Small Molecule Antagonists of the DNA Repair ERCC1/XPA Protein-Protein Interaction.

The DNA excision repair protein ERCC1 and the DNA damage sensor protein, XPA are highly overexpressed in patient samples of cisplatin-resistant solid tumors including lung, bladder, ovarian, and testicular cancer. The repair of cisplatin-DNA crosslinks is dependent upon nucleotide excision repair (NER) that is modulated by protein-protein binding interactions of ERCC1, the endonuclease, XPF, and XPA. Thus, inhibition of their function is a potential therapeutic strategy for the selective sensitization of tumors to DNA-damaging platinum-based cancer therapy. Here, we report on new small-molecule antagonists of the ERCC1/XPA protein-protein interaction (PPI) discovered using a high-throughput competitive fluorescence polarization binding assay. We discovered a unique structural class of thiopyridine-3-carbonitrile PPI antagonists that block a truncated XPA polypeptide from binding to ERCC1. Preliminary hit-to-lead studies from compound 1 reveal structure-activity relationships (SAR) and identify lead compound 27 o with an EC50 of 4.7 µM. Furthermore, chemical shift perturbation mapping by NMR confirms that 1 binds within the same site as the truncated XPA67-80 peptide. These novel ERCC1 antagonists are useful chemical biology tools for investigating DNA damage repair pathways and provide a good starting point for medicinal chemistry optimization as therapeutics for sensitizing tumors to DNA damaging agents and overcoming resistance to platinum-based chemotherapy.

Evaluating Scalable Supervised Learning for Synthesize-on-Demand Chemical Libraries.

Traditional small-molecule drug discovery is a time-consuming and costly endeavor. High-throughput chemical screening can only assess a tiny fraction of drug-like chemical space. The strong predictive power of modern machine-learning methods for virtual chemical screening enables training models on known active and inactive compounds and extrapolating to much larger chemical libraries. However, there has been limited experimental validation of these methods in practical applications on large commercially available or synthesize-on-demand chemical libraries. Through a prospective evaluation with the bacterial protein-protein interaction PriA-SSB, we demonstrate that ligand-based virtual screening can identify many active compounds in large commercial libraries. We use cross-validation to compare different types of supervised learning models and select a random forest (RF) classifier as the best model for this target. When predicting the activity of more than 8 million compounds from Aldrich Market Select, the RF substantially outperforms a naïve baseline based on chemical structure similarity. 48% of the RF's 701 selected compounds are active. The RF model easily scales to score one billion compounds from the synthesize-on-demand Enamine REAL database. We tested 68 chemically diverse top predictions from Enamine REAL and observed 31 hits (46%), including one with an IC50 value of 1.3 µM.

Inflammatory mediators act at renal pericytes to elicit contraction of vasa recta and reduce pericyte density along the kidney medullary vascular network.

Introduction: Regardless of initiating cause, renal injury promotes a potent pro-inflammatory environment in the outer medulla and a concomitant sustained decrease in medullary blood flow (MBF). This decline in MBF is believed to be one of the critical events in the pathogenesis of acute kidney injury (AKI), yet the precise cellular mechanism underlying this are still to be fully elucidated. MBF is regulated by contractile pericyte cells that reside on the descending vasa recta (DVR) capillaries, which are the primary source of blood flow to the medulla. Methods: Using the rat and murine live kidney slice models, we investigated the acute effects of key medullary inflammatory mediators TNF-α, IL-1ß, IL-33, IL-18, C3a and C5a on vasa recta pericytes, the effect of AT1-R blocker Losartan on pro-inflammatory mediator activity at vasa recta pericytes, and the effect of 4-hour sustained exposure on immunolabelled NG2+ pericytes. Results and discussion: Exposure of rat and mouse kidney slices to TNF-α, IL-18, IL-33, and C5a demonstrated a real-time pericyte-mediated constriction of DVR. When pro-inflammatory mediators were applied in the presence of Losartan the inflammatory mediator-mediated constriction that had previously been observed was significantly attenuated. When live kidney slices were exposed to inflammatory mediators for 4-h, we noted a significant reduction in the number of NG2+ positive pericytes along vasa recta capillaries in both rat and murine kidney slices. Data collected in this study demonstrate that inflammatory mediators can dysregulate pericytes to constrict DVR diameter and reduce the density of pericytes along vasa recta vessels, further diminishing the regulatory capacity of the capillary network. We postulate that preliminary findings here suggest pericytes play a role in AKI.

