Structural and metabolic specificity of methylthiocoformycin for malarial adenosine deaminases

What make malarial adenosine deaminase from PLASMODIUM VIVAX recognise adenosine and 5'-methylthioadenosine: simulation studies

Malaria is a life-threatening disease in humans caused by Plasmodium parasites. Plasmodium vivax (P. vivax) is one of the prevalent species found worldwide. An increase in an anti-malarial drug resistance suggests the urgent need for new drugs. Zn²⁺-containing adenosine deaminase (ADA) is a promising drug target because the ADA inhibition is fatal to the parasite. Malarial ADA accepts both adenosine (ADN) and 5'-methylthioadenosine (MTA) as substrates. The understanding of the substrate binding becomes crucial for an anti-malarial drug development. In this work, ADA from P. vivax (pvADA) is of interest due to its prevalence worldwide. The binding of ADN and MTA are studied here using Molecular Dynamics (MD) simulations. Upon binding, the open and closed states of pvADA are captured. The displacement of α7, linking loops of β3/α12, β4/α13, β5/α15, and α10/α11 is involved in the cavity closure and opening. Also, the inappropriate substrate orientation induces a failure in a complete cavity closure. Interactions with D46, D172, S280, D310, and D311 are important for ADN binding, whereas only hydrogen bonds with D172 and D311 are sufficient to anchor MTA inside the pocket. No Zn²⁺-coordinated histidine residues is acquired for substrate binding. D172 is found to play a role in ribose moiety recognition, while D311 is crucial for trapping the amine group of an adenine ring towards the Zn²⁺ site. Comparing between ADN and MTA, the additional interaction between D310 and an amine nitrogen on ADN supports a tighter fit that may facilitate the deamination.Communicated by Ramaswamy H. Sarma.

Transition Path Sampling Based Calculations of Free Energies for Enzymatic Reactions: The Case of Human Methionine Adenosyl Transferase and Plasmodium vivax Adenosine Deaminase

Transition path sampling (TPS) is widely used for the calculations of reaction rates, transition state structures, and reaction coordinates of condensed phase systems. Here we discuss a scheme for the calculation of free energies using the ensemble of TPS reactive trajectories in combination with a window-based sampling technique for enzyme-catalyzed reactions. We calculate the free energy profiles of the reactions catalyzed by the human methionine S-adenosyltransferase (MAT2A) enzyme and the Plasmodium vivax adenosine deaminase (pvADA) enzyme to assess the accuracy of this method. MAT2A catalyzes the formation of S-adenosine-l-methionine following a S_N2 mechanism, and using our method, we estimate the free energy barrier for this reaction to be 16 kcal mol^-1, which is in excellent agreement with the experimentally measured activation energy of 17.27 kcal mol^-1. Furthermore, for the pvADA enzyme-catalyzed reaction we estimate a free energy barrier of 21 kcal mol^-1, and the calculated free energy profile is similar to that predicted from experimental observations. Calculating free energies by employing our simple method within TPS provides significant advantages over methods such as umbrella sampling because it is free from any applied external bias, is accurate compared to experimental measurements, and has a reasonable computational cost.

TLR8 is activated by 5'-methylthioinosine, a Plasmodium falciparum-derived intermediate of the purine salvage pathway

The innate immune recognition of the malaria-causing pathogen Plasmodium falciparum (P. falciparum) is not fully explored. Here, we identify the nucleoside 5'-methylthioinosine (MTI), a Plasmodium-specific intermediate of the purine salvage pathway, as a pathogen-derived Toll-like receptor 8 (TLR8) agonist. Co-incubation of MTI with the TLR8 enhancer poly(dT) as well as synthetic or P. falciparum-derived RNA strongly increase its stimulatory activity. Of note, MTI generated from methylthioadenosine (MTA) by P. falciparum lysates activates TLR8 when MTI metabolism is inhibited by immucillin targeting the purine nucleoside phosphorylase (PfPNP). Importantly, P. falciparum-infected red blood cells incubated with MTI or cultivated with MTA and immucillin lead to TLR8-dependent interleukin-6 (IL-6) production in human monocytes. Our data demonstrate that the nucleoside MTI is a natural human TLR8 ligand with possible in vivo relevance for innate sensing of P. falciparum.

