Robert C. Reynolds

Combining Computational Methods for Hit to Lead Optimization in Mycobacterium Tuberculosis Drug Discovery

ABSTRACTPurposeTuberculosis treatments need to be shorter and overcome drug resistance. Our previous large scale phenotypic high-throughput screening against Mycobacterium tuberculosis (Mtb) has identified 737 active compounds and thousands that are inactive. We have used this data for building computational models as an approach to minimize the number of compounds tested.MethodsA cheminformatics clustering approach followed by Bayesian machine learning models (based on publicly available Mtb screening data) was used to illustrate that application of these models for screening set selections can enrich the hit rate.ResultsIn order to explore chemical diversity around active cluster scaffolds of the dose–response hits obtained from our previous Mtb screens a set of 1924 commercially available molecules have been selected and evaluated for antitubercular activity and cytotoxicity using Vero, THP-1 and HepG2 cell lines with 4.3%, 4.2% and 2.7% hit rates, respectively. We demonstrate that models incorporating antitubercular and cytotoxicity data in Vero cells can significantly enrich the selection of non-toxic actives compared to random selection. Across all cell lines, the Molecular Libraries Small Molecule Repository (MLSMR) and cytotoxicity model identified ~10% of the hits in the top 1% screened (>10 fold enrichment). We also showed that seven out of nine Mtb active compounds from different academic published studies and eight out of eleven Mtb active compounds from a pharmaceutical screen (GSK) would have been identified by these Bayesian models.ConclusionCombining clustering and Bayesian models represents a useful strategy for compound prioritization and hit-to lead optimization of antitubercular agents.

Open Source Bayesian Models. 1. Application to ADME/Tox and Drug Discovery Datasets

On the order of hundreds of absorption, distribution, metabolism, excretion, and toxicity (ADME/Tox) models have been described in the literature in the past decade which are more often than not inaccessible to anyone but their authors. Public accessibility is also an issue with computational models for bioactivity, and the ability to share such models still remains a major challenge limiting drug discovery. We describe the creation of a reference implementation of a Bayesian model-building software module, which we have released as an open source component that is now included in the Chemistry Development Kit (CDK) project, as well as implemented in the CDD Vault and in several mobile apps. We use this implementation to build an array of Bayesian models for ADME/Tox, in vitro and in vivo bioactivity, and other physicochemical properties. We show that these models possess cross-validation receiver operator curve values comparable to those generated previously in prior publications using alternative tools. We have now described how the implementation of Bayesian models with FCFP6 descriptors generated in the CDD Vault enables the rapid production of robust machine learning models from public data or the user’s own datasets. The current study sets the stage for generating models in proprietary software (such as CDD) and exporting these models in a format that could be run in open source software using CDK components. This work also demonstrates that we can enable biocomputation across distributed private or public datasets to enhance drug discovery.

A Chemical Proteomics Approach to Profiling the ATP-binding Proteome of Mycobacterium tuberculosis

Tuberculosis, caused by Mycobacterium tuberculosis, remains one of the leading causes of death worldwide despite extensive research, directly observed therapy using multidrug regimens, and the widespread use of a vaccine. The majority of patients harbor the bacterium in a state of metabolic dormancy. New drugs with novel modes of action are needed to target essential metabolic pathways in M. tuberculosis; ATP-competitive enzyme inhibitors are one such class. Previous screening efforts for ATP-competitive enzyme inhibitors identified several classes of lead compounds that demonstrated potent anti-mycobacterial efficacy as well as tolerable levels of toxicity in cell culture. In this report, a probe-based chemoproteomic approach was used to selectively profile the M. tuberculosis ATP-binding proteome in normally growing and hypoxic M. tuberculosis. From these studies, 122 ATP-binding proteins were identified in either metabolic state, and roughly 60% of these are reported to be essential for survival in vitro. These data are available through ProteomeXchange with identifier PXD000141. Protein families vital to the survival of the tubercle bacillus during hypoxia emerged from our studies. Specifically, along with members of the DosR regulon, several proteins involved in energy metabolism (Icl/Rv0468 and Mdh/Rv1240) and lipid biosynthesis (UmaA/Rv0469, DesA1/Rv0824c, and DesA2/Rv1094) were found to be differentially abundant in hypoxic versus normal growing cultures. These pathways represent a subset of proteins that may be relevant therapeutic targets for development of novel ATP-competitive antibiotics.

