Clarifications, Misconceptions and Discussion

EVB clarification

The tar file below includes all the information required to verify the EVB calculations we present in our recent work:

Shina C. L. Kamerlin, Jie Cao, Edina Rosta and Arieh Warshel, "On Unjustifiably Criticizing the EVB While Simultaneously Adopting It", J. Phys. Chem. B, 113 (2009), 10905-10915.

This tar file is 5.6MB in size, and available for free download. It includes a comprehensive README that explains what is available in the package as well as how to run the necessary calculations to the user.

Please note that the README has been included as both an MS Word document and a textfile for convenience.

Please note also that in light of continued factually incorrect criticism of our approach, we have now added further clarification to complement our paper in the pdf available for download below, and we hope that you find this helpful.

In the event of further queries please contact either Lynn Kamerlin ( or Prof. Arieh Warshel (, and we would be happy to help.

  • Sample Paper
  • Response to critisism of EVB
  • Kemp Elimination Reaction

    Here we deposited the PDB structures of the reactant state (KE70_RS.pdb) and transition state (KE70_TS.pdb) of the KE70 (wild type) designed variant with 5-nitrobenzisoxazole substrate that has been studied in our recent work:

    Maria P. Frushicheva, Jie Cao and Arieh Warshel "Challenges and Advances in Validating Enzyme Design Proposals: The Case of the Kemp Eliminase Catalysis".

    Additionally we provide the tar file below that includes all necessary information required to run the EVB calculations for the Kemp Elimination reaction in solution as well as in 1A53-2 protein.This tar file is 20MB in size, and available for free download. It includes a comprehensive README.txt that explains what is available in the package as well as how to run the necessary calculations to the user.

    If you need an additional information please contact either Maria P. Frushicheva ( or Prof. Arieh Warshel ( and we would be happy to help.

  • KE70_RS.pdb
  • KE70_TS.pdb
  • kemp_EVB.tar.gz

  • Insightful Discussion

    Testing geometrical discrimination within an enzyme active site: constrained hydrogen bonding in the ketosteroid isomerase oxyanion hole.

    Sigala PA, Kraut DA, Caaveiro JM, Pybus B, Ruben EA, Ringe D, Petsko GA, Herschlag D
    J Am Chem Soc 2008 Oct 15 130(41):13696-708

    Selected by: Shina Caroline Lynn Kamerlin and Arieh Warshel
    University of Southern California, United States of America
    Chemical Biology

    In this work, Sigala et al. provide an instructive analysis of the catalytic effects in ketostero id isomerase (KSI). Here, they argue that the active site constrains sidechain orientations in such a way as to prevent hydrogen bonds from shortening by 0.1� or less (which is indistinguishable by mo st X-ray studies), which in turn is assumed to provide a major catalytic effect. This paper provides an interesting departure from the previous works of the same authors {1}, where in the current work , it is implied that preorganization plays an important role in catalysis and can account for the ac tion of KSI. We find this to be an important work, as it brings together a wealth of information fro m a combination of experimental sources (such as X-ray crystallography, both 1H and 19F NMR spectros copy and TSA [transition state analogue] calculations), as well as from computational work. Here, th e authors present a crystal structure of 2,6-difluorophenolate (DFP) bound to KSI, as well as experi mental data on the binding of a range of di-ortho-F-phenolates, which provides a potential picture o f the hydrogen bonding networks between these TSAs and the active site residues, as well as some ide a of the possible energetics of these interactions. These data are also an excellent starting point for more detailed computational studies that will answer many of the questions brought up in this pa per.

