FEP Augmentation as a Means to Solve Data Paucity Problems for Machine Learning in Chemical Biology.

In the realm of medicinal chemistry, the primary objective is to swiftly optimize a multitude of chemical properties of a set of compounds to yield a clinical candidate poised for clinical trials. In recent years, two computational techniques, machine learning (ML) and physics-based methods, have evolved substantially and are now frequently incorporated into the medicinal chemist's toolbox to enhance the efficiency of both hit optimization and candidate design. Both computational methods come with their own set of limitations, and they are often used independently of each other. ML's capability to screen extensive compound libraries expediently is tempered by its reliance on quality data, which can be scarce especially during early-stage optimization. Contrarily, physics-based approaches like free energy perturbation (FEP) are frequently constrained by low throughput and high cost by comparison; however, physics-based methods are capable of making highly accurate binding affinity predictions. In this study, we harnessed the strength of FEP to overcome data paucity in ML by generating virtual activity data sets which then inform the training of algorithms. Here, we show that ML algorithms trained with an FEP-augmented data set could achieve comparable predictive accuracy to data sets trained on experimental data from biological assays. Throughout the paper, we emphasize key mechanistic considerations that must be taken into account when aiming to augment data sets and lay the groundwork for successful implementation. Ultimately, the study advocates for the synergy of physics-based methods and ML to expedite the lead optimization process. We believe that the physics-based augmentation of ML will significantly benefit drug discovery, as these techniques continue to evolve.

Selective targeting of CD38 hydrolase and cyclase activity as an approach to immunostimulation.

The ectoenzyme CD38 is highly expressed on the surface of mature immune cells, where they are a marker for cell activation, and also on the surface of multiple tumor cells such as multiple myeloma (MM). CD38-targeted monoclonal antibodies (MABs) such as daratumumab and isatuximab bind to CD38 and promote cancer cell death by stimulating the antitumor immune response. Although MABs are achieving unprecedented success in a percentage of cases, high rates of resistance limit their efficacy. Formation of the immunosuppressive intermediate adenosine is a major route by which this resistance is mediated. Thus there is an urgent need for small molecule agents that boost the immune response in T-cells. Importantly, CD38 is a dual-function enzyme, serving as a hydrolase and a nicotinamide adenine dinucleotide (NAD+) cyclase, and both of these activities promote immunosuppression. We have employed virtual and physical screening to identify novel compounds that are selective for either the hydrolase or the cyclase activity of CD38, and have demonstrated that these compounds activate T cells in vitro. We are currently optimizing these inhibitors for use in immunotherapy. These small molecule inhibitors of the CD38-hydrolase or cyclase activity can serve as chemical probes to determine the mechanism by which CD38 promotes resistance to MAB therapy, and could become novel and effective therapeutic agents that produce immunostimulatory effects. Our studies have identified the first small molecule inhibitors of CD38 specifically for use as immunostimulants.

Distinct GluN1 and GluN2 Structural Determinants for Subunit-Selective Positive Allosteric Modulation of N-Methyl-d-aspartate Receptors.

N-Methyl-d-aspartate receptors (NMDARs) are ionotropic ligand-gated glutamate receptors that mediate fast excitatory synaptic transmission in the central nervous system (CNS). Several neurological disorders may involve NMDAR hypofunction, which has driven therapeutic interest in positive allosteric modulators (PAMs) of NMDAR function. Here we describe modest changes to the tetrahydroisoquinoline scaffold of GluN2C/GluN2D-selective PAMs that expands activity to include GluN2A- and GluN2B-containing recombinant and synaptic NMDARs. These new analogues are distinct from GluN2C/GluN2D-selective compounds like (+)-(3-chlorophenyl)(6,7-dimethoxy-1-((4-methoxyphenoxy)methyl)-3,4-dihydroisoquinolin-2(1H)-yl)methanone (CIQ) by virtue of their subunit selectivity, molecular determinants of action, and allosteric regulation of agonist potency. The (S)-enantiomers of two analogues (EU1180-55, EU1180-154) showed activity at NMDARs containing all subunits (GluN2A, GluN2B, GluN2C, GluN2D), whereas the (R)-enantiomers were primarily active at GluN2C- and GluN2D-containing NMDARs. Determination of the actions of enantiomers on triheteromeric receptors confirms their unique pharmacology, with greater activity of (S) enantiomers at GluN2A/GluN2D and GluN2B/GluN2D subunit combinations than (R) enantiomers. Evaluation of the (S)-EU1180-55 and EU1180-154 response of chimeric kainate/NMDA receptors revealed structural determinants of action within the pore-forming region and associated linkers. Scanning mutagenesis identified structural determinants within the GluN1 pre-M1 and M1 regions that alter the activity of (S)-EU1180-55 but not (R)-EU1180-55. By contrast, mutations in pre-M1 and M1 regions of GluN2D perturb the actions of only the (R)-EU1180-55 but not the (S) enantiomer. Molecular modeling supports the idea that the (S) and (R) enantiomers interact distinctly with GluN1 and GluN2 pre-M1 regions, suggesting that two distinct sites exist for these NMDAR PAMs, each of which has different functional effects.

