Human genetic studies have provided lists of genes and loci associated with risk for neuropsychiatric
disorders including autism spectrum disorder and neurodevelopmental delay (ASD/ND), but to systematically
evaluate their functions and mechanisms remains a dauting task. What brain regions, cell types, and neural
circuits are involved in the pathological state? Do different risk genes and genetic variants affect the same sets
of brain regions, cell types, neuronal projection networks and circuits? To address these questions, we will
develop scalable genomic technologies to allow high-throughput, in vivo genetic screening with single-cell and
meso-scale spatial resolution readout from intact tissues. First, we aim establish a method, named Perturb-map,
to readout the spatial location, cytoarchitecture, and projectome of each perturbed cell from intact, whole brains
without sectioning (Aim 1). We will perform sparse perturbation using the CRISPR-Cas9 system in vivo and
harvest the brain with pooled genetic perturbations. Next, we will use hybridization-based detection and whole-
brain tissue clearing methods, to allow iterative rounds of light-sheet imaging to deduce gRNA perturbation
identities within an intact tissue. We will further incorporate imaging of fluorescent reporters and endogenous
markers to ask how the perturbation changes the cell type identity, anatomical distribution, morphology and
neuronal projectome from intact tissues. In parallel, this approach will be complemented with an improved in
vivo Perturb-seq with unprecedented scalability (Aim 2a-b). We will establish an AAV-based Perturb-seq system
to allow high scalability and flexible tissue-specificity. We will then apply it to study a panel of ASD/ND risk genes
across two brain regions. Pooled, sparse perturbation will be introduced during embryogenesis, and single-cell
multi-omic analysis and in situ Perturb-map analysis will reveal how each perturbation affect gene expression,
cellular migration, cytoarchitecture, and neuronal projectome for each specific cell type across brain regions
including neocortex and striatum, which are implicated in the disorders. This will allow the first in vivo screen with
3-dimensional, meso-scale optical readout with high spatial resolution. The PI has extensive experience and has
pioneered the in vivo Perturb-seq technology, and she has brought together a diverse team and expertise with
long-standing relationships and collaborations across the fields of tissue clearing and high-throughput imaging,
computational genomics, and human genetics (including non-Caucasian populations) for equitable biomedicine
(co-I Dr. Zhuhao Wu, LoS from Drs. Pejman Mohammadi, Joshua Levin, and Mireille Kamariza). Altogether, we
will establish Perturb-Map as a scalable and generalizable approach to study gene function in complex tissues.
We expect to identify the cell types, molecular networks, and projectomes affected by diverse ASD/ND risk gene
variants with unprecedented scale and resolution. Perturb-Seq and Perturb-Map will be modular and broadly
applicable genomic tools contributing to not only neuroscience but also many other tissue types and fields as we
move from gene variant to their mechanisms and functions.