Structural basis for the dual thymidine and thymidylate kinase activity of herpes thymidine kinases

Optimization of a lentivirus-mediated gene therapy targeting HIV-1 RNA to eliminate HIV-1-infected cells

Persistence of HIV-1 in cellular reservoirs results in lifelong infection, with cure achieved only in rare cases through ablation of marrow-derived cells. We report on optimization of an approach that could potentially be aimed at eliminating these reservoirs, hijacking the HIV-1 alternative splicing process to functionalize the herpes simplex virus thymidine kinase (HSVtk)/ganciclovir (GCV) cell suicide system through targeted RNA trans-splicing at the HIV-1 D4 donor site. AUG1-deficient HSVtk therapeutic pre-mRNA was designed to gain an in-frame start codon from HIV-1 tat1. D4-targeting lentiviral vectors were produced and used to transduce HIV-1-expressing cells, where trans-spliced HIV-1 tat/HSVtk mRNA was successfully detected. However, translation of catalytically active HSVtk polypeptides from internal AUGs in HSVtk ΔAUG1 caused GCV-mediated cytotoxicity in uninfected cells. Modifying these sites in the D4 opt 2 lentiviral vector effectively mitigated this major off-target effect. Promoter choice was optimized for increased transgene expression. Affinity for HIV-1 RNA predicted in silico correlated with the propensity of opt 2 payloads to induce HIV-1 RNA trans-splicing and killing of HIV-1-expressing cells with no significant effect on uninfected cells. Following latency reversing agent (LRA) optimization and treatment, 45% of lymphocytes in an HIV-1-infected latency model could be eliminated with D4 opt 2/GCV. Further development would be warranted to exploit this approach.

Identification and Characterization of Entamoeba histolytica Choline Kinase

PURPOSE

Entamoeba histolytica is one of the death-causing parasites in the world. Study on its lipid composition revealed that it is predominated by phosphatidylcholine and phosphatidylethanolamine. Further study revealed that its phosphorylated metabolites might be produced by the Kennedy pathway. Here, we would like to report on the characterizations of enzymes from this pathway that would provide information for the design of novel inhibitors against these enzymes in future.

METHODOLOGY

E. histolytica HM-1:IMSS genomic DNA was isolated and two putative choline/ethanolamine kinase genes (EhCK1 and EhCK2) were cloned and expressed from Escherichia coli BL21 strain. Enzymatic characterizations were further carried out on the purified enzymes.

RESULTS

EhCK1 and EhCK2 were identified from E. histolytica genome. The deduced amino acid sequences were more identical to its homologues in human (35-48%) than other organisms. The proteins were clustered as ethanolamine kinase in the constructed phylogeny tree. Sequence analysis showed that they possessed all the conserved motifs in choline kinase family: ATP-binding loop, Brenner's phosphotransferase motif, and choline kinase motif. Here, the open reading frames were cloned, expressed, and purified to apparent homogeneity. EhCK1 showed activity with choline but not ethanolamine. The biochemical characterization showed that it had a Vmax of 1.9 ± 0.1 µmol/min/mg. Its Km for choline and ATP was 203 ± 26 µM and 3.1 ± 0.4 mM, respectively. In contrast, EhCK2 enzymatic activity was only detected when Mn2+ was used as the co-factor instead of Mg2+ like other choline/ethanolamine kinases. Highly sensitive and specific antibody against EhCK1 was developed and used to confirm the endogenous EhCK1 expression using immunoblotting.

CONCLUSIONS

With the understanding of EhC/EK importance in phospholipid metabolism and their unique characteristic, EhC/EK could be a potential target for future anti-amoebiasis study.

Unusual metal ion cofactor requirement of Entamoeba histolytica choline and ethanolamine kinase isoforms

