Non-volatile signals and redox mechanisms are required for the responses of Arabidopsis roots to Pseudomonas oryzihabitans

Abstract

Soil bacteria promote plant growth and protect against environmental stresses, but the mechanisms involved remain poorly characterized, particularly when there is no direct contact between the roots and bacteria. Here, we explored the effects of Pseudomonas oryzihabitans PGP01 on the root system architecture (RSA) in Arabidopsis thaliana seedlings. Significant increases in lateral root (LR) density were observed when seedlings were grown in the presence of P. oryzihabitans, as well as an increased abundance of transcripts associated with altered nutrient transport and phytohormone responses. However, no bacterial transcripts were detected on the root samples by RNAseq analysis, demonstrating that the bacteria do not colonize the roots. Separating the agar containing bacteria from the seedlings prevented the bacteria-induced changes in RSA. Bacteria-induced changes in RSA were absent from mutants defective in ethylene response factor (ERF109), glutathione synthesis (pad2-1, cad2-1, and rax1-1) and in strigolactone synthesis (max3-9 and max4-1) or signalling (max2-3). However, the P. oryzihabitans-induced changes in RSA were similar in the low ascorbate mutants (vtc2-1and vtc2-2) to the wild-type controls. Taken together, these results demonstrate the importance of non-volatile signals and redox mechanisms in the root architecture regulation that occurs following long-distance perception of P. oryzihabitans.

Keywords: Pseudomonas oryzihabitans; Ascorbate; ethylene-responsive transcription factor 109; glutathione; plant growth-promoting rhizobacteria; reactive oxygen species; root system architecture.

Fig. 1.

Representative images of wild-type Arabidopsis seedlings that had been grown for 6 d in the absence of P. oryzihabitans and then for a further 7 d in either the absence (control) or the presence of bacteria (PGP01).

Fig. 2.

Differentially expressed transcripts in the roots of the wild type (A) and number of transcripts significantly increased and decreased (B). Seedlings had been grown for 6 d in the absence of P. oryzihabitans and then for a further 7 d in either the absence or presence of bacteria.

Fig. 3.

Gene Ontology (GO) analysis showing the biological processes involved in root responses to P. oryzihabitans.

Fig. 4.

Transcripts that were most increased (A) or decreased in abundance (B) in response to the presence of P. oryzihabitans PGP01. (A) AT5G45950 (GGL28, GDSL-motif esterase/acetyltransferase/lipase); AT3G22830 (HSFA6B, heat stress transcription factor A-6b), AT4G13420 (HAK5, potassium channel); AT2G39980 (HP, hypotetical unknown protein); AT5G28610 (DRS1-like, ATP-dependent RNA helicase); AT5G47220 (ERF2, ethylene response factor 2); AT1G71030 (MYBL2, myb family transcription factor); AT1G19530 (RGAT1, RGA target 1); AT4G28850 (XTH26, xyloglucan endotransglucosylase 26); AT2G42250 (CYP12A1, cytochrome P450); AT3G26740 (CCL, circadian control of mRNA stability); AT2G33830 (DRM2, dormancy/auxin associated protein 2); AT4G06746 (RAP2.9, ERF/AP2 transcription factor family); AT1G69490 (NAC-TF, transcription factor); AT5G02020 (SIS, salt-induced serine rich). (B) AT2G22590 (UGT91A1, UDP-glucosyltransferase 91A1); AT5G17040 (UGT78D4, UDP-glucosyltransferase 78D4); AT4G15480 (UGT84A1, UDP-glucosyltransferase 84A19); AT5G07990 (TT7, flavonoid 3ʹ hydroxylase activity); AT5G17220 (GSTF12, glutathione S-transferase 12); AT1G65060 (4CL, 4-coumarate:CoA ligase); AT3G22840 (ELIP1, early light inducible 1); AT2G23910 [NAD(P) binding, Rossmann-fold superfamily]; AT3G12900 (S8H, scopoletin 8 hydrolase); AT5G08640 (FLS1, flavonol synthase 1); AT1G80340 (GA3OX2, gibberellin 3-oxidase 2); AT1G30530 (UGT78D1, UDP-glucosyl transferase 78D1); AT4G17680 (EBS1, exclusivly sensitive to bicarbonate 1); AT5G62210 (ATS3, embryo-specific protein 3); AT3G51240 (F3H, flavanone 3-hydroxylase).

Fig. 5.