Negative allosteric modulation of GluN1/GluN3 NMDA receptors.

NMDA receptors are ligand-gated ion channels that mediate excitatory neurotransmission. Most native NMDA receptors are tetrameric assemblies of two glycine-binding GluN1 and two glutamate-binding GluN2 subunits. Co-assembly of the glycine-binding GluN1 with glycine-binding GluN3 subunits (GluN3A-B) creates glycine activated receptors that possess strikingly different functional and pharmacological properties compared to GluN1/GluN2 NMDA receptors. The role of GluN1/GluN3 receptors in neuronal function remains unknown, in part due to lack of pharmacological tools with which to explore their physiological roles. We have identified the negative allosteric modulator EU1180-438, which is selective for GluN1/GluN3 receptors over GluN1/GluN2 NMDA receptors, AMPA, and kainate receptors. EU1180-438 is also inactive at GABA, glycine, and P2X receptors, but displays inhibition of some nicotinic acetylcholine receptors. Furthermore, we demonstrate that EU1180-438 produces robust inhibition of glycine-activated current responses mediated by native GluN1/GluN3A receptors in hippocampal CA1 pyramidal neurons. EU1180-438 is a non-competitive antagonist with activity that is independent of membrane potential (i.e. voltage-independent), glycine concentration, and extracellular pH. Non-stationary fluctuation analysis of neuronal current responses provided an estimated weighted mean unitary conductance of 6.1 pS for GluN1/GluN3A channels, and showed that EU1180-438 has no effect on conductance. Site-directed mutagenesis suggests that structural determinants of EU1180-438 activity reside near a short pre-M1 helix that lies parallel to the plane of the membrane below the agonist binding domain. These findings demonstrate that structural differences between GluN3 and other glutamate receptor subunits can be exploited to generate subunit-selective ligands with utility in exploring the roles GluN3 in neuronal function.

Accelerated Discovery of Potent Fusion Inhibitors for Respiratory Syncytial Virus.

A series of five benzimidazole-based compounds were identified using a machine learning algorithm as potential inhibitors of the respiratory syncytial virus (RSV) fusion protein. These compounds were synthesized, and compound 2 in particular exhibited excellent in vitro potency with an EC50 value of 5 nM. This new scaffold was then further refined leading to the identification of compound 44, which exhibited a 10-fold improvement in activity with an EC50 value of 0.5 nM.

NMDA receptor channel gating control by the pre-M1 helix.

The NMDA receptor (NMDAR) is an ionotropic glutamate receptor formed from the tetrameric assembly of GluN1 and GluN2 subunits. Within the flexible linker between the agonist binding domain (ABD) and the M1 helix of the pore-forming transmembrane helical bundle lies a two-turn, extracellular pre-M1 helix positioned parallel to the plasma membrane and in van der Waals contact with the M3 helix thought to constitute the channel gate. The pre-M1 helix is tethered to the bilobed ABD, where agonist-induced conformational changes initiate activation. Additionally, it is a locus for de novo mutations associated with neurological disorders, is near other disease-associated de novo sites within the transmembrane domain, and is a structural determinant of subunit-selective modulators. To investigate the role of the pre-M1 helix in channel gating, we performed scanning mutagenesis across the GluN2A pre-M1 helix and recorded whole-cell macroscopic and single channel currents from HEK293 cell-attached patches. We identified two residues at which mutations perturb channel open probability, the mean open time, and the glutamate deactivation time course. We identified a subunit-specific network of aromatic amino acids located in and around the GluN2A pre-M1 helix to be important for gating. Based on these results, we are able to hypothesize about the role of the pre-M1 helix in other NMDAR subunits based on sequence and structure homology. Our results emphasize the role of the pre-M1 helix in channel gating, implicate the surrounding amino acid environment in this mechanism, and suggest unique subunit-specific contributions of pre-M1 helices to GluN1 and GluN2 gating.

