Skip main navigation
Previous
Next
RESEARCHOpen Access icon OPENOpen Access license

Glutamate Positively Regulates Chitinase Activity and the Biocontrol Efficacy of Pseudomonas protegens

    Affiliations
    Authors and Affiliations
    • Kasumi Takeuchi1 †
    • Masayo Ogiso1
    • Tomohiro Morohoshi2
    • Shigemi Seo1
    1. 1Division of Plant Molecular Regulation, Institute of Agrobiological Sciences, National Agriculture and Food Research Organization, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8518, Japan
    2. 2Graduate School of Regional Development and Creativity, Utsunomiya University, 7-1-2 Yoto, Utsunomiya, Tochigi 321-8585, Japan
    Published Online:28 Jul 2023https://doi.org/10.1094/MPMI-09-22-0178-R

    Abstract

    Broad-spectrum biocontrol by Pseudomonas protegens CHA0 and other fluorescent pseudomonads is achieved through the generation of various secondary metabolites with antibiotic activities against not only other microbes but, also, nematodes and insects present in the rhizosphere. A previous metabolomic study demonstrated that intracellular low–molecular weight effectors, such as guanosine tetraphosphate and γ-aminobutyrate, function as important signals in niche adaptation by strain CHA0 to plant roots. We investigated the role of amino acids in the biocontrol trait of P. protegens Cab57 towards Pythium damping off and root rot in cucumber. Among the 11 amino acids tested, only glutamate markedly enhanced the efficacy of biocontrol. An RNA-Seq analysis revealed that glutamate upregulated the expression of a chitinase gene cluster (c21370-c21380, in which the c21370 gene was annotated as a gene encoding the chitin-binding protein cbp and the c21380 gene encoded chitinase chiC) in strain CHA0. Glutamate upregulated the expression of the regulatory small RNA rsmZ but reduced the production levels of other Gac/Rsm-regulated biocontrol factors, such as 2,4-diacetylphloroglucinol and pyoluteorin. The promoter activity of cbp and chitinase activity were characterized in detail; their activities were up-regulated in response to glutamate and their expression was under the control of GacA. Therefore, glutamate appears to be essential for biocontrol activity in which chitinase production is regulated in response to glutamate.

    Copyright © 2023 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license.

    Broad-spectrum biocontrol by Pseudomonas protegens and other root-colonizing fluorescent pseudomonads is achieved by the generation of extracellular enzymes and secondary metabolites with antibiotic activities that suppress not only plant diseases but, also, nematodes and insects. Previous studies that investigated the biosynthesis of extracellular enzymes, including AprA protease, and secondary metabolites, such as 2,4-diacetylphloroglucinol (DAPG), pyrrolnitrin, and pyoluteorin (Plt), used the model P. protegens strains Pf-5 and CHA0 (also called P. fluorescens Pf-5 and CHA0, respectively) (Haas and Keel 2003). We recently isolated P. protegens Cab57 from the rhizosphere of shepherd's purse in Japan as an effective biocontrol strain and its genomic features were characterized (Takeuchi et al. 2014).

    The expression of these biocontrol factors depends on the Gac/Rsm signal transduction pathway (Kidarsa et al. 2013; Lapouge et al. 2008), which is elicited by the GacS/GacA two-component system (Lapouge et al. 2008; Valentini and Filloux 2016). The autophosphorylation of the GacS sensor kinase occurs at high cell-population densities and activates the cognate GacA response regulator, which, in turn, induces the transcription of two or more noncoding small RNAs (sRNAs) in Pseudomonas spp. (Kay et al. 2005; 2006; Lalaouna et al. 2012). The expression of three sRNAs, called RsmX, RsmY, and RsmZ, was shown to be positively regulated by GacA in the P. protegens strains CHA0 and Pf-5 and in P. brassicacearum (Kay et al. 2005; Kidarsa et al. 2013; Lalaouna et al. 2012). The affinity of these sRNAs for the RNA-binding proteins RsmA and RsmE was previously found to be high (Heeb et al. 2002; Kay et al. 2005; Reimmann et al. 2005; Valverde et al. 2003). The translation of genes with a conserved ANGGAN motif in or near the Shine-Dalgarno sequence was repressed by the RsmA and RsmE proteins (Lapouge et al. 2008; Reimmann et al. 2005). However, the translational repression of target genes was attenuated by the induction of RsmX, RsmY, and RsmZ sRNAs at high cell-population densities, which resulted in the RsmA and RsmE proteins being sequestered. Therefore, the synthesis of biocontrol factors, inhibition of plant diseases, and production of insecticides was impaired in mutants that were defective for the Gac/Rsm signal transduction pathway (Haas and Défago 2005; Kim et al. 2011; Kupferschmied et al. 2013).

    Previous studies demonstrated that other phenotypes of multicellular behavior, including the formation of biofilms and swarming motility, were influenced by the Gac/Rsm signal transduction pathway (Kidarsa et al. 2013; Lapouge et al. 2008; Song et al. 2016). The switch between planktonic and biofilm lifestyles has been shown to involve the Gac/Rsm system (Valentini and Filloux 2016). During the biofilm mode of growth by biocontrol strains of Pseudomonas spp., this switch promotes their proliferation on plant roots as well as competition with other organisms in the rhizosphere (Barahona et al. 2011; Lalaouna et al. 2012). In our metabolomic study, we revealed the positive regulation of this switch by the intracellular signal guanosine tetraphosphate, which is characteristic of limited nutrient and stress conditions (Takeuchi et al. 2012). γ-Aminobutyrate also functions as an important intracellular signal in niche adaptation by strain CHA0 to plant roots, leading to the planktonic mode of growth, the decreased formation of biofilms, and root colonization without showing the highly pleiotropic gacA phenotypes (Takeuchi 2018).

