Cell type specific AAVs to study reward and cognition

Adeno-Associated Viruses (AAVs) are potent gene delivery vectors for neuroscience studies and gene therapy applications. However, naturally occurring AAVs are not cell type specific: they must be combined with other technologies, such as transgenic animals, to achieve cell type specific gene expression. This requirement limits

the use genetically coded ‘circuit-breaking’ tools to study behavior in nonhuman primates (NHPs) – the experimental animal model with the greatest similarity to humans – and hinders development of cell type specific targeting strategies for achieving direct clinical benefits. To expand cell type specific access in NHPs and lay the

foundation for circuit specific gene therapy, we propose to create, test, and validate next generation, cell type specific AAVs. First, we will define cell type specific enhancers – distal regulatory elements that have demonstrated considerable promise as cell type specific AAV drivers. In preliminary data, we collected

transcriptomic and chromatin accessibility (i.e. “multi-omic”) single cell data from the striatum, dorsolateral prefrontal cortex (dlPFC), primary motor cortex (M1), insula, and ventral midbrain of 2 rhesus macaque monkeys. We combined the rhesus monkey data set with existing human and mouse data and used convolutional neural

networks (CNNs) to rank open chromatin sequences according to their potential as cell type specific enhancers. We packaged AAVs with the top candidate enhancers, injected them into NHP striatum, and we have observed cell type specific, enhancer driven expression in striosomes – a cell type specific striatal compartment related to

reward processing. To broadly advance this agenda and develop AAVs that drive robust, cell type-specific expression, we propose to expand our multi-omic single cell database with additional data from macaque and marmoset. We will leverage this updated, sex-balanced database, with will include data from 8 NHPs to identify

cell type specific enhancers that are likely to drive robust expression in primates. In parallel, we will use our validated scAAVengr pipeline to screen AAV capsid mutants for cell-type biased infection patterns in the NHP cognitive and reward systems. We will combine the top cell type specific enhancers with the most biased AAV

capsids to generate new, cell type specific AAVs for targeting neurons in the NHP cognitive and reward systems. We will validate AAV specificity using Fluorescent in situ hybridization (FISH). This data will be combined with ultra-high resolution MRI scans to create a rhesus macaque brain atlas, and the validated vectors will be stored

and distributed by The University of Pittsburgh BioForge Initiative. NHPs are critical for studying human cognition and disease, and thus there is a pressing need to define the molecular properties of NHP cell types and study their behavioral functions. This proposal will generate a unique NHP multi-omic single cell database, provide cell

type specific AAVs for neuron types in cognitive and reward systems, and establish a new multimodal rhesus brain atlas. These contributions will significantly advance circuit manipulation capabilities in the primate brain and promote fundamental research in basic and preclinical science.