Tapping into the maize root microbiome to identify bacteria that promote growth under chilling conditions

Abstract

Background: When maize (Zea mays L.) is grown in the Northern hemisphere, its development is heavily arrested by chilling temperatures, especially at the juvenile phase. As some endophytes are beneficial for plants under stress conditions, we analyzed the impact of chilling temperatures on the root microbiome and examined whether microbiome-based analysis might help to identify bacterial strains that could promote growth under these temperatures.

Results: We investigated how the maize root microbiome composition changed by means of 16S rRNA gene amplicon sequencing when maize was grown at chilling temperatures in comparison to ambient temperatures by repeatedly cultivating maize in field soil. We identified 12 abundant and enriched bacterial families that colonize maize roots, consisting of bacteria recruited from the soil, whereas seed-derived endophytes were lowly represented. Chilling temperatures modified the root microbiome composition only slightly, but significantly. An enrichment of several chilling-responsive families was detected, of which the Comamonadaceae and the Pseudomonadaceae were the most abundant in the root endosphere of maize grown under chilling conditions, whereas only three were strongly depleted, among which the Streptomycetaceae. Additionally, a collection of bacterial strains isolated from maize roots was established and a selection was screened for growth-promoting effects on juvenile maize grown under chilling temperatures. Two promising strains that promoted maize growth under chilling conditions were identified that belonged to the root endophytic bacterial families, from which the relative abundance remained unchanged by variations in the growth temperature.

Conclusions: Our analyses indicate that chilling temperatures affect the bacterial community composition within the maize root endosphere. We further identified two bacterial strains that boost maize growth under chilling conditions. Their identity revealed that analyzing the chilling-responsive families did not help for their identification. As both strains belong to root endosphere enriched families, visualizing and comparing the bacterial diversity in these communities might still help to identify new PGPR strains. Additionally, a strain does not necessarely need to belong to a high abundant family in the root endosphere to provoke a growth-promoting effect in chilling conditions. Video abstract.

Keywords: Chilling temperatures; Maize; Microbiome; PGPR; Plant growth-promoting rhizobacteria; Root endosphere.

Fig. 1.

Identification of the main maize root endosphere families of field-grown and pot-grown maize. a Principal coordinate analysis of the microbial communities in bulk soil and endosphere in field- and pot-grown maize (experiments I and II). PCoA plots are based on Bray-Curtis dissimilarity indices. b Analysis of the bacterial communities in both experiments at the family level. The first two panels show the mean relative abundance of families (> 0.5% in the root endosphere) in bulk soil and endosphere. The third panel shows the relative abundance of the identified main microbiome families (highly abundant and enriched in both experiments) in the root endosphere. c Overlap of the enriched (P < 0.05) and abundant families (relative abundance > 0.5%) in the root endosphere of experiments I (blue) and II (red) representing the families of the main microbiome. d Heatmap of the enriched (P < 0.05) and abundant families (relative abundance > 0.5%) in the root endosphere of each experiment. Bulk soil and root endosphere samples are presented separately

Fig. 2

Contribution of the seed-inherited root microbiome to the maize root microbiome. a PCoA plot based on Bray-Curtis dissimilarity indices of the microbial communities in bulk soil and root endosphere of soil-grown and of in vitro-grown maize in experiment III. b Presence and abundance of the 21 ASVs belonging to 11 different families, detected in the root endosphere of in vitro-grown maize plants and compared with the abundance of the root endosphere of soil-grown maize and bulk soil samples

Fig. 3

Bacterial community shifts upon chilling temperature treatment in experiments IV and V. a Four-week-old maize plants grown in field soil. Plants on the left and the right are grown under chilling (16 h/8 h light/dark regime and 17 °C/12 °C) and under normal (16 h/8 h light/dark regime and constant 21 °C) temperature conditions, respectively. b Principal coordinate analysis of the microbial communities in bulk soil and root endosphere. Effects of the variables temperature and compartment are given by differences in color and shape, respectively. PCoA plots are based on Bray-Curtis dissimilarity indices. c Overlap between the chilling-responsive families in experiments IV and V and the main microbiome. d Root endosphere bacterial families enriched (top) and depleted (bottom) in chilling experiments IV and V (relative abundance > 0.5%). The relative abundance of the families is presented; error bars represent the standard error

Fig. 4

Bacterial collection of maize endophytes. The number of detected families in the collection, the families in the main microbiome, and chilling-responsive families is given together with the overlap of the three groups

Fig. 5

Screening assay for the effect of different bacterial isolates. a Inoculated (right tray) and mock-inoculated (left tray) juvenile maize plants grown for 30 days under chilling conditons (16 h/8 h light/dark regime and 17 °C/12 °C) (a) treated with RHG12 (Pseudoduganella sp.) (right tray) and (b) treated with RHG17 (Rhizobium sp.) (right tray). b Different bacterial isolates screened for growth-promoting effects on juvenile maize grown under chilling stress conditions. In total, 28 different isolates belonging to three different phyla were screened. Total fresh weights of bacterial- and mock-inoculated (n = 15) plants were measured and compared in two or three repeats. The figure illustrates the 95% confidence interval, based on a two-sample Student’s t test, of the treated plants compared with the mock-inoculated control. When the confidence interval does not cross the dashed line at zero, the effect is significant