Andrew Stewart - US grants

? DESCRIPTION (provided by applicant): Both Types 1 and 2 diabetes result from reductions in pancreatic beta cell mass and function. Thus, a major goal of the NIH/NIDDK is to develop novel drugs and tools that can lead to replacement and/or regeneration of human beta cells. This is has proven difficult, because adult human beta cells are refractory to engagement in cell cycle progression. Recently, we designed and performed a unique high-throughput screen (HTS) of two small molecule libraries, a 2000 compound FDA library and a second 100,000 compound library, and have identified a novel and effective small molecule, harmine, that is able to activate mouse, rat and human beta cell replication at rates that approach those required for therapeutic human beta cell replication. We have also identified additional compounds that share structural and functional features with harmine, and refer to them as harmalogs. Ongoing structure-activity studies suggest that the common pathway employed by these compounds is a calcineurin-NFaT-DYRK1A pathway, but additional pathways and intracellular targets remain possible. As for the broad field of beta cell biology in general, targeting harmalogs to beta cells is challenging. Accordingly, in this application, we assemble a team of experienced beta cell biologists and medicinal chemists to pursue three Specific Aims: 1. To Fully Define the Mechanism of Action of the Harmalogs on Human Beta Cell Proliferation. 2. To Document In Vivo Effects of the Harmalogs on Human Beta Cell Expansion and Function. 3. To Synthesize Modified Harmalogs with Chemical Linkers That Allow Both Retention of Bioactivity and Conjugation to Beta Cell-Targeting Ligands. We believe these studies are highly significant because they document that adult human beta cells can be induced to proliferate at therapeutically relevant rates using small molecule approaches; they will define the molecular mechanism of action of the harmalog class of compounds; and, because they explore novel methodologies to target these effective regenerative compounds to the human beta cell.

All forms of diabetes are due to a relative deficit of functional beta cells, yet adult human beta cells are resistant to attempts at replication, and serve as a poor experimental model for replicating human beta cells. Because of this paucity of models for adult human beta cell replication, we have developed a large Biorepository of rare human insulinomas, benign pancreatic adenomas that grow and over-secrete insulin, believing that they may provide information that will inform attempts at clues and drug targets and pathways for therapeutic human beta cell regeneration. Recently, we have provided the results of intensive genomic, transcriptomic and bioinformatic analysis of human insulinomas, comparing them to FACS-sorted pure human beta cells. Remarkably, we find that insulinomas display three cardinal features: 1) they almost universally contain recurring mutations, copy number alterations and gene expression abnormalities in members of the Trithorax Group of chromatin modifying enzymes, notably including KDM6A, MLL3 and/or MEN1; 2) they also almost universally display alterations in the Polycomb Repressive Complex of chromatin modifying enzymes and their targets, particularly YY1, EZH2 and H3F3A; 3) they almost universally display abnormalities in the chromosome 11, such as allelic loss of all or part of chromosome 11, and/or allele-specific expression and/or DNA methylation/imprinting abnormalities of the imprinted 11p15 region of chromosome 11, reminiscent of the other beta cell proliferative disorders, such as the Focal Variant of Hyperinsulinism and Beckwith-Wiedemann Syndrome. On the other hand, the mechanisms through with these events lead to beta cell proliferation and while maintaining the beta cell phenotype are unknown. More specifically, exactly how MEN1, KDM6A, MLL3, YY1, EZH2 and H3F3A modulate beta cell function and proliferation are largely unknown. The three Specific Aims of this application address this important knowledge gap. Aim 1. To Elucidate the Abnormal Pathobiology of Three Key Trithorax Members In Insulinoma vs. Beta Cells. Aim 2. To Define Abnormal Biology of Three Key Polycomb Repressive Complex Members in Insulinomas vs. Beta Cells. Aim 3. To Define 3-D Chromosome 11p15 Architecture in Insulinomas and Beta Cells.

Project Summary: This supplemental grant application is a translational drug discovery proposal in response to NOT-AG-18-039 for NIH Funding Opportunities - Alzheimer's-focused administrative supplements for NIH grants that are not focused on Alzheimer's disease. Our approach covers specific areas of interest including medicinal and computational chemistry, structural biology and pharmacological validation of novel therapeutics to treat AD and ADRD. Our ultimate goal is to develop novel, selective, CNS penetrant small molecule DYRK1A inhibitor therapeutics to treat AD/ADRD with data generated to apply for future funding opportunities. DYRK1A inhibitors in our current research program are directed for peripheral use to treat diabetes. We request supplemental funding to characterize our new lead scaffolds for CNS penetrance and activity in AD/ADRD cell based and in vivo assays of AD mouse models. The data generated will be useful to optimize our inhibitors neurological diseases such as ADRD and then apply for future funding for further AD drug development. Our premise is that novel leads, developed for the specific use in the CNS will provide further validation of this target for ADRD. Our strategy of structure based drug design and medicinal chemistry kinase optimization, coupled with powerful in vitro and in vivo assays will lead to improved CNS penetrant DYRK1A inhibitors. To that end, we have identified novel drug-like leads that target DYRK1A. These results provide the basis for a drug discovery effort that includes a molecular target, in vitro/in vivo ADRD biology, and viable starting points for medicinal chemistry optimization. Based on these complementary capabilities and important objectives and implications for DYRK1A in AD therapeutics, we propose three specific aims: Specific Aim 1. Develop two novel, structurally distinct DYRK1A inhibitors with kinase/off-target selectivity, oral bioavailability and brain penetration for AD therapy using structure-based drug design. Specific Aim 2. Evaluate novel compounds for DYRK1A inhibition, in vitro ADME-PK, optimized off- target selectivity, effects on cellular tau phosphorylation using biochemical and cell-based assays. Specific Aim 3. Evaluate optimized DYRK1A inhibitors for in vivo ADME PK, brain penetration, effects on tau and other AD associated proteins using in vivo microdialysis and efficacy in AD mouse models.

