Rapid circuit-level dendritic spine morphometric analysis for in vivo pharmacology in CNS drug discovery

Period of Performance: 09/15/2017 - 02/28/2018


Phase 1 SBIR

Recipient Firm

Afraxis 2.0, Inc.
Principal Investigator


In vivo phenotypic CNS assays that successfully identify hits and leads in academic translational research are often challenging to scale up and standardize for commercial SAR and MedChem. Afraxis' Enhanced Spine Profiling platform quantifies brain-wide dendritic spine morphometry and plasticity, providing insights into neural circuits and changes resulting from disease, pharmacological and behavioral manipulations. ESP enables dendritic spine analysis for drug discovery using patented methods to obtain redundant human observer-based analyses (Expert Crowdsourcing) via automated processes that demonstrate exceptional repeatability and require no intervention from investigators and analysts. We proposes to further align ESP with the needs of medium/high-throughput compound screens and optimization through three key improvements: (a) implementation of superresolution microscopy; (b) design novel implementations of image recognition software to aid observer-based spine quantification; and (c) development of more robust hierarchical, multi-dimensional statistical methods appropriate to multi-observation image and feature analysis for the larger and more complex datasets produced by superresolution imaging. We will complete a series of validation studies to empirically refine the optimizations and demonstrate reliable and stable signal. Due to time limits of this initial project, we chose ketamine as a fast-turnaround validation tool compound because of its rapid well-studied structural plasticity effects in rodents and clinical relevance to cognition and disease. We will next profile five procognitive drugs as reference compounds to verify multivariate signal dynamic ranges across multiple compounds. Following these evaluations, ESP will be a robust commercially validated phenotypic tool for batch compound analysis to identify hits and guide SAR and MedChem optimization. Specific Aim 1: Optimize the Afraxis dendritic spine analysis platform for Lead Identification criteria. We propose to optimize the performance of the Afraxis ESP platform to meet the fundamental criteria for inclusion in traditional structure-activity relationship (SAR) paradigms, rendering the assay viable for lead identification and lead optimization in CNS drug discovery programs. We propose to meet these operational criteria through implementation superresolution microscopy, adaptation of our novel distributed corroborative human observer analytical paradigm to facilitate novel image segmentation spine morphometric quantification, and novel hierarchical statistical methodologies. Specific Aim 2: Validate Dendritic Spine analysis using the rapid anti-depressant ketamine. Considering the long lead times required for investigating aged animals, we propose to validate the optimized version of the assay using the rapid anti-depressant ketamine. Ketamine rapidly triggers structural modifications that are relevant to aging and age-related diseases and Afraxis has previously and multiply replicated (from numerous published studies) the effect whereby low doses of ketamine induce increased cortical spine density. Afraxis has expanded on the original work to show diverse responses across doses, dosing paradigms, cortical regions, and cortical lamina. Here we will validate the optimized assay across these contexts by evaluating ketamine's effect in 9 independent pathways within 3 brain regions. In this manner we will attempt to characterize a high-resolution multivariate dose response for ketamine's anti-depressant effect on dendritic spines. Central to this effort is empirical establishment of statistical tools optimized for this novel evaluation design (see discussion under Approach: Specific Aim 1). Specific Aim 3: Validate Dendritic Spine Analysis across known clinical compounds. We propose to build on our understanding of dendritic spine responses in SA2 by testing 5 compounds classified as pro-cognitive from clinical evaluations. This panel will consist of clinically tested compounds of mixed efficacious results across cognitive domains (e.g. working memory, reversal learning). The statistical procedures developed in SA1 and validated in SA2 will be broadened to evaluate ?hits? within a screening panel based on response profiles across the diverse set of neuroanatomical pathways tested in SA2.