Arabidopsis thaliana Early Foliar Proteome Response to Root Exposure to the Rhizobacterium Pseudomonas simiae WCS417

Plant growth–promoting rhizobacteria (PGPR) can enhance plant development and defense through their antagonist actions against soil plant pathogens (Wang et al. 2021). Pseudomonas is a competitive bacterial genus in the rhizosphere (de Weert et al. 2002; Simons et al. 1996). In particular, Pseudomonas simiae WCS417 (previously known as Pseudomonas fluorescens WCS417) (Berendsen et al. 2015) is one of the most characterized PGPR for the activation of the induced systemic response (ISR) (Pieterse et al. 2020). The molecular basis of ISR has been thoroughly investigated in Arabidopsis thaliana roots colonized by WCS417 (Stringlis et al. 2018a; Zamioudis et al. 2014); ISR response partially overlaps with the iron (Fe)-deficiency response in A. thaliana (Romera et al. 2019). Volatile organic compounds (VOCs) and Fe-chelating siderophores produced by PGPR trigger plant Fe uptake pathways (Trapet et al. 2021; Verbon et al. 2019). Plants suffering from Fe deficiency may recruit more siderophore-producing bacteria than plants growing under normal nutritional conditions (Jin et al. 2006, 2010), and VOCs generated by ISR-inducing bacteria may relieve nutritional stress induced by low Fe levels (Zamioudis et al. 2015; Zhang et al. 2009). In addition, WCS417 promotes the expression of plant genes activated by low Fe levels, such as MYB72 (Palmer et al. 2013; Zamioudis et al. 2015).

A recent transcriptional analysis of roots and shoots of A. thaliana seedlings inoculated with WCS417 suggested that the beneficial effects of WCS417 on plant growth and development are achieved by modulation of sugar transport; the sucrose transporters SWEET11 and SWEET12 are, indeed, involved in the growth-promoting effects of WCS417 (Desrut et al. 2020).

Formate dehydrogenase (FDH) enzymes are found in bacteria, fungi, and plants (Alekseeva et al. 2011); plant FDH enzymes, localized in mitochondria (Choi et al. 2014; Herman et al. 2002) and chloroplasts (Lee et al. 2022; Olson et al. 2000), catalyze the oxidation of formate (HCOO⁻) to CO₂ with the reduction of NAD⁺ to NADH. FDH enzymes are referred to as stress enzymes, because their expression is upregulated in response to several abiotic stresses (Ambard-Bretteville et al. 2003; Andreadeli et al. 2009; David et al. 2010; Hourton-Cabassa et al. 1998; Kurt-Gür et al. 2018; Li et al. 2002; Lou et al. 2016; Murgia et al. 2020; Suzuki et al. 1998). Only a few studies have shown FDH involvement in the response to bacterial infections (Choi et al. 2014; David et al. 2010; Lee et al. 2022; Marzorati et al. 2021).

WCS417 represents a model for studying the interactions between plants and beneficial rhizobacteria (Pieterse et al. 2020). Still, little is known about the molecular changes occurring at the foliar level upon plant root exposure to WCS417, as most studies have focused on the rhizobacterial effects on roots several days after exposure (Trapet et al. 2016; Verbon et al. 2019; Wintermans et al. 2016; Zamioudis et al. 2013).

Given these premises, the goal of the present study is to explore the early foliar responses of A. thaliana to WCS417 and the involvement of FDH in such responses; such an approach could shed light not only on the plant defense strategies but also on the growth-promoting effects of WCS417 on the aerial parts of plants.

For that, the activity of the FDH promoter was first investigated in response to WCS417. The colonization of roots by WCS417 in seedlings of wild type (WT) Col and of an FDH knock-out mutant (atfdh1-5) was also investigated. Proteomic analysis of leaves of both WT Col and atfdh1-5 mutant, after exposure of their roots to the rhizobacterium, was then performed.

Such analysis shows that FDH levels increase after exposure to WCS417, thus confirming FDH involvement in the early A. thaliana foliar responses to WCS417, not only in terms of FDH promoter activity but also at the protein level. Moreover, proteomic analysis reveals that chloroplasts, in particular some proteins of photosystems I and II, as well as various members of the glutathione S-transferase (GST) family, are early targets of the adaptive plant response to WCS417. Finally, the comparison of WT Col and atfdh1-5 leaf proteomes reveals a different regulation of some stress-responsive genes between the two lines, particularly after WCS417 treatment, suggesting FDH involvement in the early stages of the interactions between A. thaliana and the rhizobacterium WCS417.

Results

P. simiae WCS417 rapidly induces FDH promoter activity in hydathodes

FDH is expressed in the leaves of A. thaliana, especially in hydathodes (Murgia et al. 2020), where it is involved in early defense responses against the pathogen Xanthomonas campestris pv. campestris. It recently emerged that X. campestris pv. campestris infection reduces FDH expression in WT leaves and that the spread of an X. campestris pv. campestris strain expressing β-glucuronidase (GUS) is more pronounced in the atfdh1-5 mutant (Marzorati et al. 2021). Consistent with these results, the atfdh1-5 mutant is more susceptible to the virulent strain X. campestris pv. campestris 8004ΔxopAC; the disease index (DI), scored 7 days after wound-inoculation with bacterial suspensions of X. campestris pv. campestris 8004ΔxopAC, is significantly higher in atfdh1-5 than in WT plants (Supplementary Fig. S1).

The effect of the PGPR WCS417 on FDH expression was hence explored. To this end, 7-day-old seedlings of A. thaliana Vu FDH::GUS (an A. thaliana transgenic line transformed with the construct of Vigna umbellata FDH promoter fused to the reporter GUS) (Murgia et al. 2020) were co-cultivated in vitro with WCS417 for up to 7 more days, avoiding, however, any direct contact between the seedlings and the rhizobacterium (or the mock solution) (Fig. 1A). After 2 or 7 days of co-cultivation, GUS staining of the seedlings revealed an increased FDH promoter activity in WCS417-treated ones, in both roots and shoots; the latter were stained in the vascular tissue and hydathodes of the rosette leaves (Fig. 1B and D) and were compared with mock-treated seedlings (Fig. 1C and D). Hence, beneficial rhizobacteria can induce FDH expression in A. thaliana grown in vitro. Interestingly, such an effect is opposite to the described effect of the pathogen X. campestris pv. campestris on FDH expression (Marzorati et al. 2021).

