Root transcriptional dynamics induced by beneficial rhizobacteria and microbial immune elicitors reveal signatures of adaptation to mutualists

Abstract

Below ground, microbe-associated molecular patterns (MAMPs) of root-associated microbiota can trigger costly defenses at the expense of plant growth. However, beneficial rhizobacteria, such as Pseudomonas simiae WCS417, promote plant growth and induce systemic resistance without being warded off by local root immune responses. To investigate early root responses that facilitate WCS417 to exert its plant-beneficial functions, we performed time series RNA-Seq of Arabidopsis roots in response to live WCS417 and compared it with MAMPs flg22⁴¹⁷ (from WCS417), flg22^Pa (from pathogenic Pseudomonas aeruginosa) and fungal chitin. The MAMP transcriptional responses differed in timing, but displayed a large overlap in gene identity. MAMP-upregulated genes are enriched for genes with functions in immunity, while downregulated genes are enriched for genes related to growth and development. Although 74% of the transcriptional changes inflicted by live WCS417 overlapped with the flg22⁴¹⁷ profile, WCS417 actively suppressed more than half of the MAMP-triggered transcriptional responses, possibly to allow the establishment of a mutually beneficial interaction with the host root. Interestingly, the sector of the flg22⁴¹⁷ -repressed transcriptional network that is not affected by WCS417 has a strong auxin signature. Using auxin response mutant tir1afb2afb3, we demonstrate a dual role for auxin signaling in finely balancing growth-promoting and defense-eliciting activities of beneficial microbes in plant roots.

Keywords: Arabidopsis thaliana; Pseudomonas simiae WCS417; auxin; chitin; flagellin; growth-defense trade-off; plant growth-promoting rhizobacteria; root microbiome; time series RNA-Seq; transcriptomics.

Figure 1

Expression profiles of microbe‐associated molecular pattern (MAMP)‐triggered immunity (MTI) marker genes MYB51 and CYP71A12, and the number of differentially expressed genes (DEGs) in roots in response to WCS417, flg22⁴¹⁷, flg22^Pa and chitin. (a) Gene expression profiles of MYB51 and CYP71A12 in WCS417‐, flg22⁴¹⁷‐, flg22^Pa‐ and chitin‐treated Arabidopsis Col‐0 roots quantified by quantitative reverse transcriptase‐polymerase chain reaction (qRT‐PCR; top) and RNA‐Seq (bottom). For qRT‐PCR, transcript levels were normalized to that of reference gene PP2AA3 (At1g13320). Data are means of three biological replicates. Error bars represent SE. Different letters represent statistically significant differences between treatments (two‐way anova, Tukey's test; P < 0.05; ns, not significant). (b) Number of induced (red bars) and repressed (blue bars) DEGs from RNA‐Seq analysis of WCS417‐, flg22⁴¹⁷‐, flg22^Pa‐ and chitin‐treated Arabidopsis Col‐0 roots at indicated time points after treatment [false discovery rate (FDR) < 0.05; > twofold].

Figure 2

Dynamics and comparative analysis of the root transcriptome in response to live WCS417 cells and the microbe‐associated molecular patterns (MAMPs) flg22⁴¹⁷, flg22^Pa and chitin. (a) Dynamics of the expression patterns of the 3559 WCS417‐, 5934 flg22⁴¹⁷‐, 6955 flg22^Pa‐ and 3342 chitin‐responsive differentially expressed genes (DEGs) in Arabidopsis Col‐0 roots. Gene expression is plotted in yellow–blue color scale, with yellow indicating upregulation and blue indicating downregulation in comparison to controls at the same time point. DEGs were clustered using SplineCluster. (b) Venn diagrams showing the pair‐wise overlap between the upregulated and downregulated DEGs of the indicated treatments. Enriched gene ontology (GO) terms associated with the core sets of upregulated (1 307) and downregulated (437) DEGs in all treatments are shown with P‐values indicated on x‐axes (full data in Dataset S3). (c) Venn diagrams showing the overlap between the upregulated and downregulated genes of the indicated treatments. Percentages indicate the percentage of the smallest group of genes that is shared with the other group of genes in the comparison. The colors of circles in the Venn diagrams are as follows: green = WCS417; orange = flg22⁴¹⁷; red = flg22^Pa; blue = chitin.

Figure 3

Time course gene ontology (GO) term enrichment analysis of the WCS417, flg22⁴¹⁷, flg22^Pa and chitin differentially expressed genes (DEGs) sets. Heatmaps represent the P‐values of GO term overrepresentation of up‐ or down‐regulated DEGs (corresponding to DEGs provided in Dataset S1) in response to treatment of the Arabidopsis roots with WCS417, flg22⁴¹⁷, flg22^Pa or chitin at given time points. On the right, overrepresented GO terms. The significance was plotted in gray–red color scale, with gray indicating no significance and red high significance.