Error Tolerance of Machine Learning Algorithms across Contemporary Biological Targets.

Machine learning continues to make strident advances in the prediction of desired properties concerning drug development. Problematically, the efficacy of machine learning in these arenas is reliant upon highly accurate and abundant data. These two limitations, high accuracy and abundance, are often taken together; however, insight into the dataset accuracy limitation of contemporary machine learning algorithms may yield insight into whether non-bench experimental sources of data may be used to generate useful machine learning models where there is a paucity of experimental data. We took highly accurate data across six kinase types, one GPCR, one polymerase, a human protease, and HIV protease, and intentionally introduced error at varying population proportions in the datasets for each target. With the generated error in the data, we explored how the retrospective accuracy of a Naïve Bayes Network, a Random Forest Model, and a Probabilistic Neural Network model decayed as a function of error. Additionally, we explored the ability of a training dataset with an error profile resembling that produced by the Free Energy Perturbation method (FEP+) to generate machine learning models with useful retrospective capabilities. The categorical error tolerance was quite high for a Naïve Bayes Network algorithm averaging 39% error in the training set required to lose predictivity on the test set. Additionally, a Random Forest tolerated a significant degree of categorical error introduced into the training set with an average error of 29% required to lose predictivity. However, we found the Probabilistic Neural Network algorithm did not tolerate as much categorical error requiring an average of 20% error to lose predictivity. Finally, we found that a Naïve Bayes Network and a Random Forest could both use datasets with an error profile resembling that of FEP+. This work demonstrates that computational methods of known error distribution like FEP+ may be useful in generating machine learning models not based on extensive and expensive in vitro-generated datasets.

A Machine Learning Approach for Predicting HIV Reverse Transcriptase Mutation Susceptibility of Biologically Active Compounds.

HIV resistance emerging against antiretroviral drugs represents a great threat to the continued prolongation of the lifespans of HIV-infected patients. Therefore, methods capable of predicting resistance susceptibility in the development of compounds are in great need. By targeting the major reverse transcription residues Y181, K103, and L100, we used the biological activities of compounds against these enzymes and the wild-type reverse transcriptase to create Naïve Bayes Networks. Through this machine learning approach, we could predict, with high accuracy, whether a compound would be susceptible to a loss of potency due to resistance. Also, we could perfectly predict retrospectively whether compounds would be susceptible to both a K103 mutant RT and a Y181 mutant RT. In the study presented here, our method outperformed a traditional molecular mechanics approach. This method should be of broad interest beyond drug discovery efforts, and serves to expand the utility of machine learning for the prediction of physical, chemical, or biological properties using the vast information available in the literature.

The Bioactive Protein-Ligand Conformation of GluN2C-Selective Positive Allosteric Modulators Bound to the NMDA Receptor.

N-methyl-d-aspartate (NMDA) receptors are ligand-gated, cation-selective channels that mediate a slow component of excitatory synaptic transmission. Subunit-selective positive allosteric modulators of NMDA receptor function have therapeutically relevant effects on multiple processes in the brain. A series of pyrrolidinones, such as PYD-106, that selectively potentiate NMDA receptors that contain the GluN2C subunit have structural determinants of activity that reside between the GluN2C amino terminal domain and the GluN2C agonist binding domain, suggesting a unique site of action. Here we use molecular biology and homology modeling to identify residues that line a candidate binding pocket for GluN2C-selective pyrrolidinones. We also show that occupancy of only one site in diheteromeric receptors is required for potentiation. Both GluN2A and GluN2B can dominate the sensitivity of triheteromeric receptors to eliminate the actions of pyrrolidinones, thus rendering this series uniquely sensitive to subunit stoichiometry. We experimentally identified NMR-derived conformers in solution, which combined with molecular modeling allows the prediction of the bioactive binding pose for this series of GluN2C-selective positive allosteric modulators of NMDA receptors. These data advance our understanding of the site and nature of the ligand-protein interaction for GluN2C-selective positive allosteric modulators for NMDA receptors.