Download as PowerPoint

To better evaluate how rapidly WCS417 may affect FDH expression in plants grown in soil, the roots of 4-week-old A. thaliana Vu FDH::GUS plants grown in soil were exposed to WCS417 by direct inoculation of the bacterium into the soil, and their rosette leaves were stained for GUS activity. After 2 days, a higher number of stained hydathodes was observed in the rosette leaves of WCS417-treated plants than in those of mock-treated plants (Fig. 2A). To assess whether the treatment with WCS417 could have altered leaf physiology, the common parameters for photochemical efficiency, i.e., F₀ (initial), F_m (maximum), F_v (variable) fluorescence, and the maximum photochemical efficiency (F_v/F_m) were evaluated; these parameters are statistically similar in mock and P. simiae–treated samples (Supplementary Fig. S2A and B). The induction of FDH promoter activity in vitro, without contact between rhizobacteria and roots, suggests that WCS417 could affect FDH expression through the emission of volatile compounds. To investigate this possibility in vivo, plants were organized in rows in Aratrays to keep those positioned at the edges (named ‘close to WCS417’) fully isolated from those positioned at the tray center; indeed, an entire row of empty baskets was positioned between the two groups (Supplementary Fig. S3A). The roots of the plants positioned in the central rows were exposed to WCS417 by direct inoculation of WCS417 into the soil pots; the whole tray was then covered with a lid without holes, to avoid dispersal of volatile compounds, and plants were not watered during the following 4 days to avoid any cross-contamination with P. simiae (Supplementary Fig. S3B). The same plant arrangement and treatment were also performed for mock-treated plants and the ‘close to mock’ ones. Sampling of the leaves from all plants (WCS417-treated, mock-treated, close to WCS417–treated, close to mock) was then performed after 4 days, i.e., after 2 more days with respect to what was described in Figure 2A, to better assess the possible effects of volatile compounds. After 4 days, the induction of FDH promoter activity could be observed with a higher number of GUS-stained hydathodes in WCS417-treated plants than in mock-treated plants (Fig. 2B); the slightly higher number of stained hydathodes observed in the leaves of the plants positioned in the external lines (which did not receive WCS417 themselves) is however not statistically different from what was observed in their mock counterpart (Fig. 2B).

Download as PowerPoint

Colonization of atfdh1-5 roots by WCS417 is more pronounced than that of WT roots

The well-established morphological root responses to WCS417, i.e., inhibition of root elongation, promotion of lateral root formation, and root-hair development (Zamioudis et al. 2013), were observed in 7-day-old WT seedlings, co-cultivated in vitro with WCS417 for 7 more days, without any initial direct contact between the seedlings and the rhizobacterium (Supplementary Figs. S4, S5, and S6). The atfdh1-5 mutant, in the same experimental condition, also showed enhanced root-hair development with respect to its mock counterpart (Supplementary Figs. S4, S7, and S8), whereas the inhibition of root elongation and the promotion of lateral root formation were less obvious in such mutants (Supplementary Figs. S4, S7, and S8), likely due to its short roots phenotype, as previously described (Murgia et al. 2020). Nonetheless, roots of atfdh1-5 seedlings grown on Murashige and Skoog medium are colonized by WCS417 even better than WT roots (Fig. 3).

Download as PowerPoint

The chloroplasts, GSTs, and stress-responsive proteins are early targets in the metabolic changes induced by WCS417

To investigate the effect of WCS417 on aerial parts of plants, seedlings of both WT Col and atfdh1-5 were grown in either control or alkaline soil (pH 7.6) (Supplementary Fig. S9); WCS417 was then inoculated into the soil in the proximity of the roots, and chlorophyll content and fresh weight (FW) of single rosettes were evaluated after 8 days. As expected, the FW of both lines in the control soil was higher than at pH 7.6 (Supplementary Fig. S10A). The chlorophyll content, for each line, was similar in both growth conditions (Supplementary Fig. S10D); this lack of difference is partly because very small plants, i.e., those more affected by growth in alkaline soil (Supplementary Fig. S9), impact chlorophyll values less than healthier plants.

Notably, WCS417 treatment reduced the differences of FW in WT plants grown in control soil, suggesting that WCS417 has a genuinely positive effect on growth, at least in control soil (Supplementary Fig. S10B); this WCS417 growth-promoting effect is not observed in the atfdh1-5 mutant (Supplementary Fig. S10C) nor on the chlorophyll content of both lines (Supplementary Fig. S10E and F). The observed early changes in FDH expression suggest that WCS417 can also affect the metabolic and signaling pathways in A. thaliana. To uncover early rearrangements of these pathways and the specific role of FDH in these WCS417-induced networks, the roots of 4-week-old WT Col plants and atfdh1-5 mutant were exposed to WCS417; after 2 days, rosette leaves were sampled for proteomic analysis. WCS417 treatment slightly decreases the weight of the rosettes in the WT but not in the atfdh1-5, which increased in weight (Supplementary Fig. S11).

