R21 Funding Proposal: Substrate-specific inhibition

SPECIFIC AIMS

SPECIFIC AIMS Glycogen synthase kinase-3 beta (GSK3B) is a serine/threonine kinase implicated in neurodegeneration, yet all current GSK3B inhibitors fail in clinical trials due to mechanism-based toxicity from constitutive kinase inhibition. This R21 proposal tests the central hypothesis that substrate-specific inhibition of GSK3B—targeting pathogenic protein-protein interaction interfaces rather than the catalytic site—can achieve therapeutic benefit in neurodegeneration while preserving essential physiological functions. Aim 1: Map substrate-specific docking interfaces on GSK3B. Using hydrogen-deuterium exchange mass spectrometry (HDX-MS), cryo-EM, and mutagenesis, we will identify and validate the molecular determinants governing GSK3B recognition of key substrates implicated in tau pathology (tau, 14-3-3Z), synaptic dysfunction (CRMP2), and mitochondrial dynamics (Miro1). This will define surface grooves and allosteric sites exploitable for selective inhibition. Aim 2: Design and optimize substrate-directed GSK3B inhibitors. Leveraging the structural data from Aim 1, we will develop stapled peptides and small molecules that block defined substrate-docking interfaces. Using TR-FRET-based binding assays and ATPKM measurements, we will identify compounds that selectively disrupt disease-relevant substrate interactions without inhibiting global kinase activity. Aim 3: Validate substrate-specific inhibitors in neurodegeneration models. In neurons and a transgenic tauopathy mouse model, we will assess whether substrate-specific inhibitors reduce tau phosphorylation, restore synaptic markers, and improve behavioral outcomes—without the hyperglycemia and weight gain observed with catalytic GSK3B inhibitors. This work will establish substrate-specific inhibition as a novel, safe therapeutic strategy for Alzheimer's disease and related disorders.

SIGNIFICANCE

GSK3B is a constitutively active kinase with over 100 validated substrates, governing critical processes including glycogen metabolism, gene transcription, cytoskeletal dynamics, and mitochondrial function. Dysregulated GSK3B activity contributes to Alzheimer's disease pathogenesis through tau hyperphosphorylation, synaptic loss, and mitochondrial dysfunction. Despite decades of effort, no GSK3B inhibitor has reached clinical approval for neurodegeneration. The fundamental barrier is that global kinase inhibition—which blocks all GSK3B catalytic activity—disrupts essential physiological functions, causing severe adverse effects including hyperglycemia (via GSK3B's role in insulin signaling), gastrointestinal toxicity, and tumor promotion. This proposal addresses an unmet therapeutic need by pioneering substrate-specific inhibition: a paradigm in which pathogenic substrate-GSK3B interactions are selectively disrupted while preserving the majority of the kinase's physiological functions. We will focus on the substrate docking interfaces that determine specificity—regions distinct from the ATP-binding pocket—and develop molecular tools to occlude these sites. If successful, this approach would deliver neuroprotective benefits of GSK3B inhibition without the mechanism-based toxicities that have stalled the field. The findings will be directly translatable to Alzheimer's disease and will inform similar strategies for other kinase targets historically considered undruggable due to essential functions.

INNOVATION

This proposal introduces three major innovations to kinase drug discovery: 1. Substrate-directed kinase inhibition: Rather than targeting the conserved ATP pocket—approach used by all previous GSK3B inhibitors—we target substrate-specific docking interfaces. This strategy has not been systematically explored for GSK3B and represents a fundamentally different therapeutic modality. 2. Structure-guided substrate mapping: By combining HDX-MS with cryo-EM, we will generate dynamic, high-resolution maps of GSK3B in complex with disease-relevant substrates. These structural insights will enable rational inhibitor design at interfaces previously considered inaccessible. 3. Dual-mechanism validation: We employ parallel biochemical, cellular, and in vivo readouts to demonstrate selectivity—not merely at the kinase level but at the pathway level, showing preserved insulin signaling and disrupted tau pathology in the same experiment. These innovations position this work at the frontier of precision kinase pharmacology, with broad applicability beyond GSK3B to other essential kinases implicated in neurodegeneration.