    Now, the focus of this paper on the preorganization effect is clearly important (even though the aut hors incorrectly attribute the idea of transition state (TS) stabilization by the preorganization ef fect [see ref {2} on which Arieh Warshel is an author] to the proposals of prominent workers, such a s, for instance, Jencks, who actually proposed ground state destabilization). This is a relevant iss ue, as the authors correctly wonder how the preorganization can act as an aid to catalysis when the substrate structural changes during the reaction are typically very small. Proving that the structur al changes are very small is stimulating, as it draws clear attention to the question of what is act ually going on in this enzyme. However, we believe that the authors. conclusion that constraints on sub-0.1� changes are the source of these catalytic effects overlooks the fact that the largest chang es in the substrate (that are recognized by the enzyme) mainly occur in the charge distribution, and not in its structure, and that the preorganization effect is associated with these charge changes. Overall, we find that the most important point of this work is the fact that it brings to light the difficulty of identifying the source of the catalytic effect when looking at structural changes with out exploring the corresponding larger structural changes in water. It would seem to us that the rea son for not observing the structural signature of the catalytic effect (and, thus, assuming that it is a very subtle change) is due to the fact that the authors are looking at the charged system rathe r than the uncharged state (see a discussion of this issue in one of our recent works {3}). However, the difference actually lies in the uncharged state, and, therefore, the main difficulty arises whe n looking for the preorganization effect in the enzyme with the charged TS or TSA. This preorganizat ion effect is associated with the reduction of the reorganization energy in water upon charging the TS, and, thus, the reorganization energy is very large in water, and comparatively small in the enzy me.s active site. Furthermore, in the case of KSI, the difference in the preorganization between the TSA (where there is no large electrostatic effect) and the real TS is associated with the freedom o f motion in the uncharged state of the TSA (see ref {3} on which Arieh Warshel is a co-author). Fina lly, Sigala et al. argue that rigid active site constraints promote catalysis by preventing favourab le interactions in the ground state. In our view, this seems to present a ground state destabilizati on idea (rather than that of transition state stabilization). It is proposed that the catalysis is a ssociated with preventing the hydrogen bonds from stabilizing the C=O state rather than in the extra stabilization of the charged C-O- state. Apparently (as was found by quantitative free energy calcu lations [see ref {3} on which Arieh Warshel is a co-author] of exactly the same TSAs), the hydrogen bond to the C=O state is far less important than that to the charged C-O- state. Overall, the presen t paper is an exciting work that draws clear attention to the difficulty associated with elucidating catalytic effects when using only structural information, and also helps (if combined with other wo rks) to emphasize the need for clear ways of defining and analyzing the electrostatic preorganizatio n effect.

    {1} Kraut et al. PLoS Biol 2006, 4:e99 [PMID:16602823].
    {2} Warshel A, Proc Natl Acad Sci USA 1978, 75:5250-4 [PMID:281676].
    {3} Warshel et al. Biochemistry 2007, 46:1466- 76 [PMID:17279612].

    Source: e/id/1161402

    Does Water Relay Play an Important Role in Phosphoryl Transfer Reactions? Insights from Theoretical Study of a Model Reaction in Water and tert-Butanol.

    Yang Y, Cui Q
    J Phys Chem B 2009 Apr 9 113(14):4930-9

    Selected by: Shina Caroline Lynn Kamerlin and Arieh Warshel

    Phosphate hydrolysis is one of the most important biochemical reactions, being involved in, amon gst other things, DNA and RNA synthesis, protein synthesis and energy transduction as well as many m etabolic and signalling pathways. However, despite extensive experimental and theoretical work, the precise mechanism of this key reaction and the nature of the transition states both in enzymatic sys tems and in solution remain elusive.

