Fitness consequences of genetically engineered herbicide and antibiotic resistance in Arabidopsis thaliana.

Abstract

Researchers have often invoked the concept of metabolic drain to explain the lower growth rates of bacteria containing plasmids that confer antibiotic resistance. This idea posits that the energetic input needed to produce detoxifying enzymes diverts resources from clonal reproduction. In this paper we examine whether the concept of metabolic drain can be applied successfully to plants that differ from bacteria in several key aspects including their relative genome size and reproductive rate. We have conducted a field experiment using mutant and transgenic Arabidopsis thaliana that allows the comparison of genotypes differing by a single gene conferring resistance to either the herbicide chlorsulfuron or the antibiotic kanamycin. In addition to testing whether these traits reduce fitness, this experiment was conducted at two levels of resource availability to examine whether costs of resistance are sensitive to environmental quality. We found that herbicide-resistant individuals produced 26% fewer seeds than susceptible counterparts. However, contrasting published results in bacterial systems, the fecundity of individuals was completely unaffected by the expression of an introduced antibiotic resistance gene. The fitness cost associated with chlorsulfuron resistance was greater in nutrient-poor conditions relative to nutrient-rich conditions for comparisons involving mutant, but not transgenic, genotypes.