Effect of bacterial inoculation, plant genotype and developmental stage on root-associated and endophytic bacterial communities in potato (Solanum tuberosum)

Abstract

Beneficial bacteria interact with plants by colonizing the rhizosphere and roots followed by further spread through the inner tissues, resulting in endophytic colonization. The major factors contributing to these interactions are not always well understood for most bacterial and plant species. It is believed that specific bacterial functions are required for plant colonization, but also from the plant side specific features are needed, such as plant genotype (cultivar) and developmental stage. Via multivariate analysis we present a quantification of the roles of these components on the composition of root-associated and endophytic bacterial communities in potato plants, by weighing the effects of bacterial inoculation, plant genotype and developmental stage. Spontaneous rifampicin resistant mutants of two bacterial endophytes, Paenibacillus sp. strain E119 and Methylobacterium mesophilicum strain SR1.6/6, were introduced into potato plants of three different cultivars (Eersteling, Robijn and Karnico). Densities of both strains in, or attached to potato plants were measured by selective plating, while the effects of bacterial inoculation, plant genotype and developmental stage on the composition of bacterial, Alphaproteobacterial and Paenibacillus species were determined by PCR-denaturing gradient gel-electrophoresis (DGGE). Multivariate analyses revealed that the composition of bacterial communities was mainly driven by cultivar type and plant developmental stage, while Alphaproteobacterial and Paenibacillus communities were mainly influenced by bacterial inoculation. These results are important for better understanding the effects of bacterial inoculations to plants and their possible effects on the indigenous bacterial communities in relation with other plant factors such as genotype and growth stage.

Fig. 1

Comparison between PCR-DGGE fingerprints for Paenibacillus sp. made from Robijn roots from juvenile (1-month old) (a) and flowering (two-months old) (b) plants. Dendrograms were made by Pearson correlation and arrow indicates the conspicuous band only present in E119^rp inoculated juvenile plants

Fig. 2

Ordination diagrams made by multivariate analysis of PCR-DGGE fingerprints made with bacterial, alphaproteobacterial and Paenibacillus primers. Separate analysis were performed for DGGE patterns obtained with primers for the domain Bacteria (a, b), Alphaproteobacteria (c, d) and Paenibacillus spp. (e, f). Also different communities were evaluated individually for roots (a, c, e) and endophytic (b, d, f). All environmental variables influencing in each analysis are presented, however, only those with significant effect (P < 0.05), according to the Monte Carlo permutation test, is marked with *. Values represent the percentage of the correlation species-environmental variable explained in each axis