    In recent years, a variety of computational approaches have been suggested in order to study this cr ucial reaction, one of which is the use of a "mixed" solvation model, in which one or more water mol ecules are explicitly included in the ab initio description of the reacting system, with the remaind er of the solvent being modelled implicitly as a continuum (see e.g. {1-3}). This approach has been a cause of concern as, amongst other potential problems, it is unlikely to accurately reproduce the correct boundary conditions between the solute and the bulk. In fact, in a recent study ({4}, on whi ch we are authors), it was demonstrated that the only way to obtain any sort of reliable results fro m such an approach is by carefully selecting the explicit water molecules from short QM/MM runs that act as a model for the true infinite system (while also taking the entropy of the explicit water mo lecules into account), and that anyhow such an approach appears not to provide any advantage over us ing a "pure" continuum model, without explicit water molecules. Additionally, it was demonstrated th at even when carefully selecting the position of the water molecules this approach becomes increasin gly inaccurate with increasing numbers of water molecules. Cui and coworkers have performed a carefu l computational study of 3'-m-nitrobenzyl phosphate (UNP) in both aqueous and tert-butanol solutions , in order to investigate whether water relay places an important role in phosphoryl transfer reacti ons. Here, they have analyzed solute flexibility and solvent structure near the solute using equilib rium molecular dynamics simulations, and a combined QM/MM potential function for the solute. Snapsho ts from these simulations were used in minimum energy path calculations in order to compare the ener getics of direct nucleophilic attack as well as water-mediated nucleophilic attack pathways. From th ese calculations, they demonstrate that water relay appears not to have any major energetic impact o n phosphoryl transfer in the system that they have chosen. This corroborates the previous finding th at there is no energetic advantage to using a "mixed" as opposed to a "pure" continuum model when mo delling phosphate hydrolysis; i.e. whilst the ideal situation would be to of course used hybrid QM/M M calculations in which the solvent is modelled explicitly, even the use of a continuum appears to b e surprisingly adequate, and increasing explicit water molecules into the system (while still using a continuum model) only creates problems, unless of course one is specifically interested in the che mistry related to the inclusion of these water molecules. Cui et al.'s finding is important, because while theoretical studies can be very useful (and in some cases crucial) for resolving mechanistic questions, their accuracy is only ever as good as the accuracy of the approach used. Finally, it sho uld be noted that, in recent years, several notable workers (see e.g. {5-8}, amongst others) have su ggested that proton relay plays a significant role in enzyme catalysis. However, the current work of Cui and co-workers casts significant doubt on this proposition.

    {1} Liu et al. Biochemistry 2006, 45:10043-53 [PMID:16906762].
    {2} Zhang et al. Chem Commun (Cam b) 2007, 28:1638-40 [PMID:17530085].
    {3} Hu and Brinck, J Phys Chem A 1999, 103:5379-86 [DOI:10. 1021/jp9835061].
    {4} Kamerlin et al. Chemphyschem 2009, 10:1125-34 [PMID:19301306].
    {5} Agar wal et al. J Am Chem Soc 2000, 122:4803.12 [DOI:10.1021/ja994456w].
    {6} Dittrich et al. Biophys J 2004, 87:2954-67 [PMID:15315950].
    {7} Wang et al. J Am Chem Soc 2007, 129:4731-7 [PMID:1737592 6].
    {8} Wang and Schlick, J Am Chem Soc 2008, 130:13240-50 [PMID:18785738].

    Source: /id/1160434

    Inhibition of protein tyrosine phosphatase 1B and alkaline phosphatase by bis(maltolato)oxovanadi um (IV)

    Li M, Ding W, Baruah B, Crans DC, Wang R
    J Inorg Biochem 2008 Oct 102(10):1846-53

    Shina Caroline Lynn Kamerlin and Arieh Warshel
    University of Southern California, United States of America
    Chemical Biology

    This important paper reports a highly promising model for the design of potent and selective prot ein tyrosine phosphatase 1B (PTP1B) inhibitors for the treatment of diabetes.

