Massively parallel functional dissection of schizophrenia-associated noncoding genetic variants.

Schizophrenia (SCZ) is a highly heritable mental disorder with thousands of associated genetic variants located mostly in the noncoding space of the genome. Translating these associations into insights regarding the underlying pathomechanisms has been challenging because the causal variants, their mechanisms of action, and their target genes remain largely unknown. We implemented a massively parallel variant annotation pipeline (MVAP) to perform SCZ variant-to-function mapping at scale in disease-relevant neural cell types. This approach identified 620 functional variants (1.7%) that operate in a highly developmental context and neuronal-activity-dependent manner. Multimodal integration of epigenomic and CRISPRi screening data enabled us to link these functional variants to target genes, biological processes, and ultimately alterations of neuronal physiology. These results provide a multistage prioritization strategy to map functional single-nucleotide polymorphism (SNP)-to-gene-to-endophenotype relations and offer biological insights into the context-dependent molecular processes modulated by SCZ-associated genetic variation.

Calcium Channel Subunit α2δ4 Is Regulated by Early Growth Response 1 and Facilitates Epileptogenesis.

Transient brain insults, including status epilepticus (SE), can trigger a period of epileptogenesis during which functional and structural reorganization of neuronal networks occurs resulting in the onset of focal epileptic seizures. In recent years, mechanisms that regulate the dynamic transcription of individual genes during epileptogenesis and thereby contribute to the development of a hyperexcitable neuronal network have been elucidated. Our own results have shown early growth response 1 (Egr1) to transiently increase expression of the T-type voltage-dependent Ca2+ channel (VDCC) subunit CaV3.2, a key proepileptogenic protein. However, epileptogenesis involves complex and dynamic transcriptomic alterations; and so far, our understanding of the transcriptional control mechanism of gene regulatory networks that act in the same processes is limited. Here, we have analyzed whether Egr1 acts as a key transcriptional regulator for genes contributing to the development of hyperexcitability during epileptogenesis. We found Egr1 to drive the expression of the VDCC subunit α2δ4, which was augmented early and persistently after pilocarpine-induced SE. Furthermore, we show that increasing levels of α2δ4 in the CA1 region of the hippocampus elevate seizure susceptibility of mice by slightly decreasing local network activity. Interestingly, we also detected increased expression levels of Egr1 and α2δ4 in human hippocampal biopsies obtained from epilepsy surgery. In conclusion, Egr1 controls the abundance of the VDCC subunits CaV3.2 and α2δ4, which act synergistically in epileptogenesis, and thereby contributes to a seizure-induced "transcriptional Ca2+ channelopathy."SIGNIFICANCE STATEMENT The onset of focal recurrent seizures often occurs after an epileptogenic process induced by transient insults to the brain. Recently, transcriptional control mechanisms for individual genes involved in converting neurons hyperexcitable have been identified, including early growth response 1 (Egr1), which activates transcription of the T-type Ca2+ channel subunit CaV3.2. Here, we find Egr1 to regulate also the expression of the voltage-dependent Ca2+ channel subunit α2δ4, which was augmented after pilocarpine- and kainic acid-induced status epilepticus. In addition, we observed that α2δ4 affected spontaneous network activity and the susceptibility for seizure induction. Furthermore, we detected corresponding dynamics in human biopsies from epilepsy patients. In conclusion, Egr1 orchestrates a seizure-induced "transcriptional Ca2+ channelopathy" consisting of CaV3.2 and α2δ4, which act synergistically in epileptogenesis.

Polygenic risk for schizophrenia converges on alternative polyadenylation as molecular mechanism underlying synaptic impairment.

Schizophrenia (SCZ) is a genetically heterogenous psychiatric disorder of highly polygenic nature. Correlative evidence from genetic studies indicate that the aggregated effects of distinct genetic risk factor combinations found in each patient converge onto common molecular mechanisms. To prove this on a functional level, we employed a reductionistic cellular model system for polygenic risk by differentiating induced pluripotent stem cells (iPSCs) from 104 individuals with high polygenic risk load and controls into cortical glutamatergic neurons (iNs). Multi-omics profiling identified widespread differences in alternative polyadenylation (APA) in the 3' untranslated region of many synaptic transcripts between iNs from SCZ patients and healthy donors. On the cellular level, 3'APA was associated with a reduction in synaptic density of iNs. Importantly, differential APA was largely conserved between postmortem human prefrontal cortex from SCZ patients and healthy donors, and strongly enriched for transcripts related to synapse biology. 3'APA was highly correlated with SCZ polygenic risk and affected genes were significantly enriched for SCZ associated common genetic variation. Integrative functional genomic analysis identified the RNA binding protein and SCZ GWAS risk gene PTBP2 as a critical trans-acting factor mediating 3'APA of synaptic genes in SCZ subjects. Functional characterization of PTBP2 in iNs confirmed its key role in 3'APA of synaptic transcripts and regulation of synapse density. Jointly, our findings show that the aggregated effects of polygenic risk converge on 3'APA as one common molecular mechanism that underlies synaptic impairments in SCZ.

Cell type and condition specific functional annotation of schizophrenia associated non-coding genetic variants.

Schizophrenia (SCZ) is a highly polygenic disease and genome wide association studies have identified thousands of genetic variants that are statistically associated with this psychiatric disorder. However, our ability to translate these associations into insights on the disease mechanisms has been challenging since the causal genetic variants, their molecular function and their target genes remain largely unknown. In order to address these questions, we established a functional genomics pipeline in combination with induced pluripotent stem cell technology to functionally characterize ~35,000 non-coding genetic variants associated with schizophrenia along with their target genes. This analysis identified a set of 620 (1.7%) single nucleotide polymorphisms as functional on a molecular level in a highly cell type and condition specific fashion. These results provide a high-resolution map of functional variant-gene combinations and offer comprehensive biological insights into the developmental context and stimulation dependent molecular processes modulated by SCZ associated genetic variation.