Development of a continuous assay for high throughput screening to identify inhibitors of the purine salvage pathway in Plasmodium falciparum

Malaria, an infectious disease caused by protozoan parasites from the genus Plasmodium, represents a serious global health threat. The continued emergence of drug resistant strains has severely decreased current antimalarial drug efficacy and led to a perpetual race for drug discovery. Most protozoan parasites, including Plasmodium spp., are unable to synthesize purines de novo and instead rely on an essential purine salvage pathway for acquisition of purines from the infected host. Because purines are essential for Plasmodium growth and survival, the enzymes of the purine salvage pathway represent promising targets for drug discovery. Target-based high-throughput screening (HTS) assays traditionally focus on a single target, which severely limits the screening power of this type of approach. To circumvent this limitation, we have reconstituted the purine salvage pathway from Plasmodium falciparum in an assay combining four drug targets. This assay was developed for HTS and optimized to detect partial inhibition of any of the four enzymes in the pathway. Inhibitors of several enzymes in the pathway were identified in a pilot screen, with several compounds exhibiting effective inhibition when provided in micromolar amounts.

1,3-Diazepine: A privileged scaffold in medicinal chemistry

Privileged structures have been widely used as effective templates for drug discovery. While benzo-1,4-diazepine constitutes the first historical example of such a structure, the 1,3 analogue is just as rich in terms of applications in medicinal chemistry. The 1,3-diazepine moiety is present in numerous biological active compounds including natural products, and is used to design compounds displaying a large range of biological activities. It is present in the clinically used anticancer compound pentostatin, in several recent FDA approved β-lactamase inhibitors (e.g., avibactam) and also in coformycin, a natural product known as a ring-expanded purine analogue displaying antiviral and anticancer activities. Several other 1,3-diazepine containing compounds have entered into clinical trials. This heterocyclic structure has been and is still widely used in medicinal chemistry to design enzyme inhibitors, GPCR ligands, and so forth. This review endeavours to highlight the main use of the 1,3-diazepine scaffold and its derivatives, and their applications in medicinal chemistry, drug design, and therapy. We will focus more particularly on the development of enzyme inhibitors incorporating this scaffold, with a strong emphasis on the molecular interactions involved in the inhibition mechanism.

Development of a target identification approach using native mass spectrometry

A key step in the development of new pharmaceutical drugs is the identification of the molecular target and distinguishing this from all other gene products that respond indirectly to the drug. Target identification remains a crucial process and a current bottleneck for advancing hits through the discovery pipeline. Here we report a method, that takes advantage of the specific detection of protein-ligand complexes by native mass spectrometry (MS) to probe the protein partner of a ligand in an untargeted method. The key advantage is that it uses unmodified small molecules for binding and, thereby, it does not require labelled ligands and is not limited by the chemistry required to tag the molecule. We demonstrate the use of native MS to identify known ligand-protein interactions in a protein mixture under various experimental conditions. A protein-ligand complex was successfully detected between parthenolide and thioredoxin (PfTrx) in a five-protein mixture, as well as when parthenolide was mixed in a bacterial cell lysate spiked with PfTrx. We provide preliminary data that native MS could be used to identify binding targets for any small molecule.

Crystal structure of Plasmodium falciparum adenosine deaminase reveals a novel binding pocket for inosine

Plasmodium falciparum (Pf), a malarial pathogen, can only synthesize purine nucleotides employing a salvage pathway because it lacks de novo biosynthesis. Adenosine deaminase (ADA), one of the three purine salvage enzymes, catalyzes the irreversible hydrolytic deamination of adenosine to inosine, which is further converted to GMP and AMP for DNA/RNA production. In addition to adenosine conversion, Plasmodium ADA also catalyzes the conversion of 5'-methylthioadenosine, derived from polyamine biosynthesis, into 5'-methylthioinosine whereas the human enzyme is not capable of this function. Here we report the crystal structure of a surface engineered PfADA at a resolution of 2.48 Å, together with results on kinetic studies of PfADA wild-type and active site variants. The structure reveals a novel inosine binding pocket linked to a distinctive PfADA substructure (residues 172-179) derived from a non-conserved gating helix loop (172-188) in Plasmodium spp. and other ADA enzymes. Variants of PfADA and human (h) ADA active site amino acids were generated in order to study their role in catalysis, including PfADA- Phe136, -Thr174, -Asp176, and -Leu179, and hADA-Met155, equivalent to PfADA-Asp176. PfADA-Leu179His showed no effect on kinetic parameters. However, kinetic results of PfADA-Asp176Met/Ala mutants and hADA-Met155Asp/Ala showed that the mutation reduced adenosine and 5'-methylthioadenosine substrate affinity in PfADA and k_cat in hADA, thereby reducing catalytic efficiency of the enzyme. Phe136Leu mutant showed increased K_m (>10-fold) for both substrates whereas Thr174Ile/Ala only affected 5'-methylthioadenosine binding affinity. Together, the structure with the novel inosine binding pocket and the kinetic data provide insights for rational design of inhibitors against PfADA.