    The previous demonstration that the knockout of PTP1B in mice results in insulin hypersensitivity ha s made this enzyme a key drug target for the treatment of diabetes, which is a problem of urgent the rapeutic relevance as current therapies for diabetes are limited and often ineffective. However, the design of PTP1B inhibitors has been impeded by the fact that even only distantly related members of the PTP superfamily share a common sequence (known as the "PTP signature motif") in the catalytic d omain, making it very hard to design potent inhibitors that are also selective for PTP1B. One potent ial solution to this problem is the use of vanadate, which has been recognised to be a potent phosph atase inhibitor (due to its structural similarity to phosphate). However, inorganic vanadium salts e xhibit poor gastrointestinal absorption and side effects in vivo making them undesirable as potentia l drugs. Additionally, an issue that needs to be taken account is the fact that vanadium compounds h ave been demonstrated to not only inhibit PTPases but also a number of other enzymes, including alka line phosphatase (ALP). ALP is a metallophosphatase that has been linked to bone and liver disease i n vivo, and is a marker of osteoblastic differentiation. Like PTP1B, ALP is also responsible for pho sphate monoester hydrolysis. In this work, Crans and coworkers have synthesized an organic vanadium compound -- bis(maltolato)oxovanadium(IV), aka BMOV -- that is more lipophilic than inorganic vanadi um salts and thus potentially has better bioavilability than the inorganic vanadium compounds. This compound has been implicated to inhibit phosphatases in two recent animal studies {1,2} and one cell {1} study of its physiological activity. Additionally, BMOV is a hypoglycaemia agent that has exhib ited high insulin-mimetic potential to diabetes, and is currently being investigated in phase 2 clin ical trials. Crans and coworkers present steady state kinetic studies of the effect of a solution of BMOV on both GST-PTP1B and ALP in vitro. Their results suggest that the inhibition pattern of GST-P TP1B by BMOV is a mixture of competitive and noncompetitive inhibition. Additionally, the activity o f GST-PTP1B appears to be reversibly inhibited by solutions of BMOV with an IC50 value of 0.86 � 0.0 2 mM, and the incubation of GST-PTP1B with BMOV shows time-dependent biphasic inactivation of the pr otein. In contrast, the inhibition pattern of ALP by BMOV appears to be reversible and competitive w ith an IC50 value of 32.1 � 0.06 mM and no biphasic inactivation of the protein. This is an importan t result, because it hints that the mechanisms for the inhibitory effects of BMOV on the two enzymes are very different, something that can be taken advantage of in designing inhibitors for PTP1B that have better bioavailability and a lower side effect profile compared to inorganic vanadium salts. < br>
    References: {1} Carr et al. Am J Physiol Heart Circ Physiol 2004, 287:H268.76 [PMID:14988069]. {2} S hah and Singh, Naunyn Schmiedebergs Arch Pharmacol 2006, 373:221.9 [PMID:16736159]. {3} Ferrer et al . J Biol Inorg Chem 2006, 11:791.801 [PMID:16821038].

    Source: 932

    Anionic charge is prioritized over geometry in aluminum and magnesium fluoride transition state a nalogs of phosphoryl transfer enzymes.

    Baxter NJ, Blackburn GM, Marston JP, Hounslow AM, Cliff MJ, Bermel W, Williams NH, Hollfelder F, Wem mer DE, Waltho JP
    J Am Chem Soc 2008 Mar 26 130(12):3952-8a

    Shina Caroline Lynn Kamerlin and Arieh Warshel
    University of Southern California, United States of America
    Chemical Biology

    Phosphoryl transfer reactions are ubiquitous in biology, and central to, amongst other things, DN A and RNA synthesis, protein synthesis as well as several signalling pathways. Thus, unsurprisingly, the enzymes that regulate this key reaction have increasingly become the subject of great therapeut ic interest and intensive research effort. Particularly, aluminium fluoride (AlFx) complexes have be en widely used as transition state analogues (TSAs) for phosphoryl transfer reactions, and provide t he closest models that have so far been obtained for the catalytic transition state. These authors s how that the physiological effects of fluoride are caused, at least in part, by aluminium, which at near-physiological ratios to magnesium, dominates in the enzyme-metal fluoride inhibitory TSA comple xes.