A novel functional role for the classic CNS neurotransmitters, GABA, glycine, and glutamate, in the kidney: potent and opposing regulators of the renal vasculature.

The presence of a renal GABA/glutamate system has previously been described; however, its functional significance in the kidney remains undefined. We hypothesized, given its extensive presence in the kidney, that activation of this GABA/glutamate system would elicit a vasoactive response from the renal microvessels. The functional data here demonstrate, for the first time, that activation of endogenous GABA and glutamate receptors in the kidney significantly alters microvessel diameter with important implications for influencing renal blood flow. Renal blood flow is regulated in both the renal cortical and medullary microcirculatory beds via diverse signaling pathways. GABA- and glutamate-mediated effects on renal capillaries are strikingly similar to those central to the regulation of central nervous system capillaries, that is, exposing renal tissue to physiological concentrations of GABA, glutamate, and glycine led to alterations in the way that contractile cells, pericytes, and smooth muscle cells, regulate microvessel diameter in the kidney. Since dysregulated renal blood flow is linked to chronic renal disease, alterations in the renal GABA/glutamate system, possibly through prescription drugs, could significantly impact long-term kidney function.NEW & NOTEWORTHY Functional data here offer novel insight into the vasoactive activity of the renal GABA/glutamate system. These data show that activation of endogenous GABA and glutamate receptors in the kidney significantly alters microvessel diameter. Furthermore, the results show that these antiepileptic drugs are as potentially challenging to the kidney as nonsteroidal anti-inflammatory drugs.

Discovery and Mechanism of Small Molecule Inhibitors Selective for the Chromatin-Binding Domains of Oncogenic UHRF1.

Chromatin abnormalities are common hallmarks of cancer cells, which exhibit alterations in DNA methylation profiles that can silence tumor suppressor genes. These epigenetic patterns are partly established and maintained by UHRF1 (ubiquitin-like PHD and RING finger domain-containing protein 1), which senses existing methylation states through multiple reader domains, and reinforces the modifications through recruitment of DNA methyltransferases. Small molecule inhibitors of UHRF1 would be important tools to illuminate molecular functions, yet no compounds capable of blocking UHRF1-histone binding in the context of the full-length protein exist. Here, we report the discovery and mechanism of action of compounds that selectively inhibit the UHRF1-histone interaction with low micromolar potency. Biochemical analyses reveal that these molecules are the first inhibitors to target the PHD finger of UHRF1, specifically disrupting histone H3 arginine 2 interactions with the PHD finger. Importantly, this unique inhibition mechanism is sufficient to displace binding of full-length UHRF1 with histones in vitro and in cells. Together, our study provides insight into the critical role of the PHD finger in driving histone interactions, and demonstrates that targeting this domain through a specific binding pocket is a tractable strategy for UHRF1-histone inhibition.

High Throughput Small Molecule Screen for Reactivation of FMR1 in Fragile X Syndrome Human Neural Cells.

Fragile X syndrome (FXS) is the most common inherited cause of autism and intellectual disability. The majority of FXS cases are caused by transcriptional repression of the FMR1 gene due to epigenetic changes that are not recapitulated in current animal disease models. FXS patient induced pluripotent stem cell (iPSC)-derived gene edited reporter cell lines enable novel strategies to discover reactivators of FMR1 expression in human cells on a much larger scale than previously possible. Here, we describe the workflow using FXS iPSC-derived neural cell lines to conduct a massive, unbiased screen for small molecule activators of the FMR1 gene. The proof-of-principle methodology demonstrates the utility of human stem-cell-based methodology for the untargeted discovery of reactivators of the human FMR1 gene that can be applied to other diseases.