Enzymatic Transition States and Drug Design

Transition state theory teaches that chemically stable mimics of enzymatic transition states will bind tightly to their cognate enzymes. Kinetic isotope effects combined with computational quantum chemistry provides enzymatic transition state information with sufficient fidelity to design transition state analogues. Examples are selected from various stages of drug development to demonstrate the application of transition state theory, inhibitor design, physicochemical characterization of transition state analogues, and their progress in drug development.

Novel Plasmodium falciparum metabolic network reconstruction identifies shifts associated with clinical antimalarial resistance

Background

Malaria remains a major public health burden and resistance has emerged to every antimalarial on the market, including the frontline drug, artemisinin. Our limited understanding of Plasmodium biology hinders the elucidation of resistance mechanisms. In this regard, systems biology approaches can facilitate the integration of existing experimental knowledge and further understanding of these mechanisms.

Results

Here, we developed a novel genome-scale metabolic network reconstruction, iPfal17, of the asexual blood-stage P. falciparum parasite to expand our understanding of metabolic changes that support resistance. We identified 11 metabolic tasks to evaluate iPfal17 performance. Flux balance analysis and simulation of gene knockouts and enzyme inhibition predict candidate drug targets unique to resistant parasites. Moreover, integration of clinical parasite transcriptomes into the iPfal17 reconstruction reveals patterns associated with antimalarial resistance. These results predict that artemisinin sensitive and resistant parasites differentially utilize scavenging and biosynthetic pathways for multiple essential metabolites, including folate and polyamines. Our findings are consistent with experimental literature, while generating novel hypotheses about artemisinin resistance and parasite biology. We detect evidence that resistant parasites maintain greater metabolic flexibility, perhaps representing an incomplete transition to the metabolic state most appropriate for nutrient-rich blood.

Conclusion

Using this systems biology approach, we identify metabolic shifts that arise with or in support of the resistant phenotype. This perspective allows us to more productively analyze and interpret clinical expression data for the identification of candidate drug targets for the treatment of resistant parasites.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-017-3905-1) contains supplementary material, which is available to authorized users.

Transition-state inhibitors of purine salvage and other prospective enzyme targets in malaria

Malaria is a leading cause of human death within the tropics. The gradual generation of drug resistance imposes an urgent need for the development of new and selective antimalarial agents. Kinetic isotope effects coupled to computational chemistry have provided the relevant details on geometry and charge of enzymatic transition states to facilitate the design of transition-state analogs. These features have been reproduced into chemically stable mimics through synthetic chemistry, generating inhibitors with dissociation constants in the pico- to femto-molar range. Transition-state analogs are expected to contribute to the control of malaria.

Deamination of 6-aminodeoxyfutalosine in menaquinone biosynthesis by distantly related enzymes

Proteins of unknown function belonging to cog1816 and cog0402 were characterized. Sav2595 from Steptomyces avermitilis MA-4680, Acel0264 from Acidothermus cellulolyticus 11B, Nis0429 from Nitratiruptor sp. SB155-2 and Dr0824 from Deinococcus radiodurans R1 were cloned, purified, and their substrate profiles determined. These enzymes were previously incorrectly annotated as adenosine deaminases or chlorohydrolases. It was shown here that these enzymes actually deaminate 6-aminodeoxyfutalosine. The deamination of 6-aminodeoxyfutalosine is part of an alternative menaquinone biosynthetic pathway that involves the formation of futalosine. 6-Aminodeoxyfutalosine is deaminated by these enzymes with catalytic efficiencies greater than 10(5) M(-1) s(-1), Km values of 0.9-6.0 μM, and kcat values of 1.2-8.6 s(-1). Adenosine, 2'-deoxyadenosine, thiomethyladenosine, and S-adenosylhomocysteine are deaminated at least an order of magnitude slower than 6-aminodeoxyfutalosine. The crystal structure of Nis0429 was determined and the substrate, 6-aminodeoxyfutalosine, was positioned in the active site on the basis of the presence of adventitiously bound benzoic acid. In this model, Ser-145 interacts with the carboxylate moiety of the substrate. The structure of Dr0824 was also determined, but a collapsed active site pocket prevented docking of substrates. A computational model of Sav2595 was built on the basis of the crystal structure of adenosine deaminase and substrates were docked. The model predicted a conserved arginine after β-strand 1 to be partially responsible for the substrate specificity of Sav2595.