The Structure-Activity Relationship of a Tetrahydroisoquinoline Class of N-Methyl-d-Aspartate Receptor Modulators that Potentiates GluN2B-Containing N-Methyl-d-Aspartate Receptors.

We have identified a series of positive allosteric NMDA receptor (NMDAR) modulators derived from a known class of GluN2C/D-selective tetrahydroisoquinoline analogues that includes CIQ. The prototypical compound of this series contains a single isopropoxy moiety in place of the two methoxy substituents present in CIQ. Modifications of this isopropoxy-containing scaffold led to the identification of analogues with enhanced activity at the GluN2B subunit. We identified molecules that potentiate the response of GluN2B/GluN2C/GluN2D, GluN2B/GluN2C, and GluN2C/GluN2D-containing NMDARs to maximally effective concentrations of agonist. Multiple compounds potentiate the response of NMDARs with submicromolar EC50 values. Analysis of enantiomeric pairs revealed that the S-(-) enantiomer is active at the GluN2B, GluN2C, and/or GluN2D subunits, whereas the R-(+) enantiomer is only active at GluN2C/D subunits. These results provide a starting point for the development of selective positive allosteric modulators for GluN2B-containing receptors.

Functional Evaluation of a De Novo GRIN2A Mutation Identified in a Patient with Profound Global Developmental Delay and Refractory Epilepsy.

The N-methyl-d-aspartate receptor (NMDAR), a ligand-gated ionotropic glutamate receptor, plays important roles in normal brain development and a wide range of neurologic disorders, including epilepsy. Here, we evaluate for the first time the functional properties of a de novo GRIN2A missense mutation (p.M817V) in the pre-M4 linker in a child with profound global developmental delay and refractory epilepsy. Electrophysiologic recordings revealed that the mutant GluN2A(M817V)-containing receptors showed enhanced agonist potency, reduced sensitivity to endogenous negative inhibitors (Mg2+, proton, and zinc), prolonged synaptic-like response time course, increased single-channel mean open time, and increased channel open probability. These results suggest that the gain-of-function M817V mutation causes overactivation of NMDAR and drives neuronal hyperexcitability, which may contribute to the patient's observed epileptic phenotype. Molecular modeling of the closed channel conformation reveals that this mutation weakens the interaction between GluN2 transmembrane helix M4 and two GluN1 transmembrane helices, and increases atomic fluctuation or movement of the pre-M1 region of GluN1 subunit, suggesting a mechanism by which channel function is enhanced. The functional changes of this mutation on agonist potency occur when the mutation is introduced into all other GluN2 subunits, suggesting a conserved role of this residue in control of NMDAR function through interactions of membrane spanning GluN2 and GluN1 helices. A number of NMDAR-targeted drugs including U.S. Food and Drug Association-approved NMDAR channel blockers were evaluated for their ability to inhibit receptors containing GluN2A(M817V) as a first step to exploring the potential for rescue pharmacology and personalized medicine.

Mechanistic Insight into NMDA Receptor Dysregulation by Rare Variants in the GluN2A and GluN2B Agonist Binding Domains.

Epilepsy and intellectual disability are associated with rare variants in the GluN2A and GluN2B (encoded by GRIN2A and GRIN2B) subunits of the N-methyl-D-aspartate receptor (NMDAR), a ligand-gated ion channel with essential roles in brain development and function. By assessing genetic variation across GluN2 domains, we determined that the agonist binding domain, transmembrane domain, and the linker regions between these domains were particularly intolerant to functional variation. Notably, the agonist binding domain of GluN2B exhibited significantly more variation intolerance than that of GluN2A. To understand the ramifications of missense variation in the agonist binding domain, we investigated the mechanisms by which 25 rare variants in the GluN2A and GluN2B agonist binding domains dysregulated NMDAR activity. When introduced into recombinant human NMDARs, these rare variants identified in individuals with neurologic disease had complex, and sometimes opposing, consequences on agonist binding, channel gating, receptor biogenesis, and forward trafficking. Our approach combined quantitative assessments of these effects to estimate the overall impact on synaptic and non-synaptic NMDAR function. Interestingly, similar neurologic diseases were associated with both gain- and loss-of-function variants in the same gene. Most rare variants in GluN2A were associated with epilepsy, whereas GluN2B variants were associated with intellectual disability with or without seizures. Finally, discerning the mechanisms underlying NMDAR dysregulation by these rare variants allowed investigations of pharmacologic strategies to correct NMDAR function.

