Microbe‐induced drought tolerance by ABA‐mediated root architecture and epigenetic reprogramming

Abstract

The use of beneficial microbes to mitigate drought stress tolerance of plants is of great potential albeit little understood. We show here that a root endophytic desert bacterium, Pseudomonas argentinensis strain SA190, enhances drought stress tolerance in Arabidopsis. Transcriptome and genetic analysis demonstrate that SA190‐induced root morphogenesis and gene expression is mediated via the plant abscisic acid (ABA) pathway. Moreover, we demonstrate that SA190 primes the promoters of target genes in an epigenetic ABA‐dependent manner. Application of SA190 priming on crops is demonstrated for alfalfa, showing enhanced performance under drought conditions. In summary, a single beneficial root bacterial strain can help plants to resist drought conditions.

The beneficial root endophytic desert bacterium Pseudomonas argentinensis sp. SA190 enhances drought stress tolerance in Arabidopsis by priming target gene promoters in an epigenetic ABA‐dependent manner.

Introduction

Climate change increases the frequency, severity, and duration of drought in many parts of the world, making crop resistance to drought stress a major goal of agricultural biotechnology. The percentage of the planet affected by drought has doubled in the last 40 years which affected humanity more than any other natural hazard. Drought stress adversely affects the plant growth, which results in reduced crop yield and ultimately to critical food shortages or famines (Seleiman et al, ²⁰²¹). Although enormous progress has been made to understand and identify important molecular players of drought stress in model plants, success in knowledge transfer to economical crops is still limited (Cominelli et al, ²⁰¹³). This is partially due to the fact that drought stress is a complex trait causing dramatic changes in many physiological plant parameters (Zhu, 2002). After all, the different stress conditions that coexist in open field agriculture go beyond simple arithmetic of the tolerance to single stresses (Suzuki et al, ²⁰¹⁴).

Drought stress drastically changes plant morphology, biochemistry and physiology leading to severe impacts on plant growth and yield. The first physiological response of plants to drought is the closure of stomata to avoid water loss via transpiration (Pirasteh‐Anosheh et al, ²⁰¹⁶). The decrease in transpiration rate leads to a decrease in relative water content (RWC) of plants (Giday et al, ²⁰¹⁴). Stomatal closure also decreases plant photosynthetic efficiency by decreasing intracellular CO₂ concentrations (Lisar et al, ²⁰¹⁶).

Abscisic acid (ABA) is one of the key phytohormones, which regulates the plant response to drought stress. ABA acts through a conserved signal transduction pathway, comprised of a PYRABACTIN RESISTANCE 1‐Like (PYL)–PROTEIN PHOSPHATASE 2C (PP2C) and SNF1‐RELATED PROTEIN KINASE 2 (SnRK2) module. ABA binding to PYL protein triggers a conformational change in the receptors, which allows it to bind and inhibit the PP2C that normally represses ABA signaling. This PYL–ABA–PP2C complex releases SnRK2 from the otherwise inhibitory complex with PP2C, initiating phosphorylation of transcription factors that regulate gene expression involved in ABA output responses (Fujii et al, ²⁰⁰⁹; Cutler et al, ²⁰¹⁰; Chen et al, ²⁰²⁰). ABA also regulates the early production of reactive oxygen species (ROS) after drought perception (Cruz de Carvalho, 2008). ROS serves as a stress signal to activate downstream processes, but accumulation of ROS can lead to cell damage and growth penalty.

Water movement across the cellular membranes is largely regulated by a family of water channel proteins called aquaporins. In Arabidopsis, there exist 35 aquaporin genes (Johanson et al, ²⁰⁰¹) which can be broadly divided into plasma membrane intrinsic proteins (PIPs) and tonoplast intrinsic proteins (TIPs) genes. Some of the Arabidopsis PIPs and TIPs proved to be active water channels in Xenopus oocytes (Quigley et al, ²⁰⁰²). The biological significance of aquaporins in plants is their ability to modulate transmembrane water transport in situations where adjustment of water flow is physiologically critical (Li et al, ²⁰¹⁴). PIPs play an important role in controlling the transcellular water transport and are subdivided into two subfamilies PIP1 and PIP2 (Maurel et al, ²⁰¹⁵). The overexpression of PIP isoforms increased the root osmotic hydraulic conductivity, transpiration and shoot‐to‐root ratio while the downregulation leads to drought stress susceptibility (Pawłowicz & Masajada, 2019). However, when expressed in heterologous systems, the overexpression of aquaporins can lead to a negative effect on stress resistance, due to the fact that the native stress response machinery may recognize them as foreign proteins (Li et al, ²⁰¹⁴).

The transcriptional responsiveness of drought stress‐regulated genes is correlated with changes in histone modifications (Kim et al, ^{2008, 2012}; To & Kim, ²⁰¹⁴). Two of the important histone modifications, H3K4me3 and H3K9ac, are enriched on drought stress‐regulated genes like RD20 and RD29A. The levels of these modifications change from mild to severe drought stress, suggesting that epigenetic responsiveness depend on the intensity of the drought stress (Kim et al, ^{2008, 2012}).

Beneficial plant microbes have been used to overcome abiotic stress challenges of plants (de Zelicourt et al, ²⁰¹³; Busby et al, ²⁰¹⁷). Indeed, plants and their rhizosphere host diverse microbial communities, selected from bulk soil (Bulgarelli et al, ²⁰¹²), and the beneficial bacteria, defined as plant growth‐promoting bacteria (PGPBs), can establish symbiotic associations to promote plant growth under optimal conditions or in response to biotic and abiotic stresses (Finkel et al, ²⁰¹⁷; de Zelicourt et al, ²⁰¹⁸; Saad et al, ²⁰²⁰; Synek et al, ²⁰²¹). A number of PGPBs have been reported to mitigate drought stress in a variety of plant species implicating a number of processes, such as modification of phytohormonal levels, production of heat‐shock proteins and dehydrins, production of osmolytes, exopolysaccharides, volatiles, detoxification of ROS, and the accumulation of compatible solutes like sugars, amino acids, and polyamines (Sati et al, ²⁰²³). However, a detailed molecular and genetic analysis of the underlying molecular mechanisms has not been done so far.

The ideal natural reservoir for the isolation of beneficial bacteria which help in drought stress tolerance is the desert. To this end, a number of rhizo‐ and endosphere bacterial strains from desert plants were isolated, and abiotic stress screens were conducted on Arabidopsis under controlled lab conditions (Lafi et al, ²⁰¹⁶; de Zelicourt et al, ²⁰¹⁸; Bokhari et al, ²⁰²⁰). Here, we report on Pseudomonas argentinensis sp. SA190, an endophytic bacterium isolated from root nodules of the indigenous desert plant Indigofera argentea (Lafi et al, ²⁰¹⁶), that massively increases drought tolerance in plants. Physiological and genetic analysis showed that SA190 maintains growth and photosynthesis by enhancing plant water use efficiency (WUE). Transcriptome analysis uncovered that SA190 induces the expression of multiple aquaporin genes in the roots of Arabidopsis. Inhibition of aquaporin function abrogates SA190‐induced plant drought tolerance. SA190 epigenetically regulates gene expression by enhanced H3K4me3 deposition. The agronomic potential of SA190 is shown in drought experiments with alfalfa (Medicago sativa).