The de novo biosynthesis of phosphatidylcholine and phosphatidylethanolamine in Entamoeba histolytica is largely dependent on the CDP-choline and CDP-ethanolamine pathways. Although the first enzymes of these pathways, EhCK1 and EhCK2, have been previously characterized, their enzymatic activity was found to be low and undetectable, respectively. This study aimed to identify the unusual characteristics of these enzymes in this deadly parasite. The discovery that EhCKs prefer Mn²⁺ over the typical Mg²⁺ as a metal ion cofactor is intriguing for CK/EK family of enzymes. In the presence of Mn²⁺, the activity of EhCK1 increased by approximately 108-fold compared to that in Mg²⁺. Specifically, in Mg²⁺, EhCK1 exhibited a V_max and K_0.5 of 3.5 ± 0.1 U/mg and 13.9 ± 0.2 mM, respectively. However, in Mn²⁺, it displayed a V_max of 149.1 ± 2.5 U/mg and a K_0.5 of 9.5 ± 0.1 mM. Moreover, when Mg²⁺ was present at a constant concentration of 12 mM, the K_0.5 value for Mn²⁺ was ~ 2.4-fold lower than that in Mn²⁺ alone, without affecting its V_max. Although the enzyme efficiency of EhCK1 was significantly improved by about 25-fold in Mn²⁺, it is worth noting that its K_m for choline and ATP were higher than in equimolar of Mg²⁺ in a previous study. In contrast, EhCK2 showed specific activity towards ethanolamine in Mn²⁺, exhibiting Michaelis-Menten kinetic with ethanolamine (K_m = 312 ± 27 µM) and cooperativity with ATP (K_0.5 = 2.1 ± 0.2 mM). Additionally, we investigated the effect of metal ions on the substrate recognition of human choline and ethanolamine kinase isoforms. Human choline kinase α2 was found to absolutely require Mg²⁺, while choline kinase β differentially recognized choline and ethanolamine in Mg²⁺ and Mn²⁺, respectively. Finally, mutagenesis studies revealed that EhCK1 Tyr129 was critical for Mn²⁺ binding, while Lys233 was essential for substrate catalysis but not metal ion binding. Overall, these findings provide insight into the unique characteristics of the EhCKs and highlight the potential for new approaches to treating amoebiasis. Amoebiasis is a challenging disease for clinicians to diagnose and treat, as many patients are asymptomatic. However, by studying the enzymes involved in the CDP-choline and CDP-ethanolamine pathways, which are crucial for de novo biosynthesis of phosphatidylcholine and phosphatidylethanolamine in Entamoeba histolytica, there is great potential to discover new therapeutic approaches to combat this disease.

Substrate recognition and mechanism revealed by ligand-bound polyphosphate kinase 2 structures

Inorganic polyphosphate is a ubiquitous, linear biopolymer built of up to thousands of phosphate residues that are linked by energy-rich phosphoanhydride bonds. Polyphosphate kinases of the family 2 (PPK2) use polyphosphate to catalyze the reversible phosphorylation of nucleotide phosphates and are highly relevant as targets for new pharmaceutical compounds and as biocatalysts for cofactor regeneration. PPK2s can be classified based on their preference for nucleoside mono- or diphosphates or both. The detailed mechanism of PPK2s and the molecular basis for their substrate preference is unclear, which is mainly due to the lack of high-resolution structures with substrates or substrate analogs. Here, we report the structural analysis and comparison of a class I PPK2 (ADP-phosphorylating) and a class III PPK2 (AMP- and ADP-phosphorylating), both complexed with polyphosphate and/or nucleotide substrates. Together with complementary biochemical analyses, these define the molecular basis of nucleotide specificity and are consistent with a Mg²⁺ catalyzed in-line phosphoryl transfer mechanism. This mechanistic insight will guide the development of PPK2 inhibitors as potential antibacterials or genetically modified PPK2s that phosphorylate alternative substrates.

Development of a cancer-marker activated enzymatic switch from the herpes simplex virus thymidine kinase

Discovery of new cancer biomarkers and advances in targeted gene delivery mechanisms have made gene-directed enzyme prodrug therapy (GDEPT) an attractive method for treating cancer. Recent focus has been placed on increasing target specificity of gene delivery systems and reducing toxicity in non-cancer cells in order to make GDEPT viable. To help address this challenge, we have developed an enzymatic switch that confers higher prodrug toxicity in the presence of a cancer marker. The enzymatic switch was derived from the herpes simplex virus thymidine kinase (HSV-TK) fused to the CH1 domain of the p300 protein. The CH1 domain binds to the C-terminal transactivation domain (C-TAD) of the cancer marker hypoxia inducible factor 1α. The switch was developed using a directed evolution approach that evaluated a large library of HSV-TK/CH1 fusions using a negative selection for azidothymidine (AZT) toxicity and a positive selection for dT phosphorylation. The identified switch, dubbed TICKLE (Trigger-Induced Cell-Killing Lethal-Enzyme), confers a 4-fold increase in AZT toxicity in the presence of C-TAD. The broad substrate specificity exhibited by HSV-TK makes TICKLE an appealing prospect for testing in medical imaging and cancer therapy, while establishing a foundation for further engineering of nucleoside kinase protein switches.

