Session 1: Genetic and molecular disease mechanisms

Autism Spectrum Disorder (ASD) is a neuropsychiatric syndrome characterized by the presence of difficulties in social communication and repetitive restrictive behaviors. Like other psychiatric disorders, ASD is a syndrome that is not defined by a particular etiology. The last decade of genetic studies has yielded substantial advances in our knowledge of ASD's genetic architecture, implicating hundreds of genes and loci. Despite this genetic heterogeneity, we have hypothesized that similar to other biomedical conditions ranging from asthma to cancer, that ASD and other psychiatric disorders might converge on specific biological processes. We have used genomic profiling and integrative methods including gene network analysis, which have demonstrated convergent molecular pathology in ASD post mortem brain, as well as convergent biological pathways implicated by risk genes. These analyses inform a framework for developing a mechanistic understanding of neuropsychiatric disorders and provide hope for new therapeutic development.

Genome-wide association studies (GWAS) have successfully identified many novel loci for neuropsychiatric traits. At the same time results showed that these traits are highly polygenic, mostly influenced by large numbers of weakly associated variants. Interpreting such results is challenging. Recent large-scale initiatives, such as those from the Allen Brain Institute and the PyschEncode consortium provide fine-scaled atlases of functional genetic elements at cellular level. This novel information can be used to interpret results from GWAS studies and facilitate biological understanding of complex traits.

In this session, I will discuss how we can leverage both GWAS results and novel functional genomic resources to formulate hypotheses that can be tested in functional experiments. I will highlight our recent analytical work that includes single cell enrichment for psychiatric disorders and local genetic correlation analyses. I will discuss the need to bridge the gap between genetics and neuroscience and which obstacles may be encountered.  

Schizophrenia is thought to be a neurodevelopmental brain disorder whose genetic risk is associated with shifting clinical phenomena across the lifespan. We investigated brain co-expression of schizophrenia risk genes in postmortem neurotypical human prefrontal cortex, hippocampus, caudate nucleus, and dentate gyrus granule cells, parsed by specific age periods. Specifically, we assessed the enrichment of PGC3 schizophrenia risk genes within co-expression modules in 562 brains across four lifespan stages (fetal to 5 years, 5-25 years, 25-50 years, over 50). Schizophrenia risk genes clustered into specific modules, especially in the perinatal prefrontal cortex, and showed predominantly neuronal cell specificity. Parsing by age periods explained more variance in a continuous measure of risk (H-MAGMA) than lumping all age periods together in a single network (Vuong test, z=7.4, p=9.1e-14). Schizophrenia risk genes exert combined effects while part of their molecular environment changes across development, potentially underwriting the shifting clinical presentation of the disorder.

The overall risk for developing major depressive disorder (MDD) is determined by a complex interaction between genetic variants  and adverse experiences such as stress exposure. Here, we use cell population-specific chromatin accessibility profiling to capture the chromatin regulatory signature in MDD orbitofrontal cortex.  We mapped genetic risk for MDD to open chromatin regions (OCRs) in non-neuronal OFC cell-types. Characterization of MDD-specific OCRs revealed a key role for chromatin remodeling protein ZBTB7A, which facilitates transcription of inflammation genes, and is enriched in astrocytes, though its role in neuropsychiatric disease is unknown.  We utilized preclinical rodent models to show that bidirectional manipulation of ZBTB7A specifically in OFC astrocytes is sufficient to alter behavioral, transcriptional, and neuronal activity responses to stress. Together, these data demonstrate that epigenetic regulation of inflammatory signaling in astrocytes impairs neuroadaptive responses to stress, with direct implications for the role of astrocyte plasticity in OFC dysfunction and MDD pathology.

The 3q29 deletion (3q29Del) is the strongest known genetic risk factor for schizophrenia, but the biological basis for this risk is not understood. We generated cortical organoids from CRISPR-engineered isogenic 3q29Del induced-pluripotent stem cell lines and employed a cross-species transcriptomic strategy to identify salient effects of 3q29Del in the developing CNS. Single-cell RNA-sequencing was performed at two time points (2-months and 12-months) in cortical organoid development (n=4 Ctrl, n=4 iso3q29Del; 54,255 cells) and in neonatal 3q29Del mouse isocortex (n=4 WT, n=4 3q29Del; 71,066 cells). We found striking overlap in the differential expression signature across model systems; human homologs of most of the differentially expressed genes (DEGs) identified in mouse isocortex were also differentially expressed in the human organoids (OR=1.53, p-value=1.64e-162). Pathway analysis of mouse-human intersecting DEGs indicated an enrichment of dysregulated neuronal genes linked to axon guidance. We will present detailed results of this analysis and in vitro validation experiments.