PROTEOMIC ANALYSIS OF CSF IN A PHASE 2 CLINICAL TRIAL FOR AD TO IDENTIFY PHARMACODYNAMIC BIOMARKERS OF THE S2R MODULATOR CT1812 N. Seyfried1 , L. Waybright2 , D. Duong1 , E. Malagise2 , K. Pandey3 , C. Williams2 , E. Dammer1 , L. Ping4 , K. Blennow5 , H. Zetterberg6 , J. Lah4 , A. Levey7 , L. Ricciardi2 , A.O. Caggiano2 , Mary E Hamby2 Affiliations: 1Emory University School of Medicine, Biochemistry, Atlanta, GA, United States of America, 2Cognition Therapeutics, Research, Pittsburgh, PA, United States of America, 3Emtherapro Inc, Systems Biology, Atlanta, GA, United States of America, 4Emory School of Medicine, Neurology, Atlanta, GA, United States of America, 5 Institute of Neuroscience and Physiology, The Sahlgrenska Academy, University of Gothenburg, Department of Psychiatry and Neurochemistry, Göteborg, Sweden INTRODUCTION The sigma-2 receptor (S2R) is encoded by TMEM97, a four-domain transmembrane protein that forms a complex with progesterone receptor membrane component 1 (PGRMC1). CT1812 is a highly brain-penetrant small molecule modulator of S2R, that displaces Aβ oligomers bound to neuronal synapses1 (Schema 1). In preclinical studies, CT1812 protects synapses, facilitates their restoration and improves cognitive performance in transgenic Alzheimer’s disease (AD) mice1. CT1812 is in clinical development for AD. COG0201, the SHINE study, is a randomized, double-blind, placebo- controlled Phase 2 clinical trial designed to enroll ~120 patients with mild-to- moderate AD to evaluate the safety and efficacy of CT1812. Participants are divided equally in two CT1812 dose groups (100 mg, 300 mg) and placebo for once daily oral dosing for 6 mo (Schema 2). Endpoints include safety, cognitive function, as measured by the AD Assessment Scale- Cognitive Subscale METHODS Tandem-mass tag mass spectrometry (TMT-MS) followed by unbiased quantification of CSF proteomes was conducted on all treatment-compliant patients for which CSF at baseline and end of study CSF was collected (N=18; Schema 3A, B). CSF proteomes were compared to within-study pooled AD and age-matched non- demented control CSF reference standards from the Emory Alzheimer’s Disease Research Center (ADRC) to compare protein levels in the SHINE-A cohort with well-characterized AD CSF and to assess treatment effects through differential expression and pathway analyses (Schema 3C). Schema 3. Following sample (A) analysis via TMT-proteomics (B), differentially expressed proteins (Cii) were mapped to a previously generated protein co- expression networks (Ci) built from 516 brain samples with healthy individuals, asymptomatic and symptomatic AD patients (Johnson et al. 2022 Nat Neuro) NCT03507790. Ciii) Pathway analysis was also performed on DE proteins using Metacore, STRING, and GO terms. REFERENCES 1. Izzo et al. Preclinical and clinical biomarker studies of CT1812: A novel approach to Alzheimer's disease modification. Alz & Dementia 2021. 2. Johnson et. al. Large-scale deep multi-layer analysis of Alzheimer’s disease brain reveals strong proteomic disease-related changes not observed at the RNA level. Nat Neurosci. 2022. 3. Blennow et al. Fluid biomarker-based molecular phenotyping of Alzheimer's disease patients in research and clinical settings. Prog Mol Biol Transl Sci. 2019;168:3-23. ClinicalTrials.gov: NCT03507790 Supported by National Institute on Aging: 3R01AG058660 Corresponding author: mhamby@cogrx.com -4 -2 0 2 4 6 8 Day ADAS-Cog 11 (LS mean change from baseline r S.E.M) worsening 42 98 182 0 placebo (n=8) 100 mg (n=8) 300 mg (n=8) Candidate PD Biomarkers Linked to Changes in Cognition Identified from SHINE-A CSF Proteomics Analyses Fig 3. A) Differentially abundant proteins