    Regarding effective ecological applications of biocontrol agents, the identification of extracellular signals that affect biocontrol activity will lead to advances in biocontrol research. Extra- and intracellular levels of intermediates of the Krebs cycle, including fumarate, succinate, and 2-oxoglutarate, were found to positively correlate with the regulation of secondary metabolism through GacA-dependent sRNA expression in P. protegens (Takeuchi et al. 2009). However, the fumA mutant, which lacks fumarase and, thus, accumulates fumarate, did not exhibit higher biocontrol activity than the wild type, although it showed higher levels of Gac/Rsm traits such as sRNA expression and antibiotic activity. These findings suggest the presence of other signals that induce biocontrol activity (Takeuchi et al. 2009).

    Amino acids have recently been reported to induce resistance in plants against pathogens (Goto et al. 2020; Kadotani et al. 2016; Seo et al. 2016). Based on these findings, amino acids appear to have crucial functions as signal molecules in plant-microbe interactions, which prompted us to investigate the effects of exogenous amino acids on the plant-beneficial bacterium P. protegens. In the present study, we identified glutamate as a potent activator of the biocontrol efficacy of P. protegens Cab57 against soil-borne disease. We then investigated the extent to which the transcriptomic profile of strain CHA0 was affected by a treatment of glutamate. To achieve this, we used a RNA-Seq–based approach, which provides an analysis of a broad range of genes. Among the genes upregulated by the glutamate treatment, the role of a chitinase gene cluster (c21370-c21380, in which the c21370 gene was annotated as a gene encoding the chitin-binding protein cbp and the c21380 gene encoded chitinase chiC) in this signaling pathway was examined in detail.

    Results

    Effects of amino acids on the biocontrol efficacy of P. protegens Cab57.

    We investigated the effects of amino acids on the biocontrol efficacy of P. protegens. To obtain an overview of the effects of amino acids, we evaluated the biocontrol efficacy of our domestic strain Cab57 (Takeuchi et al. 2014) based on a cucumber–P. ultimum pathosystem using plastic pots incubated in a growth chamber. We previously demonstrated that strain Cab57 was as effective as strain CHA0 as a biocontrol agent against Pythium damping off in cucumber and exhibited typical Gac/Rsm activities and antibiotic production (Takeuchi et al. 2014). The rationale for selecting our domestic strain Cab57 in the present study was for future potential field studies in Japan. We tested 11 amino acids in the L configuration and only glutamate markedly enhanced the biocontrol efficacy of strain Cab57 (Fig. 1). The concentrations of the amino acids were determined as previously reported on plant activators, which induces plant resistance to pathogens. To function as a plant activator, amino acids are applied to plant roots at a low millimolar level (Kadotani et al. 2016; Seo et al. 2016). Based on these findings and in consideration of initial water absorbed in soil, we applied 10 mM of each amino acid to soil to reach a final concentration of approximately 5 mM in the rhizosphere.

    Fig. 1.

    Fig. 1. Effects of amino acids on the suppression of Pythium damping off and root rot in cucumber by Pseudomonas protegens Cab57. To evaluate the effects of 11 amino acids on the biocontrol activity of P. protegens, plastic pots with a depth of 5 cm containing 30 g of nonsterile soil (vermiculite) were planted with two seedlings of cucumber each and were treated with Pythium ultimum MAFF425494 and P. protegens Cab57. P. protegens Cab57 was added to soil as a suspension of cells in sterile distilled water to give 8 × 106 CFU per gram of soil. Regarding the amino-acid treatment, 10 ml of 10 mM amino acid was added to soil. Control pots received the same amount of sterile water. After a 14-day incubation, the biocontrol activity of each treatment was assessed by measuring fresh weights per seedling. Data represent the means of 12 to 18 replicates per treatment. Relative fresh weights are given as the water control (amino acid minus) with P. protegens Cab57 being set to 1. Different letters indicate significant differences (P < 0.05) based on an analysis of variance and Tukey's honestly significant difference test.

    Download as PowerPoint

    RNA-Seq profiling of strain CHA0 treated with glutamate.

    We selected glutamate for follow-up studies because, among the 11 amino acids tested, only glutamate markedly enhanced the biocontrol efficacy of strain Cab57. To examine the transcriptomic effects of the glutamate treatment, we performed RNA-Seq profiling of strain CHA0, which enabled molecular analyses, examined its growth in glycerol-casamino acid medium (GCM) amended with 5 mM glutamate, and compared it with that in GCM without amended glutamate (the water control). Among the 45 genes specifically upregulated in glutamate-treated cells (Table 1), the chitinase gene cluster c21370-c21380, in which the c21370 gene was annotated as a gene encoding the chitin-binding protein cbp and the c21380 gene encoded chitinase chiD (Winsor et al. 2011; http://www.pseudomonas.com) was investigated because i) this cluster was strongly up-regulated by glutamate and ii) chitinase is an important biocontrol factor of chitinase-producing bacteria in the rhizosphere (Veliz et al. 2017). Chitinase C was recently identified as one of the Gac-regulated factors contributing to the toxicity of P. protegens CHA0 and Pf-5 toward Plutella xylostella and Drosophila melanogaster, respectively (Flury et al. 2016; Loper et al. 2016). The c21380 gene was annotated as chiD in the database; however, since this gene and the homologue were examined as chiC in both of these studies, we hereafter referred to c21380 as chiC (Fig. 2A).