Results

Pseudomonas argentinensis sp. SA190 enhances drought tolerance of Arabidopsis and Medicago sativa

Pseudomonas argentinensis strain SA190 is a plant endophytic bacterium that was previously isolated from root nodules of the desert plant I. argentea (Lafi et al, ²⁰¹⁶). To test the effect of SA190 under drought‐mimicking conditions, plants were germinated for 5 days on ½ MS agar medium containing 10⁸ cfu of P. argentinensis sp. SA190 before transfer to fresh ½ MS plates infiltrated with 0 or 25% polyethylene‐glycol (PEG; Fig 1A). After 16 days, plant morphology, total fresh weight (FW) and dry weight (DW), root length, and lateral root density (LRD) were determined (Fig 1). Our results show that SA190 did not significantly influence the morphology or development of Arabidopsis under non‐stress conditions (Fig 1B). Upon 25% PEG treatment, the growth of non‐colonized plants was severely inhibited (Fig 1B) resulting in a reduction of FW and DW by more than 90% (Fig 1C and D). In contrast, Arabidopsis colonized with SA190 maintained growth, resulting in a six‐fold enhancement of FW and DW on 25% PEG (Figs 1C and D, and EV1). In addition, a significant increase in the primary root length and development of the lateral root system were observed in SA190 colonized plants (Fig 1E and F).

Figure 1. Pseudomonas argentinensis sp. SA190 enhances drought tolerance in Arabidopsis and crops.

• A
  Graphical representation of workflow for plant screening assay under 25% PEG.
• B
  Growth of non‐colonized and SA190‐colonized Arabidopsis plants grown under drought stress proxy for 21 days (½ MS + 25% PEG) or ambient conditions for 14 days (½ MS).
• C–F
  (C) Total fresh weight, (D) Total dry weight, (E) Primary root length and (F) Lateral root density of 21‐day‐old, non‐colonized (Mock) and SA190‐colonized plants upon growth for 16 days on ½ MS or ½ MS + 25% PEG. Boxplot shows data from three independent biological replicates (technical replicates n = 18–24) where whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate the statistical differences based on the Student's t‐test (***P < 0.001).
• G, H
  (G) Growth and (H) Survival percentage of mock‐ and SA190‐colonized plants in jiffy pots grown upon watering for 2 weeks, no watering for 3 weeks and after 1 week of re‐watering. All plots represent the mean of three biological replicates (technical replicates n = 36). Whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate the statistical differences based on the Student's t‐test (***P < 0.001).
• I
  Alfalfa biomass yield in sandy soil open field agriculture with normal or reduced irrigation (30% reduction). The differences in biomass yield between different treatments are indicated in percent (%) and significance at P < 0.05 is indicated by asterisks (*). Non‐colonized plants are indicated in gray and SA190‐colonized in green. Values in the graph represents mean of six biological replicates with ±SEM.
• J
  Confocal microscopy showing the colonization of SA190:GFP on 5‐day‐old seedlings grown on ½ MS and 25% PEG. Green signal corresponds to GFP‐tagged SA190 while in magenta are plant cell walls stained with propidium iodide (PI). Bar indicates 100 nm.
• K
  SA190 colonization levels in Arabidopsis under normal (½ MS) and drought proxy (½ MS + 25% PEG) conditions as quantified by colony forming units of bacteria per mg of plant fresh weight. Boxplot shows data from four independent biological replicates where whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate the statistical differences based on the Student's t‐test (*P < 0.05, ***P < 0.001).

Source data are available online for this figure.

Figure EV1. SA190 enhances Arabidopsis PEG drought stress proxy tolerance.

• A, B
  (A) Shoot fresh weight and (B) Root fresh weight (b) of 16‐day‐old mock‐ and SA190‐colonized plants grown on ½ MS medium and ½ MS + PEG medium.
• C, D
  (C) Shoot dry weight and (D) Root dry weight of 16‐day‐old mock‐ and SA190‐colonized plants grown on ½ MS medium and ½ MS + PEG medium.

Data information: All plots represent the mean of three biological replicates (technical replicates n > 39). Whiskers in boxplot represents minimum and maximum values, “+” indicates mean and horizontal line represents median Error bars represent SE. Asterisks indicate the statistical differences based on the Student's t‐test (***P < 0.001).

To see the effect of SA190 on the growth of adult plants under drought conditions, 5 day‐old non‐colonized or SA190‐colonized Arabidopsis were transferred to pots for 2 weeks. Thereafter, watering was stopped, and plants were subjected to drought stress for 3 weeks. While non‐colonized (Mock) plants showed a drop in survival by more than 70%, SA190‐colonized plants were affected by less than 10% (Fig 1G and H). One week after re‐watering, less than 50% of non‐colonized plants resumed growth in comparison with a 100% of SA190‐colonized plants (Fig 1G and H). These results indicate that SA190 significantly enhances drought stress tolerance of Arabidopsis plants.

To evaluate the agronomic potential of SA190 on crops, we performed growth chamber experiments with alfalfa which is used as an important animal feed in different regions of the world. Alfalfa seeds were coated with SA190 and tested in parallel with mock‐coated seeds in control growth chamber conditions. Randomized pot experiments with a drought protocol were performed in independent biological replicates. A control set of plants that was continuously watered for 52 days was used as controls. In parallel, for the drought experiment, plants were only watered for 42 days and then evaluated at day 52. SA190‐inoculated alfalfa plants showed a 13% increase in biomass yield under continuous irrigation (Figs 1I and EV2). When the plants were exposed to 10 days without irrigation (drought stress), SA190‐inoculated plants showed 37% higher biomass to the drought conditions compared with non‐inoculated plants. These results show that SA190 can efficiently enhance the performance of alfalfa under the applied drought regime.

Figure EV2. SA190 enhances alfalfa drought stress tolerance.

Photographs of 52‐day‐old alfalfa grown on trays in green house with and without SA190 colonization and then subjected to 10 days drought stress.

SA190‐enhanced root architecture and colonization of Arabidopsis upon polyethylene‐glycol drought stress proxy

Since SA190 was isolated from the roots of the desert plant I. argentea, we characterized the interaction between SA190 and Arabidopsis in more detail. After 16 days on normal or 25% PEG medium, root length and LRD were determined. Our results show that SA190 did not significantly influence the morphology or development of Arabidopsis roots under non‐stress conditions (Figs 1B and EV1A–D). However, on 25% PEG medium and in contrast to non‐colonized plants, SA190‐inoculated plants maintained their root development at wild‐type (WT) levels, while mock‐treated plants showed a strong reduction in root growth (Fig 1B), with a massive increase in root FW and DW (Fig EV1B and D) mostly due to enhanced growth of the primary root and development of the lateral root system (Fig 1E and F).

We next investigated root colonization by SA190. To this end, SA190 was stably transformed with a GFP‐expressing cassette (SA190:GFP). Confocal microscopy revealed that SA190:GFP mainly colonized roots (Fig 1J), preferentially epidermal cells of the root elongation and differentiation zones (Fig 1J). Interestingly, the roots of plants grown on 25% PEG showed enhanced colonization compared with plants grown only on ½ MS agar (Fig 1J). At later stages of colonization, SA190:GFP was also found inside root tissues (Fig 1J). The endophytic nature of SA190 was also revealed by determining the SA190 colony forming units (CFUs) of the plants after intensive washing (Fig 1K).

