Changing the start codon context of the 30K gene of tobacco mosaic virus from "weak" to "strong" does not increase expression.

The translation initiation region of the 30K gene of tobacco mosaic virus (TMV) was modified by in vitro mutagenesis to create more optimal start codon contexts. A complicating factor was that modifications in this region also altered the 3' terminus of the 183K ORF that overlaps the 30K ORF. An insertion of GACUCGA between nucleotides 4901 and 4902 resulted in a purine (G) in position -3 relative to the AUG creating a "stronger" start codon context, but this also changed the last four amino acids of the 183K protein. This mutant was infectious, replicated efficiently, but produced reduced amounts of 30K protein. Despite the reduced amount of movement protein, this mutant spread effectively from cell to cell and had a phenotype indistinguishable from that of wild-type virus. A more conservative mutation inserted GAC between TMV nucleotides 4901 and 4902 resulting in a "strong" start codon context (ACGAUGG) and modification of the 183K protein only by insertion of an aspartic acid adjacent to a native aspartic acid. This modification did not enhance the production of 30K protein. These data demonstrate consensus sequences that are optimal for other eukaryotic systems did not cause increased expression of the 30K gene in vivo. The modified sequences of both mutants were stably maintained during relatively long periods of replication. Even though each mutant replicated efficiently, when mixed with wild-type TMV, neither mutant effectively competed with the wild-type virus. Another mutant which removed the native 30K AUG to determine whether subsequent internal start codons with "stronger" contexts would function in its absence was constructed. However, this mutant and a mutant that fused the 183K reading frame to the 30K reading frame did not replicate and move in intact plants.