    Table 1. Upregulated genes identified from RNA-Seq profiles in Pseudomonas protegens CHA0y

    Fig. 2.

    Fig. 2. Physical map of the chitinase gene and surrounding genes and the construction of cbp′-′lacZ and cbp-lacZ fusions in Pseudomonas protegens CHA0. A, A 350-bp EcoRI/BamHI fragment upstream of PFLCHA0_c21370 (cbp) that included the first four codons of the cbp gene was fused to the first codon of lacZ in the vector pME6014, with an interval of the BamHI site in frame, to give pME6014cbp. A 260-bp EcoRI/BamHI fragment upstream of PFLCHA0_c21370 (cbp) that excluded the 5′ untranslated region of the cbp gene was fused to the upstream region of lacZ in the vector pME6016 to give pME6016cbp. A putative RsmA/E-binding site is shown in the box upstream of the predicted ATG start codon of cbp. The typical GGA motif is present in bold face. The putative −35 and −10 promoter sequences are shown in gray boxes. Inset, predicted secondary structure of the mRNA leader sequence near the translation start site when complexed with RsmA/E. The typical GGA motif is present in bold face. B, Construction of the cbp′-′lacZ and cbp-lacZ fusions. The plasmids pME6014cbp and pME6016cbp contain a translational ′lacZ fusion and transcriptional lacZ fusion in the vectors pME6014 and pME6016, respectively.

    Download as PowerPoint

    The genes belonging to the ofa cluster, which encodes orfamideA and the surrounding region (locus tags are given as c21840 to c21900), were also up-regulated. Parts of the phl cluster and plt cluster, which encode DAPG and Plt, respectively, were also up-regulated (locus tags are given as c59080 to c59110 and as c28450 to c28580). The genes encoding other biocontrol factors, such as AprA and HCN, were not included in this list. Genes downregulated by the glutamate treatment are listed in Table 2, with most of the list being amino acid–related genes. It is important to note that the genes belonging to the histidine utilization (hut) gene cluster were down-regulated (locus tags are given as c04050 to c04170), suggesting that regulatory links may exist between glutamate input and histidine utilization.

    Table 2. Downregulated genes identified from RNA-Seq profiles in Pseudomonas protegens CHA0y

    Antibiotic production profiles in CHA0 treated with glutamate.

    We examined the production levels of antibiotics to clarify the effects of the glutamate treatment, using a high-performance liquid chromatography (HPLC) analysis (Table 3). The production of DAPG was reduced under the glutamate treatment. The production of Plt was also decreased, which suggests a degree of regulation between the transcriptional levels observed in RNA-Seq profiles and the production levels of these metabolites, thereby preventing the overproduction of these metabolites. Although time-course analyses of transcripts and metabolites may provide important insights, they were not conducted in the present study, the focus of which, at this stage, was chitinase.

    Table 3. 2,4-Diacetylphloroglucinol (DAPG) and pyoluteorin production levels by Pseudomonas protegens CHA0y

    Glutamate positively regulates cbp-chiC expression at the translational level.

    Based on their RNA-Seq profiles, glutamate appears to be involved in the co-transcription and coordinated induction of the cbp-chiC genes of strain CHA0. To elucidate the mechanisms by which the extracellular level of glutamate affects their expression, we constructed a translational cbp′-′lacZ fusion (Fig. 2B) and assessed its expression. The addition of glutamate to the medium dose-dependently upregulated expression, which confirmed that glutamate acted as a signaling molecule in the expression of cbp-chiC (Fig. 3A). We also examined the effects of histidine, which was previously shown to function as a plant activator (Seo et al. 2016); however, the results obtained revealed that it did not affect the expression of this fusion in strain CHA0 (Fig. 3B). The D configuration of glutamate (D-Glu) also exhibited positive activity (Fig. 3C), suggesting that CHA0 does not distinguish the DL configuration of glutamate in the regulation of cbp-chiC expression.

    Fig. 3.

    Fig. 3. Effects of the L configuration of A, glutamate (l-Glu) and B, histidine (l-His) and C, the D configuration of glutamate (d-Glu) on the expression of the cbp′-′lacZ fusion on pME6014cbp in Pseudomonas protegens CHA0 grown in glycerol-casamino acid medium. Symbols indicate the means of triplicate cultures and bars indicate standard deviations. OD600 = optical density at 600 nm.

    Download as PowerPoint

    We analyzed eight other amino acids and, among them, several amino acids, particularly tryptophan, markedly increased expression levels (Fig. 4). Arginine and aspartate were also examined but were found to attenuate the growth of CHA0 under the conditions tested; therefore, we did not include these data.

    Fig. 4.

    Fig. 4. Effects of amino acids on cbp′-′lacZ expression in Pseudomonas protegens CHA0 grown in glycerol-casamino acid medium. Bars show the means of triplicate cultures. Sampling was performed at an optical density at 600 nm of approximately 1.6. Different letters indicate significant differences (P < 0.05), based on an analysis of variance and Tukey's honestly significant difference test.

