Abstract

   Pseudomonas syringae pathovar tabaci 11528 (P. syringae 11528) is a
   phytopathogen that causes wild‐fire disease in soybean and tobacco
   plants. It utilizes a cell density‐dependent regulation system known as
   quorum sensing (QS). In its QS system, the psyI is responsible for the
   biosynthesis of N‐acylhomoserine lactones (AHLs). By comparing the
   transcripts from P. syringae 11528 wild‐type strain with those of the
   ΔpsyI mutant using RNA sequencing (RNA‐seq) technology, 1118
   AHL‐regulated genes were identified in the transition from exponential
   to stationary growth phase. Numerous AHL‐regulated genes involved in
   pathogenicity were negatively controlled, including genes linked to
   flagella, chemotaxis, pilus, extracellular polysaccharides, secretion
   systems, and two‐component system. Moreover, gene ontology and pathway
   enrichment analysis revealed that the most pronounced regulation was
   associated with bacterial motility. Finally, phenotypic assays showed
   that QS‐regulated traits were involved in epiphytic growth of pathogens
   and disease development in plants. These findings imply that the
   AHL‐mediated QS system in P. syringae 11528 plays significant roles in
   distinct stages of interactions between plants and pathogens, including
   early plant colonization and late plant infection.

   Keywords: gene expression, Pseudomonas syringae pathovar tabaci 11528,
   quorum sensing, RNA sequencing, virulence traits

1. INTRODUCTION

   Bacterial quorum sensing (QS), a cell–cell communication mechanism, is
   a gene regulation system that acts in a cell density‐dependent manner
   (Waters & Bassler, [32]2005). In QS system, signaling molecules, termed
   autoinducers, are produced and accumulated until they reach a critical
   threshold concentration, after which they bind to receptors that can
   trigger coordinated changes in gene expression (Keller & Surette,
   [33]2006). The most well‐known signaling molecules are N‐acylhomoserine
   lactones (AHLs) in Gram‐negative bacteria, which are mainly synthesized
   by LuxI‐homologs and are sensed by LuxR‐type regulator proteins,
   resulting in the subsequent regulation of bacterial gene expression and
   synergistic changes in physiological behaviors (Fuqua, Winans, &
   Greenberg, [34]1994). AHL‐mediated QS regulation plays a significant
   role in phytopathogens, as it is responsible for the expression of many
   traits involved in plant–pathogenic interactions (Loh, Pierson,
   Pierson, Stacey, & Chatterjee, [35]2002; Von Bodman, Bauer, & Coplin,
   [36]2003). The QS system in Erwinia carotovora, CarI/CarR, is required
   for the production of several exoenzymes involved in the maceration of
   plant tissues (Jones et al., [37]1993; Pirhonen, Flego, Heikinheimo, &
   Palva, [38]1993). In Pantoea (Erwinia) stewartii, exopolysaccharide as
   a virulence determinant is under the regulation of EsaI/EsaR system
   (Von Bodman, Majerczak, & Coplin, [39]1998). P. aeruginosa, a pathogen
   of both plants and animals, has two separate but interrelated QS
   systems, and the LasI/LasR system controls the expression of the
   RhlI/RhlR system (Schuster and Greenberg, [40]2006). Both QS systems
   regulate a large amount of extracellular virulence factors and
   secondary metabolites, among which some factors contribute to the
   epiphytic growth of bacteria in planta (Heurlier, Denervaud, & Haas,
   [41]2006; Rahme et al., [42]2000; Smith, [43]2003; Whitehead, Barnard,
   Slater, Simpson, & Salmond, [44]2001).

