As these GEITP pages continue to emphasize, genome-wide association studies (GWAS) have been used for ~13 years in order to find associations between specific single-nucleotide variants (SNVs) throughout the genome — and the multifactorial trait chosen for study. Multifactorial phenotypes include quantitative traits (e.g. blood pressure, weight), complex diseases (e.g. type-2 diabetes, obesity, schizophrenia), responses to drugs or a combination of drugs, and responses to environmental toxicants or a mixture of toxicants. More than 100 SNVs have currently been (statistically significantly) associated with schizophrenia.
Authors [see attached article & editorial] demonstrate that disease-relevant expression changes in schizophrenia risk genes results in pre- and postsynaptic deficits (a ‘synapse’ is the structure that permits a neuron (nerve cell) to pass an electrical or chemical signal to another neuron or to a target effector cell) in human neuron cells. Authors show that simultaneous perturbation of multiple risk genes leads to more extreme changes than expected on the basis of separate studies of each gene by itself; this is termed “synergism” (i.e. the effect of multiple genes is greater than the sum of the effect of each gene separately). Although their approach is focused on only a very small subset (i.e. only four) of the more than 100 risk genes in schizophrenia — intriguingly, this prioritized and combinational approach might be applicable to more genes and to other complex disorders.
Schizophrenia is highly polygenic (i.e. many genes contributing to the trait) and caused by SNVs with frequencies ranging from
ultra-rare (e.g. one in 100,000) to common (i.e. allele frequency >0.01; greater than 1%). And these variants individually (usually) have only a very small-effect (e.g. 0.01 or 0.0001) on disease risk. Fine-mapping with novel bioinformatic tools allows for prioritization
of the “most likely causal variants” among potentially hundreds of SNVs with complex correlation structures within GWAS- implicated loci. Causal variants — that are likely to affect schizophrenia risk through their effects on gene expression — can be further identified by using information on expression quantitative trait loci (eQTLs) — generated (e.g. by integration of GWAS data and global gene expression profiling of postmortem brains).
Authors [see attached article] integrated CRISPR-mediated gene editing, activation and repression technologies — in postmortem brain RNA-sequence analyses; they studied one putative schizophrenia eQTL for the FURIN gene (encoding a paired basic amino acid-cleaving enzyme), and compared those data to data determined from studies of four top-ranked schizophrenia eQTL genes [FURIN, SNAP91 (encoding synaptosome-associated protein-91), TSNARE1 (encoding t-SNARE domain-containing-1), and CLCN3 (chloride voltage-gated channel-3)]. Authors were able to resolve pre- and postsynaptic neuronal deficits, recapitulate genotype-dependent gene expression differences, and identify convergence downstream of schizophrenia eQTL-gene perturbations.
These findings highlight the cell-type-specific effects of common SNVs and demonstrate a synergistic effect between schizophrenia eQTL genes — that converges on synaptic function. Authors propose that the interactions between rare and common variants — that are implicated in psychiatric disease risk — might constitute a phenomenon (i.e. synergism) that might occur much more widely in complex genetic disorders (e.g. if you have ten genes contributing 0.001 to the trait, synergism might make this 0.02 contribution to the trait, instead of the 0.004 total by simple addition). Holy cow. 😊
Nat Genet Oct 2019; 51: 1475-1485 & editorial pp 1434-1436