(1126-C) The pioneering NETSseq platform leverages an automated RNA sequencing workflow to drive innovation in precision medicine against neurodegenerative diseases.

Introduction: Cerevance is generating an unparalleled catalog of novel targets for next generation treatments of patients affected by central nervous system (CNS) disorders. NETSseq™ is an innovative workflow that allows Cerevance to isolate nuclei of defined cell types from post-mortem human brain tissue, which are then sequenced at high depth and analyzed with advanced machine learning techniques for identification of previously undiscovered gene expression profiles. Therefore, the detection of rare transcripts in low-quality FFPE samples requires the integration of a reliable automated system with sensitive NGS technology.

Method: To support the NETSseq workflow, a specialized automated RNA-seq library prep method was developed in the Tecan's DreamPrep® NGS utilizing patented Tecan reagent technologies such as SPIA® (Single Primer Isothermal Amplification) for unbiased amplification, SPIA AnyDeplete® for targeted transcript depletion and DimerFree® to eliminate adaptor dimers. The developed method must be compatible with a wide range of sample quality and quantity. Due to low sample quantity, QC steps are often not feasible before the library prep, as the number of sorted nuclei is variable, ranging from less than 100 to 10000. Fifteen runs of 96 samples are analyzed herein together with K562 controls (50 ng, 5 ng and 0.5 ng) and control Nuclei (~6000 Nuclei replicates). Sizes and yields from generated libraries were measured for each of the 15 runs and libraries were sequenced with NextSeq 500 instrument. The overall sequence quality was assessed by FastQC, Picard, STAR alignment and TIN (Transcript Integrity Number) metrics.

RESULTS AND

Discussion: Despite samples being highly variable, the automated method generated good libraries from samples with as few as 100 nuclei but this was often donor/sample dependent. Overall, libraries consistently displayed very well defined peaks around 300 bp and yields were close to 5000 ng on average. Tapestation traces and sequencing data did not show any evidence of contamination, dimers nor overamplification. After downstream analysis, transcriptomic signatures revealed changes in response to disease states in pathways and targets with potential for drug discovery. Amongst these, GPR6 was confirmed to be selectively expressed in striatal medium spiny neurons, supporting the hypothesis that modulating this neurocircuitry can bring therapeutic benefits to patients with Parkinson’s disease. Cerevance’s lead compound, CVN424, is currently entering Phase 3 as an adjunctive therapy to L-DOPA.

Conclusions: The automated method can perform the library preparation completely walk-away and deliver robust results from very challenging samples. This consistency of results is critical to generate high quality data sets for downstream bioinformatic analysis. The walk-away time provided by the system enabled the staff to increase throughput and to allocate time to introduce new methods, to further enhance Cerevance’s ability to develop a broad pipeline of drug candidates for CNS disease management.