Abstract
Termination codon readthrough is utilized for extension of C-terminal protein domain. Signals in mRNA that promote translation readthrough have not been characterized in plant. The surrounding context of readthrough stop codon may contain conserved sequence feature and can be utilized for prediction of potential readthrough candidates. Using computational method designed by Dr. JK Huang and TT Huang, we analyzed the stop codon context in Arabidopsis genes and identified 547 potential candidates. The desired sequence was inserted in NAN-GUS dual reporter system and transiently expressed in Arabidopsis protoplasts to analyze readthrough efficiency. The tobacco mosaic virus (TMV) leaky UAG stop codon showed significant readthrough in Arabidopsis protoplasts. Thus the sequence context containing 12 nucleotides flanking both sides of the first stop codon of each candidate was compared to the context of the TMV leaky UAG stop codon. Ten candidates were selected with > 50% sequence identity, and readthrough activity was detected in seven sequences. We further analyzed Arabidopsis database using 5 criteria including > 50% identity to the context of TMV leaky UAG stop codon, containing a single gene model, having an in frame stop codon in 3' UTR more than 120 nucleotides, and extending a functional domain, and 9 candidates were extracted. This study provides information for identifying potential readthrough genes in Arabidopsis.