Proteomic analysis was then performed on total proteins extracted from leaves. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis from untreated (WT Col mock, atfdh1-5 mock) or exposed to WCS417 (WT Col WCS417, atfdh1-5 WCS417) samples allowed the identification of a total of 2,196 distinct proteins (Supplementary Table S1). About 16% of the total proteins had an average peptide spectrum match (PSM) higher than 1 (Fig. 4A, blue and red dots). Globally, for each condition and genotype, about 1,000 proteins were identified, half of which were shared in pairwise comparisons (Fig. 4B); 918 and 912 proteins were detected in mock-treated WT Col and atfdh1-5 plants, respectively, whereas 998 and 1,066 proteins were detected in WT and atfdh1-5 plants exposed to WCS417, respectively. A label-free semiquantitative comparison among the characterized protein profiles (WT Col mock vs. WT Col WCS417, atfdh1-5 mock vs. atfdh1-5 WCS417, WT Col mock vs. atfdh1-5 mock, and WT Col WCS417 vs. atfdh1-5 WCS417) allowed the extraction of a total of 362 differentially expressed proteins (DEPs) (Supplementary Table S2). Major differences emerge between WT Col mock and WT Col WCS417 (150 DEPs, P ≤ 0.05; 63 DEPs, P ≤ 0.01) (Fig. 5A; Supplementary Table S3) and between atfdh1-5 mock and atfdh1-5 WCS417 (268 DEPs, P ≤ 0.05; 161 DEPSs, P ≤ 0.01) (Fig. 5B; Supplementary Table S4).

Download as PowerPoint

Download as PowerPoint

The functional evaluation of the characterized proteomes reveals a major enrichment of metabolic processes (amino acid, carbon, nitrogen metabolism, and protein synthesis) in both genotypes exposed to WCS417 (Supplementary Table S5), which is more pronounced in the atfdh1-5 treated with WCS417. In this scenario, the presence of WCS417 correlates with the enrichment of other interesting pathways, including photosynthesis, stress response, immune response, and transcription and translation. Notably, FDH increases in WCS417-treated samples of WT Col (Supplementary Tables S1 and S2), thus confirming the data on FDH promoter activity at the protein level; as expected, FDH was absent in the foliar proteomes of the atfdh1-5 mutant (Supplementary Tables S1 and S2) regardless of treatment.

Among the identified WCS417-upregulated proteins, 10 are shared between WT Col and atfdh1-5 (Table 1); among the WCS417-downregulated proteins, 19 are shared between WT Col and atfdh1-5 (Table 2). Therefore, these 29 WCS417-regulated proteins shared between the WT and the mutant line are early targets of the changes mediated by WCS417 treatment in an FDH-independent manner. The group of upregulated proteins (Table 1) is composed of two glutathione transferases (GSTF7 and GSTF8), two subunits of vacuolar-type H⁺-ATPase (VHA-B1 and VHA-C), three proteins with housekeeping functions (translation initiation factor 4A1 EIF4A1 and ribosomal proteins RPS1 and RPS18C), and three plastidial proteins (the ATP synthase subunit beta atpB, a ribose-5-phosphate isomerase RPI3, and a lipid-associated protein PAP6, also known as fibrillin 4). GSTs of the Phi type, formerly known as type I, are involved in the response to abiotic and biotic stresses (Sylvestre-Gonon et al. 2019); the expression of GSTF7 and GSTF8 is modulated by salicylic acid (SA) (Sappl et al. 2004). V-type H⁺-ATPase is formed by various subunits with complex regulation and is involved in stress adaptation (Dietz et al. 2001; Li et al. 2022); VHA-B1 is involved in the modeling of the actin cytoskeleton (Ma et al. 2012). Among the four A. thaliana RPI isoforms, RPI1 is involved in actin organization (Huang et al. 2020); however, to date, no physiological functions have been assigned to RPI3. PAP6 is involved in the resistance to biotic and abiotic stresses (Singh et al. 2010).

Table 1. Proteins with increased expression in Arabidopsis thaliana wild-type (WT) Col and atfdh1-5 leaves after root exposure to WCS417^a

					F ratio		Probability > F		DAve
Protein	AGI code	UniProt	Annotation	Localization	WT Col	atfdh1-5	WT Col	atfdh1-5	WT Col	atfdh1-5
EIF4A1	AT3G13920	A8MRZ7	Translation initiation factor 4A1	Nucleus/cytosol	11	19	8.0E−03	1.3E−03	−0.41	−0.79
GSTF7	AT1G02920	Q9SRY5	GSTF7	Mitoch/cytosol	11	15	7.6E−03	3.3E−03	−2.00	−2.00
GSTF8	AT2G47730	Q96266	GSTF8	Plastid/plastid	16	14	2.6E−03	3.7E−03	−0.96	−1.09
PAP6	AT3G23400	Q9LW57	Plastid-lipid-associated protein 6	Plastid/plastid	16	51	2.5E−03	3.1E−05	−1.18	−1.60
RPI3	AT3G04790	Q9S726	Putative ribose-5-phosphate isomerase 3	Plastid/plastid	15	15	3.1E−03	3.1E−03	−0.89	−0.91
RPS1	AT5G30510	Q93VC7	30S ribosomal protein S1	Plastid/plastid	14	19	4.1E−03	1.4E−03	−1.16	−1.18
RPS18C	AT4G09800	P34788	40S ribosomal protein S18	Mitoch/cytosol	52	29	2.9E−05	2.9E−04	−1.20	−0.53
VHA-B1	AT1G76030	P11574	V-type proton ATPase subunit B1	SP/Golgi or vacuole	22	11	8.8E−04	7.1E−03	−2.00	−0.95
VHA-C	AT1G12840	Q9SDS7	V-type proton ATPase subunit C	Mitoch/Golgi or vacuole	35	43	1.4E−04	6.7E−05	−1.46	−1.72
atpB	ATCG00480	P19366	ATP synthase subunit beta	−/Plastid	18	13	1.8E−03	4.9E−03	−0.28	−0.20

^a The list of proteins identified by proteomic analysis that are upregulated in the leaves of both WT Col and atfdh1-5 after 2 days of root exposure to WCS417 (or to mock treatment), showing the respective highest scores for subcellular localization according to Aramemnon (http://aramemnon.uni-koeln.de) and SUBA5 (https://suba.live), F ratio and probability (P ≤ 0.01) according to linear discriminant analysis), and DAve index according to MAProMa. AGI = Arabidopsis Genome Initiative; GST = glutathione S-transferase; Mitoch = mitochondria; SP = secretory pathway.

