Method of preventing and treating alzheimer's disease and related dementias with drugs inhibiting the interaction between 14-3-3g protein and hexokinase-1 protein

Abstract

A method of treating and preventing Alzheimer's disease and related dementias is disclosed. The method includes administering to a patient an effective amount of a drug that is operable to inhibit an interaction between 14-3-3G protein and hexokinase-1 protein in the patient. The drug may be ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astemizole.

Description

[0001] METHOD OF PREVENTING AND TREATING ALZHEIMER’S DISEASE AND RELATED DEMENTIAS WITH DRUGS INHIBITING THE INTERACTION BETWEEN 14-3-3G PROTEIN AND HEXOKINASE-1 PROTEIN

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] This application claims the benefit of U.S. Provisional Application No. 63 / 439,664, entitled “14-3-3 Binding FDA Approved Drugs for Preventing and Treating Neurodegenerative Disorders” and filed on January 18, 2023. The complete disclosure of said provisional application is hereby incorporated by reference.

[0004] STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0005] This invention was made with government support from grant no. 101 BX001655, a Merit grant awarded by the U.S. Veterans Administration, and grant numbers R01 AG062254-01 and P01 AG012411 -17A1 awarded by the National Institute on Aging. The government has certain rights in the invention.

[0006] BACKGROUND

[0007] 14-3-3 proteins comprise small families of paralogous proteins (7 in mammals, 2 in nematodes) that serve as modulators of critical signaling and regulatory pathways, e.g. via their interactions with transcription factors including FOXO and TFEB. The 14-3-3 protein family in mammals comprises seven 30-kDa proteins designated as [3, y, e, T, q, a, and paralogs (beta, gamma, epsilon, tau, eta, sigma, and zeta). These mammalian 14-3-3 genes, situated at 7 distinct chromosomal loci, are expressed ubiquitously but their encoded proteins are especially abundant in cerebral neurons. The 14-3-3 proteins have especially high binding affinities for proteins containing phospho-serine or phospho-threonine, and serve as both molecular chaperones and regulatory modulators of their numerous protein ligands. They occur predominantly as homo- and heterodimers, forming complexes with over 200 other ligand proteins. Their highly disordered C-terminal and N-terminal domains are generally stabilized by protein-protein interactions (PPI) with their binding partners, th apoptosis, cell-cycle progression, RNA transcription, and DNA replication, inter alia.

[0008] Each 14-3-3 monomer contains a highly conserved and well-structured core region comprising 9 alpha helices. Of these, helices 3, 5, 7 and 9 form a ‘W- or ‘L’- shaped cluster with a concave amphipathic groove (combining hydrophilic and hydrophobic moieties) that interacts with ligand peptides, which are often (but not necessarily) phosphopeptides. Dimerization of 14-3-3 proteins entails interaction between helix H1 of one monomer and H3 and H4 of the other, and may drive dimerization of their binding partners. The final structural conformation of 14-3-3 paralogs is influenced by their interacting partners, believed to confer stable structures to the otherwise unstructured N- and C-terminal tails. Functions of 14-3-3 paralogs, in conjunction with their ligand proteins, span a wide range of enzymatic activities, subcellular localizations, and pathways. Many binding partners of 14-3-3 contain RSX(pS / T)XP and / or RXXX(pS / T)XP consensus sequences, where ‘X’ is any amino acid and (pS / T) indicates phosphoserine or phosphothreonine.

[0009] Expression patterns vary among the 14-3-3 paralogs. In mammals, 14-3-3E (s, epsilon) protein is highly abundant in brain, lymphoblasts, and adipocytes, while 14-3-3G (y, gamma) is most highly expressed in brain, skeletal muscle, heart, and embryonic stem cells. Both of these paralogs were shown to oppose androgen biosynthesis. The 14-3-3G (y) paralog has been implicated in neurodegenerative diseases and disorders. Several reports have proposed 14-3-3y as a CSF biomarker for the clinical diagnosis of sporadic Creutzfeldt-Jakob Disease (CJD) and several other inflammatory, pathological conditions. Hyperphosphorylated tau (hP- tau), a major constituent of neurofibrillary tangles in Alzheimer’s disease, is a critical neuropathological binding target of 14-3-3^ (zeta).

[0010] Numerous proteins have been implicated in neurodegenerative protein aggregation; A[3I-42 and hP-tau are best known as aggregate markers typical of AD, but they are also constituents of aggregates observed in other neurodegenerative diseases. Previous research showed that tau is an interacting partner of 14-3-3Z / However, the mechanisms and consequences of that interaction in AD pathology are not well understood. In the present study we found AD-specific interactions of 14-3-3 paralogs (G / y and S / o) in ag<

[0011] Proteomics of AD brain aggregates showed co-aggregation of 14-3-3 paralogs along with major seed proteins A|3 and tau. Knockdowns of these interacting partners were evaluated for amelioration of aggregate burden and rescue of associated physiological functions. Moreover, the identified regions of proximity between 14-3-3 paralogs and their ligands can be used as therapeutic targets for disruption by protein-protein interaction inhibitors (PPI Is). We identified candidate drugs to target the interface between 14-3-3G / y and hexokinase, its principal interacting partner, to reduce aggregation in human-cell and C. elegans models of neuropathic aggregation and neurodegeneration.

[0012] Coronary Heart disease (CHD) and cognitive impairment are two of the most highly prevalent and debilitating diseases in the world. Contributing factors include aging, hypertension, obesity and smoking. CHD is the most frequent cause of human morbidity and mortality worldwide, and deaths from cardiovascular disease exceed deaths from all other causes combined [AHA, CDC statistics], CHD and its sequelae including heart failure (HF) and myocardial infarction (Ml), predispose to neurodegeneration and dementias, such as Alzheimer’s disease (AD), especially in the elderly. . Hippocampal and prefrontal abnormalities similar to AD are found in HF animal models and human subjects.

[0013] We previously reported that 14-3-3 proteins are more abundant in hippocampal aggregates from AD than from AMC patients, and also increase with adult age in aggregates from human skeletal muscle. The 14-3-3G (y or gamma) paralog colocalizes with neurofibrillary tangles in AD hippocampi, but is less abundant in the frontal cortex of AD patients vs. controls, assessed at autopsy. Our published work on protein aggregation identifies several proteins, specifically including the 14-3-3 family that are significantly enriched in aggregates isolated from AD hippocampi and also from heart and brain after Ml, relative to age-matched controls. These paralogous proteins interact with several other proteins in both cardiac and non-cardiac tissues, and have been shown to regulate multiple signaling pathways, especially those relating to protein homeostasis and stress responses. The 14-3-3 interaction networks thus provide therapeutic targets involved in metabolic and critical proteos and neuropathologies. Hyperphosphorylated tau (hP-tau), a major constituent of neurofibrillary tangles in Alzheimer’s disease, is a critical neuropathological binding partner of 14-3-3Z ( or zeta). A recent study reported both hyperphosphorylated tau and a cardiac-specific, high-molecular-weight isoform of tau in aggregates isolated from hearts of HF and AD patients. We found increased levels of hexokinase-1 and several 14-3-3 paralogs, in AD hippocampus and also in hearts and brains of mice after simulated Ml. Hexokinase and 14-3-3G proteins interact in brains from AD patients, but not in control brains.

[0014] In the current study we found ezetimibe, an FDA-approved drug to block interaction between 14-3-3G and hexokinase-1 , serving as a protein-protein interaction (PPI) inhibitor. We evaluated effects of ezetimibe under normal conditions, and after hypoxia to simulate transient ischemia. We find that ezetimibe reduces aggregation and ER stress in both human-cell and C. elegans models of AD-like aggregation, while activating autophagy. By analyzing the IQVIA clinical database, we found a significant, 7-to 8-fold reduction in relative risk of Alzheimer’s disease and related dementias (ADRD) among heart-disease patients prescribed ezetimibe.

[0015] These and other features, objects and advantages of the present invention will become better understood from a consideration of the following detailed description of the preferred embodiments and appended claims in conjunction with the drawings as described following:

[0016] DISCLOSURE OF THE INVENTION

[0017] In one preferred embodiment, the present invention is directed to a method of treating and preventing Alzheimer’s disease and related dementias. The method includes administering to a patient an effective amount of a drug that is operable to inhibit an interaction between 14-3-3G protein and hexokinase-1 protein in the patient. The drug may be ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astern izole.

[0018] In another preferred embodiment, the present invention is directed to a method of treating a patient with age-related dementia. The method includes administering to the patient a an interaction between 14-3-3G protein and hexokinase-1 protein in the patient. The drug may be ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astemizole.