Methylthioadenosine deaminase in an alternative quorum sensing pathway in Pseudomonas aeruginosa

Pseudomonas aeruginosa possesses an unusual pathway for 5'-methylthioadenosine (MTA) metabolism involving deamination to 5'-methylthioinosine (MTI) followed by N-ribosyl phosphorolysis to hypoxanthine and 5-methylthio-α-d-ribose 1-phosphate. The specific MTI phosphorylase of P. aeruginosa has been reported [Guan, R., Ho, M. C., Almo, S. C., and Schramm, V. L. (2011) Biochemistry 50, 1247-1254], and here we characterize MTA deaminase from P. aeruginosa (PaMTADA). Genomic analysis indicated the PA3170 locus to be a candidate for MTA deaminase (MTADA). Protein encoded by PA3170 was expressed and shown to deaminate MTA with 40-fold greater catalytic efficiency for MTA than for adenosine. The k(cat)/K(m) value of 1.6 × 10(7) M(-1) s(-1) for MTA is the highest catalytic efficiency known for an MTA deaminase. 5'-Methylthiocoformycin (MTCF) is a 4.8 pM transition state analogue for PaMTADA but causes no significant inhibition of human adenosine deaminase or MTA phosphorylase. MTCF is permeable to P. aeruginosa and exhibits an IC(50) of 3 nM on cellular PaMTADA activity. PaMTADA is the only activity in P. aeruginosa extracts to act on MTA. MTA and 5-methylthio-α-d-ribose are involved in quorum sensing pathways; thus, PaMTADA is a potential target for quorum sensing. The crystal structure of PaMTADA in complex with MTCF shows the transition state mimic 8(R)-hydroxyl group in contact with a catalytic site Zn(2+), the 5'-methylthio group in a hydrophobic pocket, and the transition state mimic of the diazepine ring in contact with a catalytic site Glu.

The adenosine deaminases of Plasmodium vivax and Plasmodium falciparum exhibit surprising differences in ligand specificity

Plasmodium vivax and Plasmodium falciparum cause malaria, so proteins essential for their survival in vivo are potential anti-malarial drug targets. Adenosine deaminases (ADA) catalyze the irreversible conversion of adenosine into inosine, and play a critical role in the purine salvage pathways of Plasmodia and their mammalian hosts. Currently, the number of selective inhibitors of Plasmodium ADAs is limited. One potent and widely used inhibitor of the human ADA (hADA), erythro-9-(2-hydroxy-3-nonly)adenine (EHNA), is a very weak inhibitor (K(i)=120 μM) of P. falciparum ADA (pfADA). EHNA-like compounds are thus excluded from consideration as potential inhibitors of Plasmodium ADA in general. However, EHNA activity in P. vivax ADA (pvADA) has not been reported. Here we applied computational molecular modeling to identify ligand recognition mechanisms unique to P. vivax and P. falciparum ADA. Our biochemical experiments show that EHNA is at least 60-fold more potent against pvADA (K(i)=1.9 μM) than against pfADA. The D172A pvADA mutant is bound even more tightly (K(i)=0.9 μM). These results improve our understanding of the mechanisms of ADA ligand recognition and species-selectivity, and facilitate the rational design of novel EHNA-based ADA inhibitors as anti-malarial drugs. To demonstrate a practical application of our findings we have computationally predicted a novel potential inhibitor of pvADA that will not interact with the human ADA.

