Genetic compensation

Recent work by Rossi et al. shows that an unintended consequence of gene knockout may be genetic compensation that mitigates phenotypes.

Knockdown in zebrafish of egfl7, an endothelial extracellular gene, causes severe vascular defects:

However, following knockout of eglf7, there was no visible effect on vascular development, even after application of the knockdown reagent (demonstrating that the knockdown phenotype was not due to an off-target and that the knockout’s normal vascular development was not due some minor levels of egfl7):

The authors found that in egfl7 mutants, Emilin genes were upregulated.  Like egfl7, these genes are involved in elastogenesis, and thus their up regulation could be compensating for missing Egfl7.  (It seems that humans are also able to compensate for loss of Egfl7).

Work by Kok et al. also reported discrepancies between the phenotypes elicited by knockdown versus knockout experiments.  They found that the vast majority of phenotypes from knockdown experiments were not confirmed by knockout experiments.  They concluded:

Based on these results, we suggest that mutant phenotypes become the standard metric to define gene function in zebrafish, after which Morpholinos [knockdown reagent] that recapitulate respective phenotypes could be reliably applied for ancillary analyses.

People are understandably wary of knockdown phenotypes, given the prevalence of off-target effects.  But the work by Rossi et al. suggests that gene inactivation may give misleading results about gene function. The best metric for defining gene function will be gene knockdown experiments using reagents that prevent off-target effects.