    Several structures of such complexes have been reported to date, and it has been proposed that there is a correlation between the pH of the crystallization solution and the number of coordinated F in the AlFx complex that is subsequently formed {1}. That is, it was proposed that enzymes crystallized in the presence of Mg2+ above pH 7.0 contain a central trigonal planar moiety that was most likely AlF3, whereas at or below pH 7.0, with two exceptions, the enzymes contained a central square planar moiety that was most likely AlF4- (provided that the atomic resolution was high enough). It was the n proposed that this switch in coordination with pH arises from either a change in ionisation state within the complex, or from the pH-dependent switch between the different coordination states observ ed for aluminium hydroxyls and fluorides in solution, though neither proposal fully accounts for all available data. In this work, the authors have performed a detailed study of b-phosphoglucomutase f rom Lactococcus lactis using 19F NMR, and they have demonstrated that the pH switch in fluoride coor dination does not derive from the conversion of an AlF4- moiety into AlF3 as was previously suggeste d, but rather the change in coordination comes from the enzyme progressively replacing AlF4- with Mg F3- as the pH is increased. That is, interestingly, it appears that the enzyme would rather sacrific e the preferred trigonal geometry in order to maintain anionic charge, and this occurs over a wide r ange of pH. Subsequently, the authors examine two related phosphatases: phosphoserine phosphatase fr om Methanococcus jannaschii as well as a phosphoglycerate kinase from Geobacillus steraothermophilus , and demonstrate that these enzymes undergo similar behaviour. Finally, the authors also demonstrat e that, at near-physiological ratios of aluminium to magnesium, aluminium can dominate over magnesiu m in the enzyme-metal fluoride inhibitory TSA complexes, which makes it likely that any of the physi ological effects that are obtained during enzyme inhibitions by metal fluorides are actually the res ult of aluminium fluoride species. This work is particularly important because it demonstrates that not only does AlF4- out-compete AlF3 but also there is no evidence for any population of AlF3 in any of the TSAs for the three enzymes studied, suggesting that at least several of the high-pH AlF3 com plexes suggested in the literature should be approached with caution.

    References: {1} Schlichting and Reinstein, Nat Struct Biol 1999, 6:721-723 [PMID:10426946].
    Competing interests: I recently co-authored a paper with N. H. Williams on dineopentyl phosph ate hydrolysis in solution, which is not directly related to the topic of this article. Evaluated 13 Nov 2008

    Source: 855

    Mismatched dNTP incorporation by DNA polymerase beta does not proceed via globally different con formational pathways.

    Tang KH, Niebuhr M, Tung CS, Chan HC, Chou CC, Tsai MD
    Nucleic Acids Res 2008 May 36(9):2948-57

    Selected by:: Shina Caroline Lynn Kamerlin and Arieh Warshel
    University of Southern California, United States of America
    Chemical Biology

    DNA polymerases catalyze the insertion reaction with dNTP substrates according to the identity of th e templating base (in order to form Watson-Crick base pairs). Thus, DNA polymerases control the fide lity of DNA replication, which is in turn essential for maintaining the integrity of the genome. In order to understand how these enzymes control the fidelity of DNA replication, it is essential to el ucidate the mechanism of matched vs. mismatched dNTP incorporations. However, studies of the latter have been complicated by the fact that mismatched complexes do not crystallize readily. In this work , the authors employ small-angle X-ray scattering (SAXS) as well as structural modelling to probe th e conformation of different intermediate states of mammalian DNA polymerase b (Pol b) in both its wi ld-type state and in an error-prone variant (I260Q). This study indicates that the conformation of t he mismatched ternary complex of wild-type Pol b lies on the conformational pathway between the open and closed forms.

