Multifactorial traits (e.g. height, weight, blood pressure, schizophrenia, cancer, serum LDL levels, etc.) reflect the contributions of hundreds if not thousands of genes (DNA-sequence variants), plus epigenetic effects (e.g. DNA-methylation, RNA-interference, histone modifications, chromatin remodeling), plus environmental effects (i.e. adverse insults of the cell or animal over time). Drug efficacy and certain types of toxicity (adverse drug reactions) can also fall within the category of multifactorial traits. Autism spectrum disorder (ASD) is an excellent example of a very complex disorder with genetic and clinical heterogeneity.
Beyond common DNA-sequence variants, previous studies focusing on germline mutations have demonstrated a significant contribution from de novo copy number variants (CNVs), which has recently been discussed in these GEITP emails. Whole-exome sequencing (WES) analyses have highlighted the role of de novo point mutations. Exomes contain the “coding region” resulting in protein synthesis. Although the number of exonic de novo mutations is similar between affected and unaffected individuals (~1 de novo point mutation per exome), ASD probands (afflicted patients) harbor an excess of deleterious and loss-of-function (LoF) de novo mutations in exons –– compared with that of their unaffected siblings. Collectively, 4–7% of probands have a de novo CNV and ~7% of probands have a de novo point mutation that confers risk to ASD. In addition, WES studies have uncovered increased ASD risk from rare autosomal recessive (3%) and X-linked variants (2%). However, a large portion of ASD risk cannot be explained by germline de novo, recessive and X-linked variants, and this warrants investigation of other genetic contributions to ASD risk.
Authors [see attached paper] therefore sought to study postzygotic mutations (PZMs) in whole-exome sequences; PZMs represent mutations that arise AFTER the sperm and egg have formed a zygote, and the subsequent cells begin dividing and developing into various cell types, organs and tissues, i.e. mutations can be germline or they can be somatic (PZM).
Authors therefore systematically analyzed PZMs in whole-exome sequences from the largest collection of trios (having an afflicted ASD proband) (N=5,947) –– including 282 unpublished trios –– and performed resequencing using multiple independent technologies. Authors identified 7.5% of de novo mutations as PZMs, 83.3% of which were not described in previous studies. Damaging, nonsynonymous PZMs within critical exons of prenatally expressed genes were more common in ASD probands than controls (P <1 × 10–6), and genes carrying these PZMs were enriched for expression in the amygdala (P = 5.4 × 10–3). Two genes (KLF16 and MSANTD2) were significantly enriched for PZMs genome-wide, and other PZMs involved genes (SCN2A, HNRNPU and SMARCA4) whose mutations are known to cause ASD or other neurodevelopmental disorders. Authors conclude that PZMs constitute a significant proportion of de novo mutations and represent an important contribution to risk of ASD. At this time, we can only speculate that PZMs might play similar roles in some, or many, other multifactorial traits. Nat Neurosci July 2o17; [advance online publication] http://dx.doi.org/10.1038/nn.4598