TRPV4 receptor as a functional sensory molecule in bladder urothelium: Stretch-independent, tissue-specific actions and pathological implications.

The newly recognized sensory role of bladder urothelium has generated intense interest in identifying its novel sensory molecules. Sensory receptor TRPV4 may serve such function. However, specific and physiologically relevant tissue actions of TRPV4, stretch-independent responses, and underlying mechanisms are unknown and its role in human conditions has not been examined. Here we showed TRPV4 expression in guinea-pig urothelium, suburothelium, and bladder smooth muscle, with urothelial predominance. Selective TRPV4 activation without stretch evoked significant ATP release-key urothelial sensory process, from live mucosa tissue, full-thickness bladder but not smooth muscle, and sustained muscle contractions. ATP release was mediated by Ca2+-dependent, pannexin/connexin-conductive pathway involving protein tyrosine kinase, but independent from vesicular transport and chloride channels. TRPV4 activation generated greater Ca2+ rise than purinergic activation in urothelial cells. There was intrinsic TRPV4 activity without exogeneous stimulus, causing ATP release. TRPV4 contributed to 50% stretch-induced ATP release. TRPV4 activation also triggered superoxide release. TRPV4 expression was increased with aging. Human bladder mucosa presented similarities to guinea pigs. Overactive bladders exhibited greater TRPV4-induced ATP release with age dependence. These data provide the first evidence in humans for the key functional role of TRPV4 in urothelium with specific mechanisms and identify TRPV4 up-regulation in aging and overactive bladders.

Piperidine carbamate peptidomimetic inhibitors of the serine proteases HGFA, matriptase and hepsin.

Matriptase and hepsin are type II transmembrane serine proteases (TTSPs). Along with related S1 trypsin like serine protease HGFA (hepatocyte growth factor activator), their unregulated proteolytic activity has been associated with cancer including tumor progression and metastasis. These three proteases have two substrates in common, hepatocyte growth factor (HGF) and macrophage stimulating protein (MSP), the ligands for MET and recepteur d'origine nantais (RON) receptor tyrosine kinases. Mechanism-based tetrapeptide and benzamidine inhibitors of these proteases have been shown to block HGF/MET and MSP/RON cancer cell signaling. Herein, we have rationally designed a new class of peptidomimetic hybrid small molecule piperidine carbamate dipeptide inhibitors comparable in potency to much larger tetrapeptides. We have identified multiple compounds which have potent activity against matriptase and hepsin and with excellent selectivity over the off-target serine proteases factor Xa and thrombin.

A general strategy for diversifying complex natural products to polycyclic scaffolds with medium-sized rings.

The interrogation of complex biological pathways demands diverse small molecule tool compounds, which can often lead to important therapeutics for the treatment of human diseases. Since natural products are the most valuable source for the discovery of therapeutics, the derivatization of natural products has been extensively investigated to generate molecules for biological screenings. However, most previous approaches only modified a limited number of functional groups, which resulted in a limited number of skeleta. Here we show a general strategy for the preparation of a library of complex small molecules by combining state-of-the-art chemistry - the site-selective oxidation of C-H bonds - with reactions that expand rigid, small rings in polycyclic steroids to medium-sized rings. This library occupies a unique chemical space compared to selected diverse reference compounds. The diversification strategy developed herein for steroids can also be expanded to other types of natural products.

Identification of 4-Amino-Thieno[2,3-d]Pyrimidines as QcrB Inhibitors in Mycobacterium tuberculosis.