A novel inhibitor of Plasmodium falciparum spermidine synthase: a twist in the tail.

BACKGROUND: Plasmodium falciparum is the most pathogenic of the human malaria parasite species and a major cause of death in Africa. It's resistance to most of the current drugs accentuates the pressing need for new chemotherapies. Polyamine metabolism of the parasite is distinct from the human pathway making it an attractive target for chemotherapeutic development. Plasmodium falciparum spermidine synthase (PfSpdS) catalyzes the synthesis of spermidine and spermine. It is a major polyamine flux-determining enzyme and spermidine is a prerequisite for the post-translational activation of P. falciparum eukaryotic translation initiation factor 5A (elF5A). The most potent inhibitors of eukaryotic SpdS's are not specific for PfSpdS. METHODS: 'Dynamic' receptor-based pharmacophore models were generated from published crystal structures of SpdS with different ligands. This approach takes into account the inherent flexibility of the active site, which reduces the entropic penalties associated with ligand binding. Four dynamic pharmacophore models were developed and two inhibitors, (1R,4R)-(N1-(3-aminopropyl)-trans-cyclohexane-1,4-diamine (compound 8) and an analogue, N-(3-aminopropyl)-cyclohexylamine (compound 9), were identified. RESULTS: A crystal structure containing compound 8 was solved and confirmed the in silico prediction that its aminopropyl chain traverses the catalytic centre in the presence of the byproduct of catalysis, 5'-methylthioadenosine. The IC50 value of compound 9 is in the same range as that of the most potent inhibitors of PfSpdS, S-adenosyl-1,8-diamino-3-thio-octane (AdoDATO) and 4MCHA and 100-fold lower than that of compound 8. Compound 9 was originally identified as a mammalian spermine synthase inhibitor and does not inhibit mammalian SpdS. This implied that these two compounds bind in an orientation where their aminopropyl chains face the putrescine binding site in the presence of the substrate, decarboxylated S-adenosylmethionine. The higher binding affinity and lower receptor strain energy of compound 9 compared to compound 8 in the reversed orientation explained their different IC50 values. CONCLUSION: The specific inhibition of PfSpdS by compound 9 is enabled by its binding in the additional cavity normally occupied by spermidine when spermine is synthesized. This is the first time that a spermine synthase inhibitor is shown to inhibit PfSpdS, which provides new avenues to explore for the development of novel inhibitors of PfSpdS.

Structural determinants and mechanism of action of a GluN2C-selective NMDA receptor positive allosteric modulator.

NMDA receptors are tetrameric complexes of GluN1, GluN2A-D, and GluN3A-B subunits and are involved in normal brain function and neurologic disorders. We identified a novel class of stereoselective pyrrolidinone (PYD) positive allosteric modulators for GluN2C-containing NMDA receptors, exemplified by methyl 4-(3-acetyl-4-hydroxy-1-[2-(2-methyl-1H-indol-3-yl)ethyl]-5-oxo-2,5-dihydro-1H-pyrrol-2-yl)benzoate. Here we explore the site and mechanism of action of a prototypical analog, PYD-106, which at 30 µM does not alter responses of NMDA receptors containing GluN2A, GluN2B, and GluN2D and has no effect on AMPA [α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid] and kainate receptors. Coapplication of 50 µM PYD-106 with a maximally effective concentration of glutamate and glycine increases the response of GluN1/GluN2C NMDA receptors in HEK-293 cells to 221% of that obtained in the absence of PYD (taken as 100%). Evaluation of the concentration dependence of this enhancement revealed an EC50 value for PYD of 13 µM. PYD-106 increased opening frequency and open time of single channel currents activated by maximally effective concentrations of agonist but only had modest effects on glutamate and glycine EC50. PYD-106 selectively enhanced the responses of diheteromeric GluN1/GluN2C receptors but not triheteromeric GluN1/GluN2A/GluN2C receptors. Inclusion of residues encoded by GluN1-exon 5 attenuated the effects of PYD. Three GluN2C residues (Arg194, Ser470, Lys470), at which mutagenesis virtually eliminated PYD function, line a cavity at the interface of the ligand binding and the amino terminal domains in a homology model of GluN1/GluN2C built from crystallographic data on GluN1/GluN2B. We propose that this domain interface constitutes a new allosteric modulatory site on the NMDA receptor.

