A Living Organoid Biobank of Crohn's Disease Patients Reveals Molecular Subtypes for Personalized Therapeutics.

Crohn's disease (CD) is a complex, clinically heterogeneous disease of multifactorial origin; there is no perfect pre-clinical model, little insight into the basis for such heterogeneity, and still no cure. To address these unmet needs, we sought to explore the translational potential of adult stem cell-derived organoids that not only retain their tissue identity, but also their genetic and epigenetic disease-driving traits. We prospectively created a biobank of CD patient-derived organoid cultures (PDOs) using biopsied tissues from colons of 34 consecutive subjects representing all clinical subtypes (Montreal Classification B1-B3 and perianal disease). PDOs were generated also from healthy subjects. Comparative gene expression analyses enabled benchmarking of PDOs as tools for modeling the colonic epithelium in active disease and revealed that despite the clinical heterogeneity there are two major molecular subtypes: immune-deficient infectious-CD [IDICD] and stress and senescence-induced fibrostenotic-CD [S2FCD]. The transcriptome, genome and phenome show a surprising degree of internal consistency within each molecular subtype. The spectrum of morphometric, phenotypic, and functional changes within the "living biobank" reveals distinct differences between the molecular subtypes. These insights enabled drug screens that reversed subtype-specific phenotypes, e.g., impaired microbial clearance in IDICD was reversed using agonists for nuclear receptors, and senescence in S2FCD was rectified using senotherapeutics, but not vice versa . Phenotyped-genotyped CD-PDOs may fill the gap between basic biology and patient trials by enabling pre-clinical Phase '0' human trials for personalized therapeutics. In Brief: This work creates a prospectively biobanked phenotyped-genotyped Crohn's disease patient-derived organoids (CD-PDOs) as platforms for molecular subtyping of disease and for ushering personalized therapeutics. HIGHLIGHTS: Prospectively biobanked CD-organoids recapitulate the disease epithelium in patientsThe phenome-transcriptome-genome of CD-organoids converge on two molecular subtypesOne subtype shows impaired microbial clearance, another increased cellular senescencePhenotyped-genotyped PDOs are then used for integrative and personalized therapeutics.

Complete Model of Vinculin Suggests the Mechanism of Activation by Helical Super-Bundle Unfurling.

To shed light onto the activation mechanism of vinculin, we carried out a detailed refinement of chicken vinculin and compared it to the human protein which is greater than 95% identical. Refinement resulted in a complete and significantly improved model. This model includes important elements such as a pro-rich strap region (PRR) and C-terminus. The conformation of the PRR stabilized by its inter- and intra-molecular contacts shows a dynamic, but relatively stable motif that constitutes a docking platform for multiple molecules. The contact of the C-terminus with the PRR suggests that phosphorylation of Tyr1065 might control activation and membrane binding. Improved electron densities showed the presence of large solvent molecules such as phosphates/sulfates and a head-group of PIP2. The improved model allowed for a computational stability analysis to be performed by the program Corex/Best which located numerous hot-spots of increased and decreased stability. Proximity of the identified binding sites for regulatory partners involved in inducing or suppressing the activation of vinculin to the unstable elements sheds new light onto the activation pathway and differential activation. This stability analysis suggests that the activation pathway proceeds by unfurling of the super-bundle built from four bundles of helices without separation of the Vt region (840-1066) from the head. According to our mechanism, when activating proteins bind at the strap region a separation of N and C terminal bundles occurs, followed by unfurling of the super-bundle and flattening of the general shape of the molecule, which exposes the interaction sites for binding of auxiliary proteins.

Artificial intelligence guided discovery of a barrier-protective therapy in inflammatory bowel disease.

Modeling human diseases as networks simplify complex multi-cellular processes, helps understand patterns in noisy data that humans cannot find, and thereby improves precision in prediction. Using Inflammatory Bowel Disease (IBD) as an example, here we outline an unbiased AI-assisted approach for target identification and validation. A network was built in which clusters of genes are connected by directed edges that highlight asymmetric Boolean relationships. Using machine-learning, a path of continuum states was pinpointed, which most effectively predicted disease outcome. This path was enriched in gene-clusters that maintain the integrity of the gut epithelial barrier. We exploit this insight to prioritize one target, choose appropriate pre-clinical murine models for target validation and design patient-derived organoid models. Potential for treatment efficacy is confirmed in patient-derived organoids using multivariate analyses. This AI-assisted approach identifies a first-in-class gut barrier-protective agent in IBD and predicted Phase-III success of candidate agents.