in CSF from CT1812 vs placebo treated AD patients are highly interconnected (STRING analysis) and APP is a hub gene that is significantly lower in CT1812 vs placebo patients (B). C) Metacore pathway mapping shows amyloid pathways as the top most significantly associated with CT1812 treatment vs placebo (p<0.05). D) Statistically significant lowering of AE40 and AE42, as assessed via Lumipulse, by CT1812 vs placebo. Fig 1. A) Pearson correlation analysis of AD core biomarkers from clinically validated assays3 to that in the TMT proteomics dataset. B) Top two most highly correlated proteins with NFL quantitative assay (Uman Diagnostics; pg/ml). C) AD core biomarkers3 from CSF proteomes from SHINE-A at baseline (BL) were compared to pooled AD and non-demented control (CT) CSF reference standards (Emory ADRC). Candidate Disease Modification Biomarkers Identified: Proteins Dysregulated in AD Normalized with CT1812 CONCLUSIONS • Strong correlations with clinically validated assays for core AD biomarkers validate TMT- MS proteomics as a quantitative method • Brain module association and pathway analysis corroborate the mechanism of action of CT1812 in regulating synapses and AD biology • Comparisons to reference CSF standards illuminate proteins disrupted in, or genetically linked to, AD that were normalized by CT1812 • Pharmacodynamic biomarkers of CT1812 were identified, including that which may reflect disease modification and cognitive improvement Fig 5. Within-study comparisons to CSF protein levels in reference standards (ADRC AD and control (CT)) enabled comparison of the SHINE-A AD cohort to well-characterized AD and non-demented control CSF (A,B). A) Box plots illustrate two proteins significantly increased in AD compared to control CSF that are significantly downregulated in CT1812 vs placebo (SHINE-A). B) 22 proteins are significantly (p<0.05) normalized towards control with CT1812 (log2 change in abundance in AD vs control (red) and CT1812 vs placebo (blue)). A B A Pharmacodynamic (PD) Biomarkers of CT1812 (126) Identified, & Mapping to the Brain Network Supports Role at Synapses Unbiased Pathway Analysis of Differentially Abundant Proteins Implicates CT1812 in Regulating Amyloid Biology Proteomics Method Shows High Congruence with Clinically Validated Assays & Allows Benchmarking of SHINE-A Cohort FDR p<0.05 p<0.05 APP Fig 4. A) Several PD biomarkers (CT1812 vs placebo; p<0.05) are associated with GO term “cognition” (in yellow; GO:0050890; STRING). B) SHINE-A ADAS-Cog-11 cognitive score data which showed a non-significant but clinically meaningful (3- point difference) from placebo. C) Top biological processes from gene ontology of protein log2 abundance significantly correlated (Pearson, p<0.05) with ADAS-Cog11 change from baseline scores. A C B Fig 2. Differential expression analysis of CSF from AD vs control (A, left) and AD patients given CT1812 vs placebo (B, left). Differentially expressed proteins for each comparison were mapped to the AD co-expression modules2 (Schema 3C, top panel; top 12 shown here (A,B right)). AD vs control CT1812 vs placebo p<0.05) STRING PPI enrichment p<1.0x10-16; APP is a major hub protein B Cognitive Data Overall, data provide additional support that the S2R modulator CT1812 may be a promising therapeutic approach to AD log2 abundance A A C B 11-item version (ADAS-Cog-11), and biomarker evidence of disease modification. An interim analysis of the first 24 patients was conducted. No subjects were withdrawn from the study due to treatment-emergent adverse events and there were no SAEs attributed to study. A B C A D Schema 2 Schema 1 C A B 2,182 CSF proteins detected Good congruence across batches