View as image HTML

Table 2. Proteins with decreased expression in Arabidopsis thaliana wild-type (WT) Col and atfdh1-5 leaves after root exposure to WCS417^a

					F ratio		Probability > F		DAve
Protein	AGI code	UniProt	Annotation	Localization	WT Col	atfdh1-5	WT Col	atfdh1-5	WT Col	atfdh1-5
CAM7	AT3G43810	A0A1I9LPJ2	Calmodulin 7	SP/cytosol	13	26	5.3E−03	4.8E−04	1.03	0.91
CP29B	AT2G37220	Q9ZUU4	RNA-binding protein CP29B	Plastid/plastid	14	36	4.1E−03	1.3E−04	0.37	0.49
CPN10-2	AT2G44650	O80504	10-kDa Chaperonin 2	Plastid/plastid	13	19	4.6E−03	1.5E−03	1.49	1.63
CPN20	AT5G20720	O65282	20-kDa Chaperonin	Plastid/plastid	12	61	5.4E−03	1.5E−05	0.36	0.76
CSP2	AT4G38680	Q41188	Cold shock protein 2	Nucleus/nucleus	12	13	6.0E−03	4.7E−03	0.73	0.63
ERD10	AT1G20450	P42759	Early responsive to dehydration protein 10	Nucleus/cytosol or PM	34	55	1.7E−04	2.4E−05	0.98	0.95
MEE59	AT4G37300	O23157	Maternal effect embryo arrest 59	Nucleus/nucleus	25	61	5.4E−04	1.5E−05	1.44	2.0
F13I12.120	AT3G47070	Q9SD66	Thylakoid soluble phosphoprotein	SP/plastid	11	25	7.2E−03	5.7E−04	0.96	1.42
PSAE1	AT4G28750	Q9S831	PSI subunit E1	Plastid/plastid	56	262	2.2E−05	1.7E−08	0.62	0.70
PSAE2	AT2G20260	Q9S714	PSI subunit E2	Plastid/plastid	14	68	3.7E−03	9.1E−06	0.60	0.78
PSB27-1	AT1G03600	Q9LR64	PSII repair protein PSB27-H1	Plastid/plastid	32	32	2.1E−04	2.2E−04	0.63	0.94
PSBO2	AT3G50820	Q9S841	Oxygen-evolving enhancer protein 1-2	Plastid/plastid	12	155	5.4E−03	2.0E−07	0.31	0.46
PSBP1	AT1G06680	Q42029	Oxygen-evolving enhancer protein 2-1	Plastid/plastid	32	28	2.0E−04	3.3E−04	0.55	0.78
RGGA	AT4G16830	O23523	RGG repeats nuclear RNA binding protein A	Nucleus/nucleus	12	94	6.7E−03	2.2E−06	1.11	1.89
RGGC	AT5G47210	Q9LVT8	Putative RGG repeats nuclear RNA binding protein C	Plastid/cytosol	10	30	9.7E−03	2.7E−04	0.92	1.61
RRF	AT3G63190	Q9M1×0	Ribosome-recycling factor	Plastid/plastid	14	53	3.6E−03	2.7E−05	1.23	1.22
At2g21530	AT2G21530	Q8GWP4	Unknown protein	Plastid/plastid	33	13	1.8E−04	4.8E−03	1.44	2.0
At5g02570	AT5G02570	Q9LZ45	Histone H2B.9	Nucleus/nucleus	16	22	2.4E−03	8.9E−04	2.0	1.29
At5g24165	AT5G24165	Q8LDQ8	Unknown protein	Plastid/mitoch	12	14	6.7E−03	4.0E−03	1.17	1.15

^a The list of proteins identified by proteomic analysis that are downregulated in the leaves of both WT Col and atfdh1-5 after 2 days of root exposure to WCS417 (or to mock treatment, showing the respective highest scores for subcellular localization according to Aramemnon (http://aramemnon.uni-koeln.de) and SUBA5 (https://suba.live), F ratio and probability (P ≤ 0.01) according to linear discriminant analysis), and DAve index according to MAProMa. AGI = Arabidopsis Genome Initiative; GST = glutathione S-transferase; Mitoch = mitochondria; SP = secretory pathway; PM = plasma membrane.

View as image HTML

Among the WCS417-downregulated proteins (Table 2), several are localized in plastids and, in particular, are part of the photosynthetic electron transport chain. PSAE1 and PSAE2 are subunits IV A and B of photosystem I; PSBP1 is an oxygen-evolving enhancer protein required for photosystem II organization (Yi et al. 2007); PBS27-1 is a repair protein involved in photosystem II assembly (Cormann et al. 2016); PSBO2 is the oxygen-evolving enhancer protein 1-2, required for the regulation of the D1 reaction center of photosystem II (Lundin et al. 2007); CP29 is a minor monomeric component of the PSII light-harvesting complex that, when phosphorylated, contributes to PSII state transition and disassembly (Chen et al. 2013). Among the WCS417-downregulated proteins that are localized in the plastid, there are also the chaperonins CPN10-2 and CPN20, the ribosome recycling factor RRF required for chloroplast biogenesis (Wang et al. 2010), the RNA-binding protein RGGC characterized by the arginine-glycine-glycine (RGG) region, the thylakoid soluble phosphoprotein F13I12.120, and one unknown protein encoded by the At2g21530 gene (Table 2). A few WCS417-downregulated proteins are localized in the nucleus, i.e., the RNA-binding protein RGGA involved in the response to salt and drought stresses (Ambrosone et al. 2015), the negative regulator of cold acclimation cold-shock protein 2 CSP2 (Sasaki et al. 2013), the core component of the nucleosome histone H2B.9 (At5g02570) and an essential embryogenesis protein MEE59 (Pagnussat et al. 2005). Last, three more proteins were also identified as WCS417-downregulated proteins, the stress-responsive ERD10, which belongs to the dehydrin family and is expressed in particular under different abiotic stresses (Sun et al. 2021), calmodulin 7 CAM7 (Kushwaha et al. 2008), and a protein of unknown function encoded by the At5g24165 gene (Table 2).