SA190 reprograms the transcriptional response of Arabidopsis to polyethylene‐glycol stress

To understand the molecular mechanism of SA190‐induced PEG tolerance, we next performed RNA‐seq analysis of 21‐day‐old non‐ or SA190‐colonized roots upon normal and PEG stress for 16 days. The root transcriptome data were organized by hierarchical clustering into groups of differentially expressed genes (DEGs) according to their expression patterns in non‐colonized (Mock) and SA190 colonized plants (SA190) under normal conditions, as well as non‐colonized and SA190‐colonized plants under PEG stress conditions, denoted as PEG and PEG + SA190, respectively (Fig 2A). We then assessed the gene ontology enrichment of the up‐ and downregulated root DEGs. When comparing SA190‐ with mock‐colonzied plants, we only observed 30 up‐ and 145 down‐regulated DEGs under non‐stress conditions. In this relatively small gene set, we, nonetheless, found significant GO term enrichment of genes related to cell wall organization (e.g., expansins and pectin lyases) and response to water deprivation.

Figure 2. Transcriptome analysis and role of ABA signaling in SA190‐induced PEG drought proxy tolerance.

• A
  Heatmap showing hierarchical clustering of up‐ and downregulated DEGs profiles in response to PEG (25%), SA190 or both treatments based on root RNA‐Seq analysis. Quantified read counts are normalized for genes. Colors indicate normalized expression levels. q < 0.05 and a fold change cut‐off was set to log₂ > 1 for up‐ and log₂ > −1 for down regulated genes.
• B, C
  GO term analysis of differentially expressed genes of SA190‐colonized (PEG+190) compared with non‐colonized (PEG) plants under drought proxy conditions (½ MS + 25% PEG).
• D, E
  Total fresh weight and root fresh weight of 21 days old non‐colonized (Mock) and SA190‐colonized (SA190) Col‐0, aba2‐1 and qpyr/pyl plants grown under drought proxy conditions (½ MS + 25% PEG).
• F, G
  Primary root length and Lateral root density of 21‐day‐old, mock‐ and SA190‐colonized Col‐0, aba2‐1 and qpyr/pyl plants upon growth for 16 days on ½ MS + 25% PEG.

Data information: Boxplot shows data from three independent biological replicates (technical replicates n = 36) where whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate the statistical differences based on the Student's t‐test (*P < 0.05, ***P < 0.001).

Source data are available online for this figure.

Under PEG stress conditions, massive changes in DEGs were seen between SA190‐colonized and non‐colonized roots revealing 2,384 up and 1,365 downregulated genes. Whereas the upregulated genes showed enrichment in responses to hormone signaling, glucosinolate biosynthesis, water deprivation, and oxidative stress as well as root morphogenesis and developmental growth (Fig 2B), the most significant GO terms in the set of downregulated genes were related to ribosome, RNA metabolism, and DNA repair (Fig 2C). Interestingly, the hormone, abiotic stress, and water deprivation‐related GO terms were all related to ABA signaling, including the ABA receptor components PYL1, PYL3, the ABA protein phosphatases HAI1, and HAI2, as well as the key ABA‐responsive protein kinase SnRK2‐3 and a large number of aquaporin genes.

ABA pathway mutants are compromised in SA190‐induced polyethylene‐glycol stress tolerance

Since ABA plays a key role in drought stress tolerance and our transcriptome analysis showed that SA190 colonization of plants affects ABA signaling under PEG stress, we tested Arabidopsis mutants impaired in ABA biosynthesis (aba2‐1) and signaling pyr1 pyl4 pyl5 pyl8 (qpyr/pyl). Under non‐stress conditions, we observed no significant differences in the FW or DW of SA190‐colonized and non‐colonized plants between WT and both mutants (Fig EV3A). Under PEG drought proxy stress, however, SA190‐maintained plant growth was strongly compromised in both aba2‐1 and qpyr/pyl mutants (Fig 2D), indicating a central role of the ABA pathway in the SA190‐induced Arabidopsis PEG stress tolerance.

Figure EV3. ABA does not influence SA190‐induced Arabidopsis growth and development under ambient conditions.

• A, B
  Total fresh weight and root fresh weight of 21‐day‐old non‐colonized (Mock) and SA190‐colonized (SA190) Col‐0, aba2‐1 and qpyr/pyl plants grown under non‐stress conditions (½ MS).
• C, D
  Primary root length and lateral root density of 21‐day‐old, mock‐ and SA190‐colonized Col‐0, aba2‐1 and qpyr/pyl plants upon growth for 16 days on ½ MS plates.

Data information: Boxplot shows data from three independent biological replicates (technical replicates n = 18–24) where whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate the statistical differences based on the Student's t‐test (***P < 0.001).

SA190‐induced changes in root architecture are dependent on ABA signaling

ABA plays an important role in pathogenic plant‐microbe interactions. Since SA190 induces massive changes in Arabidopsis root structure, we next tested whether ABA pathway mutants affect SA190‐induced modification of Arabidopsis root morphogenesis. Under non‐stress conditions, we observed no significant differences in the FW of SA190‐colonized and non‐colonized roots comparing WT with either ABA biosynthesis (aba2‐1) or ABA signaling pyr1 pyl4 pyl5 pyl8 (qpyr/pyl) mutants (Fig EV3B–D). Under PEG drought proxy stress, however, the SA190‐induced changes in root structure, including FW, primary root length, and LRD were strongly compromised in both aba2‐1 and qpyr/pyl mutants (Fig 2E–G).

Next, we tested whether ABA‐deficient mutants might compromise the interaction of Arabidopsis with SA190 and, therefore, result in the loss of the beneficial effect on drought stress tolerance. However, similar levels of SA190 colonization were seen under non‐stress (½ MS) and drought proxy stress (½ MS + 25%PEG) conditions (Fig EV4), indicating that ABA does not influence SA190 colonization of Arabidopsis. These results show a central role of the ABA pathway in SA190‐induced changes in root morphogenesis.

Figure EV4. ABA does not influence SA190 colonization of Arabidopsis .

Efficiency of root colonization by SA190 in 21‐day‐old Col‐0 and aba2‐1 plants grown for 16 days under normal (½ MS) or drought proxy conditions (½ MS + 25% PEG). Boxplot shows data from four independent biological replicates where whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate the statistical differences based on the Student's t‐test (*P < 0.05, ***P < 0.001).

SA190 modulates the expression of aquaporin genes under polyethylene‐glycol drought proxy stress in an ABA‐dependent manner

RNAseq analysis of root tissue revealed that under PEG conditions, a large set of Arabidopsis genes including several aquaporins were differentially regulated by SA190 (Fig 3A). Since aquaporins are critical elements in adjusting the water flow in physiologically critical situations (Li et al, ²⁰¹⁴), many of which are regulated by ABA (Pawłowicz & Masajada, 2019), we analyzed these aquaporin genes as proxy for the set of SA190‐regulated genes in more detail. qRT‐PCR analysis of aquaporin transcript levels did not change under non‐stress conditions but showed significantly higher levels in roots of SA190‐colonized compared with non‐inoculated plants (Fig 3B). Both aba2‐1 and qpyr/pyl mutants strongly compromised the SA190‐enhanced expression of all tested aquaporin genes under PEG drought stress proxy conditions (Fig 3C), indicating that the ABA pathway is essential in mediating the SA190‐induced changes of aquaporin gene expression.