   The pathogenicity of phytopathogens depends upon various factors,
   including effectors secreted by a type III secretion system (TTEs),
   toxins, and epiphytic fitness (Baltrus et al., [45]2011; Studholme
   et al., [46]2009). Among these factors, epiphytic fitness, which
   represents the ability to acquire nutrients and survive on the surface
   of a leaf (Lindeberg, Myers, Collmer, & Schneider, [47]2008), is
   influenced by many determinants, such as motility (Quiñones, Dulla, &
   Lindow, [48]2005), chemotaxis (Brencic & Winans, [49]2005; Yao & Allen,
   [50]2006), extracellular polysaccharides (EPS) (Taguchi et al.,
   [51]2008; Von Bodman & Farrand, [52]1995; Von Bodman et al., [53]1998),
   and iron uptake (Cha, Lee, Oh, Choi, & Baik, [54]2008). Moreover, cell
   wall‐degrading hydrolytic enzymes (Mäe, Montesano, Koiv, & Palva,
   [55]2001) and other secretion systems (Jakob, Kniskern, & Bergelson,
   [56]2007; Preston, Studholme, & Caldelari, [57]2005) may also
   contribute to both virulence and epiphytic fitness. Multidrug
   resistance mediated by multidrug efflux pump, is involved in the
   resistance to antimicrobials produced by plants and contributes to
   bacterial virulence (Barabote et al., [58]2003; Burse, Weingart, &
   Ullrich, [59]2004). Many studies have reported that the expression of
   many genes involved in pathogenicity is regulated by QS in
   phytopathogens, including Pantoea stewartii (Von Bodman & Farrand,
   [60]1995), Ralstonia solanacearum (Clough, Flavier, Schell, & Denny,
   [61]1997), Erwinia carotovora (Mäe et al., [62]2001), and Burkholderia
   glumae PG1 (Gao et al., [63]2015).

   Quorum sensing regulation is also very common to Pseudomonas syringae,
   a widespread bacterial pathogen that causes disease in a broad range of
   economically important plant species. The production of AHL effects
   colony morphology as well as epiphytic viability in P. syringae pv.
   syringae strain B3A (Dumenyo, Mukherjee, Chun, & Chatterjee, [64]1998).
   In addition to epiphytic fitness, EPS and motility are also regulated
   by AhlI/AhlR system of P. syringae pv. syringae B728a (Quiñones et al.,
   [65]2005). P. syringae pv. tabaci, one of P. syringae pathovars, causes
   wild‐fire disease in soybeans and tobacco plants (Gasson, [66]1980;
   Ribeiro, Hagedorn, Durbin, & Uchytil, [67]1979). P. syringae pv. tabaci
   synthesizes a phytotoxin, termed tabtoxin, which generates a hydrolytic
   product tabtoxinine‐β‐lactam that can inhibit glutamine synthetase and
   contributes to the accumulation of ammonia and subsequent loss of
   chlorophyll in the host cell, ultimately resulting in the formation of
   chlorotic halos that surround necrotic spots on the leaves of infected
   plants (Durbin, [68]1991; Thomas, Langstonunkefer, Uchytil, & Durbin,
   [69]1983). The QS system of P. syringae pv. tabaci depends on psyI,
   which is responsible for the biosynthesis of AHLs, as well as the
   regulatory protein PsyR (Elasri et al., [70]2001). The production of
   N‐(3‐oxo‐hexanoyl)‐homoserine lactone (3OC6‐HSL) and N‐hexanoyl‐L‐HSL
   (C6‐HSL) has been reported in P. syringae pv. tabaci (Shaw et al.,
   [71]1997; Taguchi et al., [72]2006). Several phenotypic assays have
   been reported that the psyI‐deficiency could affect some bacterial
   behaviors including a few virulence traits in P. syringae pv. tabaci
   (Taguchi et al., [73]2006). But limited further analysis is available
   concerning the role of AHL‐mediated QS regulation on the pathogenicity
   of P. syringae pv. tabaci. Notably, we do not know the number and types
   of virulence traits under the regulation of QS system, and how AHL
   controls virulence traits has not been studied systematically. Thus,
   the comprehensive exploration of QS‐dependent regulons is urgently
   needed in P. syringae pv. tabaci.

   In this study, we investigated the AHL‐mediated QS regulation of P.
   syringae pv. tabaci 11528 (P. syringae 11528), which naturally causes
   disease in wild tobacco, an important model system for studying
   plant–pathogen interactions, and its draft complete genome sequence has
   been analyzed (Studholme et al., [74]2009). We compared the transcripts
   from the ΔpsyI mutant with those of the wild type during growth using
   RNA sequencing (RNA‐seq) technology. Phenotypic assays designed to
   assess plant–pathogen interactions, including swarming motility and
   disease symptoms, were comparatively analyzed. These findings will
   extend our understanding of AHL‐mediated regulation on plant–pathogen
   interactions and provide the molecular basis for the pathogenicity in
   P. syringae pv. tabaci.