    Download as PowerPoint

    GacA regulates cbp-chiC expression at the translational level.

    In P. protegens CHA0 and Pf-5, chitinase C was previously identified as one of the Gac-regulated factors contributing to toxicity towards P. xylostella and D. melanogaster, respectively (Flury et al. 2016; Loper et al. 2016). In the present study, expression levels were examined, using the translational cbp′-′lacZ fusion to clarify whether the Gac/Rsm system controlled the promoter activity of cbp in CHA0. A RsmA/E-binding site was detected in the predicted Shine-Dalgarno sequence of cbp (AATGGAGT) (Fig. 2A), with the central ribonucleotides (ATGGAG) forming a hexa-loop on an A-T base pair (underlined), which permitted RsmA and RsmE to bind to this site (Lapouge et al. 2008). This predicted structure is highly similar to that in the phlA mitochondrial RNA (mRNA) leader sequence of P. protegens CHA0 (AAUGGAAU) (Lapouge et al. 2013).

    The expression levels cbp′-′lacZ throughout growth were markedly lower in the gacA mutant (Fig. 5A) than in the wild type (Fig. 3A). Glutamate-induced increases in the expression of cbp′-′lacZ in the wild type were not observed in the gacA mutant (Fig. 5A), suggesting a regulatory hierarchy in which glutamate exerted positive effects on the expression of cbp′-′lacZ in a gacA-positive background.

    Fig. 5.

    Fig. 5. Effects of glutamate on cbp′-′lacZ expression in the gacA mutant and rsmZ-lacZ expression in Pseudomonas protegens CHA0. A, Expression of the cbp′-′lacZ fusion on pME6014cbp in the gacA mutant grown in glycerol-casamino acid medium (GCM). B, Expression of the rsmZ-lacZ fusion on pME6091 in P. protegens CHA0 grown in GCM. Symbols indicate the means of triplicate cultures and bars indicate standard deviations. OD600 = optical density at 600 nm.

    Download as PowerPoint

    To establish whether the function of the Gac/Rsm cascade was affected by glutamate, rsmZ-lacZ expression levels with or without amended glutamate were also assessed (Fig. 5B). Glutamate exerted positive effects on the expression of rsmZ. This result was consistent with those of the RNA-Seq analysis, which showed that Gac/Rsm-regulated genes such as ofa, phl, and plt as well as chi were up-regulated, whereas the levels of production of DAPG and Plt are not up-regulated, as described above.

    Since translational cbp′-′lacZ fusion is dependent on the native cbp-chiC promoter, we constructed a pure transcriptional cbp-lacZ fusion without the putative 5′ untranslated region (UTR) (Fig. 2B) and assessed its expression. As shown in Figure 6, the addition of glutamate to the medium did not affect expression levels in strain CHA0, suggesting that the effects of glutamate on the upregulated expression of the cbp-chiC operon observed in the RNA-Seq analysis did not occur at the transcriptional level. In other words, the increased mRNA levels of cbp and chiC in the presence of glutamate may be attributed to a reduction in the degradation of mRNAs but not to the higher transcription of the operon.

    Fig. 6.

    Fig. 6. Effects of glutamate on the expression of the transcriptional cbp-lacZ fusion on pME6016cbp in Pseudomonas protegens CHA0 (wild type) (closed symbols) and the gacA mutant (open symbols) grown in glycerol-casamino acid medium. Symbols indicate the means of triplicate cultures and bars indicate standard deviations. OD600 = optical density at 600 nm.

    Download as PowerPoint

    The expression levels of the cbp-lacZ fusion throughout growth were slightly lower in the gacA mutant than in the wild type, and the addition of glutamate to the medium did not affect expression levels in the gacA mutant (Fig. 6). These results suggest that GacA regulates cbp-chiC expression not only at the translational level but, also, at the transcriptional level; however, this regulation is subtle. The differences observed in the cbp′-′lacZ fusion between the wild type without glutamate and the gacA mutant (approximately 1,000 Miller units at optical density at 600 nm [OD600] of 2.0 in the wild type and approximately 30 Miller units at all cell densities in the gacA mutant [Figs. 3A and 5A]) were markedly larger than those in the cbp-chiC fusion (approximately 100 Miller units in the wild type and approximately 60 Miller units in the gacA mutant at OD600 of 2.0 [Fig. 6]), suggesting that GacA mainly regulates cbp-chiC expression at the translational level.

    Glutamate enhances the chitinase activity of CHA0.

    To investigate the effects of glutamate on the chitinase activity of strain CHA0 grown in GCM, we measured extracellular chitinase activity with chitin azure as a substrate. As shown in Table 4, an amendment with 5 mM glutamate increased the chitinase activity of strain CHA0. In the gacA mutant, enzyme activity was below the detection limit in the presence and the absence of amended glutamate (Table 4). These results are consistent with those of the assay on translational cbp′-′lacZ fusion.

    Table 4. Effects of glutamate on extracellular chitinase activity in Pseudomonas protegens CHA0 and its mutants

    To confirm the function of the cbp-chiC operon in chitinase activity, a chromosomal cbp-chiC deletion mutant was generated. The activity of the resultant strain, named CHA0chi, was markedly lower than that of the wild type (Table 4). Enzyme activity was weak and enhanced by amended glutamate, suggesting that a major chitinase is encoded by this region; however, an additional chitinase may exist in CHA0. We constructed a plasmid carrying the cbp-chiC operon with its promoter region and introduced it into strain CHA0chi for complementation. The chitinase activity of the resultant strain was restored, confirming the complemented function of this operon.