Figure 3. SA190‐induced aquaporin gene expression is mediated in an ABA‐dependent manner.

1. Heat map of expression profile of differentially regulated PIP and TIP genes in root RNA‐Seq analysis.
2. Relative gene expression of aquaporin PIP and TIP genes by qRT‐PCR analysis normalized to tubulin levels in 21‐day‐old mock‐ and SA190‐colonized roots of Arabidopsis grown for 16 days on ½ MS with or without 25% PEG.
3. Relative gene expression of aquaporin PIP and TIP genes by qRT‐PCR analysis in Col‐0, aba2‐1 and qpyr/pyl mutants normalized to tubulin levels in 21‐days old mock‐ and SA190‐colonized roots of Arabidopsis grown for 16 days on ½ MS with or without 25% PEG.

Data information: Values represent the means of three biological experiments. Error bars indicate SE. All plots represent the mean of three biological replicates (biological replicates n = 3). Asterisks indicate a statistical difference based on the Student's t‐test (*P < 0.05; **P < 0.01; ***P < 0.001).

Source data are available online for this figure.

Epigenetic priming of SA190‐targeted genes is mediated by the ABA pathway

Our transcriptome analysis of SA190‐colonized plants showed that the expression of many genes is not constantly increased, but only shows differential changes under drought stress proxy conditions. We, therefore, tested whether SA190 colonization might already be related to changes in the epigenetic state of the differentially regulated aquaporin genes under non‐stress conditions. As described in Fig 1, plants were germinated on SA190 containing ½ MS media for 5 days before transfer onto ½ MS + 25%PEG but containing no bacteria. As control, we used plants that were germinated in parallel but in the absence of SA190 (mock). As H3K4me3 was found as a priming mark for genes in drought‐trained plants (Ding et al, ^2012a), we tested different regions of the set of SA190‐regulated aquaporin genes for H3K4me3 enrichment. We observed a significant enrichment of the H3K4me3 mark in these aquaporin genes, especially near transcription start sites (Fig 4). Since ABA was essential for mediating SA190‐induced aquaporin gene expression and PEG drought stress proxy tolerance, we also analyzed the epigenetic status of these aquaporin genes in the ABA‐deficient aba2‐1 mutant. In non‐colonized (mock) plants, the aba2‐1 mutant showed similar levels of H3K4me3 as WT plants. In SA190‐colonized plants, however, aba2‐1 mutant plants were completely compromised in the enhancement of H3K4me3 levels in the aquaporin gene loci. We conclude that SA190 primes aquaporin genes in an ABA‐dependent manner by enhancing H3K4me3 levels.

Figure 4. SA190 primes aquaporin genes via H3K4me3 enrichment.

Relative enrichment of the H3K4me3 mark at the indicated gene loci of selected PIPs and TIPs of non‐colonized (Mock) and SA190‐colonized (SA190) plants in Col‐0 and aba2‐1 mutant as determined by chromatin immunoprecipitation‐qPCR (ChIP‐qPCR). The regions targeted for amplification are labeled as R1, R2, and R3. R1 encompasses 5′‐UTR, R2 is close to the transcription start site while R3 is outside the gene body. Amplification values were normalized to input and H3 and region 3 (R3) of non‐colonized (Mock) Col‐0 plants. The plots represent the means of two biological replicates. Error bars represent SE. (ns) non‐significant Asterisks indicate a statistical difference based on two‐way ANOVA (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001 for P value differences between the conditions compared with Col‐0 Mock using Dunnett's multiple‐comparison test).Source data are available online for this figure.

Role of aquaporins in SA190‐induced polyethylene‐glycol drought stress proxy tolerance

Our analysis suggested that aquaporins might be key targets and mediators of SA190‐induced drought stress tolerance. Aquaporins play a key role in transmembrane water transport in environmental stress responses (Li et al, ²⁰¹⁴; Maurel et al, ²⁰¹⁵) and their enhanced expression increased the root osmotic hydraulic conductivity, transpiration and shoot‐to‐root ratio while the downregulation lead to higher drought stress susceptibility (Pawłowicz & Masajada, 2019). To directly test the significance of the set of aquaporins identified in the SA190 transcriptome under PEG stress conditions, we used the quadruple pip1;1 pip1;2 pip1;3 pip1;5 (qpip) and tip2;1 Arabidopsis mutants in the absence and presence of SA190 (Fig 5A and B). No significant changes of SA190‐induced PEG drought proxy tolerance were observed when comparing the growth of WT with qpip and tip2;1 mutants under non‐stress (Fig 5A) or PEG stress conditions (Fig 5B). We also assessed whether the aquaporin mutants affect SA190 colonization. Although PEG conditions clearly induced SA190 colonization of Arabidopsis, this effect was similar for WT, qpip, and tip2;1 mutants (Fig 5C).

Figure 5. Role of SA190 colonization in aquaporin and plant PEG drought stress proxy physiology.

• A, B
  Total fresh weight of 21‐day‐old non‐colonized‐ (Mock) and SA190‐colonized (SA190) Col‐0, tip2;1 and qpip plants grown for 16 days under normal (½ MS) or drought stress proxy conditions (½ MS + 25% PEG).
• C
  Quantification of SA190 colonization efficiency in 21‐day‐old SA190‐colonized (SA190) Col‐0, tip2;1 and qpip plants grown for 16 days under normal (½ MS) or drought conditions (½ MS + 25% PEG).
• D, E
  Total fresh weight of mock‐ and SA190‐colonized 21‐day‐old Arabidopsis plants grown for 16 days under normal on (½ MS) or PEG drought stress proxy (½ MS + 25% PEG) conditions supplemented with the aquaporin inhibitors HgCl₂ (38.2 μM) or AgNO₃ (5 μM).
• F
  The percentage of leaf relative water content (RWC) of 21‐day‐old mock‐ or SA190‐colonized plants grown for 16 days under drought stress proxy (½ MS + 25% PEG) conditions. The RWC was calculated as the formula RWC (%) = [(FW‐DW)/(TW‐DW)] × 100, where FW is fresh weight, DW is dry weight, and TW is the weight of fully turgid leaves.
• G
  Transpiration rate of mock‐ or SA190‐colonized 21‐day‐ old plants after transfer of 5‐day‐old seedlings from ½ MS to ½ MS ± 25% PEG. Water loss was quantified by calculating the changes in fresh weight for 30 min with reading at every 5 min interval.
• H, I
  Comparison of stomatal aperture of SA190‐ and non‐colonized 21‐day‐old plants grown for 16 days under non‐stress (½ MS) or drought stress proxy (½ MS + 25% PEG) conditions. Scale bar = 20 μm.
• J
  Water use efficiency (WUE) of 4‐week‐old mock‐ or SA190‐colonized plants under PEG drought stress proxy conditions. WUE was calculated as mg of dry weight produced per ml of water used.