2. MATERIALS AND METHODS

2.1. Bacterial strains, plasmids, and culture conditions

   The bacterial strains and plasmids used in this study are listed in
   Table  [75]1. Agrobacterium tumefaciens A136 and Chromobacterium
   violaceum CV026 were cultivated at 30°C in LB broth (McLean, Whiteley,
   Stickler, & Fuqua, [76]1997). P. syringae pv. tabaci 11528 (P. syringae
   11528) and the ΔpsyI mutant were grown at 30°C in King's medium B (KB)
   (Cha et al., [77]2008). Escherichia coli DH5α was grown at 37°C in LB
   broth. Antibiotics were added to media as required at the following
   final concentrations: ampicillin, 50 μg/μl; spectinomycin, 50 μg/μl;
   tetracycline, 20 μg/μl; and kanamycin, 20 μg/μl.

Table 1.

   Bacterial strains and plasmids used in this study
   Strains or plasmids Relevant characteristics^a Source or reference
   Strains
   P. syringae pv. tabaci 11528 Tox^+ Tox^r, causal agent of wild‐fire of
   tobacco (ATCC type strain) ATCC, USA
   P. syringae pv. tabaci △psyI mutant ATCC 11528 △psyI (683‐bp deletion)
   Cha et al. ([78]2008)
   A. tumefaciens A136 Tet^r,Sp^r,traI‐lacZ fusion; AHL biosensor McLean
   et al. ([79]1997)
   C. violaceum CV026 Km^r,AHL biosensor McLean et al. ([80]1997)
   Escherichia coli DH5α F^‐φDH5lacZ ΔM15 Δ(lacZYA‐argF)U169 end A1 recA1
   hsdR17(r[k] ^‐m[k] ^‐) supE44 λ‐ thi‐1 gyrA96 relA1 phoA Biomed,
   Peking, China
   Pta(pBQ9) Sp^r , P. syringae pv. tabaci 11528 wild type containing
   plasmid pBQ9 Cheng et al.([81]2016)
   △psyI(pBQ9‐P[nptII]) Sp^r , P. syringae pv. tabaci 11528 △psyI mutant
   containing plasmid pBQ9‐P[nptII] This study
   Plasmids
   pBQ9 Sp^r, pPROBE‐OT derivative harboring P. syringae ahlI promoter
   upstream of GFP Quiñones et al. ([82]2005)
   pUCGMA2T[1−4] Amp^r, P [nptII]:gfp Deng et al. ([83]2014)
   pBQ9‐P[nptII] Sp^r, derivative of pBQ9 containing P [nptII] promoter
   This study
   [84]Open in a new tab

   Tox^r, tabtoxin resistance; Tet^r, tetracycline resistance; Km^r,
   kanamycin resistance; Sp^r, spectinomycin resistance; Amp^r, ampicillin
   resistance.

2.2. RNA isolation

   To prepare RNA samples, P. syringae 11528 wild‐type strain and ΔpsyI
   mutant were grown in a shaker incubator at 30°C and 200 rpm to
   exponential phase after inoculated into test tubes with 5 ml KB and
   then subcultured in conical flasks with 50 ml KB. Cells were harvested
   at lag phase (L phase; OD[600] of 0.3), exponential phase (E phase;
   OD[600] of 0.7) and a transition from exponential to stationary phase
   (T phase; OD[600] of 1.7) by centrifugation at 5,000g for 5 min. Total
   RNA was extracted from ~10 mg of cell pellets using a miRNeasy Mini Kit
   (Qiagen, Hilden, Germany) according to the previous study (Cheng, Ma,
   Zhuang, & Fray, [85]2016). Three biological replicates were processed
   per sample.

2.3. RNA‐seq library construction and sequencing

   RNA‐seq libraries were generated using the NEBNext^® Ultra^™
   Directional RNA Library Prep Kit for Illumina^® (New England BioLabs,
   Inc., Ipswich, MA, USA) following the manufacturer's recommendations.
   One hundred base‐pair, paired‐end sequencing was performed using an
   Illumina HiSeq 2000 platform. Raw reads were subjected to standard
   quality control criteria to remove all reads that fit any of the
   following parameters: reads that aligned to adaptor sequences; reads
   for which more than 10% of bases were unknown; reads for which there
   were more than 50% of low‐quality bases (quality value <5) in one read.
   All remaining clean reads were mapped to P. syringae 11528 genome
   (Studholme et al., [86]2009) (sequence download from
   [87]http://www.ncbi.nlm.nih.gov/Traces/wgs/?val=ACHU02#contigs) and
   analyzed using Bowtie2‐2.0.6 (Langmead & Salzberg, [88]2012). Those
   reads that mapped to the reference genome were used in further
   analyses. Sequence‐read mapping and genome coverage data are summarized
   in Table  [89]2. The number of reads that mapped to each gene was
   counted using HTSeq v0.6.1. Levels of gene expression were calculated
   using the RPKM (reads per kilobase transcriptome per million reads)
   method (Mortazavi, Williams, McCue, Schaeffer, & Wold, [90]2008).

