Testing GxG interactions between coinfecting microbial parasite genotypes within hosts

Abstract

Host-parasite interactions represent one of the strongest selection pressures in nature. They are often governed by genotype-specific (GxG) interactions resulting in host genotypes that differ in resistance and parasite genotypes that differ in virulence depending on the antagonist's genotype. Another type of GxG interactions, which is often neglected but which certainly influences host-parasite interactions, are those between coinfecting parasite genotypes. Mechanistically, within-host parasite interactions may range from competition for limited host resources to cooperation for more efficient host exploitation. The exact type of interaction, i.e., whether competitive or cooperative, is known to affect life-history traits such as virulence. However, the latter has been shown for chosen genotype combinations only, not considering whether the specific genotype combination per se may influence the interaction (i.e., GxG interactions). Here, we want to test for the presence of GxG interactions between coinfections of the bacterium Bacillus thuringiensis infecting the nematode Caenorhabditis elegans by combining two non-pathogenic and five pathogenic strains in all possible ways. Furthermore, we evaluate whether the type of interaction, reflected by the direction of virulence change of multiple compared to single infections, is genotype-specific. Generally, we found no indication for GxG interactions between non-pathogenic and pathogenic bacterial strains, indicating that virulence of pathogenic strains is equally affected by both non-pathogenic strains. Specific genotype combinations, however, differ in the strength of virulence change, indicating that the interaction type between coinfecting parasite strains and thus the virulence mechanism is specific for different genotype combinations. Such interactions are expected to influence host-parasite interactions and to have strong implications for coevolution.

Keywords: Bacillus thuringiensis; Caenorhabditis elegans; GxG; host; multiple infection; parasite; virulence; within-host interaction.

FIGURE 1

Schematic overview of the experimental design showing different treatments and different strain combinations within each treatment. Every strain combination was replicated four times, except for survival of the combination of 407 Cry- and B-18247 (three times).

FIGURE 2

Line graphs of (A) survival rate and (B) logarithmic population size of the non-pathogenic-pathogenic mixed treatment. Vertical lines represent the one-fold standard error.

FIGURE 3

Boxplots of (A) mean host survival rate and (B) mean logarithmic population size per strain combination of different treatments. The horizontal line indicates the median, the box the 25% quartile above and below the median, and the whiskers the data range. Circles represent outliers and stars represent extreme values.

FIGURE 4

Line graphs of mixed treatments and the mean of the corresponding two single infections. (A) Survival rate of the non-pathogenic-pathogenic mixed treatment. (B) Survival rate of the pathogenic mixed treatment. (C) Logarithmic population size of the non-pathogenic-pathogenic mixed treatment. (D) Logarithmic population size of the pathogenic mixed treatment. Dashed lines represent combinations, for which the slope differs significantly from zero, dashed-dotted lines represent combinations, for which the slope shows a tendency to differ from zero. Vertical lines represent the onefold standard error.