[0019] In yet another embodiment, the present invention is directed to a method of treating a dementia patient having heart disease. The method includes administering to the dementia patient an effective amount of a drug that is operable to inhibit an interaction between 14-3-3G protein and hexokinase-1 protein in the patient. The drug may be ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astemizole.

[0020] The mammalian 14-3-3 family comprises seven intrinsically unstructured, evolutionarily conserved proteins that bind more than 200 protein partners, thereby modulating cell-signaling pathways. Presence of 14-3-3 proteins in cerebrospinal fluid provides a sensitive and specific biomarker of neuronal damage associated with Alzheimer's disease (AD), Creutzfeldt-Jakob disease (CJD), spongiform encephalitis, brain cancers, and stroke. We observed significant enrichment of 14-3- 3 paralogs G, S, and Z in human brain aggregates diagnostic of AD. We used intraaggregate crosslinking to identify 14-3-3 interaction partners, all of which were significantly enriched in AD brain aggregates relative to controls. We screened FDA-approved drugs in silica for structures predicted to target the 14-3- 3G / hexokinase interface, an interaction specific to aggregates and AD. C. elegans possesses only two 14-3-3 orthologs, which bind diverse proteins including DAF-16 (a FOXO transcription factor) and SIR-2.1 (a sensor of nutrients and stress), influencing lifespan. Top drug candidates were tested in C. elegans models of neurodegeneration-associated aggregation, and in a human neuroblastoma cellculture model of AD-like amyloidosis. Several drugs opposed aggregation in all models assessed, and rescued behavioral deficits in C. elegans AD-like neuropathy models, suggesting that 14-3-3 proteins are instrumental in aggregate accrual, and supporting advancement of drugs targeting 14-3-3 protein complexes with their partners.

[0021] Heart-disease patients have elevated risk of subsequent cognitive impairment and dementia, including Alzheimer’s disease / AD. We found increased protein aggregation, a diagnostic net myocardial-infarction mice. We also observed hippocampal aggregates resembling AD-brain inclusions, in heart-disease brains, suggesting common mechanisms. By crosslinking brain aggregates from heart-disease and AD patients’ hippocampi, we observed numerous shared protein -protein interactions involving 14-3-3-family proteins. The 14-3-3G / y paralog, implicated in neurodegeneration and neuronal signaling, is enriched in brain aggregates from heart-disease and AD (each P<0.008). Hexokinase-1 / HXK1 adheres strongly to 14-3-3G, yielding abundant crosslinks in AD and coronary artery disease (CAD) aggregates, undetected in age-matched controls. Screening FDA-approved drugs in silico for predicted interfacial binding, we identified ezetimibe, a well-tolerated lipid-lowering medication. Diverse AD models support ezetimibe as a potent autophagy activator and in-vivo disruptor of 14-3-3G::HXK1 adhesion. Mining clinical databases supported this off-label benefit of ezetimibe, which lowers ADRD relative risk (RR) to 0.12, both overall (P<0.0001 ) and in high-risk heart-disease subjects (P<0.006).

[0022] BRIEF DESCRIPTION OF THE DRAWINGS

[0023] FIGS. 1A-1 D show molecular-dynamic simulations and interactome analyses of 14-3-3 paralogs. FIG. 1A shows RMSD trajectories of 14-3-3 paralogs in 200-ns simulations, where RMSD is calculated as the sum of root mean square deviations for all molecular atoms from their initial center of mass. FIG. 1 B shows interactome subnetworks of 14-3-3G direct contacts, within amyloid beta aggregates isolated from AMC and AD hippocampi, respectively (as indicated). FIG. 1 C shows interactome subnetworks of 14-3-3S in amyloid beta aggregates isolated from AMC and AD respectively. FIG. 1 D shows Venn diagrams illustrating the 14-3-3G- and 14- 3-3S-interacting protein counts (i.e., degrees) in hippocampal aggregates from AMC, AD, or shared in common.

[0024] FIGS. 2A-2E show that siRNA knockdowns of influential aggregate hubs reduce aggregation in human cells. FIG. 2A shows typical fluorescence images of Thioflavin T-stained SH-SY5Y-APPsw cells (human neuroblastoma cells expressing a double mutant of Amyloid Precursor Protein, observed in a familial-AD pedigree) after liposomal transfection with siRNA constructs targeting 14-3-3 paralogs or their interacting partners. FIG. 2B staining fluorescence of amyloid as illustrated in FIG. 2A. FIG. 2C is a histogram showing means ± SEM for the intensity of SYPRO-Ruby-stained gel proteins, from sarkosyl-insoluble aggregates isolated from SY5Y-APPsw cells after exposure to RNAi constructs. FIG. 2D shows a typical Western blot probed with antibodies to GRP78, 14-3-3 or GAPDH (mediators of endoplasmic reticulum [ER] stress); gels were loaded with total protein extracted from SH-SY5Y-APPsw cells treated with the RNAi constructs indicated. FIG. 2E is a histogram showing means ± SEM of band intensity for 14-3-3 or GRP78, each normalized to GAPDH (a within-lane control protein). For FIGS. 2B, 2C, and 2E, significance by 2-tailed t tests was as follows: *P <0.05; **P <0.005; ***p < 10 -7; ****P < 1Q -17For each assay, 100 < N < 150.

[0025] FIGS. 3A-3C show that knockdowns of influential aggregate hubs reduce aggregation in C. elegans. FIG. 3A is a histogram showing total intensity of aggregates in C. elegans strain AM 141 adults after exposure (starting in late larvae) to RNAi constructs; 13 < N < 18. FIG. 3B comprises images of AM141 worms showing decreased size and number of aggregates after exposure to RNAi constructs as indicated. FIG. 3C is a histogram showing chemotaxis of C. elegans strain CL2355 adults (100 < N < 200) after treatment with the indicated RNAi constructs. Significance of chemotaxis rescue (relative to control, “CON”) by 2-tailed t tests in FIG. 3A, or Fisher Exact tests in FIG. 3C: **P < 10-3; ***p < 10-7.

[0026] FIGS. 4A-4D support our prediction of an avid interaction between 14-3-3G and hexokinase-1 (HXK1 ). FIG. 4A is a graph showing AG, the predicted binding free energies of 14-3-3G with each of the principal interacting proteins. FIG. 4B shows RMSD plots (as in FIG. 1A) of 14-3-3G bound to hexokinase-1 over a 200-ns Molecular Dynamic simulation. FIG. 4C is a protein-structure diagram showing a 14- 3-3 G::hexokinase-1 complex, indicating a protein-protein interaction region (green) based on a simulation “snapshot” at 70 ns. FIG. 4D is a graph showing the volume of the 14-3-3G::HXK1 binding pocket as it increases between 0 and 100 ns.

[0027] FIGS. 5A-5D show that protein aggregation or its sequelae are reduced by drugs that target the 14-3-3G::hexokinase interface. FIGS. 5A and 5C are structural “ribbon diagrams” showing predicted binding sites of repurposed-drug candidates lumacaftor (FIG. 5A) and con and 14-3-3G. FIGS. 5B and 5D are RMSD plots (as in FIG. 1A) of the 14-3- 3G::HXK1 complex, as predicted in the absence (orange) or presence (blue) of lumacaftor (FIG. 5B) and conivaptan (FIG. 5D), as simulated over 200-ns timespans.

[0028] FIGS. 6A-6B show that drugs targeting the hexokinase:: 14-3-3 interface reduce aggregation in C. elegans models of human neuropathology. FIG. 6A is a histogram showing calculated average aggregate counts per worm in C. elegans strain AM141 (q40::yfp), a model of HD protein aggregation in which muscle expression of polyglutamine (Q40) fused in-frame to the gene encoding yellow fluorescent protein (YFP) leads to age-dependent accrual of yellow-fluorescent aggregates. YFP+aggregates were counted in adult worms on day 5 post-hatch; significances were determined by 2-tailed t tests with 13 < N < 18. FIG. 6B shows the chemotaxis Index of C. elegans strain CL2355, in which uninduced pan-neuronal expression of human A|3I-42 causes age-dependent loss of chemotaxis [15, 16], and 3 of the 5 tested drugs (here assessed on day 5 post-hatch) demonstrate rescue of chemotaxis to levels typical of younger adult worms. Significances of differences between treated and control worms, calculated by Fisher Exact tests (100< N <200): **P<0.005; ***P<0.0005.