Antibiotic resistance is a global crisis that threatens our ability to treat bacterial infections, such as tuberculosis, caused by Mycobacterium tuberculosis Of the 10 million cases of tuberculosis in 2017, approximately 19% of new cases and 43% of previously treated cases were caused by strains of M. tuberculosis resistant to at least one frontline antibiotic. There is a clear need for new therapies that target these genetically resistant strains. Here, we report the discovery of a new series of antimycobacterial compounds, 4-amino-thieno[2,3-d]pyrimidines, that potently inhibit the growth of M. tuberculosis To elucidate the mechanism by which these compounds inhibit M. tuberculosis, we selected for mutants resistant to a representative 4-amino-thieno[2,3-d]pyrimidine and sequenced these strains to identify the mutations that confer resistance. We isolated a total of 12 resistant mutants, each of which harbored a nonsynonymous mutation in the gene qcrB, which encodes a subunit of the electron transport chain (ETC) enzyme cytochrome bc1 oxidoreductase, leading us to hypothesize that 4-amino-thieno[2,3-d]pyrimidines target this enzyme complex. We found that addition of 4-amino-thieno[2,3-d]pyrimidines to M. tuberculosis cultures resulted in a decrease in ATP levels, supporting our model that these compounds inhibit the M. tuberculosis ETC. Furthermore, 4-amino-thieno[2,3-d]pyrimidines had enhanced activity against a mutant of M. tuberculosis deficient in cytochrome bd oxidase, which is a hallmark of cytochrome bc1 inhibitors. Therefore, 4-amino-thieno[2,3-d]pyrimidines represent a novel series of QcrB inhibitors that build on the growing number of chemical scaffolds that are able to inhibit the mycobacterial cytochrome bc1 complex.IMPORTANCE The global tuberculosis (TB) epidemic has been exacerbated by the rise in drug-resistant TB cases worldwide. To tackle this crisis, it is necessary to identify new vulnerable drug targets in Mycobacterium tuberculosis, the causative agent of TB, and develop compounds that can inhibit the bacterium through novel mechanisms of action. The QcrB subunit of the electron transport chain enzyme cytochrome bc1 has recently been validated to be a potential drug target. In the current work, we report the discovery of a new class of QcrB inhibitors, 4-amino-thieno[2,3-d]pyrimidines, that potently inhibit M. tuberculosis growth in vitro These compounds are chemically distinct from previously reported QcrB inhibitors, and therefore, 4-amino-thieno[2,3-d]pyrimidines represent a new scaffold that can be exploited to inhibit this drug target.

Predicting kinase inhibitors using bioactivity matrix derived informer sets.

Prediction of compounds that are active against a desired biological target is a common step in drug discovery efforts. Virtual screening methods seek some active-enriched fraction of a library for experimental testing. Where data are too scarce to train supervised learning models for compound prioritization, initial screening must provide the necessary data. Commonly, such an initial library is selected on the basis of chemical diversity by some pseudo-random process (for example, the first few plates of a larger library) or by selecting an entire smaller library. These approaches may not produce a sufficient number or diversity of actives. An alternative approach is to select an informer set of screening compounds on the basis of chemogenomic information from previous testing of compounds against a large number of targets. We compare different ways of using chemogenomic data to choose a small informer set of compounds based on previously measured bioactivity data. We develop this Informer-Based-Ranking (IBR) approach using the Published Kinase Inhibitor Sets (PKIS) as the chemogenomic data to select the informer sets. We test the informer compounds on a target that is not part of the chemogenomic data, then predict the activity of the remaining compounds based on the experimental informer data and the chemogenomic data. Through new chemical screening experiments, we demonstrate the utility of IBR strategies in a prospective test on three kinase targets not included in the PKIS.

Identification of small molecule enzyme inhibitors as broad-spectrum anthelmintics.

