Lotus japonicus alters in planta fitness of Mesorhizobium loti dependent on symbiotic nitrogen fixation

Abstract

Rhizobial bacteria are known for their capacity to fix nitrogen for legume hosts. However ineffective rhizobial genotypes exist and can trigger the formation of nodules but fix little if any nitrogen for hosts. Legumes must employ mechanisms to minimize exploitation by the ineffective rhizobial genotypes to limit fitness costs and stabilize the symbiosis. Here we address two key questions about these host mechanisms. What stages of the interaction are controlled by the host, and can hosts detect subtle differences in nitrogen fixation? We provide the first explicit evidence for adaptive host control in the interaction between Lotus japonicus and Mesorhizobium loti. In both single inoculation and co-inoculation experiments, less effective rhizobial strains exhibited reduced in planta fitness relative to the wildtype M. loti. We uncovered evidence of host control during nodule formation and during post-infection proliferation of symbionts within nodules. We found a linear relationship between rhizobial fitness and symbiotic effectiveness. Our results suggest that L. japonicus can adaptively modulate the fitness of symbionts as a continuous response to symbiotic nitrogen fixation.

Fig 1. Traits of host and symbiont in growth chamber experiment.

L. japonicus hosts were inoculated with near-isogenic M. loti strains MAFF, STM30, and STM6 and host and symbiont phenotypes were measured at times indicated. (A) Symbiont effectiveness was measured as dried shoot biomass. (B) Host investment in symbionts was measured as average individual biomass of nodules (dry weight). (C) Rhizobial fitness was estimated based on CFUs from serially diluted extracts of crushed nodules. Error bars indicate one standard error from the mean. Horizontal lines (A) represent the mean dried shoot biomass for uninoculated controls at 6 (0.003g) and 8wpi (0.002g). Data points demarked with different letters indicate significant differences between strains at a given wpi. (one-way ANOVA, post-hoc Student’s t-test, α = 0.05).

Fig 2. Traits of host and symbiont in greenhouse experiment.

L. japonicus hosts were grown in sterilized quartzite sand supplemented with N-free Jensen’s fertilizer. Hosts were singly inoculated or co-inoculated with the near-isogenic M. loti strains MAFF, STM30, and STM6 and mean host and symbiont phenotypes were measured. (A) Symbiont effectiveness measured as dried shoot biomass. (B) Host investment in symbionts was measured as average individual biomass of nodules (dry weight). (C) Rhizobial fitness was estimated based on colony forming units (CFUs) from serially diluted extracts of crushed nodules. Error bars indicate one standard error from the mean. Data points demarked with different letters indicate significant differences between strains at a given wpi (one-way ANOVA, post-hoc Student’s t-test, α = 0.05)

Fig 3. Sanctions are triggered continuously with nitrogen fixation.

Linear regression of the natural log of estimated rhizobia per nodule (counting CFUs of serially diluted extracts of crushed nodules) against the shoot biomass of the host(dry weight) Hosts were singly inoculated with near-isogenic M. loti: MAFF (pink diamond), STM30 (yellow square) and STM6 (green triangle). Each symbol represents one nodule. (A) Hosts harvested at 6wpi; F_1,38 = 8.197. (B) Hosts harvested at 8wpi; F_1,46 = 13.1525.