[0029] FIGS. 7A-7F illustrate drug effects in human cells. FIGS. 7A-7C show doseresponse curves for SY5Y-APPsw human neuroblastoma cells treated with the topranked 3 drugs at 0.01 , 0.1 , or 1 pM. Cells were stained with Thioflavin T and DAPI to quantify amyloid per cell. Significances by 2-tailed f tests (100 < N < 150): ***P<10-5. FIGS. 7D-7E show typical gels showing loaded with sarkosyl-insoluble (FIG. 7D) and sarkosyl-soluble aggregates (FIG. 7E) from SY5Y-APPsw cells. Sarkosyl-insoluble and sarkosyl-soluble aggregates from SY5Y-APPsw cells were resuspended in Laemmli buffer at >95°C, and electrophoresed on acrylamide gels. Gels were stained with SYPRO-Ruby and protein quantified as total fluorescence per lane. FIG. 7F summarizes results in a histogram showing sarkosyl-insoluble and sarkosyl-soluble aggregate proteins from gels such as are illustrated in panels 7D and 7E. Significance by 1 -tailed heteroscedastic t tests ( / V = 4): *Pt=0.01 . FIGS. 8A-8D demonst proteins in human neuroblastoma cells. FIG. 8A shows western blots probed with antibodies indicated. Acrylamide gels were loaded with 50 pg total protein per lane, extracted from SH-SY5Y-APPsw cells after a 48-hour exposure to FDA-approved drugs lumacaftor or conivaptan. Proteins were resolved by gel electrophoresis, transferred to membranes (Western blots) and probed with antibodies to the 14-3-3 conserved core, hexokinase, or GAPDH. FIG. 8B provides a histogram showing mean ± SD band intensities of hexokinase lanes ( / V = 3), normalized to the corresponding 14-3-3 band in the same lane. FIG. 8C shows western blots used to evaluate in vivo interaction between hexokinase-1 and 14-3-3 proteins. Antibody to the 14-3-3 conserved core was used for immuno-pulldown (IP) of 14-3-3 complexes, from neuroblastoma cells treated 48 hr. with drugs or untreated. Gel lanes were loaded with protein recovered from antibody-coated magnetic beads immersed in SH-SY5Y-APPsw cell lysates. Proteins were recovered from beads, electrophoresed, and western blots probed to detect hexokinase or 14-3-3 proteins and quantify their relative yields. FIG. 8D contains histograms of HXK1 recovery (signal on western blots as in 8C), normalized to 14-3-3 protein recovery in the same lane; band signal is plotted for mean ±SD band-intensity ratios (14-3-3 / HXK1 ) of replicate samples. Significances by 2-tailed heteroscedastic t tests (8B and 8D), each N-4 **P < 10-2; ***P < 10-3.

[0030] FIG. 9 presents a flowchart for in silica and in vivo screening of FDA- approved drugs.

[0031] FIG. 10 is a table identifying spectral hits for three paralogs of 14-3-3 in AD and AMC aggregates, isolated by immunoprecipitation (IP) with antibodies to seed proteins A[3i-42 or tau, or “total” aggregates isolated without IP.

[0032] FIG. 11 is a table identifying top drug candidates along with their binding scores and the conditions they were FDA-approved to treat.

[0033] FIGS. 12A-D show that immunoprecipitation with 14-3-3 antibody followed by isolation of sarcosyl-insoluble aggregates reveals significantly elevated 14-3-3 brain aggregate proteins in heart-disease tissue. *Adjusted P<0.04 by heterscedastic t test, N = 5 - 6 per group. FIG. 12A is a gel showing the insoluble aggregates isolated from 14-3-3 bound p quantification of the aggregate gel shown in FIG. 12A. FIG 12C is a SDS-PAGE gel showing that detergent-soluble and insoluble aggregates are reduced in C. elegans lacking 14-3-3 (ftt-2 KO, vs. WT (“+”). FIG. 12D is a chart showing quantitation of the soluble and insoluble aggregates in FIG. 12C. *Adjusted P<0.01 by heterscedastic t test, N = 3

[0034] FIG. 13A shows histograms illustrating enrichment of 14-3-3G and hexokinase-1 in aggregates from AD & heart disease (HtD) hippocampi. FIG. 13B shows an interactome subnetwork of 14-3-3G and its direct contacts, derived by crosslinking A[3-IP aggregates from AMC vs AD brain. FIG. 13C is a western of the immunoprecipitated protein with 14-3-3 and hexokinase antibodies. Co-immuno- precipitation (co-IP) of 14-3-3 recovers less hexokinase (lower HXK1 / 14-3-3 ratio) from cells treated 48hours with 10-pM ezetimibe. FIG. 13D is an estimate of binding free energy prioritized for one of the top 14-3-3G::HXK1 interface binding drug ezetimibe. Molecular-dynamic simulations of this protein-protein interactions confirmed a stable 14-3-3 and hexokinase-1 interaction in the absence of drug. FDA- approved drug ezetimibe binds avidly to the 14-3-3G:: hexokinase interface. Predicted AG values are shown for ezetimibe binding at the 14-3-3G::HXK1 interface. FIG. 13E is a model of 14-3-3G (blue) adhesion to HXK1 (green). The crosslink-based interface, coinciding with the druggable pocket targeted in library screens, is highlighted in maroon. Simulation snapshots at 0 and 100 ns illustrate ezetimibe disruption of 1433G::HXK1 binding. FIG. 13F is a plot of root-mean- square deviation from initial atomic positions of 14-3-3G::HXK1 over a span of 100 nsec.

[0035] FIGS. 14A-B are histograms and images that show that ezetimibe reduces amyloid aggregation in SY5Y-APPSw (FIG. 14A) and HEK-tau cells (FIG. 14B). Cells were imaged after 48 hr. of treatment with ezetimibe. **P<0.01 ; ***P<1 E - 5, by 2-tailed t tests (each N > 30). Green fluorescence reflects Thioflavin T staining; blue fluorescence shows DAPI staining of nuclear DNA. FIG. 14C is a histogram and images showing that SY5Y-Tau cells treated with Ezetimibe and stained with proteostat shows reduction in overall aggregation with a ***P <0.001. FIGS. 14D-E are a histogram (FIG. 14E) a enhances autophagy in SY5Y-APPsw neuroblastoma cells (*P<0.03).

[0036] FIGS. 15A-B are histograms that show that ezetimibe reduces aggregation in human NT2 (neuronal) (FIG. 15A) and T98G (glial) cells (FIG. 15B). Significance of differences (by 2-tailed t tests, N >30): *P<0.05; **P<0.005. FIG. 15C is a histogram showing that ezetimibe treatment of C. elegans strain CL2355 improves chemotaxis ~2-fold. *P<0.05; ***P<1 E- 4 by 2-tailed heteroscedastic t tests. FIG. 15D is the histogram showing decreased number of aggregates in Huntington model AM 141 after exposure to ezetimibe. FIG. 15E is graph showing VH255 shows a reduced paralysis after the treatment with ezetimibe. FIG. 15F are images showing that ezetimibe (10 pM) reduces amyloid burden in transgenic C. elegans expressing A|3::mCherry (P<0.008). FIG. 15G is a histogram showing that ezetimibe (10 pM) reduces amyloid burden in transgenic C. elegans expressing A[3::mCherry.

[0037] FIG. 16 is a table showing the general characteristics of populations compared for relative risk calculations.

[0038] FIG. 17 is a table showing an analysis of baseline characteristics and ADRD incidence among the subset of FIG. 16 patients diagnosed with CHD.

[0039] DETAILED DESCRIPTION OF THE INVENTION

[0040] With reference to FIGS. 1A-17, the preferred embodiments of the present invention may be described. As described below, the present invention is described in studies identified as Part 1 and Part 2.

[0041] Part 1 of the Study:

[0042] Significant enrichment of 14-3-3 paralogs Sigma, Gamma and Zeta in AD: We isolated detergent-insoluble aggregates from AD and AMC hippocampus after immunoprecipitation with antibodies to amyloid beta or tau, seed proteins implicated in AD. FIG. 10 lists the number of spectral hits showing the relative abundance of 14-3-3 paralogs in AMC and AD.

[0043] Molecular-Dynamic simulations and interactome analyses of 14-3-3 paralogs: Previous studies demonstrated that 14-3-3 proteins have disordered N- and C- termini as monomers, but attain stable conformations when bound to a target peptide or phosphopeptide. We obtained experimental structures for the conserved core regions of three of the s o / S / sigma, and ZJ Z / Zeta,) in either peptide-bound monomeric or dimeric state, from the Protein Data Bank (PDB IDs 5N10, 4046, and 6QHL, respectively). In order to explore the structural stability of these proteins as monomers, we conducted 200-ns atomistic molecular dynamic simulations in triplicate. Based on plots of root-mean- square deviation (RMSD) vs. time (FIG. 1A), none of these 14-3-3 monomeric core regions (a / alpha, y / gamma, and ^ / Zeta) attained a stable conformation within 200 ns.