Targeting chokepoint enzymes in metabolic pathways has led to new drugs for cancers, autoimmune disorders and infectious diseases. This is also a cornerstone approach for discovery and development of anthelmintics against nematode and flatworm parasites. Here, we performed omics-driven knowledge-based identification of chokepoint enzymes as anthelmintic targets. We prioritized 10 of 186 phylogenetically conserved chokepoint enzymes and undertook a target class repurposing approach to test and identify new small molecules with broad spectrum anthelmintic activity. First, we identified and tested 94 commercially available compounds using an in vitro phenotypic assay, and discovered 11 hits that inhibited nematode motility. Based on these findings, we performed chemogenomic screening and tested 32 additional compounds, identifying 6 more active hits. Overall, 6 intestinal (single-species), 5 potential pan-intestinal (whipworm and hookworm) and 6 pan-Phylum Nematoda (intestinal and filarial species) small molecule inhibitors were identified, including multiple azoles, Tadalafil and Torin-1. The active hit compounds targeted three different target classes in humans, which are involved in various pathways, including carbohydrate, amino acid and nucleotide metabolism. Last, using representative inhibitors from each target class, we demonstrated in vivo efficacy characterized by negative effects on parasite fecundity in hamsters infected with hookworms.

Identification of a novel class of RIP1/RIP3 dual inhibitors that impede cell death and inflammation in mouse abdominal aortic aneurysm models.

Receptor interacting protein kinase-1 and -3 (RIP1 and RIP3) are essential mediators of cell death processes and participate in inflammatory responses. Our group recently demonstrated that gene deletion of Rip3 or pharmacological inhibition of RIP1 attenuated pathogenesis of abdominal aortic aneurysm (AAA), a life-threatening degenerative vascular disease characterized by depletion of smooth muscle cells (SMCs), inflammation, negative extracellular matrix remodeling, and progressive expansion of aorta. The goal of this study was to develop drug candidates for AAA and other disease conditions involving cell death and inflammation. We screened 1141 kinase inhibitors for their ability to block necroptosis using the RIP1 inhibitor Necrostatin-1s (Nec-1s) as a selection baseline. Positive compounds were further screened for cytotoxicity and virtual binding to RIP3. A cluster of top hits, represented by GSK2593074A (GSK'074), displayed structural similarity to the established RIP3 inhibitor GSK'843. In multiple cell types including mouse SMCs, fibroblasts (L929), bone marrow derived macrophages (BMDM), and human colon epithelial cells (HT29), GSK'074 inhibited necroptosis with an IC50 of ~3 nM. Furthermore, GSK'074, but not Nec-1s, blocked cytokine production by SMCs. Biochemical analyses identified both RIP1 and RIP3 as the biological targets of GSK'074. Unlike GSK'843 which causes profound apoptosis at high doses (>3 µM), GSK'074 showed no detectable cytotoxicity even at 20 µM. Daily intraperitoneal injection of GSK'074 at 0.93 mg/kg significantly attenuated aortic expansion in two mouse models of AAA (calcium phosphate: DMSO 66.06 ± 9.17% vs GSK'074 27.36 ± 8.25%, P < 0.05; Angiotensin II: DMSO 85.39 ± 15.76% vs GSK'074 36.28 ± 5.76%, P < 0.05). Histologically, GSK'074 treatment diminished cell death and macrophage infiltration in aneurysm-prone aortae. Together, our data suggest that GSK'074 represents a new class of necroptosis inhibitors with dual targeting ability to both RIP1 and RIP3. The high potency and minimum cytotoxicity make GSK'074 a desirable drug candidate of pharmacological therapies to attenuate AAA progression and other necroptosis related diseases.

The antimicrobial potential of Streptomyces from insect microbiomes.