Novel S-adenosyl-L-methionine decarboxylase inhibitors as potent antiproliferative agents against intraerythrocytic Plasmodium falciparum parasites.

S-adenosyl-l-methionine decarboxylase (AdoMetDC) in the polyamine biosynthesis pathway has been identified as a suitable drug target in Plasmodium falciparum parasites, which causes the most lethal form of malaria. Derivatives of an irreversible inhibitor of this enzyme, 5'-{[(Z)-4-amino-2-butenyl]methylamino}-5'-deoxyadenosine (MDL73811), have been developed with improved pharmacokinetic profiles and activity against related parasites, Trypanosoma brucei. Here, these derivatives were assayed for inhibition of AdoMetDC from P. falciparum parasites and the methylated derivative, 8-methyl-5'-{[(Z)-4-aminobut-2-enyl]methylamino}-5'-deoxyadenosine (Genz-644131) was shown to be the most active. The in vitro efficacy of Genz-644131 was markedly increased by nanoencapsulation in immunoliposomes, which specifically targeted intraerythrocytic P. falciparum parasites.

Discovery-2: an interactive resource for the rational selection and comparison of putative drug target proteins in malaria.

BACKGROUND: Drug resistance to anti-malarial compounds remains a serious problem, with resistance to newer pharmaceuticals developing at an alarming rate. The development of new anti-malarials remains a priority, and the rational selection of putative targets is a key element of this process. Discovery-2 is an update of the original Discovery in silico resource for the rational selection of putative drug target proteins, enabling researchers to obtain information for a protein which may be useful for the selection of putative drug targets, and to perform advanced filtering of proteins encoded by the malaria genome based on a series of molecular properties. METHODS: An updated in silico resource has been developed where researchers are able to mine information on malaria proteins and predicted ligands, as well as perform comparisons to the human and mosquito host characteristics. Protein properties used include: domains, motifs, EC numbers, GO terms, orthologs, protein-protein interactions, protein-ligand interactions. Newly added features include drugability measures from ChEMBL, automated literature relations and links to clinical trial information. Searching by chemical structure is also available. RESULTS: The updated functionality of the Discovery-2 resource is presented, together with a detailed case study of the Plasmodium falciparum S-adenosyl-L-homocysteine hydrolase (PfSAHH) protein. A short example of a chemical search with pyrimethamine is also illustrated. CONCLUSION: The updated Discovery-2 resource allows researchers to obtain detailed properties of proteins from the malaria genome, which may be of interest in the target selection process, and to perform advanced filtering and selection of proteins based on a relevant range of molecular characteristics.

Exploring inhibition of Pdx1, a component of the PLP synthase complex of the human malaria parasite Plasmodium falciparum.

Malaria tropica is a devastating infectious disease caused by Plasmodium falciparum. This parasite synthesizes vitamin B6 de novo via the PLP (pyridoxal 5'-phosphate) synthase enzymatic complex consisting of PfPdx1 and PfPdx2 proteins. Biosynthesis of PLP is largely performed by PfPdx1, ammonia provided by PfPdx2 subunits is condensed together with R5P (D-ribose 5-phosphate) and G3P (DL-glyceraldehyde 3-phosphate). PfPdx1 accommodates both the R5P and G3P substrates and intricately co-ordinates the reaction mechanism, which is composed of a series of imine bond formations, leading to the production of PLP. We demonstrate that E4P (D-erythrose 4-phosphate) inhibits PfPdx1 in a dose-dependent manner. We propose that the acyclic phospho-sugar E4P, with a C1 aldehyde group similar to acyclic R5P, could interfere with R5P imine bond formations in the PfPdx1 reaction mechanism. Molecular docking and subsequent screening identified the E4P hydrazide analogue 4PEHz (4-phospho-D-erythronhydrazide), which selectively inhibited PfPdx1 with an IC50 of 43 µM. PfPdx1 contained in the heteromeric PLP synthase complex was shown to be more sensitive to 4PEHz and was inhibited with an IC50 of 16 µM. Moreover, the compound had an IC50 value of 10 µM against cultured P. falciparum intraerythrocytic parasites. To analyse further the selectivity of 4PEHz, transgenic cell lines overexpressing PfPdx1 and PfPdx2 showed that additional copies of the protein complex conferred protection against 4PEHz, indicating that the PLP synthase is directly affected by 4PEHz in vivo. These PfPdx1 inhibitors represent novel lead scaffolds which are capable of targeting PLP biosynthesis, and we propose this as a viable strategy for the development of new therapeutics against malaria.