[0044] In view of their disordered end domains, we expected the 14-3-3 proteins to bind multiple protein partners. We previously used aggregate-permeant “click” reagents to cross-link neighboring proteins within amyloid aggregates from human SY5Y-APPsw neuroblastoma cells, and subsequently employed the same protocol to aggregates immunopurified from human (AMC, AD, heart-disease) hippocampus. In the latter study, we observed over twice as many 14-3-3 interacting partners in AD as in AMC aggregates (FIGS. 1 B-C). Consistent with our previous work in defining aggregate interactomes, we found that 85 - 87% of AD |3-amyloid proteins in close proximity to 14-3-3G or 14-3-3S are absent from AMC aggregates (FIG. 1 D). Several proteins interacting with 14-3-3G predominantly in AD are shown in FIGS. 1A-D along with their relative abundances (spectral counts) in AD vs. AMC aggregates. Our analysis of the AD interactome indicates that 14-3-3G interacts directly with tau (TAU) and other aggregate proteins, including ankyrin-3 (ANK3), plectin (PLEC), kinesin heavy chain 5C (KIF5C), and hexokinase (HXK1 ) (FIG. 1 B- C) — all of which play key roles in neurodegenerative-disease aggregation.

[0045] Identification of top 14-3-3 interactions as potential targets to disrupt aggregation: We assessed siRNA knockdowns targeting 14-3-3G, 14-3-3Z, and 4 of their most abundant interacting proteins (ankyrin, hexokinase, kinesin, and plectin) with high AD / AMC ratios and interactome influence >2, predicted by neural-network analysis. A fifth interacting partner of 14-3-3, tau, could not be pursued because we were unable to find an FDA-approved drug binding to the interface of these two disordered proteins. SH-SY5Y-APPsw human neuroblastoma cells were transfected with siRNA constructs specifically targeting the genes encoding each of these 14-3-3 binding partners. The impact on amyloid aggregation of each knockdown was quantified by thioflavin T stair suppressed 20 - 50% by these siRNA treatments (FIGS. 2A-B). We also quantified levels of total sarkosyl-insoluble protein recovered from these cells, after acrylamide gel electrophoresis and SYPRO-Ruby staining; protein per lane was reduced 15 - 30% after knockdowns (FIG. 2C). Total 14-3-3 protein levels in aggregates also declined after knockdowns of interacting proteins (FIG. 2D), in contrast to levels of GRP78, a key mediator of the endoplasmic reticulum unfolded protein response (UPRER), which were not altered significantly by two of these knockdowns, and only modestly (-25%) by a third (FIGS. 2D-E).

[0046] Knockdown of key interacting partners of 14-3-3, tested in nematodes: We next assessed the impact of RNAi-mediated knockdowns targeting the closest C. elegans orthologs of the above proteins in two nematode models of neurodegenerative aggregation: strain AM141 , expressing a polyglutamine-array marker (Q40::YFP) in muscle cells as a model of huntingtin-like aggregation in Huntington’s disease (HD) neurons; and CL2355, which forms AD-like amyloid deposits in neurons expressing human Api-42. In the Huntington model, total aggregate intensity per worm was reduced 40 - 63% by RNAi-mediated knockdown relative to empty-feeding-vector controls (FIG. 3A). Worm images (FIG. 3B) illustrate the decrease in both number and fluorescence intensity (Q40::YFP content) of aggregates. We also assessed these knockdowns in worms expressing neuronal |3-amyloid (CL2355), which show reduced chemotaxis, a measure of chemosensory response. The decline in chemotaxis, due to neuron-specific induction of a human A|342 transgene and subsequent AD-like amyloid deposition, leaves just over a third of worms moving toward the n-butanol chemo-attractant. Knockdown of each target elicited substantial and significant rescue of chemotaxis relative to control worms, elevating their levels from 36% (FV controls) to 53-80% chemotaxis (FIG. 3C). These interventions thus restored 27 - 70% of the deficit relative to wild-type worms (99 ±1 % chemotaxis, not shown).

[0047] Molecular modeling of 14-3-3 interactions with key interacting proteins: We performed simulations to predict structural variation that may affect aggregate adhesions of 14-3-3G / y to either hexokinase (HXK) or kinesin heavy chains (KIF5C, KINH / KIF5B), especially the I influence scores. Structures of hexokinase and kinesin were obtained from PDB (www.rcsb.org) and l-TASSER, respectively. We performed protein-protein dockings in Hex v 8.0 (FIG. 4A), which predicted far greater stability of the 14-3- 3y:: hexokinase (HXK) complex (-930 kcal / mol) than for either 14-3-3y::KIF5C (-340 kcal / mol) or 14-3-3y::KINH (-530 kcal / mol). We simulated the 14-3-3y::HXK complex for 200 nsec, and plotted the root-mean-squared deviation (RMSD) of atoms from their initial positions (FIG. 4B). The RMSD time course predicted by molecular- dynamic modeling suggests that the complex attains a moderately stable conformation. We selected a meta-stable “snapshot” of the complex at 70 ns, and predicted its binding pocket (FIGS. 4C-D).

[0048] In silica screening identifies drugs predicted to stably bind (and likely disrupt) 14-3-3:: hexokinase binding. Since the 14-3-3:: hexokinase interaction is enriched in AD aggregates, we posited that disruption of their complex by a small molecule may relieve protein aggregation by blocking an early-stage adhesion. Considering that both 14-3-3G and hexokinase are required for normal biological functions, targeting one or both of these proteins is likely to be toxic or to have adverse side effects. However, targeting the interface between 14-3-3 and hexokinase provides a feasible alternative that may safely oppose protein aggregation and its associated functional declines, without harmful side-effects. However, targeting the interface between 14- 3-3 and hexokinase provides a feasible alternative that may safely oppose protein aggregation and its associated functional declines. We therefore targeted a druggable pocket created at the 14-3-3y- hexokinase interface in the bound structure observed at 70 ns, a metastable point in the simulation when the binding pocket had begun to increase in volume (FIG. 4D). To identify drugs that may disrupt the interaction between 14-3-3y and hexokinase, we initially screened drugs from an FDA-approved library comprising more than 2300 compounds, via in silico docking simulations. To improve screening efficiency, docking was conducted in three stages, increasing the stringency for successive screens. We began with virtual docking of the entire FDA-approved drug library, using SiBiolead to run AutoDock in high-throughput mode. The top 10% of molecules (230 drugs) were then re-docked using Glide in its standard (T

[0049] Glide were pursued by assessing the free energy of each 14-3-3y::HXK::drug complex in implicit solvent, using the Schrodinger MM-GBSA module. The top 5 candidates from this third-stage analysis (FIG. 11 ) were pursued for further study.

[0050] Drug binding to predict disruption of the 14-3-3y::HXK complex: The above FDA-approved drugs were docked to the 14-3-3: :HXK binding pocket in atomistic molecular-dynamic simulations. The predicted binding sites and poses of the predicted top two drugs (lumacaftor and conivaptan) are shown in FIGS. 5A-B. The ability of these drugs to disrupt the interaction between 14-3-3y and human hexokinase (HXK) was predicted using molecular-dynamic simulations. FIGS. 5C-D depict RMSD during a 200-nsec simulation, for the 14-3-3G::HXK complex with or without binding of lumacaftor or conivaptan at the protein-protein interface. The RMSD plots illustrate progressive expansion of the 14-3-3G::HXK interaction complex in the presence of either drug, relative to the 14-3-3 :: HXK complex alone. These predictions support pursuit of drugs conivaptan and lumacaftor as candidates to disrupt the 14-3-3 :: HXK interaction, and thereby relieve protein aggregation.

[0051] Top-ranked drugs rescue C. elegans and human-cell aggregation models: All 5 top-ranked drugs (FIG. 11 ) were tested for rescue of protein aggregation in C. elegans strain AM 141 (a model of polyglutamine aggregation in Huntington’s disease). The total intensity of aggregates was assessed after exposure of nematodes to each drug at 2 concentrations. The top 2 drugs (conivaptan and lumacaftor, each at 10 pM) reduced protein aggregation by 60-70% (FIG. 6A). A third drug, asfemilzole at 10 pM, reduced aggregation ~25%. We also tested these drugs in a C. elegans model of AD-like amyloidosis, expressing neuronal A[3I-42 and consequently suffering impaired chemotaxis. Chemoattraction to n-butanol was 38.5% for untreated worms (vs. 99±1 % for worms not expressing A|3I-42), but restored to 82% by conivaptan, 72% by digitoxin, and 60% by lumacaftor (each at 10 pM; FIG. 6B).

