Agriculture: The selector of improbable mutations

Crop domestication took place mainly through the selection for mutations that are advantageous for agriculture. In parallel, an unconscious process of selection for undesirable mutants is taking place, such as the selection for pests (weeds, insects, and pathogens) and more recently for pests that are resistant to pesticides. Experiments for selection of exceedingly rare events cannot be performed in the laboratory due to time and space constraints. Farmers, who enjoy larger facilities than scientists, are demonstrating time and again that they are capable of selecting for mutations that researchers cannot find. We are used to reports of single-nucleotide point mutations that change an amino acid, but in the report by Patzoldt et al. (1) in this issue of PNAS, we are regaled with a mutant that has lost a whole amino acid in the target protein of a xenobiotic herbicide as well as lost the gene encoding an isozyme of the target of the herbicide. People have easily made artificial deletion mutations in the laboratory by using DNA synthesizers, but naturally occurring mutations with internal deleted amino acids have not been reported. Single-nucleotide point mutations typically occur at a frequency of approximately one in a million per gene and approximately one per billion per specific nucleotide. It is thus conceivable to obtain herbicide resistance due to a single nucleotide substitution. If two independent mutations are necessary to obtain the resistance phenotype, e.g., in both alleles for recessive mutations or if two amino acids must be substituted, the frequencies go lower and jump to the square (e.g., 10−18 instead of 10−9 per nucleotide). Taking into account the possibility of gene conversion, or of partial resistance of the single mutant, the actual frequency for a double mutant might be higher. Still, these frequencies are extremely low and are probably in a range where the odds for such mutations in any mammalian species are almost nil.