Signature of coevolution between determinants of defense and life span in Arabidopsis thaliana

The selective impact of pathogen epidemics on host defenses can be strong but remains transient. By contrast, life-history shifts can durably and continuously modify the balance between costs and benefits, which arbitrates the evolution of host defenses. Their impact, however, has seldom been documented. Here, we show with a simple mathematical model that the selective advantage of the defense system is expected to decrease with decreasing life span. We further document that, in natural populations of the model plant system Arabidopsis thaliana, the expression level of defense genes correlate positively with flowering time, a proxy for the length of vegetative life span. Using a genetic strategy to partition life span-dependent and –independent defense genes, we demonstrate that this positive co-variation is not explained by the pleiotropic action of major regulatory genes controlling both defense and life span. In agreement with our model, this study reveals that natural selection has likely assembled alleles promoting lower expression of defense genes with alleles decreasing the duration of vegetative life span in natural populations of A. thaliana. This is the first study demonstrating that life history evolution has a pervasive impact on the evolution of host immunity. Overall design: Seeds of Bur-0, Col-0 and 278 Bur-0xCol-0 Recombinant Inbred Lines (RIL) obtained after 8 generations of selfing were provided by the Arabidopsis Stock Center at INRA Versailles (France). We selected the 40 RIL in the 15% and 85% quantiles of flowering time for RNA sequencing. Each RIL and the two parental lines were planted in 20 replicates in the conditions described above. At days 14 and 28, the oldest leaf was flash-frozen in liquid nitrogen. Three pools, each combining 13 RIL, were produced at each time point for early and late lines, for a total of 3 biological replicates, 2 pool types (early and late RIL) and 2 time points (14 and 28 days). For each of the two parental lines, leaves of 12 replicates were pooled for each time point.

116

Glander S, He F, Schmitz G, Witten A, Telschow A, de Meaux J