Table 2.

   Overall statistics of RNA‐seq data
   Growth phase[91]a Sample name[92]b Clean reads Uniquely mapped Mapping
   ratio
   L phase WT (1) 8,374,920 7,960,118 95.05%
   WT (2) 9,020,734 8,581,603 95.05%
   WT (3) 12,223,654 11,596,127 94.87%
   MT (1) 11,451,222 10,859,077 94.83%
   MT (2) 11,384,438 10,822,555 95.06%
   MT (3) 10,778,120 10,245,952 95.06%
   E phase WT (1) 10,536,414 10,011,486 95.02%
   WT (2) 9,466,758 8,962,413 94.67%
   WT (3) 10,374,856 9,782,692 94.29%
   MT (1) 8,628,616 8,261,892 95.75%
   MT (2) 13,608,578 12,996,703 95.5%
   MT (3) 13,823,208 13,219,185 95.63%
   T phase WT (1) 8,228,786 7,913,448 96.17%
   WT (2) 10,591,606 10,206,726 96.37%
   WT (3) 9,670,798 9,205,368 95.19%
   MT (1) 12,170,034 11,523,821 94.69%
   MT (2) 33,263,468 31,646,901 95.14%
   MT (3) 23,287,204 22,186,569 95.27%
   [93]Open in a new tab
   ^a

   Lag phase (L phase); Exponential phase (E phase); Transition from
   exponential to stationary phase (T phase).
   ^b

   Wild‐type strain (WT); ΔpsyI mutant (MT). Replicate numbers are
   indicated in parentheses.

2.4. RNA‐seq analysis

   Differential expression analysis data were displayed using the DESeq R
   package (1.10.1) (Anders & Huber, [94]2010). All resulting p‐values
   were adjusted using the Benjamini–Hochberg procedure to control the
   false discovery rate (FDR). Those genes with an adjusted p‐value <.05
   and at least a twofold change in expression were classified as
   differentially expressed genes (DEGs). To analyze DEGs, we conducted a
   gene ontology (GO; [95]http://www.geneontology.org/) enrichment assay
   using the GOseq R package (Young, Wakefield, Smyth, & Oshlack,
   [96]2010) and then calculated an adjusted p‐value. An adjusted p‐value
   <.05 was considered to indicate significance for the enriched set of
   data. For subsequent analyses of the QS‐dependent pathway, DEGs were
   mapped to the pathway database ([97]http://www.genome.jp/kegg/) using
   KOBAS 2.0 software (Mao, Tao, & Wei, [98]2005). The mapped pathway was
   enriched using an adjusted p‐value of <.05.

2.5. Quantitative real‐time PCR

   To validate the gene expression changes indicated by the RNA‐seq data,
   quantitative real‐time PCR (qPCR) analysis was performed according to
   the protocol of Cheng et al. ([99]2016). Specific primers for qPCR are
   listed in Table S1. Quantification of transcript expression was carried
   out using the 2^−ΔΔCt method using the constitutively expressed 16S
   ribosomal RNA gene as a control for normalization.

2.6. Swarming motility tests

   Swarming motility was assessed on semisolid KB plates that contained
   0.4% (wt/vol) Bacto agar (BD‐Difco) according to a previous study
   (Taguchi et al., [100]2006) with some modifications. Cells of
   P. syringae 11528 wild‐type strain and ΔpsyI mutant were grown to T
   phase and then resuspended in 10 mmol/L potassium phosphate buffer
   (PBS). Two microliter of suspensions (~10^7 cells) was dropped onto KB
   plates and then inoculated plates were incubated at 27°C for 36 hr, and
   colony diameters were measured as indicators of swarming motility. All
   experiments were conducted with three replicates and repeated three
   times.

