absence of d-ATP, CDP reductase activity shows a biphasic response to l-ATP, first increasing and then decreasing (Fig. 2). Both the maximal observed activity and the concentration of l-ATP giving maximum activity vary with [CDP], with the first increasing and the second decreasing as [CDP] is increased. In the presence of d-ATP, however, CDP reductase shows only a monotonic decline as a function of [l-ATP], with the apparent IC 50 increasing only slightly as [d-ATP] is increased (Fig. 3). The effects of l -ATP on CDP reductase activity clo- sely mirror those obtained with dATP, albeit at much higher concentrations [2], and are well rationalized by Scheme 1. The results in Fig. 2 provide evidence that, like dATP, l-ATP binds first to the s-site, inducing active R1 2 formation, and then to the a-site, inducing inactive R1 4b formation, thus accounting for the biphasic curves observed. The changes in the curves in Fig. 2 as a function of CDP concentration are consis- tent with the known induction of R1 2 formation by high levels of CDP, with the result that achieving max- imal activity requires lower [l-ATP]. In contrast, we have previously shown that CDP reductase activity has a triphasic response to increasing d-ATP concentra- tion, first increasing, then decreasing, then increasing again as d-ATP binds to the s-, a- and h-sites in sequence [2]. High concentrations of l-ATP do not lead to a third phase activation of CDP reductase, leading to the conclusion that l-ATP, like dATP, L-ATP dATP ATP dGTP dTTP L-ATP dATP ATP Scheme 1. Allosteric ligand effects on enzyme aggregation state and activity [2,3]. R2 has been omitted for simplicity – activity assays were carried out in the presence of saturating [R2]. States shown with white text on dark background have high activity; those on white background have little or no activity. Inclusion of L-ATP is based on results reported herein. Fig. 1. The structures of D- and L-adenine nucleotides. J. He et al. Enantioselectivity of ribonucleotide reductase FEBS Journal 272 (2005) 1236–1242 ª 2005 FEBS 1237 does not bind to the h-site, at least at concentrations  10 mm. With CDP as substrate, the dissociation constants of l-ATP for the s- and a-sites, calculated as described [2], are approximately 200 lm and 1000 lm, respect- ively. These values are three- to 10-fold higher than the corresponding values for d-ATP (25 lm and 300 lm, respectively [2]), demonstrating the only modest enantioselectivity at these two sites. The three-dimensional structure of mR1 has not yet been determined. However, its s-site and a-site are likely to be very similar to those determined for the E. coli class Ia R1 (eR1 [25]), as almost all of the contact residues in each of these sites in eR1 are conserved in mR1 (the s-site interactions are also largely conserved in known class Ib and class II structures [26–28]). It thus seems likely that the only modest loss in l-enantiomer affinity at these two sites reflects conservation of the known triphosphate and base interactions, with decreased affinity mainly resulting from loss of the interactions with the ribose, e.g. in the s-site, between a specific Asp (D232 in eR1, D226 in mR1) and the ribose 3¢-OH. In the presence of 0.2 mmd-ATP (Fig. 3a), mR1 should largely partition between active mR1 2 , with d-ATP bound to the s-site only, and inactive mR1 4b , with d-ATP bound to the s- and a-sites [2]. Addition of l-ATP should lead to saturation of the a-site and complete conversion to mR1 4b , leading to the observed monophasic inhibition. By contrast, at higher d-ATP (3 mm, Fig. 3c), d-ATP should be bound to all three sites, s-, a-, and h-, and mR1 should be mostly in the form of active mR1 6 [2]. Earlier we showed that, under similar conditions, addition of dATP led to inhibition of activity, largely as a result of dATP displacement of d-ATP from the a-site, which weakens