Data information: All plots represent the mean of three biological replicates. Error bars represent SE. Boxplot shows data from three independent biological replicates (technical replicates n = (A) 16–20, (B) 8–20, (C) 2–6, (D, E) 24, (F) 15, (I) 94–104, (J) 6) where whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate the statistical differences based on the Student's t‐test (*P < 0.05, ***P < 0.001). Asterisks indicate a statistical difference based on the Student's t‐test (***P < 0.001).

Source data are available online for this figure.

The maintenance of the beneficial response of Arabidopsis to SA190 in the pip and tip mutants might also be due to the redundancy of the large aquaporin gene family in Arabidopsis. To overcome this problem, we also made use of two known aquaporin inhibitors, AgNO₃ and HgCl_2, which had no effect on Arabidopsis growth under normal conditions, irrespectively whether the plants were SA190‐ or non‐colonized (Fig 5D). However, under PEG drought stress proxy conditions, the inhibitors completely compromised the beneficial effect of SA190 on Arabidopsis (Fig 5E).

The fact that one of the aquaporin inhibitors was AgNO₃, which is also an inhibitor of ethylene signaling, made us question its role in the beneficial interaction of SA190 with Arabidopsis. We, therefore, tested a number of ethylene biosynthesis and signaling mutants for a possible role in the SA190‐Arabidopsis interaction. None of the ethylene mutants affected the growth of SA190‐colonized plants in comparison with non‐colonized plants under non‐stress conditions (Fig EV5A). Compared with the massive loss of beneficial activity of SA190 in ABA mutants (Fig 2D), SA190‐induced PEG drought stress proxy tolerance in the ethylene mutants was only mildly affected under water‐limiting conditions (Fig EV5B), suggesting that ethylene does not play a major role in mediating SA190 PEG stress tolerance of Arabidopsis plants.

Figure EV5. Arabidopsis‐SA190 interaction is independent of the ethylene pathway and the role in plant physiology.

• A, B
  Total fresh weight of mock‐ and SA190‐colonized 16‐day‐old ethylene mutants on (A) ½ MS or (B) 25%PEG. Boxplot shows data from three independent biological replicates (technical replicates n = 18–24) where whiskers represents minimum and maximum values, “+” indicates mean and horizontal line represents median. Asterisks indicate a statistical difference based on the Student's t‐test (***P < 0.001).
• C
  The percentage of leaf relative water content (RWC) of 21‐day‐old mock‐ or SA190‐colonized plants grown for 16 days under normal (½ MS) conditions. The RWC was calculated as the formula RWC (%) = [(FW−DW)/(TW−DW)] × 100, where FW is fresh weight, DW is dry weight, and TW is the weight of fully turgid leaves.
• D
  Comparison of stomatal aperture of SA190‐ and non‐colonized 21‐day‐old plants grown for 16 days under non‐stress (½ MS) conditions. Scale bar = 20 μm.
• E
  Water use efficiency (WUE) of 4‐week‐old mock‐ or SA190‐colonized plants grown under non‐stress (½ MS) conditions. WUE was calculated as mg of dry weight produced per ml of water used.

Data information: All plots represent the mean of three biological replicates. Whiskers in the boxplot represent minimum and maximum values, “+” indicates mean, and the horizontal line represents median. Asterisks indicate a statistical difference based on the Student's t‐test.

SA190 colonization changes plant polyethylene‐glycol drought stress proxy physiology

The importance of aquaporins in regulating transpiration and photosynthesis in drought stress and its importance in WUE, growth and yield was recently highlighted (Moshelion et al, ²⁰¹⁵). We, therefore, addressed whether similar effects could be observed in SA190‐colonized plants when exposed to PEG stress conditions.

Since drought stress leads to a decreased water content in plants, we measured the water content of leaves in SA190‐colonized and non‐colonized plants under non‐stress and PEG drought proxy conditions. Under non‐stress conditions, no significant differences in the RWC of leaves were observed when comparing SA190‐ and non‐colonized plants (Fig EV5C). PEG stress induced reduction in the relative leaf water content in non‐colonized plants (Fig 5F) was not seen in SA190‐colonized plants, which maintained leaf RWC to the same level as seen in plants grown under non‐stress conditions (Fig EV5C).

Plants commonly react to drought stress by reducing transpiration to save water. However, reduced transpiration also means reduced photosynthesis and growth. Under non‐stress conditions, we observed a slightly lower rate of water loss in SA190‐colonized plants compared with non‐colonized plants (Fig 5G). In contrast, whereas growth on 25% PEG resulted in reduced water loss in non‐colonized plants, SA190‐colonized plants showed a much higher transpiration rate, consistent with a better overall water status under stress conditions (Fig 5G).

Stomatal closure is a common adaptation response of plants to drought conditions to reduce water loss. We, therefore, analyzed stomatal opening under non‐stress and 25% PEG in SA190‐colonized and non‐colonized plants (Fig 5H). Under non‐stress growth conditions, no statistically significant differences in stomatal aperture were observed between SA190‐colonized and non‐colonized plants (Fig EV5D). However, under drought stress proxy conditions, non‐colonized plants rapidly induced stomatal closure, whereas SA190‐colonized plants showed significantly more open stomata (Fig 5I).

The observation that SA190‐colonized plants showed higher water contents and significantly higher growth rates under water‐limiting conditions suggested that SA190 plants might make better use of the available water. We, therefore, measured whether SA190 altered the WUE of its host plant. We observed no significant change in WUE between mock‐ and SA190‐colonized plants under normal conditions (Fig EV5E). However, when plants were exposed to PEG drought stress proxy conditions, the WUE of SA190‐colonized plants was clearly superior to that of non‐colonized plants (Fig 5J). Overall, these results suggest that an enhanced water status in SA190‐colonized plants is the basis for maintaining the growth of Arabidopsis plants under water‐limiting conditions.

Discussion

Water is critical for life on Earth and is a major determinant of food production worldwide. Due to the increasing world population and climate change, water is becoming increasingly scarce on our planet. Since globally 70% of all freshwater is used by agriculture, increasing the WUE of crops as well as making crops more drought tolerant are among the most important targets in agricultural engineering. Since water uptake primarily occurs via roots, strategies to improve WUE through changes in root system architecture have recently gained an increased interest. Besides genetic engineering and molecular breeding, the application of PGPBs might offer another way to modify the root system and enhance WUE and drought tolerance of crops. We report in this work that the root endophytic bacterium P. argentinensis SA190 isolated from the roots of the desert plant I. argentea induces major changes in root architecture and maintains the growth of plants under water‐limiting conditions. Although SA190 was isolated from root nodules, it is unlikely that the enhanced plant drought tolerance is due to enhanced nitrogen fixation by SA190. First, analysis of the SA190 genome did not reveal any of the known nif genes (nifHDK) involved in nitrogen fixation nor the nodABC genes encoding the biosynthesis enzymes for the synthesis of nodulation factors (Lafi et al, ²⁰¹⁶). Moreover, we did not observe that SA190 could enhance the growth of Arabidopsis on N‐free or N‐limiting conditions. We, therefore, concluded that N‐fixation is not part of the mechanism of how SA190 induces drought tolerance in plants.