The comparison of mock-treated WT Col and atfdh1-5 proteomes identified 17 DEPs (P ≤ 0.01) (61 DEPs, P ≤ 0.05) (Fig. 5C; Supplementary Table S6A), whereas the comparison of WCS417-treated WT Col and atfdh1-5 proteomes identified 30 DEPs (P ≤ 0.01) (84 DEPs P ≤ 0.05) (Fig. 5D; Supplementary Table S6B). These results are consistent with the correlation scores of the spectral counts compared in pairs. Comparisons of the mock-treated (WT Col mock and atfdh1-5 mock) and the WCS417-treated genotypes (WT Col WCS417 and atfdh1-5 WCS417) have higher correlation values (r approximately 0.8) than the other comparisons, which show lower correlation values (Fig. 5E).

Notably, seven enzymes involved in reactive oxygen species (ROS) detoxification are differentially expressed in one or both WT Col and atfdh1-5 under mock or WCS417 treatment (Table 3), i.e., six glutathione transferases (GSTF2, GSTF7, GSTF8, GSTF9, GSTF1, GSTU19) and ascorbate peroxidase 1 (APX1), which scavenges cytosolic H₂O₂ (Hong et al. 2022). In particular, APX1 is the only ROS detoxification protein that is upregulated in atfdh1-5 leaves under both conditions (mock and WCS417-treated) with respect to WT Col counterparts (Fig. 5C and D; Supplementary Table S6A and B). Interestingly, leaves of WT Col and atfdh1-5 stained with 3,3′-diaminobenzidine (DAB) to detect any changes in the levels of hydrogen peroxide (H₂O₂) after WCS417 exposure, appear less brown-colored than their mock counterparts, suggesting that the rhizobacterium reduces the level of ROS (Supplementary Fig. S12).

Table 3. Proteins with antioxidant function, which are differentially expressed in WT Col and/or atfdh1-5 leaves after two days root exposure to WCS417^a

					Average PSMs mock		Average PSMs WCS417
Protein name	AGI code	UniProt ID	Annotation	Localization	WT Col	atfdh1-5	WT Col	atfdh1-5
APX1	AT1G07890	Q05431	L-ascorbate peroxidase 1	Mitoch	3.7	8.5	6.2	12.5*
GSTF2	AT4G02520	P46422	GSTF2	Mitoch	7.3	21.5	22.4*	19.5
GSTF7	AT1G02920	Q9SRY5	GSTF7	Mitoch/cytosol	/	/	1.8*	2.8*
GSTF8	AT2G47730	Q96266	GSTF8	Plastid/plastid	2.0	2.6	5.7*	8.9*
GSTF9	AT2G30860	O80852	GSTF9	Mitoch	2.5	9.6	7.6*	7.9
GSTF10	AT2G30870	P42761	GSTF10	Mitoch	0.9	2.2	2.6	5.1*
GSTU19	AT1G78380	Q9ZRW8	GSTU19	SP	0.3	0.2	0.4	4.5*

^a AGI = Arabidopsis Genome Initiative; GST = glutathione S-transferase; subcellular localization according to Aramemnon (http://aramemnon.uni-koeln.de) and to SUBA5 (https://suba.live/), according to their respective highest scores; PSM = peptide spectrum matches; Mitoch = mitochondrial; SP = secretory pathway. Asterisks (*) indicate statistically significant differences between mock and WCS417-treated values within the same genotype; A diagonal slash mark (/) indicates that the given protein has not been detected in the proteome analysis under that experimental condition. SP: secretory pathway.

View as image HTML

Various other proteins involved in resistance to oxidative stress were also upregulated in atfdh1-5. Pathogenesis-related protein 5 (PR5) (At1g75040), involved in the activation of the SA signaling pathway (Ali et al. 2018), is upregulated in mock-treated atfdh1-5 compared with its WT Col counterpart (Fig. 5C; Supplementary Table S6A), whereas PER34, GLO2, and ACO3 are upregulated in WCS417-treated atfdh1-5 compared with their WT Col counterparts (Fig. 5D; Supplementary Table S6B). Vice versa, the lipoxygenase LOX2 required for jasmonic acid (JA) biosynthesis in leaves (Yang et al. 2020) is downregulated in WCS417-treated atfdh1-5 leaves compared with the WT (Fig. 5D; Supplementary Table S6B).

Discussion

FDH is a nutritional hub for Fe and molybdenum (Di Silvestre et al. 2021; Vigani et al. 2017), and it also takes part in the plant response against the pathogen X. campestris pv campestris, especially in hydathodes (Marzorati et al. 2021). This latter evidence prompted us to investigate the possible involvement of FDH in the plant response to the beneficial PGPR Pseudomonas simiae WCS417. In the present work, we demonstrate that the FDH promoter is activated in both roots and shoots of seedlings exposed to WCS417 and that this activation is quite rapid, as it could be detected in the hydathodes of the rosette leaves just 2 days after root exposure to WCS417. We also demonstrate that, at least when seedlings are grown in vitro, the observed effects may be mediated by rhizobacterium-produced volatile compounds, consistent with previous studies (Wintermans et al. 2016; Zamioudis et al. 2013).

Interestingly, FDH is likely involved in the regulation of the extent of the root colonization by WCS417, as the WCS417 colonization index is, in atfdh1-5 roots, slightly higher than in WT ones.

Our proteome analysis unveils that WCS417 not only affects the production of plant proteins involved in essential metabolic processes but that the plastids, and in particular several photosynthesis-related proteins, are early targets of WCS417, regardless of the genetic background of the plant. These findings are particularly intriguing, as the link between plant-microbial pathogen interactions and chloroplasts has been already described (Littlejohn et al. 2020; Yang et al. 2021), whereas the link between chloroplasts and plant-beneficial bacteria was unexplored so far. Our results suggest that WCS417-induces rearrangements of PSI and PSII composition, caused by reduced accumulation of their proteins PSBO2, PSBP1, PSB27-1, PSAE1, and PSAE2. The physiological relevance of the WCS417-induced modulation of composition, stability, and turnover of photosystems should be the object of future investigations, in light of the role of chloroplasts in the biosynthesis of phytohormones such as JA (Wasternack and Hause 2019) and of previous findings suggesting that WCS417 stimulates the expression of genes important for plant growth (Wintermans et al. 2016; Zamioudis et al. 2013).