the binding of mR2 2 to mR1 6 [3]. It is likely that the l-ATP inhibition seen in Fig. 3c has a similar explanation. L-NDP interaction with the mR1 substrate site An HPLC assay was used to determine whether l-ADP could mimic d-ADP in being reduced by mRR in the presence of the allosteric effector dGTP bound to the s-site [2]. As seen in Fig. 4, approximately 11% of d-ADP is converted to d-dADP when d-ADP is incubated for 2 h in the presence of mRR and dGTP. Fig. 2. The effect of L-ATP on CDP reductase in the absence of D-ATP. Assays were carried out by preincubating 1.2 lM R1, 2.0 lM R2, and varying concentrations of L-ATP for 7 min prior to [8– 3 H]CDP addition. Fig. 3. The effect of L-ATP on CDP reductase in the presence of D-ATP. Activities were carried out as in Fig. 2, but with a fixed amount D-ATP in the preincubation buffer. Fig. 4. L-ADP is not a substrate for RR. Samples were incubated for 2 h at 25 °C in buffer A containing dGTP (300 l M), R1 (0.84 l M), R2 (4 lM) and either (a) D-ADP (400 lM) or (b) L-ADP (400 l M). Quenched reaction mixtures were analyzed by HPLC. Retention times: dGTP 25 ± 2 min ( L or D)-ADP 24 ± 1 min, D-dADP 30 ± 1 min. The peaks at 29 ± 1 min and 33 ± 1 min are from D-dAMP and oxidized dithiothreitol, respectively. Enantioselectivity of ribonucleotide reductase J. He et al. 1238 FEBS Journal 272 (2005) 1236–1242 ª 2005 FEBS By contrast, no conversion of l-ADP to l-dADP is detectable under comparable conditions, or even employing 14 times as much enzyme for periods up to 8 h (Fig. 4B). These results lead to the conclusion that the specific activity of mRR for reduction of l-ADP is < 1% of that for d-ADP. Similarly, under conditions in which preincubation with the known suicide inhib- itor N 3 -d-dUDP [29] at a concentration of 20 lm redu- ces CDP reductase activity by  80%, incubation with up to 250 lm N 3 -l-dUDP leads to no reduction in activity (Fig. 5). These results parallel those we obtained with E. coli RR [24]. Although the results in Figs 4 and 5 demonstrate the failure of l-NDPs to act as a substrate or as a mechan- ism-based inhibitor, l-ADP can clearly bind to the substrate site, as shown by its inhibition of d-ADP reduction (Fig. 6). Measured in the presence of 2 mm d-ATP, which is sufficient to saturate the a-site and most of the h-site, l-ADP is a classic competitive inhibitor of dGTP-dependent d-ADP reductase, with a K i of 1.0 mm, about 11-fold higher than the apparent K m for d-ADP of 91 lm (Fig. 6A,B). As above, this probably reflects retention of the base and especially b-phosphate interactions at the active site and loss of the specific ribose interactions that are crucial for sub- strate activation [25]. l-ADP also displays competitive inhibition in the absence of d-ATP (Fig. 6C), but the nonlinearity of the secondary plot (Fig. 6D) is an indica- tion of interaction at more than just the substrate site. One possibility is that, in the absence of d-ATP, l-ADP not only acts as a competitive inhibitor, but also binds to the otherwise empty a-site at high concentrations, permitting it to act as an allosteric inhibitor as well. Here we note that high concentrations of d-ADP induce mR1 4 formation [2], providing evidence for NDP bind- ing to the a-site. L-Nucleosides as potential therapeutic agents directed against ribonucleotide reductase We detect no substrate activity for l-ADP and no irre- versible inhibitory activity for N 3 -l-dUDP, demonstra- ting the considerable stringency at the active site of mR1 with respect to catalytic activity. On the other hand, our results show that l-ADP binds to the active site and l-ATP binds to both the s- and a- allosteric sites of mR1 with affinities that are only three- to 10-fold weaker than the values for the corresponding d-enantiomers, demonstrating the potential of l-nucleo- tides for interacting with and modulating the activity of