Adenine aminohydrolase from Leishmania donovani: unique enzyme in parasite purine metabolism

Adenine aminohydrolase (AAH) is an enzyme that is not present in mammalian cells and is found exclusively in Leishmania among the protozoan parasites that infect humans. AAH plays a paramount role in purine metabolism in this genus by steering 6-aminopurines into 6-oxypurines. Leishmania donovani AAH is 38 and 23% identical to Saccharomyces cerevisiae AAH and human adenosine deaminase enzymes, respectively, catalyzes adenine deamination to hypoxanthine with an apparent K(m) of 15.4 μM, and does not recognize adenosine as a substrate. Western blot analysis established that AAH is expressed in both life cycle stages of L. donovani, whereas subcellular fractionation and immunofluorescence studies confirmed that AAH is localized to the parasite cytosol. Deletion of the AAH locus in intact parasites established that AAH is not an essential gene and that Δaah cells are capable of salvaging the same range of purine nucleobases and nucleosides as wild type L. donovani. The Δaah null mutant was able to infect murine macrophages in vitro and in mice, although the parasite loads in both model systems were modestly reduced compared with wild type infections. The Δaah lesion was also introduced into a conditionally lethal Δhgprt/Δxprt mutant in which viability was dependent on pharmacologic ablation of AAH by 2'-deoxycoformycin. The Δaah/Δhgprt/Δxprt triple knock-out no longer required 2'-deoxycoformycin for growth and was avirulent in mice with no persistence after a 4-week infection. These genetic studies underscore the paramount importance of AAH to purine salvage by L. donovani.

Sampling protein motion and solvent effect during ligand binding

An exhaustive description of the molecular recognition mechanism between a ligand and its biological target is of great value because it provides the opportunity for an exogenous control of the related process. Very often this aim can be pursued using high resolution structures of the complex in combination with inexpensive computational protocols such as docking algorithms. Unfortunately, in many other cases a number of factors, like protein flexibility or solvent effects, increase the degree of complexity of ligand/protein interaction and these standard techniques are no longer sufficient to describe the binding event. We have experienced and tested these limits in the present study in which we have developed and revealed the mechanism of binding of a new series of potent inhibitors of Adenosine Deaminase. We have first performed a large number of docking calculations, which unfortunately failed to yield reliable results due to the dynamical character of the enzyme and the complex role of the solvent. Thus, we have stepped up the computational strategy using a protocol based on metadynamics. Our approach has allowed dealing with protein motion and solvation during ligand binding and finally identifying the lowest energy binding modes of the most potent compound of the series, 4-decyl-pyrazolo[1,5-a]pyrimidin-7-one.

Discovery of a cytokinin deaminase

An enzyme of unknown function within the amidohydrolase superfamily was discovered to catalyze the hydrolysis of N-6-substituted adenine derivatives, several of which are cytokinins. Cytokinins are a common type of plant hormone and N-6-substituted adenines are also found as modifications to tRNA. Patl2390, from Pseudoalteromonas atlantica T6c, was shown to hydrolytically deaminate N-6-isopentenyladenine to hypoxanthine and isopentenylamine with a k(cat)/K(m) of 1.2 × 10(7) M(-1) s(-1). Additional substrates include N-6-benzyl adenine, cis- and trans-zeatin, kinetin, O-6-methylguanine, N-6-butyladenine, N-6-methyladenine, N,N-dimethyladenine, 6-methoxypurine, 6-chloropurine, and 6-thiomethylpurine. This enzyme does not catalyze the deamination of adenine or adenosine. A comparative model of Patl2390 was computed using the three-dimensional crystal structure of Pa0148 (PDB code 3PAO ) as a structural template, and docking was used to refine the model to accommodate experimentally identified substrates. This is the first identification of an enzyme that will hydrolyze an N-6-substituted side chain larger than methylamine from adenine.

Computed structures of point deletion mutants and their enzymatic activities

Point deletions in enzymes can vary in effect from negligible to complete loss of activity; however, these effects are not generally predictable. Deletions are widely observed in nature and often result in diseases such as cancer, cystic fibrosis, or osteogenesis imperfecta. Here, we have developed an algorithm to model the perturbed structures of deletion mutants with the ultimate goal of predicting their activities. The algorithm works by deleting the specified residue from the wild-type structure, creating a gap that is closed using a combination of local and global moves that change the backbone torsion angles of the protein structure. On a set of five proteins for which both wild-type and deletion mutant x-ray crystal structures are available, the algorithm produces deep, narrow energy funnels within 1.5 Å of the crystal structure for the deletion mutants. To assess the ability of our algorithm to predict activity from the predicted structures, we tested the correlation of experimental activity with several measures of the predicted structure ensemble using a set of 45 point deletions from ricin. Estimates incorporating likely prevalence of active and inactive deletion sites suggest that activity can be predicted correctly over 60% of the time from the active site root-mean squared deviation of the lowest energy predicted structures. The predictions are stronger than simple sequence organization measures, but more fundamental work is required in structure prediction and enzyme activity determination to allow consistent prediction of activity.

