Fig. 1: AAV capsid library enrichment and candidate selection.

A Schematic depicting the AAV capsid selection pipeline used in this work. AAV peptide display libraries were designed to contain a semi-random heptapeptide insert in multiple parental AAV serotypes (AAV1, AAV2, and AAV9), yielding an initial ~ 6.8 million capsid variants. Enrichment and selection of capsids is achieved by sequential passages through primate brain. The first round of enrichment (black arrows) involved a single infusion of ‘Round 1’ input library composed of peptide-modified AAV capsids from three parental serotypes. The input library was injected by bilateral intraparenchymal infusion into the globus pallidus of a rhesus macaque. After two days in life, CNS regions of interest were collected. DNA was extracted from collected tissues and recovered AAV variant sequences were used to generate a new (Round 2) AAV input library. A second round of screening (gray arrows) was initiated by bilateral GP infusion of the Round 2 input library into rhesus macaque. For round 2 screening, both viral DNA and RNA were collected from tissues. Peptide modifications of successfully enriched capsids were identified using Illumina sequencing of viral amplicons. Capsid performance was assessed using a combination of total detection in target tissues and enrichment, which measures the ratio of a capsid’s relative abundance in a target region vs the input library. B 161 top-ranking capsids across all parental serotypes were identified to generate a third input vector library, hereafter referred to as the optimized pool. A final round of screening was initiated by injection of the optimized pool input vector into the GP of an adult rhesus macaque and African green monkey. Tissues collected from these primates became the basis for DNA- and RNA-based amplicon sequencing. C Three highly enriched capsid variants were identified based on a combination of overall detection across brain regions (AVG enrichment), and/or targeted enrichment in caudate and putamen (AVG CP enrichment), and were selected for fluorescence validation. Rd, round; NGS, next-generation sequencing; DB1, AAV-DB-1; DB2, AAV-DB-2; DB3, AAV-DB-3; Cd, caudate; Pu, putamen; STh, subthalamic nucleus; SN, substantia nigra; Th, thalamus; Cx, cortex. Panels (A and B) created in BioRender. Lewandowski, B. (2025) https://BioRender.com/y3ikqcc.