RR. Particularly intriguing in this regard is the allosteric ligand dATP, which inhibits RR through a very high affinity interaction with the a-site (K d , 0.3– 1 lm [2]). Efforts to explore the interaction with RR of l-enantiomers of dATP and dATP analogues are currently underway. Experimental procedures Materials N 3 -d-dUDP and N 3 -l-dUDP were synthesized as described [24]. l-AMP was obtained in 74% yield by selective phos- phorylation of l-adenosine with POCl 3 in triethylphosphate [30]. After being converted to its tri-n-butylammonium salt, l-AMP was first reacted with 1,1¢-carbonyldiimidazole in 1,3-dimethyl-3,4,5,6-tetrahydropyrimidin-2(1H)-one and then with tri-n-butylammonium phosphate to give l-ADP in 15% yield [31]. l-ATP was obtained in 41% yield according to the Ludwig procedure [32]. l-ADP and l-ATP were purified to homogeneity, as judged by HPLC and HRMS, by DEAE-Sephadex A-25 and RP18 chromatogra- phy, and converted to the respective sodium salts by passage through a Dowex-AG 50WX2-400 column. The l-nucleotides were fully characterized by NMR ( 1 H, 13 C, 31 P), fast-atom-bombardment MS, UV spectroscopy and polarimetry. l-ADP: [a] 20 D +29 (c 1.0, H 2 O); 31 P NMR (D 2 O, 300 MHz) d )8.2 (d, J P–P ¼ 20.6), )10.8 (d, J P–P ¼ 20.6); MS FAB + m ⁄ z 472 (M + H) + ; HRMS (C 10 H 14 O 10 N 5 P 2 Na 2 ), calculated 472.0011, found 471.9999. l-ATP: [a] 20 D +22 (c 1.0, H 2 O); 31 P NMR (D 2 O, 300 MHz) d )6.2 (d, J P–P ¼ 18.2), )10.7 (d, J P–P ¼ 18.2), )20.5 (t, J P–P ¼ 18.2); MS FAB + m ⁄ z 574 (M + H) + ; HRMS (C 10 H 14 O 13 N 5 P 3 Na 3 ), calculated 573.9494, found 573.9453. Cloned murine R1 and R2 were prepared as described [33]. Protein concentration was estimated by Bradford assay [34] Fig. 5. L-N 3 -UDP does not inactivate RR. Solutions of 4 lM R1, 8 l M R2, and 2 mMD-ATP were preincubated in buffer A with the indicated the amounts of N 3 -L-dUDP or N 3 -D-dUDP for 15 min (total volume, 10 lL). Ninety microliters of [8- 3 H]CDP (278 lM)andD-ATP (2 m M) in buffer B were then added, and reaction was allowed to proceed for 30 min before quenching. Reaction mixture analysis was by phenyl boronate chromatography. J. He et al. Enantioselectivity of ribonucleotide reductase FEBS Journal 272 (2005) 1236–1242 ª 2005 FEBS 1239 using bovine serum albumin as standard. [2,8- 3 H]ADP was purchased from Perkin Elmer (Shelton, CT, USA), [8- 3 H]- CDP, and [8- 3 H]-GDP was purchased from Amersham (Pis- cataway, NJ, USA). Radioactive nucleotides were purified by aminophenyl boronate agarose column chromatography before use. All the other nucleotides are from Amersham or Roche (Mannheim, Germany). Ribonucleotide reductase activity assay RR was assayed by measuring the formation of [ 3 H]dNDP from [ 3 H]NDP at 25 °C, as described [33]. Reactions were carried out in buffer A (50 mm Hepes, pH ¼ 7.6, 10 mm KCl, 10 mm MgCl 2 ,25mm dithiothreitol, 7 mm NaF, 0.05 mm FeCl 3 )at25°C in a total volume of 100 lL, with final protein concentrations (given as monomer) of mR1, 0.12–3.3 lm, and of mR2, 1.0–11 lm and the indicated amount of nucleotides. mRR was preincubated for 7 or 10 min prior to [ 3 H]NDP addition. Samples were quenched (boiling water) after 10–40 min and analyzed by phenyl- boronate-agarose chromatography. Under these conditions, the amount of product formation was proportional to reac- tion time. HPLC analysis of RR-catalyzed reduction Following quenching, reaction mixtures were centrifuged through a Millipore Microcon YM-10 centrifugal filter to remove proteins and filtrates were loaded onto a Higgins C 18 RP-HPLC analytical column. The gradient was 0–2% (v ⁄ v) acetonitrile in 20 min followed by 2–5% (v ⁄ v) aceto- nitrile in 20 min, containing 10 mm triethylammonium bicarbonate pH 7.7. Acknowledgements This work was supported by NIH grant CA 58567 (BSC) and by the CNRS (BR). We thank Gilles Gos- selin for technical