Antimicrobial resistance is a global health crisis and few novel antimicrobials have been discovered in recent decades. Natural products, particularly from Streptomyces, are the source of most antimicrobials, yet discovery campaigns focusing on Streptomyces from the soil largely rediscover known compounds. Investigation of understudied and symbiotic sources has seen some success, yet no studies have systematically explored microbiomes for antimicrobials. Here we assess the distinct evolutionary lineages of Streptomyces from insect microbiomes as a source of new antimicrobials through large-scale isolations, bioactivity assays, genomics, metabolomics, and in vivo infection models. Insect-associated Streptomyces inhibit antimicrobial-resistant pathogens more than soil Streptomyces. Genomics and metabolomics reveal their diverse biosynthetic capabilities. Further, we describe cyphomycin, a new molecule active against multidrug resistant fungal pathogens. The evolutionary trajectories of Streptomyces from the insect microbiome influence their biosynthetic potential and ability to inhibit resistant pathogens, supporting the promise of this source in augmenting future antimicrobial discovery.

Practical Model Selection for Prospective Virtual Screening.

Virtual (computational) high-throughput screening provides a strategy for prioritizing compounds for experimental screens, but the choice of virtual screening algorithm depends on the data set and evaluation strategy. We consider a wide range of ligand-based machine learning and docking-based approaches for virtual screening on two protein-protein interactions, PriA-SSB and RMI-FANCM, and present a strategy for choosing which algorithm is best for prospective compound prioritization. Our workflow identifies a random forest as the best algorithm for these targets over more sophisticated neural network-based models. The top 250 predictions from our selected random forest recover 37 of the 54 active compounds from a library of 22,434 new molecules assayed on PriA-SSB. We show that virtual screening methods that perform well on public data sets and synthetic benchmarks, like multi-task neural networks, may not always translate to prospective screening performance on a specific assay of interest.

The renal and blood pressure response to low sodium diet in P2X4 receptor knockout mice.

In the kidney, purinergic (P2) receptor-mediated ATP signaling has been shown to be an important local regulator of epithelial sodium transport. Appropriate sodium regulation is crucial for blood pressure (BP) control and disturbances in sodium balance can lead to hypo- or hypertension. Links have already been established between P2 receptor signaling and the development of hypertension, attributed mainly to vascular and/or inflammatory effects. A transgenic mouse model with deletion of the P2X4 receptor (P2X4-/- ) is known to have hypertension, which is thought to reflect endothelial dysfunction and impaired nitric oxide (NO) release. However, renal function in this model has not been characterized; moreover, studies in vitro have shown that the P2X4 receptor can regulate renal epithelial Na+ channel (ENaC) activity. Therefore, in the present study we investigated renal function and sodium handling in P2X4-/- mice, focusing on ENaC-mediated Na+ reabsorption. We confirmed an elevated BP in P2X4-/- mice compared with wild-type mice, but found that ENaC-mediated Na+ reabsorption is no different from wild-type and does not contribute to the raised BP observed in the knockout. However, when P2X4-/- mice were placed on a low sodium diet, BP normalized. Plasma aldosterone concentration tended to increase according to sodium restriction status in both genotypes; in contrast to wild-types, P2X4-/- mice did not show an increase in functional ENaC activity. Thus, although the increased BP in P2X4-/- mice has been attributed to endothelial dysfunction and impaired NO release, there is also a sodium-sensitive component.

Age, menopausal status and the bladder microbiome.

OBJECTIVES: The bladder is not sterile but contains a healthy community of microbes termed the microbiome. Alterations in the bladder microbiome have been demonstrated in disease states such as the overactive bladder. The microbiome in other anatomical niches is known to alter with age eg the vagina. The objective of this study was to identify if the bladder microbiome in healthy women varies with age and menopausal status. STUDY DESIGN: Urine from 79 healthy women attending secondary care gynaecological clinics with no urinary symptoms provided clean catch mid-stream urine specimens. Urine was centrifuged and the resultant pellet was re-suspended and inoculated onto chocolate agar plates and cultured under either aerobic or anaerobic conditions. Morphologically different colonies were purity plated and 16 s rRNA gene sequencing was performed. A microbe genomic basic local alignment search tool (BLAST) was used to identify the genus of the bacteria. RESULTS: There was no significant correlation between the age of a woman and the number of different genera identified (r=-0.034, p = 0.79). There were few significant differences in the frequency with which the majority of organisms were found in pre and post-menopausal women. The exceptions however were lactobacillus, which was more common in pre-menopausal women (31 vs 3 p = 0.002) and Mobiluncus, which was more common in post-menopausal women (0 vs 3 p = 0.02). CONCLUSIONS: There was no significant correlation between patient age and diversity of the bladder microbiome but large numbers of different organisms were identified. Significant differences were however observed for Lactobacillus which is more common in pre-menopausal women and Mobiluncus which is more common in post-menopausal women.