Identification of shikimate kinase inhibitors among anti-Mycobacterium tuberculosis compounds by LC-MS

Increasing drug resistance has challenged the control and treatment of tuberculosis, sparking recent interest in finding new antitubercular agents with different chemical scaffolds and mechanisms of action. Mycobacterium tuberculosis shikimate kinase (MtSK), an enzyme present in the shikimate pathway in bacteria, is essential for the survival of the tubercle bacillus, representing an ideal target for therapeutic intervention given its absence in mammals. In this study, a small library of 404 synthetic antimycobacterial compounds identified and supplied through the NIH Tuberculosis Antimicrobial Acquisition and Coordinating Facility (TAACF) high throughput screening program against whole cell M.xa0tuberculosis H37Rv was further screened using a mass spectrometry-based functional assay in order to identify a potential enzymatic target. Fourteen compounds containing an oxadiazole-amide or a 2-aminobenzothiazole core scaffold showed MtSK inhibitory activity at 50xa0μM, with the lowest giving an IC50 of 1.94xa0μM. Induced fit docking studies suggested that the scaffolds shared by these compounds fit well in the shikimate binding pocket of MtSK. In summary, we report new early discovery stage lead scaffolds targeting the essential protein MtSK that can be further pursued in a rational drug design program for the discovery of more selective antitubercular drugs.

Biochemical and structural characterization of mycobacterial aspartyl-tRNA synthetase AspS, a promising TB drug target.

The human pathogen Mycobacterium tuberculosis is the causative agent of pulmonary tuberculosis (TB), a disease with high worldwide mortality rates. Current treatment programs are under significant threat from multi-drug and extensively-drug resistant strains of M. tuberculosis, and it is essential to identify new inhibitors and their targets. We generated spontaneous resistant mutants in Mycobacterium bovis BCG in the presence of 10× the minimum inhibitory concentration (MIC) of compound 1, a previously identified potent inhibitor of mycobacterial growth in culture. Whole genome sequencing of two resistant mutants revealed in one case a single nucleotide polymorphism in the gene aspS at 535GAC>535AAC (D179N), while in the second mutant a single nucleotide polymorphism was identified upstream of the aspS promoter region. We probed whole cell target engagement by overexpressing either M. bovis BCG aspS or Mycobacterium smegmatis aspS, which resulted in a ten-fold and greater than ten-fold increase, respectively, of the MIC against compound 1. To analyse the impact of inhibitor 1 on M. tuberculosis AspS (Mt-AspS) activity we over-expressed, purified and characterised the kinetics of this enzyme using a robust tRNA-independent assay adapted to a high-throughput screening format. Finally, to aid hit-to-lead optimization, the crystal structure of apo M. smegmatis AspS was determined to a resolution of 2.4 Å.

Development of drug targeting based on recombinant expression of the chicken avidin gene

The chemistry required for covalent biotinylation of drugs, radiopharmaceuticals and other ligands is highly developed, and a large number of biotinylated reagents can be readily synthesized. In order to investigate whether expression of avidin cDNA in mammalian cells might be useful as part of a drug targeting strategy, we transiently expressed the avidin gene in two human tumor cell lines (the cervical carcinoma cell line, HeLa, and the liver derived line, Hep G2). Avidin protein as detected by either immunohistochemistry or binding of streptavidin-biotin complexes was present and functional following transient expression. This result indicated that the mechanisms underlying avidin oligomerization which are necessary for proper protein folding are present within mammalian carcinoma cell lines. Next, we generated a producer cell line (derived from psi2) capable of releasing a recombinant retrovirus encoding chicken avidin, and a tumorigenic murine breast cancer cell line (16/C) with stable avidin expression. We show that these cell lines are suitable for conferring functional expression of avidin in vitro. These experiments establish a means by which avidin gene expression can be explored as a mechanism for targeted gene delivery of biotin-derivitized drugs in vitro, and have important implications for utilization of this strategy in vivo.