Pa0148 from Pseudomonas aeruginosa catalyzes the deamination of adenine

Four proteins from NCBI cog1816, previously annotated as adenosine deaminases, have been subjected to structural and functional characterization. Pa0148 (Pseudomonas aeruginosa PAO1), AAur1117 (Arthrobacter aurescens TC1), Sgx9403e, and Sgx9403g have been purified and their substrate profiles determined. Adenosine is not a substrate for any of these enzymes. All of these proteins will deaminate adenine to produce hypoxanthine with k(cat)/K(m) values that exceed 10(5) M(-1) s(-1). These enzymes will also accept 6-chloropurine, 6-methoxypurine, N-6-methyladenine, and 2,6-diaminopurine as alternate substrates. X-ray structures of Pa0148 and AAur1117 have been determined and reveal nearly identical distorted (β/α)(8) barrels with a single zinc ion that is characteristic of members of the amidohydrolase superfamily. Structures of Pa0148 with adenine, 6-chloropurine, and hypoxanthine were also determined, thereby permitting identification of the residues responsible for coordinating the substrate and product.

Mechanism of growth inhibition of intraerythrocytic stages of Plasmodium falciparum by 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR)

Purine nucleotide synthesis in Plasmodium falciparum takes place solely by the purine salvage pathway in which preformed purine base(s) are salvaged from the host and acted upon by a battery of enzymes to generate AMP and GMP. Inhibitors of this pathway have a potent effect on the in vitro growth of P. falciparum and are hence, implicated as promising leads for the development of new generation anti-malarials. Here, we describe the mechanism of inhibition of the intraerythrocytic growth of P. falciparum by the purine nucleoside precursor, 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR). Our results show that AICAR toxicity is mediated through the erythrocyte in which AICAR is phosphorylated to its nucleotide, ZMP. Further, purine metabolite labeling of the parasitized erythrocytes by [(3)H]-hypoxanthine, in the presence of AICAR, showed a significant decrease in radioactive counts in adenylate fractions but not in guanylate fractions. The most dramatic effect on parasite growth was observed when erythrocytes pretreated with AICAR were used in culture. Pretreatment of erythrocytes with AICAR led to significant intracellular accumulation of ZMP and these erythrocytes were incapable of supporting parasite growth. These results implicate that in addition to the purine salvage pathway in P. falciparum, AICAR alters the metabolic status of the erythrocytes, which inhibits parasite growth. As AICAR and ZMP are metabolites in the human serum and erythrocytes, our studies reported here throw light on their possible role in disease susceptibility, and also suggests the possibility of AICAR being a potential prophylactic or chemotherapeutic anti-malarial compound.

Reconstruction and flux-balance analysis of the Plasmodium falciparum metabolic network

Genome-scale metabolic reconstructions can serve as important tools for hypothesis generation and high-throughput data integration. Here, we present a metabolic network reconstruction and flux-balance analysis (FBA) of Plasmodium falciparum, the primary agent of malaria. The compartmentalized metabolic network accounts for 1001 reactions and 616 metabolites. Enzyme-gene associations were established for 366 genes and 75% of all enzymatic reactions. Compared with other microbes, the P. falciparum metabolic network contains a relatively high number of essential genes, suggesting little redundancy of the parasite metabolism. The model was able to reproduce phenotypes of experimental gene knockout and drug inhibition assays with up to 90% accuracy. Moreover, using constraints based on gene-expression data, the model was able to predict the direction of concentration changes for external metabolites with 70% accuracy. Using FBA of the reconstructed network, we identified 40 enzymatic drug targets (i.e. in silico essential genes), with no or very low sequence identity to human proteins. To demonstrate that the model can be used to make clinically relevant predictions, we experimentally tested one of the identified drug targets, nicotinate mononucleotide adenylyltransferase, using a recently discovered small-molecule inhibitor.