    LuxR solo homologue mutants have increased expression levels of cbp′-′lacZ.

    The LuxR solo homologue PsoR, encoded by PFL_5298, has been reported to play a role in the regulation of chitinase in strain Pf-5, with the overexpression of PsoR upregulating the expression of the chitinase gene (Subramoni et al. 2011). In that study, another luxR solo homologue was identified as PFL_3627 by an in-silico analysis. To investigate the relevance of the expression of these regulator proteins and chitinase in the present study, we identified these homologues in the genome of strain CHA0 as PFLCHA0_c52710 and PFLCHA0_c36680, respectively, and generated chromosomal c52710 and c36680 deletion mutants. We also constructed a c36680 c52710 double mutant.

    The expression levels of cbp′-′lacZ were assessed in each strain (Fig. 7). No significant differences were observed in its expression levels between the c52710 single mutant and the wild type. In contrast, higher levels of expression were detected in the c36680 mutant and c36680 c52710 double mutant, suggesting that a LuxR solo homologue encoded by PFLCHA0_c36680 negatively regulates the expression of chitinase.

    Fig. 7.

    Fig. 7. Effects of c36680 and c52710 mutations on the expression of the cbp′-′lacZ fusion in Pseudomonas protegens CHA0. The expression of the cbp′-′lacZ fusion on pME6014cbp was assessed in P. protegens CHA0 (wild type) (open circles), the c36680 mutant (closed triangles), c52710 mutant (closed squares), and c36680 c52710 mutant (closed circles) grown in glycerol-casamino acid medium amended with 5 mM glutamate. Symbols indicate the means of triplicate cultures and bars indicate standard deviations. OD600 = optical density at 600 nm.

    Download as PowerPoint

    Glutamate enhances the biocontrol efficacy of strain CHA0.

    We investigated whether glutamate affected the biocontrol efficacy of strain CHA0 based on a cucumber–P. ultimum pathosystem. The addition of glutamate and CHA0 did not influence plant growth in the absence of Pythium. When Pythium was infested, the addition of glutamate itself had no effect on the suppression of disease. In contrast, when added together with strain CHA0, the level of biocontrol efficacy was higher than that with strain CHA0 itself (with 95% confidence for shoot weight and root weight) (Table 5), which was consistent with the results obtained on strain Cab57. Although it currently remains unclear whether glutamate-induced increases in chitinase activity directly resulted in the greater biocontrol efficacy of strain CHA0, the role of glutamate as an extracellular positive effector on the expression of Gac/Rsm-regulated biocontrol factors observed in the RNA-Seq analysis is consistent with the biocontrol trait of P. protegens. Regarding the composition of the cell walls of pathogens, in contrast to fungi, oomycetes are diploid with cell walls composed mainly of β-1,3-d-glucans, β-1,6-d-glucans, and cellulose with a small amount of chitin or chitosaccharides. In Pythium ultimum, chitin and cellulose are both present in the cell wall (Cherif et al. 1993). Legume chitinase has been reported to exhibit potent inhibitory activity against Pythium spp. (Wang et al. 2009), suggesting a function of chitinase as a biocontrol factor against Pythium spp. in plant-microbe and microbe-microbe interactions.

    Table 5. Effects of glutamate on the inhibition of Pythium damping off and root rot in cucumber by Pseudomonas protegens CHA0

    Discussion

    The present results revealed that exogenous glutamate positively regulated the expression of chitinase at the translational level as well as biocontrol activity in P. protegens. The GacS/GacA two-component system is at the top of the hierarchy of master regulators, similar to other biocontrol factors. In the present study, glutamate itself exerted positive effects on the expression of the sRNA rsmZ; however, the promoter region of cbp-chi may be regulated by one or more additional regulators other than the Gac/Rsm cascade. Although the LuxR solo homologue encoded by c36680 is a candidate, it is considered to play a role in negative regulation. Therefore, one or more positive regulators responding to glutamate may be involved in the expression of chitinase. Numerous extra- and intracellular signals have been proposed to play a role in the switch under GacS/GacA control (Takeuchi et al. 2012). However, the control of each exoproduct appears to be more complex. For example, the production of DAPG is under positive feedback regulation and phloroglucinol is involved in the regulation of Plt biosynthesis (Clifford et al. 2016; Schnider-Keel et al. 2000). The decreases observed in the levels of DAPG and Plt produced in the present study were attributed to this feedback regulation. There is another precedent that needs to be considered regarding the fitness trade-off associated with antibiotic production; Yan et al. (2018) reported that Plt biosynthesis, which poses a metabolic burden on producer cells, leads to the accumulation of spontaneous Gac-negative mutants. Therefore, a metabolic burden may occur under conditions in which glutamate accelerates the expression of these metabolite genes, which reduces production levels.