Autocrine fibroblast growth factor 18 signaling mediates Wnt‐dependent stimulation of CD44‐positive human colorectal adenoma cells

Expansion of a stem‐like subpopulation with increased growth and survival potential is thought to drive colorectal tumor growth and progression. We investigated a CD44‐positive (CD44(+)) subpopulation with extended growth and survival capacity in the human colon adenoma cell line LT97. This subpopulation expressed elevated levels of fibroblast growth factor 18 (FGF18) and fibroblast growth factor receptor FGFR3‐IIIc. Expression levels of the FGFR3‐IIIb, which does not bind FGF18, were similar in CD44(+) and CD44(−). Addition of FGF18 to the medium or its overexpression from an adenoviral vector increased the colony formation capacity of CD44(+) threefold, and stimulated phosphorylation of ERK and GSK3β in both total LT97 populations and CD44(+) cells. FGFR3 signaling blockade by expression of a dominant‐negative FGFR3‐IIIc mutant led to inhibition of both colony formation and down‐stream signaling in the CD44(+) cells. CD44(−) cells did not respond. Blockade of the wnt‐pathway by a dominant‐negative Tcf4‐mutant inhibited FGFR3 activation in LT97 cells as well as in HT29 colorectal cancer cells. The chemical wnt‐inhibitor sulindac sulfide amide inhibited expression of FGF18 and FGFR3‐IIIc and led to inhibition of receptor activation to less than 30% of control treated cells, both in LT97 and HT29 cultures. Our results demonstrate that an FGF18/FGFR3‐IIIc autocrine growth and survival loop is up‐regulated in a wnt‐dependent manner and drives tumor cell growth in a subpopulation of colon adenoma cells. This subpopulation can be regarded as a precursor of colon cancer development and can be targeted for CRC‐prevention by blocking either wnt‐ or FGFR3‐signaling.

Diverse amide analogs of sulindac for cancer treatment and prevention

Sulindac is a non-steroidal anti-inflammatory drug (NSAID) that has shown significant anticancer activity. Sulindac sulfide amide (1) possessing greatly reduced COX-related inhibition relative to sulindac displayed in vivo antitumor activity that was comparable to sulindac in a human colon tumor xenograft model. Inspired by these observations, a panel of diverse sulindac amide derivatives have been synthesized and their activity probed against three cancer cell lines (prostate, colon and breast). A neutral analog, compound 79 was identified with comparable potency relative to lead 1 and activity against a panel of lymphoblastic leukemia cell lines. Several new series also show good activity relative to the parent (1), including five analogs that also possess nanomolar inhibitory potencies against acute lymphoblastic leukemia cells. Several new analogs identified may serve as anticancer lead candidates for further development.

Screening and Development of New Inhibitors of FtsZ from M-Tuberculosis

A variety of commercial analogs and a newer series of Sulindac derivatives were screened for inhibition of M. tuberculosis (Mtb) in vitro and specifically as inhibitors of the essential mycobacterial tubulin homolog, FtsZ. Due to the ease of preparing diverse analogs and a favorable in vivo pharmacokinetic and toxicity profile of a representative analog, the Sulindac scaffold may be useful for further development against Mtb with respect to in vitro bacterial growth inhibition and selective activity for Mtb FtsZ versus mammalian tubulin. Further discovery efforts will require separating reported mammalian cell activity from both antibacterial activity and inhibition of Mtb FtsZ. Modeling studies suggest that these analogs bind in a specific region of the Mtb FtsZ polymer that differs from human tubulin and, in combination with a pharmacophore model presented herein, future hybrid analogs of the reported active molecules that more efficiently bind in this pocket may improve antibacterial activity while improving other drug characteristics.

Global Human-Kinase Screening Identifies Therapeutic Host Targets against Influenza.

During viral infection of human cells, host kinases mediate signaling activities that are used by all viruses for replication; therefore, targeting of host kinases is of broad therapeutic interest. Here, host kinases were globally screened during human influenza virus (H1N1) infection to determine the time-dependent effects of virus infection and replication on kinase function. Desthiobiotin-labeled analogs of adenosine triphosphate and adenosine diphosphate were used to probe and covalently label host kinases in infected cell lysates, and probe affinity was determined. Using infected human A549 cells, we screened for time-dependent signal changes and identified host kinases whose probe affinities differed significantly when compared to uninfected cells. Our screen identified 10 novel host kinases that have not been previously shown to be involved with influenza virus replication, and we validated the functional importance of these novel kinases during infection using targeted small interfering RNAs (siRNAs). The effects of kinase-targeted siRNA knockdowns on replicating virus levels were measured by quantitative reverse-transcription PCR and cytoprotection assays. We identified several novel host kinases that, when knocked down, enhanced or reduced the viral load in cell culture. This preliminary work represents the first screen of the changing host kinome in influenza virus–infected human cells.