Tolerance to environmental stresses can be achieved via different modes of cellular and biochemical changes conferred by beneficial microbes. Among the physiological and molecular responses to drought conditions, the plant phytohormone ABA plays a central role in drought adaptation of plants. ABA levels rapidly rise upon water limitation, resulting in stomatal closure to reduce transpiration and save water at the cost of growth (Zhu, 2002). Interestingly, under drought stress proxy conditions, SA190‐colonized plants showed an improved leaf water status and transpiration rate as well as maintaining plant growth. These experiments were conducted under maximally controlled conditions using the genetic model plant Arabidopsis. To test whether SA190 might also be able to enhance drought tolerance to a crop plant, we performed drought stress experiments on the important fodder plant alfalfa. SA190‐colonized alfalfa showed significantly enhanced drought resilience compared with non‐colonized plants. These results support the concept that beneficial microbes such as SA190 might be a powerful tool to sustain agriculture production during drought events.

The analysis of the interaction between SA190 and Arabidopsis in the context of limited water availability revealed many of the known players of drought stress tolerance in plants. In particular, we found many genes involved in the ABA pathway, and our subsequent genetic analysis showed that ABA is essential for mediating SA190‐induced PEG drought stress proxy tolerance. Transcriptome and qRT‐PCR analysis showed that at least seven aquaporin genes are differentially regulated in SA190‐colonized plants under PEG conditions and this regulation by SA190 was found to be entirely dependent on the ABA pathway. Because aquaporins play a key role in adjusting water relations in plants and many are regulated by ABA (Jang et al, ²⁰⁰⁴), we used these genes as a proxy for the set of SA190‐regulated ABA‐dependent genes. Intriguingly, expression of all identified aquaporin genes was strongly suppressed under PEG drought stress proxy conditions in non‐colonized plants, but none of these aquaporin genes were found to have significantly altered expression under non‐stress conditions in SA190‐colonized or non‐colonized plants. These results suggested a potentially epigenetic priming mechanism. Since H3K4me3 enrichment is associated with transcriptionally active gene regions of drought‐responsive genes (van Dijk et al, ²⁰¹⁰), we investigated the set of differentially regulated aquaporin genes in the context of SA190 colonization. We found that SA190 primes aquaporin genes by enhancing H3K4me3 levels in the regions near the transcriptional start sites. Priming by SA190 occurs already in the absence of PEG drought stress conditions. Therefore, SA190 interaction functions via priming genes for enhancing their expression under water‐limiting conditions. The H3K4me3 mark is broadly distributed on many ABA inducible genes (van Dijk et al, ²⁰¹⁰) and mutants in the key enzyme of the ABA biosynthesis pathway NCED3 show enriched H3K4me3 levels under drought stress conditions (Ding et al, ²⁰¹¹). This modification is mediated by the histone methyl transferase ATX1 (Arabidopsis trithorax‐like 1), as transcript levels of several ABA and drought‐upregulated genes like RD29A and RD29B were reduced during drought treatment in the atx1 mutant (Ding et al, ^2012b). It will be interesting to see if ATX1 and/or other methyl transferases are also involved in mediating H3K4me3 priming of SA190‐targeted genes.

Aquaporins play a key role in transmembrane water transport in transpiration and environmental stress responses (Maurel et al, ²⁰¹⁵). In Arabidopsis, there exist 35 aquaporin genes (Johanson et al, ²⁰⁰¹) which can be broadly divided into PIPs and TIPs genes. Some of the Arabidopsis PIPs and TIPs proved to be active water channels in Xenopus oocytes (Quigley et al, ²⁰⁰²). The biological significance of aquaporins in plants is their ability to modulate transmembrane water transport in situations where adjustment of water flow is physiologically critical (Li et al, ²⁰¹⁴). PIPs play an important role in controlling the transcellular water transport and are subdivided into two subfamilies PIP1 and PIP2 (Maurel et al, ²⁰¹⁵). The overexpression of PIP isoforms increased the root osmotic hydraulic conductivity, transpiration and shoot‐to‐root ratio while the downregulation leads to drought stress susceptibility (Pawłowicz & Masajada, 2019). However, when expressed in heterologous systems, overexpression of aquaporins can also have negative effects on stress resistance, due to the fact that the native stress response machinery may recognize them as foreign proteins (Li et al, ²⁰¹⁴). Aquaporins are also regulated by post‐translational modification including phosphorylation, methylation, acetylation, glycosylation, and deamination. Although fine‐tuning aquaporin expression plays a key role in all functions related to the water status of plants, the multitude of genes and their potential redundancy makes genetic analyses a complicated matter. Redundancy of aquaporins in Arabidopsis might also be limiting our success in analyzing aquaporin mutants with respect to SA190‐induced tolerance to PEG drought stress proxy (Fig 5B). However, using aquaporin inhibitors (AgNO₃ and HgCl₂), although probably not highly specific, independently suggested that aquaporins might mediate part of SA190‐induced PEG stress tolerance (Fig 5D). These results require further research to unequivocally clarify the role of aquaporins in SA190‐induced drought tolerance.

Drought stress massively alters many physiological parameters in plants, with ABA playing one of the primary roles in abiotic stress tolerance (Zhu, 2002; Nakashima & Yamaguchi‐Shinozaki, ²⁰¹³). A cluster of ABA‐related genes was found in the transcriptome of SA190‐colonized plants. Our genetic analysis showed that the ABA pathway is essential for mediating SA190‐induced PEG drought stress proxy tolerance by priming of target genes. We conclude that SA190‐induced drought proxy tolerance is mediated by enhancing H3K4me3 levels of target genes in an ABA‐dependent manner (Fig 6). Priming by beneficial microbes bears an enormous potential in agricultural engineering which suffers from the problem that expressing stress‐related genes often comes with a cost in growth and yield. Given that desert plant‐associated beneficial microbes, such as Enterobacter sp. SA187 can enhance salt and heat tolerance in plants (de Zelicourt et al, ²⁰¹⁸; Shekhawat et al, ²⁰²¹), desert beneficial microbes, such as SA190, might present an affordable and readily available technology to ensure food production in arid countries with limited water availability. On the other hand, considering the effects of global climate change, desert‐beneficial microbes might also help to limit yield losses in less arid regions suffering from intermittent drought and heat stress periods.

Figure 6. Graphical model of SA190 mediated aquaporin priming.

SA190 constitutively deposits H3K4me3 histone marks on aquaporin promoters in an ABA‐dependent manner to mediate gene expression for enhancing water use efficiency and drought stress tolerance.