Subsets of transcripts involved in (A) phytohormone signalling, (B) hypoxia, and (C) nutrient status that were increased in abundance in the presence of P. oryzihabitans PGP01. (A) Responses to hormones: AT5G47220 (ERF2, ethylene-responsive transcription factor 2); AT2G33830 (DRM2, dormancy/auxin associated protein 2); AT4G06746 (RAP2.9, ethylene responsive RAP2.9); AT3G23150 (ETR2, ethylene response 2); AT5G61590 (ETR107, ethylene responsive transcription factor 107); AT3G59940 (KMD4, kiss me deadly 4, controls cytokinin signalling); AT1G21130 (IGMT4, indole glucosinolate-O-methyltransferase 4); AT4G39780 (ERF060, ethylene responsive factor 1); AT1G48690 (GH3-type, auxin responsive GH3-type protein); AT1G56220 (DRMH3, dormancy-associated protein homologue 3), AT4G30270 (SEN4, senescence 4, brassinosteroid response); AT1G43160 (RAP2.6, ethylene responsive factor RAP2.6). (B) Responses to hypoxia: AT1G19530 (RGAT1, RGA Target 1); AT3G23150 (ETR2; ethylene response 2); AT2G15890 (MEE14, maternal effect embryo arrest 14); AT1G33055 (HUP32, hypoxia response protein 32); AT5G65207 (HP, hypothetical protein responsive to hypoxia); AT3G10020 (HUP26, hypoxia response protein 26); AT1G10140 (UP, uncharacterized protein responsive to hypoxia); AT2G40000 (HSPRO2, orthologue sugar beet HSPRO2); AT4G38470 (STY46, serine/threonine kinase); AT4G27450 (HUP54, hypoxia response protein 54); AT1G26800 (MPSR1, misfolded protein sensing ring E3 ligase); AT5G66650 (CMCU, chloroplast-localized mitochondrial calcium uniporter); AT4G24230 (ACBP3, acyl-CoA-binding domain 3). (C) Transport facilitation and root growth: AT4G13420 (HAK5, potassium channel transporter 5); AT1G54970 (RHS7, root hair specific 7, ethylene regulated); AT4G36670 (PMT6, POLYOL/monosaccharide transporter 6); AT5G17860 (CCX4, cation/calcium exchanger); AT1G08430 (ALMT1, aluminium activated malate transporter); AT5G66650 (CMCU, chloroplast-localized mitochondrial calcium uniporter 3); AT2G47160 (BOR1, boron transporter 1); AT5G22410 (RHS18, root hair specific 18); AT4G38390 (RHS17, root hair specific 17); AT1G22710 (SUC2, sucrose protein symporter 2); AT2G32270 (ZIP3, zinc transporter 3 precursor).

Fig. 6.

Representative images of wild-type Arabidopsis seedlings, seedlings overexpressing ERF109 (ov32), and erf109 mutants. Seedlings had been grown for 6 d in the absence of P. oryzihabitans and then for a further 7 d in either the absence (control) or the presence of bacteria (PGP01).

Fig. 7.

The effect of the presence of P. oryzihabitans PGP01 on primary root length (A) and lateral root density (B) in wild-type A. thaliana, a transgenic line overexpressing redox-responsive transcription factor 1 (ov32), and a erf109 mutant line. Samples of bacterial inoculum was placed 5 cm away for the tips of the primary roots of 6-day-old seedlings that had been grown on agar plates. Root parameters were measured 7 d after inoculation. Data show the mean ±SE of three independent biological samples. Asterisks indicate significant differences according to t-test (P<0.05).

Fig. 8.

The effect of the presence of P. oryzihabitans PGP01 on primary root length (A) and lateral root density (B) in wild-type A. thaliana and mutants that are defective in ascorbate (vtc2-1 and vtc2-2). Samples of bacterial inoculum were placed 5 cm away for the tips of the primary roots of 6-day-old seedlings that had been grown on agar plates. Root parameters were measured 7 d after inoculation. Data show the mean ±SE of three independent biological samples. Asterisks indicate significant differences according to t-test (P<0.05).

Fig. 9.

The effect of the presence of P. oryzihabitans PGP01 on primary root length (A) and lateral root density (B) in wild-type A. thaliana and mutants that are defective in glutathione (cad2-1, pad2-1, and rax1-1). Samples of bacterial inoculum were placed 5 cm away for the tips of the primary roots of 6-day-old seedlings that had been grown on agar plates. Root parameters were measured 7 d after inoculation. Data show the mean ±SE of three independent biological samples. Asterisks indicate significant differences according to t-test (P<0.05).

Fig. 10.

The effect of the presence of P. oryzihabitans PGP01 on primary root length (A) and lateral root density (B) in wild-type A. thaliana and mutants that are defective in SL synthesis (max3-9 and max4-1) and signalling (max2-3). Samples of bacterial inoculum were placed 5 cm away for the tips of the primary roots of 6-day-old seedlings that had been grown on agar plates. Root parameters were measured 7 d after inoculation. Data show the mean ±SE of three independent biological samples. Asterisks indicate significant differences according to t-test (P<0.05).

Fig. 11.

The effect of removal of the agar between P. oryzihabitans PGP01 and Arabidopsis seedings. Arabidopsis seedlings were either separated by a 1 cm gap in the agar (Control), separated by a 1 cm gap from seedlings grown in the presence of P. oryzihabitans (Plants and PGP01), or separated from agar on which P. oryzihabitans was grown (Plants/PGP01). Seedlings were grown for 6 d in the absence of P. oryzihabitans and then for a further 7 d in either the absence or presence of bacteria. Primary root length (A) and lateral root density (B).