Notably, an increase in the enzymes responsible for ROS detoxification, such as GSTs, could be observed in both WT and atfdh1-5 lines upon WCS417 exposure (Gullner et al. 2018). GST gene induction or increased GST activity has been reported in plants that interact with beneficial bacteria (Kandasamy et al. 2009). GST7 and GST8 are upregulated in both plant lines after WCS417 exposure; one possible explanation is that WCS417 triggers a temporary antioxidant response, as supported by the DAB staining of the leaves exposed to WCS417, which appear to accumulate less H₂O₂ than mock controls.

The lipoxygenase LOX2, responsible for JA synthesis (Yang et al. 2020), is one of the WCS417-upregulated proteins in WT leaves. JA is involved in plant development and, along with other lipoxygenases, is important during defense responses against biotic stress (Singh et al. 2022). Moreover, LOX2 possesses a versatile enzymatic function, as it is also essential for the biosynthesis of a group of C6 aliphatic aldehydes, alcohols, and esters, known as green leaf volatiles and involved in plant defense (Mochizuki et al. 2016).

GSTF2, GSTF7, and LOX2 proteins were upregulated and PSAE1 was downregulated in shoots of A. thaliana exposed to Fe-deficient growth conditions (Zargar et al. 2013). This suggests that such proteins, also identified in the present proteomic work and with a regulation similar to that described by Zargar et al. (2013), might represent nodes of convergence between Fe-deficiency and WCS417-induced responses in the aerial parts of the plants.

We also noticed differences in the response triggered by WCS417 exposure in WT and atfdh1-5. The antioxidant enzyme APX1 is upregulated in the atfdh1-5 mutant in both experimental conditions (mock or P. simiae–treated) with respect to the WT. This, in turn, might imply that atfdh1-5 is in an ‘alert state’ with respect to the stress response. Both ROS and antioxidants are linked to SA signaling (Saleem et al. 2021) and the accumulation of SA and the expression of PR proteins are linked to the defense response systemic acquired resistance (SAR) (Vallad and Goodman 2004; Vlot et al. 2021). PR5 is considered a marker for SAR triggering (Sharon et al. 2011) and, surprisingly, we discovered that this protein is increased in atfdh1-5 leaves in the mock condition. In the WCS417-treated atfdh1-5 leaf proteome, several dehydrins were downregulated (HIRD11, COR15B, COR47, ERD14, and ERD10). In fact, the levels of several members of this protein family increased when A. thaliana was exposed to beneficial microorganisms that colonized its roots for defense (Baek et al. 2020; Kovacs et al. 2008; Liu et al. 2020). Last, LOX2 is not upregulated in atfdh1-5 WCS417-treated leaves. Taken together, these results suggest that the lack of FDH function in the atfdh1-5 mutant alters systemic defense mechanisms in leaves, particularly after WCS417 treatment. FDH protein levels indeed increase in WT leaves exposed to WCS417, corroborating our results on FDH promoter activity induction in A. thaliana leaves exposed to the rhizobacterium as well as FDH involvement in an early leaf defense response against pathogens (Marzorati et al. 2021). Overall, our findings on the atfdh1-5 leaf proteome suggest that FDH may have a relevant role in the early WCS417-induced responses.

In conclusion, the results presented in this work can stimulate further investigations for a better understanding of the signaling pathways triggered, at the foliar level, by growth-promoting rhizobacteria that enhance plant resistance to environmental challenges, such as nutritional stress and pathogen infections.

Materials and Methods

Plant growth

Arabidopsis thaliana WT Col, atfdh1-5 mutant (Choi et al. 2014; Murgia et al. 2020), and Vu FDH::GUS (Lou et al. 2016; Murgia et al. 2020) were stratified at 4°C and were grown on Technic n.1 DueEmme soil by using the Arasystem (Betatech BVBA), i.e., the Aratrays and the Arabaskets, in a greenhouse at 23°C and 150 μE m⁻²s⁻¹, with a 12-h-light and 12-h-dark photoperiod. Vu FDH::GUS seeds were surface-sterilized as described (Van Wees et al. 2013), were maintained in the dark for 3 days at 4°C, were then transferred on square dishes (100 × 100 × 20 mm) (Sarstedt, Australia Ltd.) containing one-half Murashige and Skoog medium supplemented with 1% sucrose, and were maintained vertically in a plant growth chamber at 22 to 25°C, with a 16-h-light and 8- h-dark photoperiod.

Bacterial growth

The X. campestris pv. campestris 8004ΔxopAC (Guy et al. 2013) was grown on MOKA-rich medium (Blanvillain et al. 2007) at 28°C. Rifampicin was used at 50 µg ml⁻¹. The Pseudomonas simiae WCS417 bacterial strain was grown overnight at 28°C on King's B medium agar supplemented with 50 μg ml⁻¹ rifampicin, suspended in 10 ml of 10 mM MgSO₄ and centrifuged for 5 min at 3,200 × g; the pellet was washed twice in 10 mM MgSO₄, with a 5-min centrifugation at 3,200 × g (Wintermans et al. 2016). The cell density was adjusted to 2 × 10⁶ or 2 × 10⁸ CFU ml⁻¹ in 10 mM MgSO₄.

Pathogenicity assays

A. thaliana plants were grown in short days (8-h of light and 16 h of dark) at 22°C (60% relative humidity, 125 µE m⁻²s⁻¹) for 4 weeks. Inoculations were performed as previously described (Meyer et al. 2005). Fully expanded leaves were wound-inoculated by piercing the central vein (from the middle to the tip of the leaf) three times with a needle dipped in a bacterial suspension at 10⁸ CFU ml⁻¹ in 1 mM MgCl₂. Disease development was scored using the following disease index: 0 = no symptoms, 1 = chlorosis at the inoculation point, 2 = extended chlorosis, 3 = necrosis, and 4 = leaf death.