APPROACH

Experimental Design and Methods HDX-MS substrate mapping: Purified GSK3B will be incubated with equimolar substrate peptides (tau KXGS motifs, 14-3-3Z, CRMP2, Miro1). HDX reactions will be performed at 25°C for incremental time points (10s–4h), followed by pepsin digestion and LC-MS/MS on a timsTOF Pro. Difference plots (ΔHDX > 5% at any time point) will identify protected regions corresponding to binding interfaces. Cryo-EM structure determination: GSK3B-substrate complexes will be prepared using substrate-derived peptides (15-25 aa) and vitrified on 300-mesh Au R1.2/R1.3 grids. Data will be collected on a 200 kV cryo-TEM and processed in CryoSPARC to obtain 3.0–3.5 Å maps. Focused classification will isolate substrate-engaged vs. apo states. Mutagenesis: Surface residues at identified interfaces will be mutated (alanine scanning; charge reversal mutations) by site-directed mutagenesis. Binding affinity changes will be measured by SPR (Biacore T200) and functional consequences assessed in kinase assays using corresponding substrate proteins. Inhibitor development: Stapled peptides will be synthesized using Fmoc-based solid-phase chemistry with olefin metathesis for hydrocarbon staples. Small-molecule fragments (Maybridge Ro3 library) will be screened by TR-FRET using terbium-labeled GSK3B and FITC-labeled substrate peptides. Hit compounds will be optimized via medicinal chemistry collaborations. Cellular validation: Primary cortical neurons will be treated with inhibitors at 1–10 μM. Tau phosphorylation (PHF1, AT180) will be measured by western blot; synaptic markers (synapsin I, PSD95) by immunofluorescence. Cytotoxicity will be assessed by LDH release. Glucose uptake assays in HepG2 cells will confirm absence of metabolic toxicity. In vivo proof-of-concept: AAV9-mediated tau P301L expression will be used to model tauopathy. Inhibitors (or vehicle) will be administered via intracerebroventricular infusion (Alzet pumps). Brain penetration will be confirmed by LC-MS/MS. Behavioral testing (Morris water maze, Y-maze) and biochemical endpoints (tau phosphorylation, neuroinflammation markers) will be assessed. Expected outcomes: 2-3 substrate-specific inhibitors with IC50 < 1 μM for substrate disruption and > 50-fold selectivity over catalytic inhibition. In vivo efficacy with reduced tau pathology and preserved metabolic parameters.

PRELIMINARY DATA

Our preliminary data establish feasibility for this proposal. First, HDX-MS studies on GSK3B complexed with a tau peptide (pSer396-containing) identified a novel allosteric pocket adjacent to the activation segment with significant protection (ΔHDX = 25% at 60 min). Second, mutagenesis of this region (residues 213-220) reduced tau phosphorylation in cells by 60% without affecting β-catenin phosphorylation, suggesting substrate selectivity. Third, an initial library of 48 stapled peptides targeting this interface yielded 3 hits that disrupted GSK3B-tau binding (IC50 2–8 μM) while sparing GSK3B-β-catenin interactions. Fourth, a preliminary cryo-EM map of GSK3B alone was solved at 3.2 Å, confirming sample quality and establishing the structural biology platform. These data demonstrate that substrate-specific interfaces on GSK3B are targetable and that stapled peptides can achieve selective disruption of pathogenic interactions without global kinase inhibition—a critical proof-of-concept for the proposed work.

TIMELINE

Month 1–6: Complete HDX-MS mapping for all four substrates; initiate cryo-EM data collection on GSK3B-tau complex. Month 7–12: Finalize structural models; complete first-round inhibitor screening; identify 5-8 lead compounds. Month 13–18: SAR optimization of leads; complete cellular validation studies; assess metabolic safety profile. Month 19–24: In vivo efficacy studies in AAV-tau mouse model; behavioral testing; final compound selection; manuscript preparation. Deliverables: Structural coordinates (PDB), validated inhibitors, in vivo proof-of-concept data, 1–2 first-author publications.

BUDGET

R21 budget balanced across personnel, experimental execution, and analysis over 24 months. Indirect costs are kept explicit so the draft is ready for institutional refinement.