    The concentrations of amino acids in the rhizosphere in natural soil are an important factor that needs to be considered from a biocontrol point of view. Although few studies have examined its concentration, glutamate is naturally present in the rhizosphere at a low micromolar level but is highly abundant among proteinogenic amino acids (Moe 2013). The effects of glutamate on biocontrol traits were examined herein, using two concentrations, 2 and 5 mM. To function as plant activators that induce plant resistance to pathogens, amino acids have also been applied to plant roots at a low millimolar level (Kadotani et al. 2016; Seo et al. 2016), with histidine and glutamate inducing resistance towards the bacterial pathogen Ralstonia solanacearum and rice blast, respectively. In the present study, Pythium damping off in cucumber was not inhibited by the glutamate treatment itself, suggesting that the intensity of a vigorous pathogen may overcome the resistance of cucumber induced by glutamate because cucumber seedlings encountered Pythium ultimum at the earliest stage under the conditions tested.

    The concentrations of endogenous glutamate in plants were previously reported to be at a low micromolar level and reached approximately 50 mM at damaged sites (Toyota et al. 2018). Assuming that similar conditions occur in natural roots, glutamate released from damaged roots may trigger the function of chitinase in root-colonizing pseudomonads. Previous findings indicated that glutamate functioned as a signal to exert chitinase activity in plants. Exogenous glutamate was found to enhance the expression of defense-related genes, including chitinase, in rice (Kadotani et al. 2016), and it has also been reported to prime chitin-induced responses in Arabidopsis (Goto et al. 2020). Taken together with the present results, these findings indicate the relevance of glutamate and chitin-related responses not only in plants but also in root-colonizing pseudomonads. Therefore, glutamate may function as a cue for effective and specific disease suppression by both sides of plants and pseudomonads. An RNA-Seq analysis of strain CHA0 in planta will provide insights into the dynamics of its response in the rhizosphere. Exogenous glutamate was also found to increase the population density of Streptomyces spp. in the rhizosphere microbiota of tomato plants, suggesting a positive effect of glutamate on plant health (Kim et al. 2021).

    It currently remains unclear whether glutamate modulates the synthesis of one or more extracellular signals activating GacS, because there is no precise biochemical assay for these signals and signal biosynthetic genes have not yet been identified. Therefore, further studies that use a bioassay to quantify the production of these signals are warranted (Dubuis et al. 2006; Dubuis and Haas 2007). Glutamate may be a precursor or building block of signaling molecules that enhance chitinase promoter activity. The nonselectivity of the DL configuration of glutamate for chitinase expression in the present study also provides some insights. There is a precedent that needs to be considered regarding histidine. The overexpression of hut genes has been shown to abolish the ability of P. aeruginosa PAO1 to induce type III–mediated cytotoxicity (Rietsch et al. 2004), suggesting that the amino acid state of the cell influences the expression of the type III regulon. Although strain CHA0 lacks a type III secretion system apparatus, the addition of amino acids may induce extensive phenotypic changes in pseudomonads other than chitinase expression.

    It is important to note that, in addition to glutamate, tryptophan also exerts positive effects on chitinase expression. Although the common properties of these two amino acids are unclear, tryptophan is a molecule of interest in plant-microbe interactions because it is a precursor of indole-3-acetic acid in both root colonizing pseudomonads (represented by CHA0) and plants (Moe 2013; Oberhänsli et al. 1991). Therefore, an abundance of tryptophan will contribute to a rhizosphere that is preferable not only for root development but also for disease suppression through the induction of chitinase production by P. protegens CHA0. In addition to disease suppression aspects, a plant morphological study on chemical signaling will provide novel insights into the modulation of an optimal rhizosphere.

    With the expectation that the upregulated expression of antibiotic secondary metabolites through the promotion of the Gac/Rsm system will contribute to increases in the biocontrol activity of CHA0, some mutants have been subjected to biocontrol tests; however, optimal disease suppression has not yet been achieved. For example, a mutation in fumA (for a fumarase isoenzyme) upregulated the expression of three sRNAs and secondary metabolism but did not enhance biocontrol efficacy under natural soil conditions (Takeuchi et al. 2009). This is consistent with the result of the constitutively active form of GacS leading to the overproduction of antifungal secondary metabolites in vitro but not to the better biocontrol of Fusarium crown and root rot of tomato (Voisard et al. 1988). In contrast, the positive effects of glutamate on the biocontrol efficacy of CHA0 observed in the present study may be due to the selective effects of glutamate on chitinase activity, which did not induce the full expression of other exoproducts, such as DAPG and Plt, thereby preventing the exhaustion of strain functions. This inexpensive carbon source has potential as an ingredient in formulation mixtures of fluorescent biocontrol pseudomonads for the optimal suppression of root diseases.

    Materials and Methods

    Bacterial strains and growth conditions.

    The bacterial strains and plasmids used in the present study are listed in Supplementary Table S1. Escherichia coli and P. protegens strains were routinely grown in NYB (2.5% [wt/vol] nutrient broth, 0.5% [wt/vol] yeast extract) and Luria-Bertani medium, with shaking, or on nutrient agar plates (4% [wt/vol] blood agar base, 0.5% [wt/vol] yeast extract) amended with the following antibiotics when required: ampicillin, 100 μg/ml, kanamycin, 25 μg/ml, or tetracycline, 25 μg/ml (100 μg/ml for the selection of P. protegens). Inoculation temperatures of 30 and 37°C were used for P. protegens and E. coli, respectively.

    Bacteria grown in GCM (Maurhofer et al. 1998) were used for RNA-Seq assays, the detection of exoproducts, and β-galactosidase assays.

    DNA manipulation.