advice regarding l-nucleotide syn- thesis. Fig. 6. L-ADP inhibition of dGTP-dependent ADP reductase. Assay mixtures contained 0.21 lM R1, 1 lM R2, 100 lM dGTP, ± 2 mMD-ATP, variable amounts of [8- 3 H]D-ADP and the indicated amounts of L-ADP. Preincubation time was 10 min prior to substrate addition. Reaction was allowed to proceed for 40 min before quenching and analysis by phenyl boronate chromatography. (A) Lineweaver–Burke plot of L-ADP inhibition of dGTP-dependent D-ADP reductase in the presence of 2 mMD-ATP. (B) Secondary plot of the slope values of A. (C) Lineweaver- Burk plot of L-ADP inhibition of dGTP-dependent D-ADP reductase in the absence of D-ATP. (D) Secondary plot of the slope values of (C). Enantioselectivity of ribonucleotide reductase J. He et al. 1240 FEBS Journal 272 (2005) 1236–1242 ª 2005 FEBS References 1 Eklund H, Uhlin U, Fa ¨ rnegordh M, Logan DT & Nor- dlund P (2001) Structure and function of the radical enzyme ribonucleotide reductase. Prog Biophys Mol Biol 77, 177–268. 2 Kashlan OB, Scott CP, Lear JD & Cooperman BS (2002) A comprehensive model for the allosteric regula- tion of mammalian ribonucleotide reductase: functional consequences of ATP- and dATP-induced oligomeriza- tion of the large subunit. Biochemistry 41, 462–474. 3 Kashlan OB & Cooperman BS (2003) Comprehensive model for allosteric regulation of mammalian ribonu- cleotide reductase: refinements and consequences. Bio- chemistry 42, 1696–1706. 4 Weber G (1983) Biochemical strategy of cancer cells and the design of chemotherapy: G. H. A. Clowes Memorial Lecture. Cancer Res 43, 3466–3492. 5 Cory JG (1988) Ribonucleotide reductase as a chemo- therapeutic target. Adv Enzyme Regul 27, 437–455. 6 Szekeres T, Fritzer-Szekeres M & Elford HL (1997) The enzyme ribonucleotide reductase: target for antitumor and anti-HIV therapy. Crit Rev Clin Lab Sci 34, 503–528. 7 Robins MJ (1999) Mechanism-based inhibition of ribo- nucleotide reductases: new mechanistic considerations and promising biological applications. Nucleosides Nucleotides 18, 779–793. 8 Stevens MR (1999) Hydroxyurea: an overview. J Biol Regul Homeost Agents 13, 172–175. 9 Heinemann V (2003) Role of gemcitabine in the treat- ment of advanced and metastatic breast cancer. Oncol- ogy 64, 191–206. 10 Tsimberidou AM, Alvarado Y & Giles FJ (2002) Evol- ving role of ribonucleotide reductase inhibitors in hema- tologic malignancies. Expert Rev Anticancer Ther 2, 437–448. 11 Sommadossi JP (2002) Antiviral b-l -nucleosides specific for hepatitis B and virus infection. In Recent Advances in Nucleosides: Chemistry and Chemotherapy (Chu CK, ed), pp. 417–432. Elsevier, Amsterdam, the Netherlands. 12 Gumina G, Song GY, Chu CK (2001) l-Nucleosides as chemotherapeutic agents. FEMS Microbiol Lett 202, 9–15. 13 Zemlicka J (2000) Enantioselectivity of the antiviral effects of nucleoside analogues. Pharmacol Ther 85, 251–266. 14 Wang P, Hong JH, Cooperwood JS & Chu CK (1998) Recent advances in l-nucleosides: chemistry and biol- ogy. Antiviral Res 40, 19–44. 15 Cheng YC (2001) Potential use of antiviral l(–) nucleo- side analogues for the prevention or treatment of viral associated cancers. Cancer Lett 162, S33–S37. 16 Focher F, Spadari S & Maga G (2003) Antivirals at the mirror: the lack of stereospecificity of some viral and human enzymes offers novel opportunities in antiviral drug development. Curr Drug Targets Infect Disord 3, 41–53. 17 Maury G (2000) The enantioselectivity of enzymes involved in current antiviral therapy using nucleoside analogues: a new strategy? Antiviral Chem Chemother 11, 165–190. 18 Tomikawa A, Kohgo S, Ikezawa H, Iwanami N, Shudo K, Kawaguchi T, Saneyoshi M, Yamaguchi T (1997) Chiral discrimination of 2¢-deoxy-l-cytidine, l-nucleo- tides by mouse deoxycytidine kinase: low stereospecifici- ties for substrates and effectors. Biochem Biophys Res Commun 239, 329–333. 