Plant exposure to WCS417

For plants grown in soil, 1 ml of 10⁸ CFU ml⁻¹ bacterial suspension or 1 ml of 10 mM MgSO₄ (mock condition) was pipetted into each Arabasket containing single 4-week-old plants, with equal distribution of the liquid around the plant roots. To optimize an even distribution of the bacterial inoculum for each single-root apparatus, plants were not watered for 2 days before treatment, so that the Aratrays remained dry before and after inoculation. The trays, closed with transparent lids without holes, were then maintained at 25°C. For seedlings grown in vitro, 7-day-old seedlings grown on Murashige and Skoog were exposed to WCS417, avoiding any direct contact between seedlings and bacteria, as previously described (Wintermans et al. 2016). Briefly, 240 μl of 2 × 10⁶ CFU ml⁻¹ WCS417 suspension (or 240 μl of 10 mM MgSO₄ for mock treatment) was pipetted onto the Murashige and Skoog medium approximately 5 cm below the seedling roots. The plates were briefly dried under laminar flow, were closed with a lid and two layers of parafilm, and were placed again vertically in a growth chamber for 2 or 7 days.

WCS417 root colonization assay

The assay was performed according to Stringlis et al. (2018b). A. thaliana WT Col and atfdh1-5 seeds were soaked in 0.1% Tween 20 for 30 min, were sterilized for 90 s in 50% commercial bleach, and were then rinsed thoroughly six times with sterile distilled water. WT Col and atfdh1-5 seeds were then plated on one-half Murashige and Skoog square plates (10 cm × 10 cm), around 30 seeds per line and both lines in each plate, and were then maintained vertically in a growth chamber at 22 to 25°C, with a 16-h-light and 8-h-dark photoperiod.

After 13 days, 100 μl of a freshly prepared WCS417 suspension (10⁷ CFU ml⁻¹) were evenly distributed at the bottom of the hypocotyl of each seedling line, under sterility. Plates were then allowed to dry and were maintained again in a vertical position in the growth chamber. After 2 more days, roots were cut with a sharp sterile blade at the root base, and the root samples were inserted into a 1.5-ml Eppendorf tube of known weight; root samples were weighed again, and 1 ml of 10 mM MgSO₄ was then added in each tube. After vortexing for 1 min, a series of bacterial dilutions was prepared from 10⁰ (initial suspension) to 10⁻⁸ in 10 mM MgSO₄; such dilutions were plated on King's B medium agar supplemented with rifampicin at 50 μg ml⁻¹ and plates were maintained overnight at 28°C. Colonies were then counted and CFU per gram for each sample was calculated as (colony number × 10 × dilution fold) divided by root fresh weight.

Leaves staining

For GUS staining, leaves and seedlings were surface-sterilized by immersion in 70% EtOH and were washed twice with sterile water, as described (van Hulten et al. 2019). Staining for GUS activity was performed according to Elorza et al. (2004). For DAB staining, H₂O₂ staining with DAB was performed as described by Murgia et al. (2004).

Protein extraction from leaves

Rosette leaves were sampled from 4-week-old plants after 2 days of exposure to either WCS417 or mock treatment, as described above. In detail, the rosette leaves from one single plant were sampled, were weighed (0.15 to 0.4 g each), were packed in alufoil, were frozen in liquid nitrogen, and were stored at –80°C, and total proteins were then extracted, essentially according to the protocol published by Wu et al. (2014), omitting the trichloroacetic acid and acetone precipitation steps (steps 2 to 9 in the published protocol). As starting material for each extraction representing a biological sample, leaves from two different rosettes were used. The protein pellets were maintained at −80°C. Pellets were then resuspended in up to a 100- to 120-μl final volume of 10 mM phosphate buffered saline by heating for 15 min at 37°C and vortexing, followed by a 2-min centrifugation at 15,000 × g; the supernatant contained the solubilized proteins, and if a remaining pellet could still be observed, it was heated, vortexed, and centrifuged again, for one or two more cycles, for thorough solubilization of all the proteins in the starting frozen pellet.

Enzymatic digestion of protein extracts

The total protein extract of each sample was concentrated from 100 to 50 µl in a vacuum concentrator at 60°C and was treated with 0.25% (wt/vol) RapiGest SF reagent (Waters Co.). The resulting suspensions were incubated with stirring at 100°C for 20 min, were cooled to room temperature, and were centrifuged for 10 min at 2,200 × g. The protein concentration was assayed using the Invitrogen Qubit Protein BR assay kit (Life Technologies Corporation, Thermo Fisher), and 50 μg of protein from each sample was digested overnight at 37°C by adding sequencing-grade modified trypsin (Promega Inc.) at a 1:50 (wt/wt) enzyme/substrate ratio. An additional aliquot of trypsin (1:100 wt/wt) was then added in the morning, and the digestion continued for 4 h. The enzymatic digestion was chemically stopped by acidification with 0.5% trifluoroacetic acid (Sigma-Aldrich Inc.), and a subsequent incubation at 37°C for 45 min completed the RapiGest acid hydrolysis. Water-immiscible degradation products were removed by centrifugation at 13,000 rpm for 10 min. Finally, the tryptic digest mixtures were desalted using Pierce C-18 spin columns (Thermo Fisher Scientific, Pierce Biotechnology), according to the manufacturer protocol, and were resuspended in 0.1% formic acid (Sigma-Aldrich Inc.) in water (LC-MS Ultra CHROMASOLV, Honeywell Riedel-de Haen) at a concentration of 0.2 µg µl⁻¹.