[0052] These drugs were next tested on human neuroblastoma (SH-SY5Y-APPsw) cells, exposed to each drug for 48 h at 3 concentrations (0.01 , 0.1 , and 1 pM). Conivaptan at 1 pM reduced amyloid fluorescence per cell by 52% as quantified by thioflavin staining, while lume

[0053] C). We pursued these results by isolating aggregates from SY5Y-APPsw cells treated with 0.1 -pM conivaptan or lumacaftor, which reduced sarkosyl-insoluble and -soluble aggregates 30-40% and 35-45%, respectively (FIGS. 7D-F).

[0054] Hexokinase is recovered from neuroblastoma cells by 14-3-3 immunopulldown: We then asked whether human neuroblastoma (SH-SY5Y-APPsw) cells expressing the APPsw double mutant observed in familial AD contain aggregates in which any 1 -3-3 protein interacts with hexokinase. We first quantified 14-3-3 paralogs and hexokinase in cell lysates to ensure that their levels were not severely depleted by the drug. This appears to be the case (FIGS. 8A-B); although several treatment groups differed significantly from controls, the difference was less than 20%. We used a co-immunopulldown (co-IP) strategy to recover and quantify binding of 14-3-3 to hexokinase. Complexes or aggregates isolated by IP using magnetic beads coated with antibody to the conserved 14-3-3 core were recovered and resuspended in hot loading buffer, electrophoresed on acrylamide gels, and their western blots probed with antibody to hexokinase or the 14-3-3 conserved core. Co-IP of hexokinase (normalized to 14-3-3 recovery) was reduced more than 45% by treating with drugs that are specific for binding the interface of 14-3-3 and hexokinase (FIGS. 8C-D).

[0055] Discussion: Previous studies have documented critical roles played by 14-3- 3 proteins in diverse neurological and other age-associated disorders. In eukaryotes, 14-3-3 paralogs are conserved adapter proteins involved in multiple physiological processes such as signal transduction, translation, protein trafficking and apoptosis. Deficiencies of specific 14-3-3 paralogs in knockout mice result in neurotransmitter imbalance and altered behavior that has been likened to schizophrenia.

[0056] In the current study we used computational methods to predict disordered regions of 14-3-3 paralogs that fail to attain stable conformations on their own, resulting in indeterminate (or partner-determined) structures. These paralogs are rich in basic amino acid residues, aromatic amino acids and amphipathic amino acids relative to acidic amino acids. We observed an increased sequestration of 14- 3-3 proteins into sarkosyl-insi states, both in heart and brain. In order to target protein-protein interaction interfaces for the treatment of neurodegenerative diseases, including Alzheimer’s disease, we proposed to use small molecules as protein-protein interaction inhibitors, to counteract aggregate progression by breaking critical interactions needed for aggregate growth.

[0057] We first identified 14-3-3 interacting partners by performing crosslinking analysis of hippocampal aggregate proteins isolated from Alzheimer’s disease vs. age-matched controls. A total of 85 interacting proteins were associated with 14-3-3 only in tissue from AD, whereas only 26 protein partners were unique to controls. Previous studies showed that 14-3-30 (theta) acts as a chaperone to assist in refolding of disordered proteins such as a-synuclein, thus reducing its toxicity; 14-3- 30-overexpressing mice are protected from toxic effects of a-synuclein fibrils. AD is characterized by tau hyperphosphorylation, contributing to elevated tau in paired helical filaments (PHF) and tangles in neurons. Prior studies showed 14-3-3 interaction with hyperphosphorylated tau (hP-tau), preventing access to phosphatases and thus indirectly promoting its aggregation and PHF formation. Several 14-3-3 paralogs were also shown to interact with polyglutamate (polyQ), and 14-3-3^ (zeta) plays an essential pro-aggregation role in a cell culture model of Huntington’s disease.

[0058] Since disruption of 14-3-3 protein results in the development of diabetic cardiomyopathy, and neurodegeneration, we used an approach that target proteinprotein interactions unique to the disease state. Pharmacological disruption of protein-protein interactions should not affect the normal biological functions mediated by either of the interacting proteins, and thus hold the promise of very limited side effects. The validity of our approach is supported by the screening of an FDA-approved drug library, leading to discovery of existing drugs that could be repurposed to prevent specific pro-aggregative interactions of 14-3-3 paralogs (See FIG. 9). This provides an express route to novel therapeutics available immediately, but at the same time offers the promise that screening of large structural libraries of small molecules is likely to lead to even better drug candidates with highly specific (and non-essential) targets, paper.

[0059] In conclusion, we have demonstrated that drugs targeting the interfaces of 14-3-3 paralogs with their interacting partners show promise to reduce aggregation and improve associated physiological functions. Such disease-specific proteinprotein interaction inhibitors have the potential to prevent, slow, or reverse aggregation associated with neurodegenerative diseases and other age-progressive disorders.

[0060] Materials and Methods: C. elegans strains: Transgenic C. elegans strains used in this study were obtained from the Caenorhabditis Genetics Center (CGC; Minneapolis, MN), and maintained at 20°C on 2% (w / v) agar plates containing 1x nematode growth medium, and overlaid with a central lawn of E. coli strain OP50 (unless otherwise noted). Two nematode models of neuropathic aggregation used in this study were ( / .) strain CL2355 (dvls50 [pCL45 (snb-1 ::Ap 1 -42::3' UTR(long) + mtl-2::GFP]), which drives pan-neuronal expression of human A|3I-42 peptide, and ( / / .) strain AM141 (rmls133 [unc-54p::Q40::YFP]), in which Q40::YFP synthesis in body wall muscle cells serves as a model of protein aggregation dependent on expansion of polyglutamine (polyQ) arrays as occurs in the huntingtin protein, resulting in amyloid-aggregate deposition.

[0061] Chemotaxis assays in an Ap-transgenic C. elegans strain: Transgenic C. elegans expressing AP1-42 in neurons (strain CL2355) were grown at 20°C with ample E. coli (OP50) bacteria, and adult worms were lysed to release unlaid eggs and thus to initiate a synchronized-aging cohort. Eggs were then placed on 100-mm NGM-agar Petri dishes seeded with RNAi-expressing bacteria (strain HT115) targeting orthologs of some of the interacting partners of 14-3-3 specific to AD, or empty-vector control bacteria. Worms at the L3 - L4 transition were upshifted to 25.5°C to induce human AP1-42 transgene expression and assayed after 48 hours. Alternatively (FIGS. 6A-B), worms were aged to adult day 5 without upshift. Chemotaxis to butanol and paralysis assays were performed as described previously. RNAi in C. elegans'. G interference (RNAi), by feeding dsRNA-expressing bacteria as described. Briefly, synchronized worms were placed, either immediately after hatch or at the late-L4 larval stage, onto IPTG-containing NGM plates seeded with bacteria (E. coli HT115[DE3]) carrying the empty L4440 vector (pPD 129.36), or with similar bacterial clones expressing double-strand RNA corresponding to exonic regions of unc-44 (encoding an ortholog of human ANK3), vab-10 (PLEC), exc-7 (NLICL) or unc-116 (KIF5C). Day-3 adult worms were imaged for aggregate count, assessed for paralysis, or evaluated for chemotaxis toward n-butanol. siRNA knockdowns of human cells and Thioflavin T staining to quantify amyloids: Exponentially growing cultures of SY5Y-APPsw (human neuroblastoma) cells were trypsinized and replated at 8,000 - 10,000 cells / well in 96-well plates, and grown for 16 h at 37°C in “DMEM+F12” (Life Technologies) medium supplemented with 10% (v / v) fetal bovine serum. When cells reached ~40% confluence, siRNAs were transfected by lipofection (RNAiMax, Life Technologies) to target genes encoding ANK3 (SAS1_Hs_00065571 ), YWHAZ (SASI_Hs01_00210839), YWHAG (SASI_Hs01_00201711 ), KIF5C (SAS1_Hs_ 00065571 ), PLEC (SAS1_Hs_00039321 ), or NUCL (SAS 1_Hs_00217223), all obtained from Millipore- Sigma (St. Louis, USA) and used as directed by the manufacturer. Transfected cells were assayed for protein aggregation 48 hours later, by fixation in 4% formaldehyde and staining with 0.1 % w / v Thioflavin-T. After 4 washes in PBS, cells were covered with Antifade + DAPI (Life Technologies) and fluorescence was captured in both blue and green channels (Keyence fluorescence microscope with motorized stage) for automated well-by-well imaging, 9 fields per well. Thioflavin-T fluorescence intensity was divided by the number of DAPI+nuclei to obtain normalized values (amyloid fluorescence per cell), summarized as mean ± SD.