19 Verri A, Montecucco A, Gosselin G, Boudou V, Imbach JL, Spadari S, Focher F (1999) ATP is recognized by some cellular and viral enzymes: does chance drive enzy- mic enantioselectivity? Biochem J 337, 585–590. 20 Wang J, Choudhury D, Chattopadhyaya J & Eriksson S (1999) Stereoisomeric selectivity of human deoxyribo- nucleoside kinases. Biochemistry 38, 16993–16999. 21 Liou JY, Dutschman GE, Lam W, Jiang Z & Cheng YC (2002) Characterization of human UMP ⁄ CMP kinase and its phosphorylation of d- and l-form de- oxycytidine analogue monophosphates. Cancer Res 62, 1624–1631. 22 Krishnan P, Fu Q, Lam W, Liou JY, Dutschman G & Cheng YC (2002) Phosphorylation of pyrimidine deoxynucleoside analog diphosphates: selective phosphorylation of 1-nucleoside analog diphosphates by 3-phosphoglycerate kinase. J Biol Chem 277, 5453– 5459. 23 Gallois-Montbrun S, Faraj A, Seclaman E, Sommadossi J-P, Deville-Bonne D & Ve ´ ron M (2004) Broad specificity of human phosphoglycerate kinase for antiviral nucleo- side analogs. Biochem Pharmacol 68, 1749–1756. 24 Roy B, Verri A, Lossani A, Spadari S, Focher F, Aub- ertin A-M, Gosselin G, Mathe ´ C&Pe ´ rigaud C (2004) Enantioselectivity of ribonucleotide reductase: a first study using stereoisomers of pyrimidine 2¢-azido- 2¢-deoxynucleosides. Biochem Pharmacol 68, 711–718. 25 Eriksson M, Uhlin U, Ramaswamy S, Ekberg M, Regn- strom K, Sjoberg BM & Eklund H (1997) Binding of allosteric effectors to ribonucleotide reductase protein R1: reduction of active-site cysteines promotes substrate binding. Structure 5, 1077–1092. 26 Sintchak MD, Arjara G, Kellogg BA, Stubbe J & Dren- nan CL (2002) The crystal structure of class II ribo- nucleotide reductase reveals how an allosterically regulated monomer mimics a dimer. Nat Struct Biol 9, 293–300. 27 Uppsten M, Fa ¨ rnegordh M, Jordan A, Eliasson R, Eklund H & Uhlin U (2003) Structure of the large sub- unit of class Ib ribonucleotide reductase from Salmo- nella typhimurium and its complexes with allosteric effectors. J Mol Biol 330, 87–97. J. He et al. Enantioselectivity of ribonucleotide reductase FEBS Journal 272 (2005) 1236–1242 ª 2005 FEBS 1241 28 Larsson KM, Jordan A, Eliasson R, Reichard P, Logan DT & Nordlund P (2004) Structural mechan- ism of allosteric substrate specificity regulation in a ribonucleotide reductase. Nat Struct Mol Biol 11, 1142–1149. 29 Thelander L & Larsson B (1976) Active site of ribonu- cleoside diphosphate reductase from Escherichia coli: inactivation of the enzyme by 2¢-substituted ribonucleo- side diphosphates. J Biol Chem 251, 1398–1405. 30 Yoshikawa M, Kato T & Takenishi T (1967) A novel method for phosphorylation of nucleosides to 5¢-nucleo- tides. Tet Lett 50, 5065–5068. 31 Hampton A, Kappler F, Maeda M & Patel AD (1978) Use of adenine nucleotide derivatives to assess the potential of exo-active-site-directed reagents as species- or isozyme-specific enzyme inactivators. 2. Isozyme-specific inactivation of a mammalian enzyme and its significance in the possible design of fetal iso- zyme targeted antineoplastic agents. J Med Chem 21, 1137–1140. 32 Ludwig J (1981) A new route to nucleoside 5¢-triphos- phates. Acta Biochim Biophys Acad Sci Hung 16, 131– 133. 33 Scott CP, Kashlan OB, Lear JD & Cooperman BS (2001) A quantitative model for allosteric control of purine reduction by murine ribonucleotide reductase. Biochemistry 40, 1651–1661. 34 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein util- izing the principle of protein-dye binding. Anal Biochem 72, 248–254. Enantioselectivity of ribonucleotide reductase J. He et al. 1242 FEBS Journal 272 (2005) 1236–1242 ª 2005 FEBS . The enantioselectivities of the active and allosteric sites of mammalian ribonucleotide reductase Jian He 1 ,Be ´ atrice. examine the enantioselectivity of the allosteric and substrate bind- ing sites of murine ribonucleotide reductase (mRR). l-ADP binds to the active site and