    A QIAprep spin miniprep kit (Qiagen) was used to conduct small-scale plasmid extraction, while a Qiagen plasmid midi kit was employed for large-scale preparations. The preparation of chromosomal DNA from P. protegens was performed using Qiagen genomic tips. A QIAquick gel extraction kit (Qiagen) was used to purify DNA fragments from agarose gels. A list of the oligonucleotides used is shown in Supplementary Table S2.

    RNA-Seq analysis.

    P. protegens CHA0 was cultured in GCM medium without glutamate or supplemented with glutamate at 5 mM until reaching an OD600 of 2.0. A bacterial culture (250 μl) was collected by centrifugation. RNA purification was conducted with an RNeasy minikit (Qiagen). We used a NanoDrop ND-1000 spectrophotometer to estimate the concentration of RNA. RNA samples were maintained at −80°C until sequencing. RNA quality was verified using Agilent TapeStation 2200. After the removal of ribosomal RNA (rRNA), using the Illumina Ribo-Zero rRNA removal kit, mRNA was used to generate a complementary DNA library, according to a custom protocol of Eurofins Genomics K. K., which was then sequenced using the HiSeq 2500 system (Illumina). Eurofins Genomics K. K. constructed the libraries and performed sequencing reactions. Reads were cleaned by Trimmomatic (ver.0.32) (Bolger et al. 2014) to remove adaptor sequences and low-quality reads and were then mapped to the P. protegens CHA0 genome (NC_021237.1), from the National Center for Biotechnology Information, using BWA (ver.0.7.17) (Li et al. 2009).

    Differentially expressed genes among the samples were identified using edgeR (ver.3.16.5) (Robinson et al. 2010), with trimmed mean of M values normalization methods (Robinson and Oshlack 2010) to normalize for the RNA composition by finding a set of scaling factors for library sizes. Up- and downregulated genes were defined with a log2 fold change ≥1.5 and ≤−1.5, respectively, with a false discovery rate cutoff of 1%.

    RNA-Seq data are available from the DDBJ Sequence Read Archive (https://www.ddbj.nig.ac.jp/index-e.html) under accession number DRA013598.

    Construction of the cbp′-′lacZ and cbp-lacZ fusions.

    The amplification of a 350-bp EcoRI-BamHI fragment upstream of PFLCHA0_c21370 (cbp), in which the first four codons of the cbp gene were present, was performed with the primers chiProFEcoRI and chiProRBamHI (Supplementary Table S2). Following sequencing, pME6014cbp containing a translational cbp′-′lacZ fusion was obtained by cloning the fragment into pME6014 cut with EcoRI and BamHI. This fusion was dependent on the native cbp-chiC promoter (Fig. 2).

    The amplification of a 260-bp EcoRI-BamHI fragment upstream of PFLCHA0_c21370 (cbp), in which the 5′ UTR of the cbp gene was excluded, was performed with the primers chiProFEcoRI and chiProRtranscBamHI (Supplementary Table S2). Following sequencing, pME6016cbp containing the transcriptional cbp-lacZ fusion was obtained by cloning the fragment into pME6016, carrying the lacZ gene with its ribosome-binding site, cut with EcoRI and BamHI (Fig. 2).

    Generation of cbp-chiC– and luxR solo–negative mutants.

    Primers are listed in Supplementary Table S2. A deletion in the chromosomal cbp-chiC genes of P. protegens CHA0 was created as follows. Fragments of approximately 700 bp located on each side of the cbp-chiC genes were amplified by PCR with the primer pairs c21370UF/c21370UR and c21380DF/c21380DR. Two corresponding fragments were annealed and amplified as a 1.4-kb fragment, using primer pair c21370UF/c21380DR. Following sequencing, pME3087chi was obtained by cloning the 1.4-kb fragment into pME3087 cut with BamHI and HindIII. Triparental mating with E. coli HB101/pME497 was conducted to mobilize the plasmid from E. coli DH5α to P. protegens CHA0. The vector was excised by a second crossover following the enrichment of tetracycline-sensitive cells (Humair et al. 2009), thereby generating the cbp-chiC mutant.

    The creation of in-frame deletions in the chromosomal luxR solo homologue genes of P. protegens CHA0 was performed as follows. The amplification of fragments of approximately 700-bp that were present on either side of luxR solo homologue genes was conducted by PCR with the primer pairs c36680UF/c36680UR and c36680DF/c36680DR for c36680 and c52710UF/c52710UR and c52710DF/c52710DR for c52710. The annealing and amplification of each set of corresponding fragments as 1.4-kb fragments were performed using the primer pairs c36680UF/c36680DR and c52710UF/c52710DR, respectively. Following sequencing, the cloning of these 1.4-kb fragments into pME3087 cut with BamHI and HindIII provided pME3087c36680 and pME3087c52710, respectively. Triparental mating with E. coli HB101/pME497 mobilized these plasmids from E. coli DH5α to P. protegens CHA0. The vector was excised by a second crossover following the enrichment of tetracycline-sensitive cells, resulting in the generation of the c36680 or c52710 mutant. The c36680 gene was deleted from the c52710 mutant by mobilizing the pME3087c36680 plasmid to the c52710 mutant as a recipient strain.

    Expression of cbp-chiC genes by a plasmid.