Materials and Methods

Plant material, seedling colonization, and stress assays

Seeds of Arabidopsis thaliana ecotype Col‐0 were surface sterilized for 10 min in 70% ethanol with 0.05% Triton X‐100, washed three times with 100% ethanol and dried in a laminar flow hood. Sterilized seeds were scattered on half strength MS (½MS) medium supplemented with either 100 μl of Luria Broth (LB) medium (Mock) or a fresh culture of SA190 grown in LB to an optical density of 0.21, at final concentration of 10⁸ cfu/ml. Seeds were stratified in the dark for 2 days at 4°C and then transferred in a growth chamber set to 22°C with a long‐day photoperiod (16 h light, 8 h dark) for 5 days. The germinated seedlings (∼1.0–1.5 cm root length) were then transferred to ½MS as a normal condition or to ½MS infiltrated with 25% PEG 8000 (Fisher Scientific, Belgium) to induce drought stress proxy. Six seedlings were used per square petri plate. The number of lateral roots was evaluated under a stereomicroscope, and the root length was measured using ImageJ on the ninth day of stress treatment. LRD was calculated by dividing the number of lateral roots by the primary root length. The FW of shoots and roots were taken after 16 days of stress. DW was measured after drying the shoot and root tissues for 2 days at 80°C. All assays were performed in three biological replicates (bacterial colonies) and two technical replicates (petri plates). The following mutant lines were used in this study: ABA mutants (aba2‐1, pyr1pyl4,5,8), ethylene mutant (ein2‐1, ein3‐1 and acs1‐1), and Aquaporin mutants tip2;1 (insertion line SM_3_1820) and pip1;1 pip1;2 pip1;3 pip1;5 (qpip, quadruple pip mutant combining pip1;1‐2 and pip1;5‐1 being CRISPR/Cas9‐induced deletion mutants, pip1;2‐2 (insertion line SALK 019794), and pip1;‐3 (insertion line SALK 051107)). Aquaporin inhibitors AgNO₃ (silver nitrate, Sigma) and HgCl₂ (mercury chloride, Sigma) were added to pre‐cooled ½MS agar medium together with 25% PEG.

Laser scanning confocal microscopy imaging

The SA190:GFP strain was generated as described by de Zelicourt et al (2018). Briefly, the GFP was introduced by the E. coli SM10λpir strain carrying the GFP donor plasmid, pUX‐BF13 and the pRK600 mobiliser plasmid. To determine the colonization of SA190:GFP on plant tissue the 5 day old seedlings, mounted in 100 μg/ml propidium iodide (PI), were used and visualized using confocal microscope ZEISS LSM880 with Airyscan with the Plan‐Apochromat 10× (n.a. 0.45) objective lense for small magnifications (100 nm) or the Plan‐Apochromat 63× (n.a. 1.4) oil‐immersive objective lense for higher magnifications (20 nm). For excitation of both GFP and PI, we used the argon‐based laser with power output 6.5 or 9% for 10× or 63× objective lense, respectively, and samples were excited at wavelength 493–584 nm for GFP and 604–718 nm for PI. The ZEN black edition 3.0 SR was used for assembly of images, and the ZEN blue lite edition 3.0 was used for image editing.

Bacterial colonization

Shoots and roots of 21‐day‐old mock‐ and SA190‐colonized shoots of Arabidopsis grown on ½ MS with or without 25% PEG were collected in Eppendorf tubes. FW was recorded using sensitive balance (METTLER TOLEDO). Samples were ground using Qiagen Tissue Lyser II Sample (Disruption −11,843) for 2 min with 500 μl of extraction buffer (10 mM MgCl₂+ 0.01% silwet 77), and then were incubated for 1 h at 28°C with shaking at 300 rpm (Eppendorf, ThermoMixer C). Samples were diluted 10‐fold, and then spread on LB agar and CFUs were counted after overnight incubation at 28°C. Calculated number of CFUs was normalized to plant FW.

Stomatal aperture measurement assay

To measure the stomatal parameters, 5‐day‐old seedlings inoculated with ±SA190 on ½MS were transferred and grown further for 16 days on ½MS ± 25% PEG plates. The leaves from same developmental stage were excised from the plants and a section from middle part of the lamina excluding the midrib was cut and mounted on a double‐sided tape, sticked to a microscopic glass slide, with the abaxial surface in contact with the tape. The upper green tissues were then scraped off by scalpel and the remaining lower epidermis was imaged using an Axio Imager Z2 microscope (Zeiss) equipped with DIC optics and EC Plan‐Neofluar. For measuring the stomatal aperture, the aperture width and total stomatal length of each stoma were manually measured using ImageJ and their ratios were analyzed statistically. A minimum of 30 stomata was measured from each leaf. The experiment was repeated thrice and samples were prepared from at least three leaves from each genotype per biological replicate.

Transpiration rates

To conduct the transpiration rate assays, the rosettes of 21‐day‐old mock‐ and SA190‐colonized shoots of Arabidopsis grown on ½ MS with or without 25% PEG were weighed at regular 5 min intervals. The water loss rates were calculated using the formula: WL (mg) = FWi (initial FW) − desiccated weight/min (Cohen et al, ²⁰¹⁵).

Leaf relative water content

The fourth leaf from the rosette of six 21‐day‐old non‐colonized or SA190‐colonized plants treated with 0 or 25% PEG was used to measure the RWC. After measuring the FW, leaves were placed in distilled water for 3 h in the dark and the turgid weight was recorded. The samples were oven‐dried at 70°C for 24 h and the LRWC was measured according to the formula: LRWC (%) = [(FW − DW)/(TW − DW)] × 100 (Sade et al, ²⁰¹⁵).

Water use efficiency

To determine the WUE for plants, sterilized 50 ml tubes were filled with a sterilize soil‐perlite mixture (w:w 1:1), and 35 ml of water was added to each tube. Single 5‐day‐old Arabidopsis seedling that had been cultured on plates with or without SA190 were transferred to the hole in the lid of individual tubes. The plants were treated as described in Wituszynska & Karpiński (2014). For the drought condition samples, the plants were cultivated for an extra 14 days.

Arabidopsis drought assays in soil

Five‐day‐old mock‐ and SA190‐colonized Arabidopsis seedlings were transferred to jiffy pots in four biological replicates of nine plants each. The plants were grown in Percival growth chambers and watered twice a week for 2 weeks, then watering was stopped for 3 weeks to induce drought conditions, before rewatering again for 7 days to allow the plants to recover. The plants were photographed at each step using Canon EOS 6D.

RNA extraction

Total RNA was extracted from the roots of 21‐day‐old mock‐ and SA190‐colonized Arabidopsis grown on ½ MS with or without 25% PEG with the Nucleospin RNA plant kit (Macherey‐Nagel) following the manufacturer's recommendations. The quality and quantity of the RNA was assessed using Nanodrop‐6000 spectrophotometer, 2100‐Bioanalyzer (RNA integrity number greater than 8.0), and QubitTM 2.0 Fluorometer with the RNA BR assay kit (Invitrogen).

RNAseq library construction and sequencing

To prepare the cDNA library, we used 1 μg of total RNA per sample. The ribosomal RNA was removed using a Ribo‐Zero Magnetic Kit with a 1:1 mixture of Ribo‐Zero Magnetic Kit (Bacteria) and Ribo‐Zero Magnetic Kit (Plant) following the manufacturer's recommendations. The quality of the library was assessed on the Agilent Bioanalyzer 2100 system. Sequencing was performed using Illumina HiSeq deep sequencing (Illumina HiSeq 2000, Illumina).