Conservation and Role of Electrostatics in Thymidylate Synthase

Conservation of function across families of orthologous enzymes is generally accompanied by conservation of their active site electrostatic potentials. To study the electrostatic conservation in the highly conserved essential enzyme, thymidylate synthase (TS), we conducted a systematic species-based comparison of the electrostatic potential in the vicinity of its active site. Whereas the electrostatics of the active site of TS are generally well conserved, the TSs from minimal organisms do not conform to the overall trend. Since the genomes of minimal organisms have a high thymidine content compared to other organisms, the observation of non-conserved electrostatics was surprising. Analysis of the symbiotic relationship between minimal organisms and their hosts, and the genetic completeness of the thymidine synthesis pathway suggested that TS from the minimal organism Wigglesworthia glossinidia (W.g.b.) must be active. Four residues in the vicinity of the active site of Escherichia coli TS were mutated individually and simultaneously to mimic the electrostatics of W.g.b TS. The measured activities of the E. coli TS mutants imply that conservation of electrostatics in the region of the active site is important for the activity of TS, and suggest that the W.g.b. TS has the minimal activity necessary to support replication of its reduced genome.

A novel viral thymidylate kinase with dual kinase activity

Nucleotide phosphorylation is a key step in DNA replication and viral infections, since suitable levels of nucleotide triphosphates pool are required for this process. Deoxythymidine monophosphate (dTMP) is produced either by de novo or salvage pathways, which is further phosphorylated to deoxythymidine triphosphate (dTTP). Thymidyne monophosphate kinase (TMK) is the enzyme in the junction of both pathways, which phosphorylates dTMP to yield deoxythymidine diphosphate (dTDP) using adenosine triphosphate (ATP) as a phosphate donor. White spot syndrome virus (WSSV) genome contains an open reading frame (ORF454) that encodes a thymidine kinase and TMK domains in a single polypeptide. We overexpressed the TMK ORF454 domain (TMKwssv) and its specific activity was measured with dTMP and dTDP as phosphate acceptors. We found that TMKwssv can phosphorylate dTMP to yield dTDP and also is able to use dTDP as a substrate to produce dTTP. Kinetic parameters K M and k cat were calculated for dTMP (110 μM, 3.6 s(-1)), dTDP (251 μM, 0.9 s(-1)) and ATP (92 μM, 3.2 s(-1)) substrates, and TMKwssv showed a sequential ordered bi-bi reaction mechanism. The binding constants K d for dTMP (1.9 μM) and dTDP (10 μM) to TMKwssv were determined by Isothermal Titration Calorimetry. The affinity of the nucleotidic analog stavudine monophosphate was in the same order of magnitude (K d 3.6 μM) to the canonical substrate dTMP. These results suggest that nucleotide analogues such as stavudine could be a suitable antiviral strategy for the WSSV-associated disease.

A cyclic HSV1-TK reporter for real-time PET imaging of apoptosis

The coordination of cell proliferation and programmed death (apoptosis) is essential for normal physiology, and imbalance in these two opposing processes is implicated in various diseases. Objective and quantitative noninvasive imaging of apoptosis would significantly facilitate rapid screening as well as validation of therapeutic chemicals. Herein, we molecularly engineered an apoptosis switch-on PET-based cyclic herpes simplex virus type 1-thymidine kinase reporter (cTK266) containing a caspase-3 recognition domain as the switch. Translation of the reporter and protein splicing in healthy mammalian cells produce an inactive cyclic chimera. Upon apoptosis, caspase-3-specific cleavage of the circular product occurs, resulting in the restoration of the thymidine kinase activity, which can be detected in living cells and animals by noninvasive PET imaging. Our results showed the high sensitivity of this reporter in dynamic and quantitative imaging of apoptosis in living subjects. This reporter could be applied as a valuable tool for high-throughput functional screening of proapoptotic and antiapoptotic compounds in preclinical models in drug development, and monitoring the destination of therapeutic cells in clinical settings.