[0062] Western-blot analysis of SH-SY5Y-APPsw aggregate proteins for 14-3-3, GRP78, GAPDH and Hexokinase: Human SH-SY5Y-APPsw neuroblastoma cells were maintained in cell culture medium (DMEM; Invitrogen / Life Technologies, Grand Island, NY), supplemented with 10% v / v fetal bovine serum (FBS). After treatment with either shRNA or drug candidate, cells were harvested and their proteins extracted in lysis buffer contc

[0063] Nonidet P40, 0.1 % SDS, and 0.5% sodium deoxycholate. Protein was quantified with Bradford reagent (Bio-Rad), and 50-pg protein aliquots were electrophoresed 2 hr at 100 V on 4-20% gradient bis-tris acrylamide gels (BioRad Life Science, Hercules, CA), and transferred to nitrocellulose membranes. Blots were blocked with BSA blocker (Pierce), and incubated overnight at 4°C with primary antibody to 14-3- 3 (ThermoFisher “14-3-3 Pan Antibody” 1 :5000 dilution), GRP78 (ThermoFisher, diluted to 2 pg / mL), GAPDH (ThermoFisher, 1 :2000 dilution), or hexokinase (ThermoFisher, 1 :1 ,000 dilution). After four washes of 6 minutes each, membranes were incubated 1 h at ~20°C with HRP-conjugated secondary antibody, either goat anti-rabbit IgG (AbCam, 1 :10,000 dilution), rabbit anti-goat IgG (Rockland Immunochemicals, Gilbertsville, PA; 1 :10,000 dilution), or goat anti-rabbit IgG (AbCam, 1 :3000 dilution), and developed using an ECL chemiluminescence detection kit (Pierce). Data were digitized, and analyzed using Imaged software (NIH).

[0064] Isolation of protein aggregates: Human cells treated with siRNA or drugs were collected, flash frozen in liquid nitrogen, and homogenized on ice in buffer containing nonionic detergent (1 % v / v NP40, 20-mM HEPES pH 7.4, 300-mM NaCI, 2-mM MgCL, and protease / phosphatase inhibitors [CalBiochem]). Lysates were centrifuged at 3000 rpm for 5 mins at 4°C to remove debris. Protein was sonicated to disrupt cell membranes and membrane-bound organelles. Following removal of cytosolic proteins (soluble in 1 % NP40 nonionic detergent) by centrifugation (18 min, 13,000 x g at 4°C), protein pellets were brought to pH 7.4 with 0.1 -M HEPES buffer containing 1 % v / v sarkosyl (sodium lauryl sarcosinate) and 5-mM EDTA, heated to 95°C for 10 min, and centrifuged 30 min at 100,000 x g. Sarkosyl-insoluble proteins (pellet fraction) were resuspended in Laemmli loading buffer (containing 50-mM dithiothreitol and 2% v / v SDS, sodium dodecyl sulfate), heated 5 min at >95°C, and resolved by electrophoresis on 4-20% polyacrylamide gradient gels with 1 % SDS. Gels were stained with SYPRO Ruby (ThermoFisher) or Coomassie Blue to visualize protein.

[0065] Molecular dynamic (MD) simulations of 14-3-3 paralogs: The x-ray crystallographic structures of 14-3-3 paralogs (14-3-3S / O, 14-3-3G / y, 14-3-3Z / ^) were obtained from the PDB

[0066] TASSER (https: / / zhanggroup.org / l-TASSER / ), an online server that uses foldrecognition and ab initio modeling

[0035] ;

[0036] , To understand the dynamic behavior of protein structure, target proteins interacting with 14-3-3 paralogs (e.g. hexokinase for 14-3-3G) were simulated using Schrodinger Maestro (version 11.9.011 ). Each protein goes through a preparation stage of preprocessing with pH constrained to 7.0 ± 2.0, after which an orthorhombic box is created around the protein with a minimized volume that varies with the protein. The system is neutralized using appropriate salt concentration (Na+, Cl-), and temperature and pressure are held constant (300°K; 1 .1023 bar). A random seed number is entered prior to starting each simulation, which is maintained for 200 ns (or as indicated), and replicated >3 times with new seeds.

[0067] MD simulation identifies a druggable target for 14-3-3G interaction with hexokinase: Docking studies were conducted using Hex, an interactive proteindocking and molecular-superposition program. Hex identifies the most stable interaction of two proteins (i.e., with the most negative AG). The complex of 14-3-3G with hexokinase was exported and simulated using Schrodinger Maestro. After MD simulation, the complex is captured, usually at an RMSD “plateau”. The druggable pocket or interface is determined using the Discovery Studio “Receptor Cavities” plug-in to predict likely druggable sites.

[0068] High-throughput computational screen to identify novel PPII molecules to disrupt the 14-3-3G - hexokinase interaction: The interface at which 14-3-3G and hexokinase interact was screened against two structural drug libraries, each at three successive stages of increasing stringency. The FDA-approved drug library (https: / / www.fda.gov / drugs) was prepared using Biovia Discovery Studio and was virtually screened as follows. The first screen was conducted with SIBiolead (www.sibiolead.com) running high-throughput screening with AutoDock. This program accepts proteins in PDB format as inputs, along with amino acid numbers that define a grid box around the targeted binding region. The user specifies the library to screen, and initiates the search. The time to completion depends on the sizes of the library and the protein. We retrieved the top 1 docking screen, as inputs for second-stage docking at higher stringency. This stage uses high-precision Glide docking, implemented within Schrodinger Maestro. Inputs for Glide docking include a protein specified in PDB format, and ligands in *.mae format. To convert ligands in the library from *.mol2 to *.mae format, we used the Ligprep plug-in under Maestro, which works in “batch” mode, returning ligands as a single file in *.mae format. Glide docking predicts binding poses for the drug with high precision. Grid boxes are formed around the receptor region, by specifying amino acid residue numbers; default values were used for all other parameters.

[0069] The top 10% of drugs from the second-stage screen were advanced to stage three, in which we use the MM-GBSA plug-in (molecular mechanics with generalized Born surface area) from Schrodinger Maestro, to predict binding free energies for all ligands based on inputs from Glide docking, employing core and ligand settings. The output lists top candidates for specific binding to the target — which can be either a protein or a protein -protein interface (FIG. 9).

[0070] Statistical analyses: For assays in which N was <10, differences between control and experimental groups were assessed for significance by the Fisher- Behrens heteroscedastic f-test (appropriate to samples of unequal or unknown variance). In some cases, significance of experimental reproducibility was also evaluated in this way, treating each experiment as a single point. For N>10, 2-tailed t tests were utilized. Within experiments, differences in proportions (fractional paralysis or chemotaxis) were assessed by chi-squared or Fisher exact tests, as appropriate based on sample counts. For assays with multiple end-points, the threshold for significance was adjusted to P<0.01 to reduce the frequency of type I errors.

[0071] Part 2 of the Study:

[0072] Aggregate proteins specific to 14-3-3 are enriched in heart disease and AD: Members of the 14-3-3 family of disordered proteins are depleted and / or sequestered into aggregates, and are also associated with diverse pathologies in mouse models of heart disease, diabetes, and diabetic cardiomyopathy, as well as AD. Using neural networks, we predicted 14-3-3 proteins to be influential in AD- specific aggregation. Disorde proteins and disruption of 14-3-3 interactions reduce protein aggregation and improves physiological functions. Since 14-3-3 proteins are enriched in AD aggregates, we isolated detergent-insoluble aggregates after immunoprecipitation with antibody to the 14-3-3 central core conserved among all paralogs. We found significant elevation of 14-3-3 aggregates in heart-disease brains relative to age- matched controls (FIGS. 12A-B). AD brain aggregates appear similarly elevated but the increase was not significant.

[0073] Deletion of a 14-3-3 gene in C. elegans reduces soluble and insoluble aggregates: Since 14-3-3 proteins are elevated in heart disease and AD, we asked whether genetic deletion (knockout) of one of the two C. elegans 14-3-3 genes would reduce protein aggregation. Both sarcosyl-soluble and sarcosyl-insoluble aggregates were isolated from a C. elegans 14-3-3 / / ft- 2 knockout strain (CGC). Protein aggregates were reduced 40 - 60% in day-3 adults relative to an isogenic wild-type control strain (Bristol-N2 [DRM]) (FIGS. 12C-D).