    This finding is consistent with a recent study from my lab {1}. We deduced (by comparing both simula tions and experiments), that the right (R) nucleotide is incorporated through a transition state in a closed conformation and the wrong (W) nucleotide is incorporated through a transition state in one or perhaps even several conformations that are somewhat shifted towards the open structure. It shou ld be noted, however, that this finding is fundamentally different from the "pre-chemistry" concept suggested by other workers (e.g. in Alberts et al. {2}), who propose that additional structural re-a rrangements (in the form of a series of subtle stochastic alterations in the metal ion and phosphate coordination distances) after the binding conformational change (but prior to the chemical reaction ) are somehow a major contributor to fidelity. This last idea is at odds with the pre-organization i dea (see e.g. our review {3}), in that while there is indeed a need for electrostatic pre-organizati on for effective catalysis, this is not a "pre-chemistry" rate-determining step, but simply a part o f the formation of the ES complex. True "pre-chemistry" must involve a rate-determining barrier befo re the chemical step. However, it has been shown that the rate-determining step in this reaction is the chemical step itself. Thus, if the barrier that corresponds to kobs is lower than the chemical b arrier, the motion from the open to closed form as well as any structural re-arrangements occurring during the binding process and before the chemical step (in the ES) will have no influence on the ra te constant and thus no effect on fidelity (see {1}). This point is related to the present proposal that the conformational closing is not a major contributor to fidelity. Finally, it should be clarif ied that the possible structural changes in the transition state are not the reason for fidelity but in fact reflect a relaxation of the bad interaction in the base binding site of W, that could have led to even larger fidelity without the structural changes (see {1}).

    References: {1} Xiang et al. Proteins 2008, 70: 231 [PMID:17671961]; {2} Alberts et al. J Am Chem Soc 2007, 129:11100 [PMID:17696533]; {3} Olsson et al. Chem Rev 2006, 105: 1737-56.

    Source: 449

    The associative nature of adenylyl transfer catalyzed by T4 DNA ligase.

    Cherepanov AV, Doroshenko EV, Matysik J, de Vries S, de Groot HJ
    Proc Natl Acad Sci U S A 2008 Jun 24 105(25):8563-8

    Selected by: Shina Caroline Lynn Kamerlin and Arieh Warshel
    University of Southern California, United States of America
    Chemical Biology

    Understanding the chemical mechanics of biocatalysis is a fundamental goal of life sciences. The development of both high resolution x-ray diffraction and solution NMR techniques have allowed for s ignificant advances to be made towards solving protein structure, which in turn allows for more real istic descriptions of protein electrostatic energies in computational studies. Additionally, time-re solved x-ray experiments at low temperature have recently made it possible to follow reaction interm ediates by means of kinetic crystallography. In this work, the authors introduce time-resolved low t emperature magic-angle-spinning (cryo-MAS) NMR spectroscopy as a complementary .non-invasive. techni que to study the catalytic dynamics of biochemical reactions.

    As an example of the applications of this approach, the authors have studied the nucleotidyl transfe r reaction catalysed by DNA ligase, an ATP-dependent enzyme responsible for sealing nicks in dsDNA. Understanding the molecular details of ligase catalysis is important for furthering our knowledge of the mechanism of metal-promoted phosphoryl transfer reactions. Here, the authors report a single-tu rnover 31P solid-state NMR study of bacteriophage T4, where the formation of a high-energy covalent ligase-nucleotide complex is triggered in situ by the photo release of caged Mg2+, allowing sequenti ally formed intermediates to be monitored by NMR. By using this approach, the authors detect a react ion intermediate in the adenylyl transfer reaction of T4 DNA ligase, which has been attributed to a penta-coordinated phosphorane protein-ligand complex. Ultimately, the key challenge lies in determin ing the structure of the key transition states for phosphate hydrolysis. Apparently, at present, thi s is not possible by any conceivable experiment due to the fact that such transition states are very high in energy, though being able to characterize key intermediates presents an important step in t his direction. Over the past twenty years, the most powerful approach towards .describing. transitio n states with the constraint of reproducing structural information about low energy states has been provided by means of computational simulates. This is also the case for phosphate hydrolysis, which has been extensively studied by means of simulation approaches, as was for instance done by our grou p in the case of the evaluation of the structure and energy of the transition state of DNA Polymerae s {1}. The work presented by the authors here is, however, a good starting point for further theoret ical studies that can characterize the geometry and energetics of the relevant transition state. Ove rall, the NMR approach presented by the authors is very impressive, and shows promise as a powerful analytical tool for molecular studies of biocatalytic processes.

    Reference: {1} Florian et al. PNAS 2005, 102:6819-24 [PMID:15863620].

    Source: 298

    back to homepage