A designed equine herpes thymidine kinase (EHV4 TK) variant improves ganciclovir-induced cell-killing

The limitations of the ganciclovir (GCV)/herpes simplex virus thymidine kinase (HSV1 TK: EC 2.7.1.21) system as a suicide gene therapy approach have been extensively studied over the years. In our study, we focused on improving the cytotoxic profile of the GCV/equine herpes virus-4 thymidine kinase (EHV4 TK: EC 2.7.1.21) system. Our approach involved the structure-guided mutagenesis of EHV4 TK in order to switch its ability to preferentially phosphorylate the natural substrate deoxythymidine (dT) to that of GCV. We performed steady-state kinetic analysis, genetic complementation in a thymidine kinase-deficient Escherichia coli strain, isothermal titration calorimetry, and analysis of GCV-induced cell killing through generation of HEK 293 stable cell-lines expressing EHV4 TK mutants and wild-type EHV4 TK. We found that the EHV4 TK S144H-GFP mutant preferentially phosphorylates GCV and confers increased GCV-induced cytotoxicity compared to wild-type EHV4 TK.

In vitro-selected drug-resistant varicella-zoster virus mutants in the thymidine kinase and DNA polymerase genes yield novel phenotype-genotype associations and highlight differences between antiherpesvirus drugs

Varicella zoster virus (VZV) is usually associated with mild to moderate illness in immunocompetent patients. However, older age and immune deficiency are the most important risk factors linked with virus reactivation and severe complications. Treatment of VZV infections is based on nucleoside analogues, such as acyclovir (ACV) and its valyl prodrug valacyclovir, penciclovir (PCV) as its prodrug famciclovir, and bromovinyldeoxyuridine (BVDU; brivudin) in some areas. The use of the pyrophosphate analogue foscarnet (PFA) is restricted to ACV-resistant (ACV(r)) VZV infections. Since antiviral drug resistance is an emerging problem, we attempt to describe the contributions of specific mutations in the viral thymidine kinase (TK) gene identified following selection with ACV, BVDU and its derivative BVaraU (sorivudine), and the bicyclic pyrimidine nucleoside analogues (BCNAs), a new class of potent and specific anti-VZV agents. The string of 6 Cs at nucleotides 493 to 498 of the VZV TK gene appeared to function as a hot spot for nucleotide insertions or deletions. Novel amino acid substitutions (G24R and T86A) in VZV TK were also linked to drug resistance. Six mutations were identified in the "palm domain" of VZV DNA polymerase in viruses selected for resistance to PFA, PCV, and the 2-phophonylmethoxyethyl (PME) purine derivatives. The investigation of the contributions of specific mutations in VZV TK or DNA polymerase to antiviral drug resistance and their impacts on the structures of the viral proteins indicated specific patterns of cross-resistance and highlighted important differences, not only between distinct classes of antivirals, but also between ACV and PCV.

Human and viral nucleoside/nucleotide kinases involved in antiviral drug activation: structural and catalytic properties

Antiviral nucleoside and nucleotide analogs, essential for the treatment of viral infections in the absence of efficient vaccines, are prodrug forms of the active compounds that target the viral DNA polymerase or reverse transcriptase. The activation process requires several successive phosphorylation steps catalyzed by different kinases, which are present in the host cell or encoded by some of the viruses. These activation reactions often are rate-limiting steps and are thus open to improvement. We review here the structural and enzymatic properties of the enzymes that carry out the activation of analogs used in therapy against human immunodeficiency virus and against DNA viruses such as hepatitis B, herpes and poxviruses. Four major classes of drugs are considered: thymidine analogs, non-natural L-nucleosides, acyclic nucleoside analogs and acyclic nucleoside phosphonate analogs. Their efficiency as drugs depends both on the low specificity of the viral polymerase that allows their incorporation into DNA, but also on the ability of human/viral kinases to provide the activated triphosphate active forms at a high concentration at the right place. Two distinct modes of action are considered, depending on the origin of the kinase (human or viral). If the human kinases are house-keeping enzymes that belong to the metabolic salvage pathway, herpes and poxviruses encode for related enzymes. The structures, substrate specificities and catalytic properties of each of these kinases are discussed in relation to drug activation.

Engineering Kinases to Phosphorylate Nucleoside Analogs for Antiviral and Cancer Therapy

Enzyme engineering by directed evolution presents a powerful strategy for tailoring the function and physicochemical properties of biocatalysts to therapeutic and industrial applications. Our laboratory's research focuses on developing novel molecular tools for protein engineering, as well as on utilizing these methods to customize enzymes and to study fundamental aspects of their structure and function. Specifically, we are interested in nucleoside and nucleotide kinases which are responsible for the intracellular phosphorylation of nucleoside analog (NA) prodrugs to their biologically active triphosphates. The high substrate specificity of the cellular kinases often interferes with prodrug activation and consequently lowers the potency of NAs as antiviral and cancer therapeutics. A working solution to the problem is the co-adminstration of a promiscuous kinase from viruses, bacteria, and other mammals. However, further therapeutic enhancements of NAs depend on the selective and efficient prodrug phosphorylation. In the absence of true NA kinases in nature, we are pursuing laboratory evolution strategies to generate efficient phosphoryl-transfer catalysts. This review summarizes some of our recent work in the field and outlines future challenges.

