Combining explainable machine learning, demographic and multi-omic data to inform precision medicine strategies for inflammatory bowel disease.

Inflammatory bowel diseases (IBDs), including ulcerative colitis and Crohn's disease, affect several million individuals worldwide. These diseases are heterogeneous at the clinical, immunological and genetic levels and result from complex host and environmental interactions. Investigating drug efficacy for IBD can improve our understanding of why treatment response can vary between patients. We propose an explainable machine learning (ML) approach that combines bioinformatics and domain insight, to integrate multi-modal data and predict inter-patient variation in drug response. Using explanation of our models, we interpret the ML models' predictions to infer unique combinations of important features associated with pharmacological responses obtained during preclinical testing of drug candidates in ex vivo patient-derived fresh tissues. Our inferred multi-modal features that are predictive of drug efficacy include multi-omic data (genomic and transcriptomic), demographic, medicinal and pharmacological data. Our aim is to understand variation in patient responses before a drug candidate moves forward to clinical trials. As a pharmacological measure of drug efficacy, we measured the reduction in the release of the inflammatory cytokine TNFα from the fresh IBD tissues in the presence/absence of test drugs. We initially explored the effects of a mitogen-activated protein kinase (MAPK) inhibitor; however, we later showed our approach can be applied to other targets, test drugs or mechanisms of interest. Our best model predicted TNFα levels from demographic, medicinal and genomic features with an error of only 4.98% on unseen patients. We incorporated transcriptomic data to validate insights from genomic features. Our results showed variations in drug effectiveness (measured by ex vivo assays) between patients that differed in gender, age or condition and linked new genetic polymorphisms to patient response variation to the anti-inflammatory treatment BIRB796 (Doramapimod). Our approach models IBD drug response while also identifying its most predictive features as part of a transparent ML precision medicine strategy.

Studies of human pancreatic elastase treatment of rabbit and human vein rings to predict human therapeutic doses.

Vascular tissue contains abundant elastic fibers that contribute to vessel elasticity. Vonapanitase (formerly PRT-201) is a recombinant human chymotrypsin-like elastase family member 1 (CELA1) shown to cleave the elastin component of elastic fibers, resulting in increased vessel diameter. The purpose of these current studies was to determine vein diameter, wall thickness, elastin content, and vonapanitase potency in veins used in a model of arteriovenous fistula (AVF) and in patients undergoing AVF creation for hemodialysis access to guide dose selection for human trials. Rabbit linguofacial, maxillary, and external jugular veins, and human basilic and upper and lower arm cephalic veins were dissected postmortem and sectioned into 2 mm length rings. Rings were incubated in vonapanitase at 37°C at varying concentrations and times. Elastin content was estimated histologically and by quantifying desmosine, a protein cross-link unique to elastin. Rabbit veins were substantially thinner and contained less elastin than human veins. In human veins, elastin content was greatest in basilic and least in lower arm cephalic. Vonapanitase removed elastin in a time- and concentration-dependent manner in all vein types. A lower concentration of vonapanitase was required to remove elastin from rabbit relative to human veins. In summary, vonapanitase reduced the elastin content of rabbit and human veins but did so at a lower concentration in the rabbit veins. Rabbit models may overestimate the potency of vonapanitase in humans. These results indicate that human dose selection should be guided by human vein ring experiments.

Recombinant Human Elastase Treatment of Cephalic Veins.