    The 2.5-kb fragment carrying the cbp-chiC genes was amplified by PCR with primers chiFBamHI and ChiRHindIII (Supplementary Table S2), high-fidelity DNA polymerase KOD Plus (Toyobo), and the genomic DNA of P. protegens CHA0 as a template. The fragment was cloned into pCR-Blunt II-TOPO (ThermoFisher Scientific). The insert obtained was confirmed by sequencing. After sequencing, the fragment was cloned into pME6031 cut with BamHI and HindIII. Electroporation was then performed to introduce pME6031chi, the plasmid obtained, into strain CHA0.

    β-Galactosidase assays.

    The Miller method was performed to quantify the activity of β-galactosidase (Miller 1972). P. protegens strains were grown at 30°C, with shaking at 180 rpm, in 50-ml flasks containing 15 ml of GCM supplemented with 0.05% Triton X-100. Cell aggregation was prevented by Triton X-100.

    Detection of antibiotics.

    To detect DAPG and Plt production, spent culture supernatant extracts of strains were prepared. After the inoculation, these strains were cultivated in 15 ml of GCM for 15 h (scaling up from an overnight culture to a fresh culture 50 times and grown in Erlenmeyer flasks at 30°C at 180 rpm), which corresponded to an OD600 of approximately 2.5. The extraction of 2 ml of the supernatant was performed using 1.7 ml of ethyl acetate and was followed by evaporation to dryness. Following the dissolution of crude extracts in 300 μl of the initial solvent, 20 μl was subjected to HPLC. Samples were chromatographed on a HPLC system (Jasco), with a UV/VIS detector set at 254 nm, using a Mightysil RP-18GP column (250 × 4.6 mm, particle diameter of 5 µm; Kanto Kagaku). Samples were isocratically eluted at 2 ml/min with water/acetonitrile/acetic acid (50:50:0.1 [vol/vol/vol]).

    Detection of chitinase activity.

    Assessments of the chitinase activity of P. protegens strains grown in GCM were performed using the substrate chitin azure (Sigma-Aldrich). Following inoculation, strains were grown in GCM for 24 h (scaling up from an overnight culture to a fresh culture 50 times and grown in Erlenmeyer flasks at 30°C at 180 rpm), which corresponded to an OD600 of approximately 3.5. The supernatant (5 ml) was concentrated to 300 μl by ultrafiltration (Vivaspin 500; cut-off: 10,000 Da, Sartorius AG) and was then subjected to the reaction as a crude enzyme. The enzyme sample (200 μl), 200 μl of 0.2 M sodium phosphate buffer (pH 7), and 1 mg of chitin-azure were mixed and were incubated at 37°C for 24 h. The tubes were centrifuged and the absorbance of supernatants was read at 560 nm. One unit of chitinase was the amount of enzyme that produced an increase of 0.001 in absorbance per OD600 of the original culture under the specified conditions.

    Plant disease suppression assays.

    To assess the effects of 11 amino acids on the biocontrol activity of P. protegens, two seedlings each of cucumber (Cucumis sativus L. cv. Tokiwajibai) were planted in plastic pots with a depth of 5 cm containing 30 g of nonsterile soil (vermiculite) and were then treated with P. ultimum MAFF425494 and P. protegens Cab57. P. protegens Cab57 was added to soil as a suspension (4 ml per pot) of cells washed once in sterile distilled water to give 8 × 106 CFU per gram of soil, and 10 ml of 10 mM amino acids was added to soil as the amino acid treatment. The same amount of sterile water was added to control pots. After covering seedlings with 15 g of nontreated soil, microcosms were incubated in a growth chamber at 26°C, with 60% relative humidity, under a 16-h-light and 8-h dark-cycle. Plant pots were watered every 3 to 4 days. After 14 days, the biocontrol activity of each treatment was evaluated based on the fresh weights (root plus shoot) of each seedling.

    To investigate the effects of glutamate in more detail, three seedlings each of cucumber (Cucumis sativus L. cv. Tokiwajibai) were planted in flasks containing 20 g of vermiculite and were then treated with one or both P. ultimum MAFF425494 and P. protegens CHA0. P. protegens CHA0 was added to soil as a suspension (4 ml per flask) of cells washed once in sterile distilled water to give 2 × 107 CFU per gram of soil, and 1 ml of 10 mM glutamate was added to soil as the glutamate treatment. The same amount of sterile water was added to control flasks. After covering seedlings with 5 g of nontreated soil, flasks were sealed with aerated silicon caps and microcosms were incubated in a growth chamber at 26°C, with 60% relative humidity, under a 16-h-light and 8-h-dark cycle. Watering was not necessary. After 12 days, the biocontrol activity of each strain was evaluated based on counts of the number of surviving plants and measurements of shoot and root fresh weights per flask. Data in Table 5 represent the means of two individual repetitions of the same experiment. Data from both experiments were initially analyzed for trial-by-treatment interactions by an analysis of variance; therefore, data from two independent trials were pooled. Means were separated using Tukey's honestly significant difference test (at P ≤ 0.05). Statistical analyses were performed using R (version 4.1.1).

    Acknowledgments

    We thank M. Yokoyama and M. Inaba for their technical assistance.

    Author-Recommended Internet Resource

    Pseudomonas Genome database: www.pseudomonas.com

    The author(s) declare no conflict of interest.

    Literature Cited

    RNA-Seq data are available from the DDBJ Sequence Read Archive under accession number DRA013598.

    Funding: This work was supported, in part, by Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research (JSPS KAKENHI) grants 18H02209 and 21H02200.

    The author(s) declare no conflict of interest.