Both Type 1 and Type 2 diabetes (T1D and T2D) result from inadequate numbers of normally functioning beta cells. Small molecule drugs that inhibit the kinase, DYRK1A, such as harmine and others, are reproducibly able to induce adult human beta cells to replicate, but at low rates (~2%/day). More recently, we have shown that adding any small molecule DYRK1A inhibitor to either a TGF-beta superfamily inhibitor (TGF?I's) or to a GLP1 receptor agonist such as GLP1 or exendin-4 markedly enhances this replicative induction to rates averaging 5-8%/day. This has been documented not only by ?markers? of replication such as Ki67 and BrdU, both in vitro and in vivo, but also by enhancing actual numbers of beta cells. Unfortunately, these drugs may not exclusively affect the beta cell, but instead may have ?off target? effects as well. Thus, diabetes researchers have effective regenerative drugs to deliver to the human beta cell, but may require a ?targeting molecule? to bring them to the beta cell. Accordingly, over the past few years, diabetes researchers have identified two classes of ?prototype beta cell targeting molecules?: a monoclonal antibody raised against the beta cell surface molecule called ?ENTPD3?; and, the GLP1-receptor class of molecules. While these molecules may or may not be perfect for beta cell targeting, they are unquestionably ?prototype? targeting molecules with which to work while the field attempts to identify more perfect beta cell targeting molecules. We have also synthesized numerous DYRK1A inhibitors and TGFbeta inhibitors and cleavable chemical linkers that enable their conjugation to any potential beta cell targeting molecule. Thus, the Aims of this application are: 1. Synthesis of TGF-beta Inhibitors With Chemical Linkers, to Complement our Novel DYRK1A inhibitor Linker Compounds, For Conjugation To Prototype Targeted Delivery Vehicles. 2. Conjugation of Harmine-Linker and TGF-beta-Inhibitor-Linker Compounds to Two Prototype Targeting Molecules: GLP1 Receptor Agonists and ENTPD3 Monoclonal Antibodies. 3. Definition of Long Term Efficacy, Specificity and Safety of the Harmine-Linker and TGF-beta Inhibitor- Linker Conjugates in vivo in Human Islet Engraftment Models. These goals are both achievable and directly responsive to the goals of aims of the NIDDK. If GLP1 receptor agonists and/or ENTPD3 MAbs prove suboptimal for beta cell targeting, the approaches and molecules developed here can readily be extended and adapted to any future more specific human beta cell targeting molecule.

Summary Types 1 and 2 diabetes result entirely or in part from a reduction in numbers of normally functioning pancreatic beta cells. Residual beta cells are present in most people with diabetes, suggesting that regenerative therapies may be uniquely helpful. Inducing human beta cells to regenerate has proven impossible until recently. This has changed with the discovery by several labs of drugs that inhibit the kinase, DYRK1A, and which induce human beta cells to replicate at ?rates?, or more properly, ?labeling indices?, of ~2%. Proliferation can be augmented by combining DYRK1A inhibitors with TGF-beta superfamily inhibitors or with GLP1 receptor agonists, generating labeling indices of 5-8%. Although these are exciting advances, they also demonstrate that >90% of human beta cells are recalcitrant to cell cycle entry. The cause of this remarkable refractoriness to proliferation is poorly understood. DYRK1A inhibition alters intracellular trafficking of the NFaT family of transcription factors. In the course of our studies on human insulinomas, we have uncovered a parallel pathway, the DREAM-MMB complex, that also restricts human beta cell proliferation. We show here that the extended family of ~200 canonical DREAM-MMB complex genes and proteins appear to be present in human beta cells, and that DREAM complex is switched from a repressive to a proliferative configuration by DYRK1A inhibition. In addition, we have also observed potential overlapping roles for Trithorax and Polycomb complexes with the DREAM complex in restraining human beta cell proliferation. Accordingly, in this application, we propose three Specific Aims: Aim 1. Complete The Characterization of the Functional DREAM-MMB Complex in the Human Beta Cell. Aim 2. Delineate Candidate DREAM-MMB Complex Target Genes in the Human Beta Cell. Aim 3. Aim 3. Define Genome-Wide Integration of Repressive and Proliferative DREAM-MMB Mediators with Trithorax and Polycomb Chromatin Regulators and DREAM Target Genes in Human Beta Cells. Our overarching goals are: 1) to clearly and comprehensively define the fundamental mechanisms that enforce quiescence in the adult human beta cell; 2) to more clearly elucidate the fundamental mechanisms through which DYRK1A inhibitors, TGF? inhibitors and GLP1RA?s synergize to induce their remarkable rates of human beta cell proliferation; and 3) to reveal novel pathways and further expand therapeutic targets for human beta cell regeneration for diabetes.