Dodging the crisis of folding proteins with knots

Proteins with nontrivial topology, containing knots and slipknots, have the ability to fold to their native states without any additional external forces invoked. A mechanism is suggested for folding of these proteins, such as YibK and YbeA, that involves an intermediate configuration with a slipknot. It elucidates the role of topological barriers and backtracking during the folding event. It also illustrates that native contacts are sufficient to guarantee folding in approximately 1-2% of the simulations, and how slipknot intermediates are needed to reduce the topological bottlenecks. As expected, simulations of proteins with similar structure but with knot removed fold much more efficiently, clearly demonstrating the origin of these topological barriers. Although these studies are based on a simple coarse-grained model, they are already able to extract some of the underlying principles governing folding in such complex topologies.

Crystal structure of poxvirus thymidylate kinase: an unexpected dimerization has implications for antiviral therapy

Unlike most DNA viruses, poxviruses replicate in the cytoplasm of host cells. They encode enzymes needed for genome replication and transcription, including their own thymidine and thymidylate kinases. Some herpes viruses encode only 1 enzyme catalyzing both reactions, a peculiarity used for prodrug activation to obtain maximum specificity. We have solved the crystal structures of vaccinia virus thymidylate kinase bound to TDP or brivudin monophosphate. Although the viral and human enzymes have similar sequences (42% identity), they differ in their homodimeric association and active-site geometry. The vaccinia TMP kinase dimer arrangement is orthogonal and not antiparallel as in human enzyme. This different monomer orientation is related to the presence of a canal connecting the edge of the dimer interface to the TMP base binding pocket. Consequently, the pox enzyme accommodates nucleotides with bulkier bases, like brivudin monophosphate and dGMP; these are efficiently phosphorylated and stabilize the enzyme. The brivudin monophosphate-bound structure explains the structural basis for this specificity, opening the way to the rational development of specific antipox agents that may also be suitable for poxvirus TMP kinase gene-based chemotherapy of cancer.

Epstein-barr virus encodes three bona fide ubiquitin-specific proteases

Manipulation of the ubiquitin proteasome system (UPS) is emerging as a common theme in viral pathogenesis. Some viruses have been shown to encode functional homologs of UPS enzymes, suggesting that a systematic identification of these products may provide new insights into virus-host cell interactions. Ubiquitin-specific proteases, collectively known as deubiquitinating enzymes (DUBs), regulate the activity of the UPS by hydrolyzing ubiquitin peptide or isopeptide bonds. The prediction of viral DUBs based on sequence similarity with known enzymes is hampered by the diversity of viral genomes. In this study sequence alignments, pattern searches, and hidden Markov models were developed for the conserved C- and H-boxes of the known DUB families and used to search the open reading frames (ORFs) of Epstein-Barr virus (EBV), a large gammaherpesvirus that has been implicated in the pathogenesis of a broad spectrum of human malignancies of lymphoid and epithelial cell origin. The searches identified a limited number of EBV ORFs that contain putative DUB catalytic domains. DUB activity was confirmed by functional assays and mutation analysis for three high scoring candidates, supporting the usefulness of this bioinformatics approach in predicting distant homologues of cellular enzymes.

Intrinsic motions along an enzymatic reaction trajectory

The mechanisms by which enzymes achieve extraordinary rate acceleration and specificity have long been of key interest in biochemistry. It is generally recognized that substrate binding coupled to conformational changes of the substrate-enzyme complex aligns the reactive groups in an optimal environment for efficient chemistry. Although chemical mechanisms have been elucidated for many enzymes, the question of how enzymes achieve the catalytically competent state has only recently become approachable by experiment and computation. Here we show crystallographic evidence for conformational substates along the trajectory towards the catalytically competent 'closed' state in the ligand-free form of the enzyme adenylate kinase. Molecular dynamics simulations indicate that these partially closed conformations are sampled in nanoseconds, whereas nuclear magnetic resonance and single-molecule fluorescence resonance energy transfer reveal rare sampling of a fully closed conformation occurring on the microsecond-to-millisecond timescale. Thus, the larger-scale motions in substrate-free adenylate kinase are not random, but preferentially follow the pathways that create the configuration capable of proficient chemistry. Such preferred directionality, encoded in the fold, may contribute to catalysis in many enzymes.