LC-MS/MS analysis

Peptide mixtures were analyzed using the Eksigent nanoLC-Ultra two-dimensional system (Eksigent, part of AB SCIEX) configured in trap-elute mode. Briefly, samples (0.8 µg injected) were first loaded on a trap (200 µm × 500 µm ChromXP C18-CL, 3 µm, 12 nm) and were washed with the loading pump running in isocratic mode, with 0.1% formic acid in water for 10 min at a flow of 3 µl min⁻¹. The automatic switching of the autosampler ten-port valve then eluted the trapped mixture on a nano reversed-phase column (75 µm × 15 cm ChromXP C18-CL, 3 µm, 12 nm) through a 145-min gradient of eluent B (eluent A, 0.1% formic acid in water; eluent B, 0.1% formic acid in acetonitrile) at a flow rate of 300 nl min⁻¹. In depth, the gradient was as follows: from 5 to 10% B in 3 min, 10 to 30% B in 104 min, 30 to 95% B in 26 min, and holding at 95% B for 12 min. The eluted peptides were directly analyzed on an LTQ-OrbitrapXL mass spectrometer (Thermo Fisher Scientific) equipped with a nanospray ion source. The spray capillary voltage was set at 1.7 kV and the ion transfer capillary temperature was held at 220°C. Full MS spectra were recorded over a 400 to 1,600 m/z range in positive ion mode, with a resolving power of 60,000 (full width at half maximum) and a scan rate of 2 spectra per second. This step was followed by five low-resolution MS/MS events that were sequentially generated in a data-dependent manner on the top five ions selected from the full MS spectrum (at 35% collision energy), using dynamic exclusion of 0.5 min for MS/MS analysis. Mass spectrometer scan functions and high-performance liquid chromatography solvent gradients were controlled by the Xcalibur data system version 1.4 (Thermo Fisher Scientific).

LC-MS/MS spectra processing and data handling

The Proteome Discoverer software 2.5 using SEQUEST HT search engine (Thermo Fisher Scientific) was used to process all LC-MS/MS runs against Arabidopsis thaliana counting 39,256 entries downloaded in July 2022 (www.uniprot.org). The following criteria were used for peptide and related protein identification: trypsin as enzyme with two missed cleavage per peptide, mass tolerance of ±50 ppm mass tolerance for the precursor, and ±0.8 Da for fragment ions. Validation was performed by Percolator node with a target-decoy search and a false discovery rate ≤0.01 and maximum ΔCN of 0.05. The minimum peptide length of 7 amino acids at confidence ‘Medium’ level was set. PSMs were used in a label-free quantification approach to compare protein lists (n = 24) and identify DEPs, as previously reported (Palma et al. 2021). Briefly, data matrix complexity was reduced by linear discriminant analysis and in a pairwise comparison (WT Col mock-treated vs. WT Col WCS417-treated; atfdh1-5 mock-treated vs. atfdh1-5 WCS417-treated; WT Col mock-treated vs. atfdh1-5 mock-treated; WT Col WCS417-treated vs. atdh1-5 WCS417-treated) and proteins with P ≤ 0.05 were retained. Pairwise comparisons were further evaluated by DAve index [(PSMs_A-PSMs_B)/(PSMs_A + PSMs_B)]/0.5, where A and B represent the samples compared; specifically, positive DAve values indicate proteins upregulated in A (and downregulated in B), while negative DAve values indicate proteins upregulated in B (and downregulated in A) (Mauri and Dehò 2008). Finally, DEPs were processed by hierarchical clustering using Ward's method and the Euclidean distance metric. All data processing was performed by JMP15.2 SAS. Using STRING Cytoscape APP (Doncheva et al. 2019), the protein profiles characterized for WT Col mock-treated, WT Col WCS417-treated, atfdh1-5 mock-treated, and atfdh1-5 WCS417-treated phenotypes were evaluated at the functional level, and the most-enriched Kyoto Encyclopedia of Genes and Genomes pathways and biological processes were extracted and compared (WT Col mock-treated vs. WT WCS417-treated; atfdh1-5 mock-treated vs. atfdh1-5 WCS417-treated; WT Col mock-treated vs. atfdh1-5 mock-treated; WT Col WCS417-treated vs. atfdh1-5 WCS417) by unpaired t test (P ≤ 0.01).

Statistical analysis

To test for significant differences between WCS417-treated plants and mock-treated plants for the in-vivo data experiments, an unpaired t test was run, establishing for each comparison equal or unequal variances before the analysis by an F-test two sample for variances. In detail, for in vitro mock-treated 2 days (sample size = 54) vs. in vitro WCS417-treated 2 days (sample size = 63), unequal variances were assumed, P = 0.04512639 (P < 0.05); for in vitro mock-treated 7 days (sample size = 70) vs. in vitro WCS417-treated 7 days (sample size = 135), equal variances were assumed, P = 3.052E⁻⁰⁸ (P < 0.01); for in vivo mock-treated 2 days (sample size = 36) vs. in vivo WCS417-treated 2 days (sample size = 36), equal variances were assumed, P = 0.001576 (P < 0.01); for in vivo mock-treated 4 days (sample size = 82) vs. in vivo WCS417-treated 4 days (sample size = 100), unequal variances were assumed, P = 0.01385 (P ≤ 0.01); for close to mock (sample size = 90) vs. close to WCS417 (sample size = 99), equal variances were assumed, P = 0.4325 (not statistically significant); for Col mock-treated weights (sample size = 25) vs. Col WCS417-treated weights (sample size = 25), unequal variances were assumed, P = 0.002381 (P < 0.01); for atfdh1-5 mock-treated weights (sample size = 25) vs. atfdh1-5 WCS417-treated weights (sample size = 25), unequal variances were assumed, P = 0.0002032 (P < 0.01); for the WCS417 colonization test (sample size = approximately 30) equal variances were assumed, P = 0.039 (P < 0.05). All data processing was performed using R (ver. 4.1.0) packages ggpubr and dplyr.

Photochemical parameters

The photochemical parameters F₀, F_m, F_v, and maximal photochemical efficiency F_v/F_m were measured in dark-adapted leaves (20 min) as previously described (Murgia et al. 2020).

Acknowledgments

We are grateful to P. Bakker for donating the Pseudomonas simiae WCS417 strain.