[0074] 14-3-3G and HXK1 are enriched in brain aggregates from heart-disease and AD patients: AD-specific aggregates, purified by immunoprecipitation (IP) with antibodies to A(3I-42 or tau, contain ~65% more protein in AD hippocampal samples than in age-matched control (AMC) aggregates. Post-translational modifications (PTMs: oxidation, phosphorylation, and acetylation) are also enriched in AD aggregates by 2.8- to 3.5-fold over AMC. We isolated total hippocampal aggregates from heart disease patients and observed many proteins and PTMs enriched to levels similar to or exceeding those seen in AD. Total 14-3-3 protein (the sum of 7 paralogs) was enriched 9-fold in AD, and 3.5-fold in heart-disease brains; 14-3-3G is elevated 2-fold in AD but ~3-fold in heart-disease brain aggregates (FIG. 13A). Hexokinase-1 , a prominent interacting partner of 14-3-3G, was absent from control aggregates, but was significantly enriched in AD brain aggregates, and was especially elevated (>10-fold the amount observed in AD) in brain aggregates from heart-disease patients (FIG. 13A).

[0075] Machine-learning analyses of AD interactomes identify aggregate “lynchpin” PPIs: Interactome modeling identified numerous PPIs unique to AD (manuscript in press). To enable the identify initiating, augmenting, and stabilizing AD aggregates, we developed a graph / network simulation technique called “leave-one-out” analysis (LOOA) for PPI hubs (manuscript in review). Key PPI hubs in the AD interactome are sequentially removed (with replacement), and at each step connectivity is recalculated for the entire interactome. The contribution of a PPI to aggregate burden is thus reflected in the impact of PPI removal on the overall complexity of the entire interactome. The fold-decline in complexity (e.g., total degree) provides a valuable measure of influence for each node or interaction. LOOA predicts that 14-3-3G subnetworks (comprising all their direct contacts) include many proteins that are themselves influential for heart-disease or AD aggregate interactomes, implying therapeutic potential as drug targets to reduce aggregation (FIG. 13B).

[0076] Screening FDA-approved drugs in silica. We used our own in-house implementation of molecular-modeling tools to create a 3D model of the adherent protein pair, 14-3-3G:: HXK1 (FIG. 13E). Molecular-dynamic simulations of this PPI confirm stable protein-protein interaction in the absence of drug (FIG. 13E). The 14- 3-3G::HXK1 model then served as a target for computational multi-stage screening of structures for more than 1800 FDA-approved drugs (i.e. , all available drugs except chemotherapy agents, omitted due to their general genotoxicity) for predicted affinity to this protein-interface target. Estimates of binding free energy prioritized top drugs, ranked by avidity of binding to the PPI-interfacial surfaces of 14-3-3G::HXK1. The top-ranked candidates were pursued by literature and patent research, and in silica prediction of ADMET properties (e.g., solubility, mouse and human pharmacokinetics, and extent of blood-brain-barrier penetration) using Discovery Studio™. Based on these data, ezetimibe was selected for validation and pursuit as a candidate inhibitor of aggregation via disruption of 14-3-3G::HXK1 interaction (FIGS. 13D-F).

[0077] Ezetimibe blocks 14-3-3 interaction with HXK1 : To test the proposed mechanism of drug action, we asked whether ezetimibe impedes 14-3-3 interaction with HXK1 in aggregates from human cells. SH-SY5Y-APPsw neuroblastoma cells were treated with ezetimibe under normal and hypoxic conditions. Complexes containing 14-3-3 were recovered from ( pulldown complex of proteins was separated on SDS-PAGE and immobilized on a nitrocellulose membrane. Western blots were then probed with antibodies to 14-3-3 or HXK1. Ezetimibe blocked 14-3-3::HXK1 interaction so that 14-3-3 IP yielded only 36% as much bound HXK1 relative to 14-3-3 recovery (FIG. 13C, “14-3-3 bound”). This drug also reduced 14-3-3 interaction with HXK1 by >50% in nonhypoxic cells.

[0078] Human cultured cells are protected by ezetimibe from amyloid aggregate accrual, and the targeted PPI is disrupted: To evaluate protection of human cells by ezetimibe, SY5Y-APPsw, SY5Y-Tau, and HEK-tau cells were treated with ezetimibe at several concentrations, and then stained with thioflavin T to quantify amyloid aggregation. SY5Y-APPsw cells treated with 0.1 -pM ezetimibe had only ~40% as much aggregate as control cells exposed to vehicle without drug. HEK-tau and SY5Y — Tau cells exposed to 0.01 -pM ezetimibe had -60% as much aggregate accrual as control cells (FIGS. 14A-C shows data and images for optimal drug concentrations).

[0079] Ezetimibe enhances autophagy impacting protein aggregate levels: Autophagy plays a key role in the clearance of aggregates and damaged organelles. Defective autophagy has been repeatedly implicated as a mechanistic contributor to the development and progression of neurodegenerative disorders, due to failure to clear toxic protein aggregates from cells in which it is deficient. 14-3-3 proteins modulate autophagy by binding to regulatory proteins such as Beclin-1 . Given ample evidence that aggregation is alleviated by ezetimibe, acting as an inhibitor of 14-3-3G::hexokinase-1 interaction (FIGS. 13A-13E), we assessed its impact on autophagy. (FIG. 14D). Ezetimibe at 0.1 pM significantly enhanced autophagy in human SY5Y-APPsw neuroblastoma cells. Combined with our earlier evidence that this same cell line accrues less aggregate in the presence of ezetimibe, this suggests that ezetimibe either activates autophagy (perhaps by liberating 14-3-3G from aggregates), or reduces the demand for it through reduction in the aggregate load. Physiological levels of 14-3-3 were shown to be depleted in AD, thereby diminishing autophagy by disruption of 14-3-3 signaling. Ezetimibe reduces pre under hypoxia: Since 14-3-3G and HXK1 are enriched in heart-disease brain aggregates, we evaluated the aggregation-limiting effects of ezetimibe in human neuronal and glial cells. NT2 (neuronal) and T98G (glial) cells were exposed to ezetimibe at 0.01 - 1 pM for 48 hr, either with or without prior hypoxia (94% N2, 6% CO2, 0% O2 for 7 hr), and then stained aggregates with thioflavin T. While hypoxia itself increased aggregation by 18 - 40%, ezetimibe conferred significant protection against protein aggregation to both neuronal and glial cells, with or without hypoxia (FIGS. 15A and 15B).

[0080] Ezetimibe protects nematodes from amyloid aggregate accrual: C. elegans strain CL2355 is a model of amyloid aggregation due to neuronal expression of human AP1-42. This results in a chemotaxis defect, with age or upon late-larval induction. While chemotaxis is greater than 95% in young WT worms, it drops to 37% in AD model aged worms; exposure to 10-pM ezetimibe restores it to 72% (FIG. 15C; P<10-4). Strain VH255, a model of Alzheimer’s disease, expresses human tau in muscle show paralyze progressively with age. Ezetimibe rescues the paralysis in VH255 (FIG 15D). Aggregate count (FIG. 15E) and average YFP content (not shown) are each reduced ~20% by 10-pM ezetimibe.

[0081] Ezetimibe reduces aggregate burden in transgenic C. elegans expressing AP1-42:: mCherry: We previously showed that expression of human AP1-42 in C. elegans neurons induces amyloid-like deposits and reduces chemotaxis toward n- butanol, a potent chemoattractant. In a transgenic C. elegans model expressing human A|3I-42 fused in-frame to mCherry, ezetimibe reduced A|3 aggregation (FIG. 15F). A[3::mCherry worms were treated from hatch for 5 days with 10-pM ezetimibe or vehicle and then assessed for A|3-mCherry amyloid deposits, quantified in fluorescence images (see histogram in FIG. 15G). Ezetimibe significantly reduced A 42:: mCherry accrual in worms, and also elicited a 10% reduction in age- associated neuronal loss in wild-type C. elegans (data not shown).

[0082] Ezetimibe significantly reduces ADRD incidence in patients with Coronary Artery Disease (CAD) and among normal elderly subjects: We analyzed data in the PharMetrics-Plus IQVIA database, comprising 2006-2020 clinical data. From this database, we found 4361 pat matched controls (mean 65 ± 5 yr). Untreated and treated groups have comparable sex ratios (46 vs. 54% males), and prevalence of hypertension (37 vs. 33%) and diabetes (18 vs. 17%). Remarkably, the incidence of ADRD during follow-up differed by 8-fold (0.8 vs. 0.1 %), reflecting a relative risk (RR) of 0.12 for those on ezetimibe (FIG. 16, 95% confidence interval 0.02-0.88; P<0.0001 ). Among those with CHD, we identified 547 patients prescribed ezetimibe and 73,387 CAD / CHD age matched controls. In this cohort with a known elevated likelihood of subsequent AD or ADRD, the relative risk for ADRD diagnosis during followup was 0.122 for ezetimibe recipients relative to controls (FIG. 17, 95% Cl: 0.02 - 0.88; P<0.006). These findings suggest highly significant protection from ADRD, associated with ezetimibe use in individuals with and without CHD. While the precise mechanisms warrant further investigation, this data provides compelling evidence supporting further study of ezetimibe for neuroprotection in those at increased risk of ADRD.

