Integrating inflammatory biomarker analysis and artificial intelligence-enabled image-based profiling to identify drug targets for intestinal fibrosis

Abstract

Intestinal fibrosis is a common complication of several enteropathies with inflammatory bowel disease being the major cause. The progression of intestinal fibrosis may lead to intestinal stenosis and obstruction. Even with an increased understanding of tissue fibrogenesis, there are no approved treatments for intestinal fibrosis. Historically, drug discovery for diseases like intestinal fibrosis has been impeded by a lack of screenable cellular phenotypes. Here we applied Cell Painting, a scalable image-based morphology assay, augmented with machine learning algorithms to identify small molecules that were able to morphologically reverse the activated fibrotic phenotype of intestinal myofibroblasts under pro-fibrotic TNF stimulus. In combination with measuring CXCL10, a common pro-inflammatory cytokine in intestinal fibrosis, we carried out a high-throughput small molecule chemogenomics screen of approximately 5000 compounds with known targets or mechanisms, which have achieved clinical stage or approval by the FDA. Through the use of two divergent analytical methods, we identified several compounds and target classes that are potentially able to treat intestinal fibrosis. The phenotypic screening platform described here represents significant improvements in identifying a wide range of drug targets over conventional methods by integrating morphological analyses and artificial intelligence using pathologically-relevant cells and disease-relevant stimuli.