2.7. GFP‐labeling of P. syringae 11528 ΔpsyI mutant

   GFP‐labeled P. syringae 11528 wild‐type strain, Pta(pBQ9), and its
   observation on tobacco leaves were reported in our previous works
   (Cheng et al., [101]2016). The plasmid pBQ9 encodes a gfp marker gene
   that is fused to the promoter of P [ahlI] and yields inducible
   expression of green fluorescence in P. syringae (Quiñones, Pujol, &
   Lindow, [102]2004). The promoter for P [nptII] is a neomycin
   phosphotransferase promoter that constitutively expresses the gfp gene
   (Stiner & Halverson, [103]2002). For GFP‐labeling of P. syringae 11528
   ΔpsyI mutant, we constructed a plasmid pBQ9‐P[nptII], which yields
   constitutive expression of green fluorescence. A fragment of the
   promoter P [nptII] was amplified from plasmid pUCGMA2T[1‐4] (Deng,
   Zhuang, Ma, Yu, & Zhuang, [104]2014) using primers nptII‐1
   (CCCGTCGACGTCAGGCTGTAACAGCTCAGA) and nptII‐2
   (GGCGAATTCATCCTGTCTCTTGATCAGATCTTG) that contained engineered SalI and
   EcoRI restriction enzyme sites (italicized with underline), and the
   fragment was then cloned into plasmid pBQ9 as a SalI‐EcoRI fragment at
   the 5ʹ‐terminus of the gfp gene, replacing the entire P [ahlI] promoter
   and creating plasmid pBQ9‐P[nptII]. ΔpsyI(pBQ9‐P[nptII]) was
   constructed by electroporation of the pBQ9‐P[nptII] plasmid into the
   ΔpsyI mutant. Thus, GFP‐labeled ΔpsyI mutant was achieved.

2.8. Epifluorescence microscopy on tobacco surfaces

   To study the ability of both P. syringae 11528 wild‐type strain and
   ΔpsyI mutant to colonize plants, tobacco leaves (Nicotiana tabacum L.)
   were spray‐inoculated to wetness with a suspension that contained
   10^6 CFU/ml of GFP‐labeled P. syringae 11528 wild‐type or ΔpsyI mutant
   cells. After inoculation, tobacco plants were placed in a greenhouse
   (25°C, ~80% relative humidity, 12 hr/day photoperiod). GFP‐labeled P.
   syringae 11528 strains were observed using a confocal laser scanning
   microscope (CLSM; LSM 780, Carl Zeiss, Oberkochen, Germany) at 1 and
   3 days following inoculation and bacterial population was determined
   following Cheng et al. ([105]2016).

2.9. Pathogenicity tests in tobacco leaves

   The pathogenicity of both P. syringae 11528 wild‐type strain and ΔpsyI
   mutant was evaluated in tobacco plants. Bacterial cells were grown to T
   phase and 10 μl suspensions of ~10^8 CFU/ml were infiltrated into
   tobacco leaves according to a previously published protocol (Cha
   et al., [106]2008). Inoculated tobacco plants were incubated in a
   greenhouse. Plants were evaluated for disease symptoms and diameters of
   the necrotic lesions that formed at each inoculation site were measured
   as indicators of virulence at 1, 3, 5, 7, and 9 days after inoculation.
   For each treatment, 20 leaves with two inoculation sites per leaf were
   randomly selected.

2.10. Statistical analysis

   The results of bacterial swarming motility, epiphytic population size,
   and lesion size on tobacco leaves are expressed as the means ± Standard
   Error of Mean. One‐way analysis of variance (ANOVA) and t test were
   performed using software Graphpad Prism V6.0 (GraphPad Software, San
   Diego, CA) for data analysis. *^* p < .01; *^** p < .001; *^***
   p < .0001.

2.11. SRA accession number

   The clean reads of RNA‐seq have been deposited in the NCBI
   sequence‐read archive (SRA) database under accession no.
   [107]SRP078136.

3. RESULTS

3.1. Overview of RNA‐seq data

   Before RNA‐seq, we assayed the AHL‐producing potential of P. syringae
   11528 ΔpsyI mutant using AHL‐biosensor strains C. violaceum CV026 and
   A. tumefaciens A136 (McLean et al., [108]1997; Steindler & Venturi,
   [109]2007) according to the previous studies (Lv et al., [110]2013).
   The ΔpsyI mutant did not induce the production of pigments in
   AHL‐biosensor strains (Figure S1), indicating the absence of bioactive
   AHL. These data imply that the ΔpsyI mutant was extremely defective in
   AHL production. Moreover, transcriptome analysis established that the
   psyI was only transcribed in P. syringae 11528 wild‐type strain, as no
   psyI mRNA could be detected in P. syringae 11528 ΔpsyI mutant (Gene
   locus: PSYTB_23601, Table S2). Thus, we suggest that the psyI in P.
   syringae 11528 is the only gene that encodes AHL synthases.