Figure 4

Expression profile of selected genes responsive to WCS417, flg22⁴¹⁷ and flg22^Pa treatments. (a) Expression profile of SRO4,SAD6,JAZ8,WRKY30,CYP71A12 and MYB51 in Col‐0 roots by RNA‐Seq analysis. Heatmap showing the expression profile in Col‐0 roots at indicated time points after treatment with live WCS417 cells, flg22⁴¹⁷ or flg22^Pa as obtained by RNA‐Seq analysis. Gene expression was plotted in blue–yellow color scale, with blue indicating downregulation or low expression, and yellow indicating high expression. Gene expression was calculated relative to mock (log₂‐fold change). (b) Expression profile of SRO4,SAD6,JAZ8,WRKY30,CYP71A12 and MYB51 in Col‐0 and fls2 roots by quantitative reverse transcriptase‐polymerase chain reaction (qRT‐PCR). Graphs show gene expression levels in Col‐0 or fls2 roots at indicated time points after treatment with live WCS417, flg22⁴¹⁷ or flg22^Pa, as quantified by qRT‐PCR. Gene expression was normalized to the expression level of reference gene At1g13320, using the 2^−ΔCt method. The shown data are means of three replicates. Error bars represent SE. Different letters represent statistically significant differences between treatments (two‐way anova, Tukey's test; P < 0.05; ns, not significant).

Figure 5

Comparative analysis of the root transcriptome to live WCS417 and flg22⁴¹⁷. Venn diagrams showing the overlap between the genes upregulated and downregulated upon treatment with WCS417 and flg22⁴¹⁷. The bottom section shows the most highly enriched gene ontology (GO) terms associated with the sets of differentially expressed genes (DEGs) that are specifically up‐ or downregulated after root exposure to flg22⁴¹⁷, but not upon colonization of the roots by live WCS417 bacteria. P‐values are indicated on the x‐axes.

Figure 6

Effect of WCS417, flg22⁴¹⁷, flg22^Pa and chitin on shoot and root growth in Col‐0 and the fls2 mutant. (a) Effect of flg22⁴¹⁷, flg22^Pa and chitin on shoot and root growth of Col‐0 and fls2 seedlings. Seven‐day‐old Arabidopsis seedlings were cultivated for 7 days in liquid Murashige and Skoog (MS) medium with 0.5% sucrose, supplemented with or without 500 nm flg22⁴¹⁷ or flg22^Pa, or 500 μg ml ⁻¹ chitin, after which shoot fresh weight, number of lateral roots formed and primary root length were measured. (b and c) Effect of live WCS417 bacteria on shoot and root growth of Col‐0 and fls2. Three‐day‐old Arabidopsis seedlings were cultivated for 7 days on agar‐solidified MS medium with 0.5% sucrose, and inoculated with a 10 μl suspension containing 2 × 10⁶ colony‐forming units (CFU) of WCS417 cells or not inoculated (Control) right below the root tip of each seedling, resulting in rapid colonization of the whole root system. Error bars indicate SE (n = 24). Different letters indicate statistically significant differences between Col‐0 and fls2 after control and WCS417 treatment (two‐way anova, Tukey's test; P < 0.05), and asterisks indicate significant difference between Col‐0 and fls2 plants treated with the same elicitor (two‐way anova, Sidak's test; *^*** P < 0.0001; ns, not significant).

Figure 7

Auxin signatures in the early transcriptome of WCS417‐colonized roots and auxin signaling‐dependent induction of the induced systemic resistance (ISR) markers MYB72 and IRT1. (a) Venn diagram showing the overlap between genes responding to indole‐3‐acetic acid (IAA; auxin) and WCS417 in Arabidopsis roots. The effect of auxin was tested 4 h after treatment of the roots with 5 μm IAA (dataset from Chaiwanon and Wang, 2015), while the effect of WCS417 was tested at all time points after application of the rhizobacteria. (b) Expression levels of MYB72, IRT1, GH3.3 and SAD6 as quantified by quantitative reverse transcriptase‐polymerase chain reaction (qRT‐PCR). Expression was tested in roots of 14‐day‐old Col‐0 and tir1afb2afb3 seedlings at 48 h after inoculation with WCS417. Gene expression levels were normalized to that of reference gene At1g13320 using the 2^−ΔCt method. The shown data are means of three biological replicates. Error bars represent SE. Different letters represent statistically significant differences between treatments (two‐way anova, Tukey's test; P < 0.05).