Both neighboring codon adjacent nucleotides and codon optimality influence mRNA decay rates

Codon-optimality-mediated decay (COMD) is a major pathway that determines mRNA decay rates in eukaryotes. Conservation of this pathway in plants has not been demonstrated. To identify codons that might influence cotranslational mRNA decay rates, we compared codon usage bias in Arabidopsis seedling-expressed mRNAs with their mRNA half-lives (t _1/2s). Finding differences in codon usage between transcripts with short and long t _1/2s led to a model of mRNA decay rate based on codon frequencies. This codon-decay rate model explained 21% of decay rate variance in Arabidopsis and was predictive of decay rates of synonymously recoded genes. In the COMD pathway, NOT3, a component of the CCR4-NOT deadenylation complex, can detect slow ribosome decoding and trigger decay cotranslationally. Because the N-terminal sensor domain of NOT3 is retained in plant genomes, it is likely that plants, yeast, and humans use the same mechanism. However, codon optimality in Arabidopsis was not reading-frame dependent, suggesting that additional sequence features also contributed to decay rates. These features were computationally identified as specific adjacent nucleotides of neighboring codons. The influence of adjacent codons appears to be a conserved feature of cotranslational decay, as published data sets from wheat (Triticum aestivum) and fission yeast (Schizosaccharomyces pombe) also showed neighboring codon adjacent nucleotides to impact RNA decay rates. These findings show that codon sequence can influence mRNA decay rates independently of charged tRNA concentrations and suggest a paradigm of selection among synonymous codons that are decoded through wobble base-pairing.