   To characterize the regulatory function of the AHL‐mediated QS system
   in P. syringae 11528, RNA‐seq analysis was conducted for P. syringae
   11528 wild‐type strain and ΔpsyI mutant. We identified transcripts of
   each sample with three biological replicates in the lag (L),
   exponential (E), and transition (T) phases, and the resulting 18
   libraries yielded 8,228,786 to 33,263,468 clean reads (Table [111]2).
   To profile the gene expression patterns, we mapped the clean reads of
   the 18 libraries to P. syringae 11528 genome. Overall, 7,913,448 to
   31,646,901 clean reads were uniquely mapped and the frequency of mapped
   reads was 94.29–96.37% (Table [112]2), indicating that effective reads
   of RNA‐seq data were well mapped to the reference genome. To validate
   the RNA‐seq data, qPCR was conducted for expression analysis using 22
   randomly selected genes (Table S1), and the relative expression levels
   of these genes between the ΔpsyI mutant and the wild type were
   compared. Data from qPCR largely confirmed the RNA‐seq data, and a
   positive correlation was detected between them (Figure S2), suggesting
   that RNA‐seq data were robust and of good quality.

3.2. QS‐dependent gene expression patterns in P. syringae 11528

   To identify AHL‐regulated genes in P. syringae 11528, we compared the
   transcripts of P. syringae 11528 wild‐type strain with those of the
   ΔpsyI mutant. As shown in Table [113]3, 31 and 41 differentially
   expressed genes (DEGs) were QS‐dependent regulated in the L and E
   phases, respectively, while a total of 1118 genes were found to be
   differentially expressed in the T phase. These data indicated that few
   genes were regulated during early growth, including L and E phases, and
   most genes were not controlled by QS until T phase. Thus, AHL‐dependent
   QS regulation was the most significant during T phase, and we focused
   on the T‐phase RNA‐seq analysis in subsequent assessments.

Table 3.

   Numbers of QS‐regulated genes in Pseudomonas syringae pv. tabaci 11528
   Growth phase[114]a No. of differentially expressed genes (DEGs)
   Up‐regulated Down‐regulated Total
   L phase 30 1 31
   E phase 38 3 41
   T phase 407 711 1118
   [115]Open in a new tab
   ^a

   Lag phase (L phase); Exponential phase (E phase); Transition from
   exponential to stationary phase (T phase).

   Among AHL‐regulated genes, many genes were associated with epiphytic
   fitness or virulence on plants (Table [116]4). Regarding bacterial
   pilus, P. syringae 11528 ΔpsyI mutant showed enhanced expression of
   genes that encode pilus assembly proteins. For secretion systems, type
   II, type III, and type VI secretion systems were down‐regulated in a
   QS‐dependent manner, with the exception of several individual genes.
   Regarding the two‐component system, six related genes were found to be
   down‐regulated. In addition, multidrug efflux pump and EPS genes were
   mostly down‐regulated by QS. Three genes related to iron transport were
   positively regulated by AHL, as was PSYTB_09751 that encodes a binary
   cytotoxin component, as shown in Table   [117]4. The most dramatic
   effect of QS regulation on gene expression was observed for genes
   linked to flagella and chemotaxis. Thirty‐eight and 45 genes
   responsible for the synthesis of bacterial flagella and chemotaxis
   proteins, respectively, were completely down‐regulated by AHL (Table
   [118]4). These data suggest that AHL‐dependent QS regulation in P.
   syringae 11528 negatively regulates the expression of a variety of
   virulence traits, and presumably plays a significant role in
   plant–pathogen interactions.

Table 4.