BACKGROUND: Vessel injury at the time of Arteriovenous Fistula (AVF) creation may lead to neointimal hyperplasia that impairs AVF maturation. Vonapanitase, a recombinant human chymotrypsin-like elastase family member 1, is an investigational drug under development to improve AVF maturation and patency. The current studies were designed to document vonapanitase effects in human cephalic veins that are used in AVF creation. METHODS: Human cephalic veins were mounted on a perfusion myograph. Vonapanitase 1.2, 4, 13.2, and 40 µg/ml or saline was applied drop wise on the vein followed by saline rinse. Vein segments were cut into rings for elastin content determination by desmosine radioimmunoassay and histology. Fluorescently-labelled vonapanitase was applied to veins and adventitial imaging was performed using laser scanning confocal microscopy. In vivo time course experiments were performed by treating rabbit jugular veins and harvesting 1 h and 4 h after vonapanitase treatment. RESULTS / CONCLUSION: Vonapanitase reduced desmosine content in a dose-related manner. Histology also confirmed a dose-related reduction in elastic fiber staining. Fluorescently-labelled vonapanitase persistently localized to elastic fibers in the vein adventitia. In vivo experiments showed a reduction in desmosine content in jugular veins from 1 h to 4 h following treatment. These data suggest that vonapanitase targets elastin in elastic fibers in a dose related manner and that elastase remains in the vessel wall and has catalytic activity for at least 1 h.

Recombinant Human Elastase Alters the Compliance of Atherosclerotic Tibial Arteries After Ex Vivo Angioplasty.

PURPOSE: This study was designed to determine whether vonapanitase (formerly PRT-201), a recombinant human elastase, treatment can fragment the protein elastin in elastic fibers and cause dilation of atherosclerotic human peripheral arteries subjected to ex vivo balloon angioplasty. MATERIALS AND METHODS: Seven patients undergoing lower limb amputation for peripheral artery disease or who died and donated their bodies to science donated 11 tibial arteries (5 anterior, 6 posterior) for this study. All arteries were atherosclerotic by visual inspection. The arteries underwent ex vivo balloon angioplasty and thereafter were cut into rings and studied on wire myographs where the rings were stretched and tension was recorded. After treatment with vonapanitase 2 mg/mL or vehicle control, myography was repeated and the rings were then subject to elastin content measurement using a desmosine radioimmunoassay and elastic fiber visualization by histology. The wire myography data were used to derive compliance, stress-strain, and incremental elastic modulus curves. RESULTS: Vonapanitase treatment reduced elastin (desmosine) content by 60% and decreased elastic fiber histologic staining. Vonapanitase-treated rings experienced less tension at any level of stretch and as a result had shifts in the compliance and stress-strain curves relative to vehicle-treated rings. Vonapanitase treatment did not alter the incremental elastic modulus curve. CONCLUSIONS: Vonapanitase treatment of atherosclerotic human peripheral arteries after ex vivo balloon angioplasty fragmented elastin in elastic fibers, decreased tension in the rings at any level of stretch, and altered the compliance and stress-strain curves in a manner predicting arterial dilation in vivo. Based on this result, local treatment of balloon angioplasty sites may increase blood vessel diameter and thereby improve the success of balloon angioplasty in peripheral artery disease.

Feasibility of Using the Mosquito Blood Meal for Rapid and Efficient Human and Animal Virus Surveillance and Discovery.

Mosquito blood meals taken from humans and animals potentially represent a useful source of blood for the detection of blood-borne pathogens. In this feasibility study, Anopheles stephensi mosquitoes were fed with blood meals spiked with dengue virus type 2 (DENV-2) and harvested at serial time points. These mosquitoes are not competent vectors, and the virus is not expected to replicate. Ingested blood was spotted on Whatman FTA cards and stored at room temperature. Mosquito abdomens were removed and stored at -80°C. Control blood meal aliquots were stored in vials or applied onto FTA cards. After 4 weeks of storage, the samples were extracted using beadbeating and QIAamp Viral RNA kit (Qiagen Sciences, Germantown, MD). Recovered viral RNA was analyzed by DENV-2 TaqMan RT-PCR assay and next-generation sequencing (NGS). Overall viral RNA recovery efficiency was 15% from the directly applied dried blood spots and approximately 20% or higher for dried blood spots made by blotting mosquito midgut on FTA cards. Viral RNA in mosquito-ingested blood decreases over time, but remains detectable 24 hours after blood feeding. The viral sequences in FTA-stored specimens can be maintained at room temperature. The strategy has the potential utility in expedited zoonotic virus discovery and blood-borne pathogen surveillance.