Mapping the binding of GluN2B-selective N-methyl-D-aspartate receptor negative allosteric modulators.

We have used recent structural advances in our understanding of the N-methyl-d-aspartate (NMDA) receptor amino terminal domain to explore the binding mode of multiple diaryl GluN2B-selective negative allosteric modulators at the interface between the GluN1 and GluN2B amino-terminal domains. We found that interaction of the A ring within the binding pocket seems largely invariant for a variety of structurally distinct ligands. In addition, a range of structurally diverse linkers between the two aryl rings can be accommodated by the binding site, providing a potential opportunity to tune interactions with the ligand binding pocket via changes in hydrogen bond donors, acceptors, as well as stereochemistry. The most diversity in atomic interactions between protein and ligand occur in the B ring, with functional groups that contain electron donors and acceptors providing additional atomic contacts within the pocket. A cluster of residues distant to the binding site also control ligand potency, the degree of inhibition, and show ligand-induced increases in motion during molecular dynamics simulations. Mutations at some of these residues seem to distinguish between structurally distinct ligands and raise the possibility that GluN2B-selective ligands can be divided into multiple classes. These results should help facilitate the development of well tolerated GluN2B subunit-selective antagonists.

Mechanism for noncompetitive inhibition by novel GluN2C/D N-methyl-D-aspartate receptor subunit-selective modulators.

The compound 4-(5-(4-bromophenyl)-3-(6-methyl-2-oxo-4-phenyl-1,2-dihydroquinolin-3-yl)-4,5-dihydro-1H-pyrazol-1-yl)-4-oxobutanoic acid (DQP-1105) is a representative member of a new class of N-methyl-d-aspartate (NMDA) receptor antagonists. DQP-1105 inhibited GluN2C- and GluN2D-containing receptors with IC(50) values that were at least 50-fold lower than those for recombinant GluN2A-, GluN2B-, GluA1-, or GluK2-containing receptors. Inhibition was voltage-independent and could not be surmounted by increasing concentrations of either coagonist, glutamate or glycine, consistent with a noncompetitive mechanism of action. DQP-1105 inhibited single-channel currents in excised outside-out patches without significantly changing mean open time or single-channel conductance, suggesting that DQP inhibits a pregating step without changing the stability of the open pore conformation and thus channel closing rate. Evaluation of DQP-1105 inhibition of chimeric NMDA receptors identified two key residues in the lower lobe of the GluN2 agonist binding domain that control the selectivity of DQP-1105. These data suggest a mechanism for this new class of inhibitors and demonstrate that ligands can access, in a subunit-selective manner, a new site located in the lower, membrane-proximal portion of the agonist-binding domain.

Structural determinants of D-cycloserine efficacy at the NR1/NR2C NMDA receptors.

We have studied relative efficacies of NR1 agonists glycine and d-cycloserine (DCS), and found efficacy to be dependent on the NR2 subunit. DCS shows partial agonism at NR1/NR2B but has higher relative efficacy than glycine at NR1/NR2C receptor. Molecular dynamics (MD) simulations of the NR1/NR2B and NR1/NR2C agonist binding domain dimer suggest only subtle differences in the interactions of DCS with NR1 binding site residues relative to glycine. The most pronounced differences were observed in the NR1/NR2C simulation between the orientation of helices F and G of the NR1 subunit. Interestingly, Helix F was previously proposed to influence receptor gating and to adopt an orientation depending on agonist efficacy. MD simulations and site-directed mutagenesis further suggest a role for residues at the agonist binding domain dimer interface in regulating DCS efficacy. To relate the structural rearrangements to receptor gating, we recorded single-channel currents from outside-out patches containing a single active NR1/NR2C receptor. DCS increased the mean open time and open probability of NR1/NR2C receptors compared with glycine. Maximum likelihood fitting of a gating model for NR1/NR2C receptor activation to the single-channel data suggests that DCS specifically accelerates the rate constant governing a fast gating step and reduces the closing rate. These changes appear to reflect a decreased activation energy for a pregating step and increased stability of the open states. We suggest that the higher efficacy of DCS at NR1/NR2C receptors involves structural rearrangements at the dimer interface and an effect on NR1/NR2C receptor pregating conformational changes.