   QS‐regulated genes related to virulence traits in Pseudomonas syringae
   pv. tabaci 11528
   Virulence traits Gene locus[119]a Predicted function Fold change[120]b
   Flagella PSYTB_08676 Flagellar motor protein −3.28
   PSYTB_08681 Flagellar motor protein −3.54
   PSYTB_15355 Flagellar synthesis chaperone protein −2.66
   PSYTB_15365 Flagellar basal body P‐ring biosynthesis protein −6.34
   PSYTB_15395 Flagellar biosynthesis protein −5.37
   PSYTB_15400 Flagellar basal body rod protein −5.83
   PSYTB_15405 Flagellar basal body rod modification protein −6.33
   PSYTB_15410 Flagellar hook protein −5.56
   PSYTB_15415 Flagellar hook protein −3.18
   PSYTB_15420 Flagellar basal body rod protein −9.29
   PSYTB_15425 Flagellar basal body rod protein −10.23
   PSYTB_15430 Flagellar L‐ring protein −10.56
   PSYTB_15435 Flagellar P‐ring protein −11.43
   PSYTB_15440 Flagellar rod assembly protein −11.26
   PSYTB_15445 Flagellar hook protein −5.34
   PSYTB_15455 Flagellar hook‐associated protein −4.07
   PSYTB_15485 Flagellar protein −2.40
   PSYTB_15495 Flagellar protein −4.14
   PSYTB_15500 Flagellar assembly protein −4.45
   PSYTB_15520 Flagellar hook‐basal body protein −8.37
   PSYTB_15525 Flagellar M‐ring protein −10.06
   PSYTB_15535 Flagellar motor switch protein −9.86
   PSYTB_15540 Flagellar assembly protein −7.49
   PSYTB_15550 Flagellar protein −4.18
   PSYTB_15570 Flagellar hook‐length control protein −4.52
   PSYTB_15575 Flagellar basal body‐associated protein −5.69
   PSYTB_15580 Flagellar motor switch protein −9.96
   PSYTB_15585 Flagellar motor switch protein −9.88
   PSYTB_15590 Flagellar assembly protein −8.39
   PSYTB_15595 Flagellar biosynthesis protein −7.68
   PSYTB_15600 Flagellar biosynthetic protein −8.81
   PSYTB_15605 Flagellar biosynthesis protein −4.93
   PSYTB_15610 Flagellar biosynthesis protein −4.77
   PSYTB_15615 Flagellar biosynthesis protein −7.51
   PSYTB_15620 Flagellar biosynthesis regulator −9.15
   PSYTB_15670 Flagellar motor protein −3.58
   PSYTB_15665 Flagellar motor protein −3.62
   PSYTB_15885 Flagellar hook‐length control protein −2.74
   Chemotaxis PSYTB_00315 Methyl‐accepting chemotaxis protein −3.45
   PSYTB_01319 Methyl‐accepting chemotaxis protein −4.21
   PSYTB_01504 Chemotaxis protein −4.24
   PSYTB_03144 Chemotaxis protein −6.97
   PSYTB_03626 Chemotaxis protein −3.02
   PSYTB_03656 Chemotaxis protein −3.64
   PSYTB_03756 Chemotaxis protein −2.51
   PSYTB_03841 Methyl‐accepting chemotaxis protein −4.15
   PSYTB_05055 Chemotaxis protein −9.00
   PSYTB_05130 Chemotaxis protein −4.48
   PSYTB_06027 Chemotaxis protein −2.48
   PSYTB_06032 Chemotaxis protein −2.52
   PSYTB_12048 Chemotaxis protein −3.22
   PSYTB_12073 Methyl‐accepting chemotaxis protein −2.27
   PSYTB_12233 Chemotaxis protein −2.44
   PSYTB_12238 Chemotaxis protein −2.38
   PSYTB_13085 Chemotaxis protein −5.00
   PSYTB_14413 Chemotaxis protein −2.37
   PSYTB_15370 Chemotaxis protein −3.51
   PSYTB_15375 Chemotaxis protein −3.05
   PSYTB_15640 Chemotaxis protein −4.00
   PSYTB_15650 Chemotaxis protein −3.32
   PSYTB_15680 Chemotaxis protein −3.63
   PSYTB_15685 Chemotaxis protein −4.02
   PSYTB_15690 Chemotaxis protein −3.43
   PSYTB_15700 Chemotaxis protein −3.24
   PSYTB_15805 Chemotaxis protein −2.36
   PSYTB_16125 Chemotaxis protein −3.39
   PSYTB_16140 Methyl‐accepting chemotaxis protein −3.09
   PSYTB_16410 Chemotaxis protein −3.33
   PSYTB_16590 Methyl‐accepting chemotaxis protein −0.37
   PSYTB_17560 Methyl‐accepting chemotaxis protein −2.04
   PSYTB_19011 Methyl‐accepting chemotaxis protein −2.99
   PSYTB_20081 Chemotaxis protein −3.53
   PSYTB_20086 Methyl‐accepting chemotaxis protein −3.92
   PSYTB_24172 Chemotaxis protein −2.55
   PSYTB_25716 Chemotaxis protein −4.77
   PSYTB_25726 Chemotaxis protein −4.58
   PSYTB_25736 Chemotaxis protein −4.55
   PSYTB_25746 Chemotaxis protein −3.21
   PSYTB_25761 Chemotaxis protein −2.32
   PSYTB_27487 Chemotaxis protein −10.16
   PSYTB_27757 Chemotaxis protein −4.70
   PSYTB_28392 Chemotaxis sensory transducer −3.26
   Pilus PSYTB_07811 Pilus assembly protein −2.45
   PSYTB_07816 Pilus assembly protein −2.13
   PSYTB_07821 Pilus assembly protein −2.21
   PSYTB_09526 Pilus assembly protein −3.32
   PSYTB_15350 Pilus assembly protein −2.25
   PSYTB_23496 Pilus assembly protein −2.32
   Secretion system PSYTB_01314 Type VI secretion protein −2.80
   PSYTB_03239 Type VI secretion protein 4.84
   PSYTB_03249 Type VI secretion protein 2.69
   PSYTB_14001 Type II secretion system protein −3.09
   PSYTB_14013 Type II secretion system protein −3.56
   PSYTB_14018 Type II secretion system protein −3.60
   PSYTB_21360 Type VI secretion protein −5.51
   PSYTB_21365 Type VI secretion protein −5.02
   PSYTB_21370 Type VI secretion protein −5.05
   PSYTB_21375 Type VI secretion system protein −4.11
   PSYTB_21380 Type VI secretion system protein −3.41
   PSYTB_21410 Type VI secretion system protein −2.01
   PSYTB_21420 Type VI secretion system effector −2.38
   PSYTB_21425 Type VI secretion protein −2.47
   PSYTB_23396 Type II secretion system protein 2.08
   PSYTB_27882 Type III effector −2.07
   Two‐component system PSYTB_04915 Two‐component system response
   regulator −2.96
   PSYTB_05410 Two‐component system sensor histidine kinase −2.97
   PSYTB_17485 Two‐component system response regulator −3.52
   PSYTB_22030 Two‐component sensor histidine kinase −2.10
   PSYTB_25756 Two‐component system response regulator −3.47
   PSYTB_26762 Two‐component system response regulator −4.47
   Iron transport PSYTB_09216 Iron dicitrate transporter 3.94
   PSYTB_09221 Iron ABC transporter 2.85
   PSYTB_09226 Iron siderophore‐binding protein 3.81
   Multidrug efflux pump PSYTB_06721 Multidrug ABC transporter permease
   2.20
   PSYTB_26650 Multidrug ABC transporter substrate −2.10
   PSYTB_04905 Multidrug transporter −2.32
   Extracellular polysaccharide PSYTB_04635 Alginate biosynthesis protein
   −2.46
   PSYTB_15990 Alginate lyase −4.65
   Toxin PSYTB_09751 Binary cytotoxin component 2.97
   [121]Open in a new tab
   ^a