Identification of rare slipknots in proteins and their implications for stability and folding

Among the thousands of known three-dimensional protein folds, only a few have been found whose backbones are in knotted configurations. The rarity of knotted proteins has important implications for how natural proteins reach their natively folded states. Proteins with such unusual features offer unique opportunities for studying the relationships between structure, folding, and stability. Here we report the identification of a unique slipknot feature in the fold of a well-known thermostable protein, alkaline phosphatase. A slipknot is created when a knot is formed by part of a protein chain, after which the backbone doubles back so that the entire structure becomes unknotted in a mathematical sense. Slipknots are therefore not detected by computational tests that look for knots in complete protein structures. A computational survey looking specifically for slipknots in the Protein Data Bank reveals a few other instances in addition to alkaline phosphatase. Unexpected similarities are noted among some of the proteins identified. In addition, two transmembrane proteins are found to contain slipknots. Finally, mutagenesis experiments on alkaline phosphatase are used to probe the contribution the slipknot feature makes to thermal stability. The trends and conserved features observed in these proteins provide new insights into mechanisms of protein folding and stability.

Crystal structures of GMP kinase in complex with ganciclovir monophosphate and Ap5G

Guanosine monophosphate kinases (GMPK), by catalyzing the phosphorylation of GMP or dGMP, are of dual potential in assisting the activation of anti-viral prodrugs or as candidates for antibiotic strategies. Human GMPK is an obligate step for the activation of acyclic guanosine analogs, such as ganciclovir, which necessitate efficient phosphorylation, while GMPK from bacterial pathogens, in which this enzyme is essential, are potential targets for therapeutic inhibition. Here we analyze these two aspects of GMPK activity with the crystal structures of Escherichia coli GMPK in complex with ganciclovir-monophosphate (GCV-MP) and with a bi-substrate inhibitor, Ap5G. GCV-MP binds as GMP to the GMP-binding domain, which is identical in E. coli and human GMPKs, but unlike the natural substrate fails to stabilize the closed, catalytically-competent conformation of this domain. Comparison with GMP- and GDP-bound GMPK structures identifies the 2'hydroxyl of the ribose moiety as responsible for hooking the GMP-binding domain onto the CORE domain. Absence of this hydroxyl in GCV-MP impairs the stabilization of the active conformation, and explains why GCV-MP is phosphorylated less efficiently than GMP, but as efficiently as dGMP. In contrast, Ap5G is an efficient inhibitor of GMPK. The crystal structure shows that Ap5G locks an incompletely closed conformation of the enzyme, in which the adenine moiety is located outside its expected binding site. Instead, it binds at a subunit interface that is unique to the bacterial enzyme, which is in equilibrium between a dimeric and an hexameric form in solution. This suggests that inhibitors could be designed to bind at this interface such as to prevent nucleotide-induced domain closure. Altogether, these complexes point to domain motions as critical components to be evaluated in therapeutic strategies targeting NMP kinases, with opposite effects depending on whether efficient phosphorylation or inhibition is being sought after.

Binding mode prediction of cytochrome p450 and thymidine kinase protein-ligand complexes by consideration of water and rescoring in automated docking

The popular docking programs AutoDock, FlexX, and GOLD were used to predict binding modes of ligands in crystallographic complexes including X-ray water molecules or computationally predicted water molecules. Isoenzymes of two different enzyme systems were used, namely cytochromes P450 (n = 19) and thymidine kinases (n = 19) and three different "water" scenarios: i.e., docking (i) into water-free active sites, (ii) into active sites containing crystallographic water molecules, and (iii) into active sites containing water molecules predicted by a novel approach based on the program GRID. Docking accuracies were determined in terms of the root-mean-square deviation (RMSD) accuracy and, newly defined, in terms of the ligand catalytic site prediction (CSP) accuracy. Consideration of both X-ray and predicted water molecules and the subsequent pooling and rescoring of all solutions (generated by all three docking programs) with the SCORE scoring function significantly improved the quality of prediction of the binding modes both in terms of RMSD and CSP accuracy.