Academia Sinica Logo

Each human genome has been reported to contain approximately 100 loss-of-function variants, with roughly 20 genes completely inactivated.  Some of these completely inactivated genes are essential genes, and yet they are present in a homozygous state in apparently healthy individuals.  This totally unexpected lack of phenotype has also been observed in commonly studied model organisms from yeast to mammals.  Various hypotheses have been proposed to explain these findings including Genetic Compensation (GC).  GC manifests itself as altered gene/protein expression, or function, which leads to a wild-type-like phenotype in homozygous mutant or heterozygous individuals who would be predicted to exhibit clear defects.  Traditionally, GC has been thought to involve protein feedback loops such that if one component of a regulatory pathway is deficient, a compensatory rewiring within a network or the activation of a functionally redundant gene occurs.  However, not every major regulatory network has evolved to incorporate such complex features.  Another mechanism of GC is the newly identified process of Transcriptional Adaptation (TA): some deleterious mutations, but not all, trigger the transcriptional modulation of so-called adapting genes.  Depending on the nature of the modulated genes, GC can occur.  Notably, unlike other mechanisms underlying genetic robustness, TA is not triggered by the loss of protein function. 

We discovered TA while trying to understand the phenotypic differences between knockout (mutant) and knockdown (morphant) zebrafish embryos.  Further studies identified additional examples of TA in zebrafish as well as examples in C. elegans and in mammalian cell lines.  By generating and analyzing several mutant alleles for these genes, including non-transcribing alleles, we found that the mutant mRNA is required to trigger TA.  Based on these and other data, we hypothesize that all mutations that cause mutant mRNA degradation can trigger TA.  This presentation will go over our published and unpublished data on TA in several model systems.