Small Molecule Inhibitors of Metabolic Enzymes Repurposed as a New Class of Anthelmintics.

The enormous prevalence of infections caused by parasitic nematodes worldwide, coupled to the rapid emergence of their resistance to commonly used anthelmintic drugs, presents an urgent need for the discovery of new drugs. Herein, we have identified several classes of small molecules with broad spectrum activity against these pathogens. Previously, we reported the identification of carnitine palmitoyltransferases (CPTs) as a representative class of enzymes as potential targets for metabolic chokepoint intervention that was elucidated from a combination of chemogenomic screening and experimental testing in nematodes. Expanding on these previous findings, we have discovered that several chemical classes of known small molecule inhibitors of mammalian CPTs have potent activity as anthelmintics. Cross-clade efficacy against a broad spectrum of adult parasitic nematodes was demonstrated for multiple compounds from different series. Several analogs of these initial hit compounds were designed and synthesized. The compounds we report represent a good starting point for further lead identification and optimization for development of new anthelmintic drugs with broad spectrum activity and a novel mechanism of action.

A High-Throughput Screening Strategy to Identify Inhibitors of SSB Protein-Protein Interactions in an Academic Screening Facility.

Antibiotic-resistant bacterial infections are increasingly prevalent worldwide, and there is an urgent need for novel classes of antibiotics capable of overcoming existing resistance mechanisms. One potential antibiotic target is the bacterial single-stranded DNA binding protein (SSB), which serves as a hub for DNA repair, recombination, and replication. Eight highly conserved residues at the C-terminus of SSB use direct protein-protein interactions (PPIs) to recruit more than a dozen important genome maintenance proteins to single-stranded DNA. Mutations that disrupt PPIs with the C-terminal tail of SSB are lethal, suggesting that small-molecule inhibitors of these critical SSB PPIs could be effective antibacterial agents. As a first step toward implementing this strategy, we have developed orthogonal high-throughput screening assays to identify small-molecule inhibitors of the Klebsiella pneumonia SSB-PriA interaction. Hits were identified from an initial screen of 72,474 compounds using an AlphaScreen (AS) primary screen, and their activity was subsequently confirmed in an orthogonal fluorescence polarization (FP) assay. As an additional control, an FP assay targeted against an unrelated eukaryotic PPI was used to confirm specificity for the SSB-PriA interaction. Nine potent and selective inhibitors produced concentration-response curves with IC50 values of <40 µM, and two compounds were observed to directly bind to PriA, demonstrating the success of this screen strategy.

Machine Learning Consensus Scoring Improves Performance Across Targets in Structure-Based Virtual Screening.

In structure-based virtual screening, compound ranking through a consensus of scores from a variety of docking programs or scoring functions, rather than ranking by scores from a single program, provides better predictive performance and reduces target performance variability. Here we compare traditional consensus scoring methods with a novel, unsupervised gradient boosting approach. We also observed increased score variation among active ligands and developed a statistical mixture model consensus score based on combining score means and variances. To evaluate performance, we used the common performance metrics ROCAUC and EF1 on 21 benchmark targets from DUD-E. Traditional consensus methods, such as taking the mean of quantile normalized docking scores, outperformed individual docking methods and are more robust to target variation. The mixture model and gradient boosting provided further improvements over the traditional consensus methods. These methods are readily applicable to new targets in academic research and overcome the potentially poor performance of using a single docking method on a new target.