Transcriptome analysis

We performed transcriptome sequencing for each library of Arabidopsis to generate 101‐bp paired‐end reads on Illumina HiSeq4000 Genome Analyzer platform. Low‐quality reads were trimmed using the Trimmomatic version 0.32 (Bolger et al, ²⁰¹⁴; http://www.usadellab.org/cms/?page=trimmomatic) with the following parameters: Minimum length of 36 bp; Mean Phred quality score greater than 30; Leading and trailing bases removal with base quality below 3; Sliding window of 4:15. After pre‐processing the Illumina reads, the reads are mapped on to transcripts using TopHat (Trapnell et al, ²⁰⁰⁹; ver. 2.1.1; http://tophat.cbcb.umd.edu/) for aligning with the genome. For TopHat, the Reference‐Arabidopsis thaliana (TAIR10) genome (https://www.arabidopsis.org) was used as the reference sequences with maximum number of mismatches as 2. We used count‐based normalization implemented in DESEQ2 (Love et al, ²⁰¹⁴). To quantify the reads from the mapped alignment, featureCounts package was overlapped with each gene co‐ordinates (Liao et al, ²⁰¹³).

To identify the DEGs, the following parameters were used: P‐value of 0.05 with a statistical correction using Benjamini–Hochberg FDR of 0.05 in DESEQ2. A cut‐off of twofold up‐ or downregulation has been chosen to define differential expression. After processing the data, visualization of differential expression was done using cummerbund v2.14.0 (http://bioconductor.org/packages/release/bioc/html/cummeRbund.html).

Hierarchical clustering of the quantified genes was performed using MeV 4.9.0 version (TM4, https://sourceforge.net/projects/mev‐tm4/files/mev‐tm4/MeV%204.9.0/) utilizing the Pearson correlation method. Enriched GO term analysis was carried out using DAVID (Huang et al, ²⁰⁰⁹; Sherman et al, ²⁰²²).

qRT‐PCR analysis

Total RNA was extracted from 21‐day‐old mock‐ and SA190‐colonized Arabidopsis grown on ½ MS ± 25% PEG using NucleoSpin Plant RNA (Macherey Nagel) kit following the manufacturer's protocol. First strand cDNA was synthesized from 1 μg of total RNA using SuperScript III First‐Strand Synthesis SuperMix kit (Invitrogen). The diluted cDNA was used to perform quantitative RT‐PCR (qRT‐PCR) using SsoAdvanced Universal SYBR Green Supermix (Bio‐Rad). All reactions were amplified in a CFX96 Touch Real‐Time PCR Detection System (BIO‐RAD) at 50°C for 2 min, 95°C for 10 min, and 40 cycles of 95°C for 10 s and 60°C for 40 s, followed by a dissociation step to validate the PCR products. The data were analyzed using Bio‐Rad CFX manager software. Tubulin was used as a housekeeping gene for normalization of gene expression levels. Analyses were performed in triplicate and were repeated thrice with independent RNA samples. Primers used in this study are listed in Table EV1.

Chromatin immunoprecipitation qPCRs

We conducted chromatin immunoprecipitation (ChIP) as described in previous studies (Shekhawat et al, ²⁰²¹). Briefly, roughly 1 g of 21‐day‐old mock‐ and SA190‐colonized Arabidopsis grown on ½ MS with or without 25% PEG were cross‐linked by vacuum‐infiltrating 1% formaldehyde for 15 min and subsequent quenching by 2 M glycine. Nuclei were extracted from the frozen ground powder using NIB (0.4 M Sucrose, 10 mM Tris–HCl pH8, 10 mM MgCl₂, 5 mM ß‐mercaptoethaol and 1× Proteases Inhibitor cocktail). Nuclei were lysed in NLB (50 mM Tris–HCl pH8, 10 mM EDTA, 1% SDS and 1× Proteases Inhibitor cocktail). Chromatin was sonicated using a Diagenode Bioruptor (40 Hz, 14 cycles each with 30 s on/30 s off with ice cooling), yielding fragments with a size of around 250–350 bp. Antibodies (anti‐H3, ab1791; anti‐H3K4me3, from Abcam, ab8580 http://www.abcam.com) were incubated with protein A‐coated agarose beads (Invitrogen) for at least 2 h at 4°C in IP buffer (1.1% Triton X‐10, 1.2 mM EDTA, 16.7 mM Tris–HCl pH8, 167 mM NaCl), and 1× Proteases Inhibitor cocktail. Immunoprecipitations were done in IP buffer at 4°C for overnight. After washing with low (150 mM NaCl) and high salt (500 mM NaCl) IP buffers and reverse cross‐linking, resulted DNA was extracted using the phenol–chloroform method and precipitated with ice chilled ethanol and glycogen (Invitrogen), and then re‐suspended in 20 μl of water. ChIP‐PCR was performed for three regions of indicated gene loci. Amplification values were normalized to H3 (normalized signal modification/normalized signal H3). The given values in graphs are the means of two biological replicates, with each replicate was normalized to the respective Col‐0 with no treatment (mock) sample before averaging.

Alfalfa drought assays in soil

Control growth chamber experiments were conducted at the KAUST core lab facility. The experiments were performed in a randomized pots (16 × 12 cm) arrangement of six biological replicates of 250–300 plants each. The Pots were irrigated using tap water 200 ml/week as full irrigation for 4 weeks after germination, for the drought experiments the irrigation was stopped for 10 days for the treated plants. Biomass yield was recorded after 52 days from sowing. The analysis of variance (one‐way ANOVA) Including Post Hoc Tukey HSD of six replicates was performed, (P < 0.05) was calculated to test the significance of differences between means. To inoculate alfalfa (Medicago sativa var. CUF 101) seeds were done according to (Daur et al, ²⁰¹⁸) with some modification. In brief, a slurry was prepared to consist of SA190, and sterilized sugar solution (0.2%). Subsequently, alfalfa seeds were coated with the slurry at a rate of 10E8 cells per g seed. As a control, seeds were coated with a similar mixture without bacteria. The plants were sawed on stander soil and incubated in a growth chamber with temperatures 22–25°C, 16 h – 8 h light–dark, photoperiod 200–300 μmol/m²/s for 52 days until the harvest.

Author contributions

Khairiah M Alwutayd: Formal analysis; investigation. Anamika A Rawat: Formal analysis; validation; investigation. Arsheed H Sheikh: Formal analysis; validation; investigation. Marilia Almeida‐Trapp: Formal analysis; investigation. Alaguraj Veluchamy: Software; formal analysis. Rewaa Jalal: Formal analysis; investigation. Michael Karampelias: Formal analysis; investigation. Katja Froehlich: Formal analysis; investigation. Waad Alzaed: Formal analysis; investigation. Naheed Tabassum: Formal analysis; investigation. Thayssa Rabelo Schley: Formal analysis; investigation. Anton R Schäffner: Resources; supervision. Ihsanullah Daur: Formal analysis; investigation. Maged M Saad: Formal analysis; supervision; investigation. Heribert Hirt: Conceptualization; writing – original draft; project administration; writing – review and editing.

Disclosure and competing interests statement

The authors declare that they have no conflict of interest.

Supporting information

Expanded View Figures PDF

Table EV1

PDF+

Source Data for Figure 1

Source Data for Figure 2

Source Data for Figure 3

Source Data for Figure 4

Source Data for Figure 5

EMBO reports (2023) 24: e56754

Data availability

The raw data of RNA‐sequencing have been deposited in NCBI with GSE Number GSE184355 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE184355). Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request.