[0083] Discussion: AD and heart disease are the two major public-health challenges affecting the general population, and especially those with predisposing conditions such as advanced age, hypertension, heart disease, diabetes, or the presence of an AP0E4 allele. Worldwide incidence of dementia is projected to increase from 57.4 million cases in 2019 to 153 million by 2050, mainly due to aging of the population. Cardiac aging and Ml tend to elevate reactive oxygen species, implicated in chronic inflammation and AD etiology. Inflammatory serum markers are elevated in AD, and neuro-inflammation is a common sequela of Ml. These features suggest early roles of inflammation in both Ml and AD, linking these diseases mechanistically. Age- associated chronic inflammation, even of moderate severity, contributes to formation of protein aggregates in and around neurons, which in turn augments inflammation, creating a positive-feedback loop or vicious cycle. Both aggregation and inflammation have been implicated in the onset and progression of neuro- degenerative disorders. We were the first to show significantly increased aggregation in both aged and hypertensive hearts. Heart disease and AD share the same risk factors, including age, hypertension, the e4 allele of apolipoprotein E (ApoE4), diabetes, and CVD, suggesting shared etiology. Common mechanisms contribute to both CVD and d together pose daunting healthcare and economic challenges. Increased cardiovascular dysfunction may lead to declines in working memory and episodic memory, and increased vascular lesions in brain, all of which are associated with neurodegeneration. However, the key molecular mechanisms linking heart disease with subsequent cognitive impairment remain to be elucidated, and the impact of Ml on protein aggregation in heart and brain has received little or no attention.

[0084] All paralogs of the 14-3-3 protein family share a conserved central domain comprising 9 helical segments, and divergent N- and C-terminal domains that are intrinsically disordered. Through these unstructured domains, in concert with the central “cradle”, these adapter proteins interact with diverse target proteins to regulate a variety of signaling pathways, including autophagy, cellular metabolism, and stress responses, with impacts on neurodegeneration. We have previously shown that 14-3-3 paralogs are enriched in aggregates from brain, muscle, and heart. Reduced levels of 14-3-3 proteins in celrebrospinal fluid is symptomatic of many disease states associated with neuronal damage such as stroke and AD. Immunostaining of AD-patient hippocampi for tau indicates colocalization of 14-3-3 proteins with neurofibrillary tangles. Individuals with polymorphisms in both 14-3-3 and tau were at 2.5-fold lower risk of AD. Interacting partners of 14-3-3 proteins, such as FOXO and TFEB transcription factors, coordinate physiological processes governing lifespan, stress responses, and autophagy; whereas binding partners such as the p53 tumor suppressor, p21 / WAF1 , and the p85 regulatory subunit of class-l PI3K, modulate cancer, cell proliferation and senescence, insulin and insulinlike signaling, and longevity. We observed not only the 14-3-3 proteins themselves, but also their interacting partners, are enriched in AD and heart disease.

[0085] RNAi knockdown of either 14-3-3 proteins or one of their critical diseasespecific interacting partners reduces the aggregate burden. This strategy cannot be employed to remediate diseases or to slow their progression, in the light of substantial evidence that disruption of 14-3-3 proteins causes or exacerbates diabetic cardiomyopathy and neurodegeneration. We therefore opted not to target the active sites of these prote pathological aggregates — 14-3-3G::hexokinase-1 . This protein-protein interaction is abundant in AD aggregates but absent from healthy brains of the same age, implying that it is disease-specific but not symptomatic of aging per se. While smallmolecules have been shown to disrupt non-aggregate PPIs, ameliorating AD pathology by blocking functional interactions of Keapl with Nrf2 and APP with Mint2, our goal was to disrupt aggregation at an early stage through this strategy.

[0086] We performed a 3-stage computational screen of the 14-3-3G::hexokinase-1 interface, the most influential protein-protein interaction in p amyloid, for binding of FDA-approved drugs, followed by in vivo assays in human-cell and C. elegans models of neuropathic aggregation. Ezetimibe, prescribed chiefly to lower circulating lipid levels and to reduce intestinal absorption of cholesterol, emerged as the FDA- approved drug with highest predicted affinity for 14-3-3G::HXK1 . This prediction was validated by consistent and significant reduction of aggregate burden in cultured human-cell and nematode models that accumulate AD-like pathogenic aggregates. Neuroprotective effects of ezetimibe were not unheralded, as it had been previously shown to improve memory in AD-model mice fed a high-fat diet, but no mechanism had been proposed. Other pleiotropic benefits include reduction of vascular inflammation and atherosclerosis in rodents, prevention of neuronal apoptosis by activating autophagy in a rat model of arterial occlusion, amelioration of nephropathy via protection of pancreatic beta cells in diabetic mice, and rescue of muscle wasting in a mouse model of muscular dystrophy.

[0087] Ezetimibe reduction of aggregates both in C. elegans and cell-culture models of AD could be attributed to its ability to enhance autophagy (see FIG. 14D), but in view of the vicious cycle between inflammation and aggregation, an antiinflammatory mechanism (possibly secondary to rescue from aggregation) cannot be excluded. In the current study we show that ezetimibe is predicted to most effectively disrupt the 14-3-3G::HXK1 interface, supported by diminished recovery of hexokinase-1 from ezetimibe-treated cells upon immunoprecipitation with antibody to the 14-3-3 conserved core. The animal and culture that ezetimibe indeed reduces aggregates across a wide swath of biological systems, and that it does so by blocking or dissociating dysfunctional interaction between 14-3-3G and hexokinase-1 . However, the most compelling evidence that ezetimibe can prospectively prevent AD comes from our interrogation of a large clinical database. We employed a time-dependent regression model to evaluate the association between ezetimibe use and ADRD incidence, demonstrating that the relative risk of Alzheimer’s disease and related dementias (ADRD) is reduced 8-fold among those receiving ezetimibe. Although this was a retrospective analysis of previously collected data, the magnitude of the ezetimibe effect (8-fold reduction in risk, (P«10“4for the general population, and P<0.006 for CHD subset) support further study for repurposing as a prophylaxis to protect those at highest risk of ADRD. As no significant side effects of ezetimibe have been identified in either clinical studies or during its widespread use for almost 20 years post approval, we believe that potential neuroprotective benefits far outweigh the risks.

[0088] The studies described above illustrate that these drugs that serve as proteinprotein interaction inhibitors by blocking the interaction between 14-3-3G and hexokinase-1 can effectively be used to prevent and treat patients with Alzheimer’s Disease and related dementias. Patients may take these drugs at the same doses as prescribed for treating the conditions they were FDA-approved to treat. Since the interaction between 14-3-3G protein and hexokinase-1 protein occurs in both heart disease and AD brains, that interaction is expected to occur in patients with mild cognitive impairment, including changes associated with age.

[0089] The meaning of ADRD is well-understood by a person of ordinary skill in the art. As used herein, ADRD shall have the same meaning as when used by U.S. government health agencies, including the Centers for Disease Control and Prevention (CDC) and the U.S. Department of Health and Human Services (HHS).

[0090] The present invention has been described with reference to certain preferred and alternative embodiments that are intended to be exemplary only and not limiting to the full scope of the present invention as set forth in the appended claims.

Claims

WE CLAIM:1 . A method of treating a patient with Alzheimer’s disease and related dementias, said method comprising administering to said patient an effective amount of a drug that is operable to inhibit an interaction between a 14-3-3G protein and a hexokinase-1 protein in said patient.

2. The method of claim 1 , wherein said drug is ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astemizole.

3. A method of preventing Alzheimer’s disease and related dementias in a patient, said method comprising administering to said patient an effective amount of a drug that is operable to inhibit an interaction between a 14-3-3G protein and a hexokinase-1 protein in said patient.

4. The method of claim 3, wherein said drug is ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astemizole.

5. A method of treating a patient with age-related dementia, said method comprising administering to said patient an effective amount of a drug that is operable to inhibit an interaction between a 14-3-3G protein and a hexokinase-1 protein in said patient.

6. The method of claim 5, wherein said drug is ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astemizole.

7. A method of treating a dementia patient having heart disease, said method comprising administering to said dementia patient an effective amount of a drug that is operable to inhibit an interaction between a 14-3-3G protein and a hexokinase-1 protein in said patient.

8. The method of claim 7, wherein said drug is ezetimibe, conivaptan, lumacaftor, ebastine, digitoxin, or astemizole.

Images