   Gene locus corresponds to the P. syringae pv. tabaci 11528 genome.
   ^b

   Fold change of gene expression in the wild type in comparison of the
   ΔpsyI mutant; The minus sign before fold change corresponds to
   down‐regulation by AHL‐mediated QS system.

   To identify specific processes that were more prominently AHL‐dependent
   in P. syringae 11528, AHL‐regulated genes were grouped into functional
   categories by GO and pathway enrichment analysis. Overall, a set of
   1118 AHL‐regulated genes were assigned to 1054 GO biological process
   terms and enriched in 41 terms, 210 GO cellular component terms and
   enriched in 11 terms, and 558 GO molecular function terms and enriched
   in three terms. Among those enrichment terms, many terms were closely
   related to bacterial motility, such as locomotion, cell motility,
   cellular component movement, and motor activity (Figure [122]1a). The
   maximal ratio of up‐regulation of those terms was only 4%, while motor
   activity was completely repressed by AHL‐mediated QS system. The
   distribution of enrichment pathways showed that three pathways were
   enriched, including two‐component system, bacterial chemotaxis, and
   flagellar assembly; the two latter pathways were involved in bacterial
   motility, of which flagellar assembly was completely down‐regulated in
   an AHL‐dependent manner (Figure [123]1b). To represent the effect of QS
   on flagella in a different format, expression profiles of genes linked
   to flagellar assembly were comparatively assessed using a heat‐map
   analysis (Figure [124]2), which revealed a significantly higher
   expression of flagellum protein in P. syringae 11528 ΔpsyI mutant than
   that in the wild type.

Figure 1.

   Figure 1
   [125]Open in a new tab

   Analysis of Quorum sensing (QS)‐dependent genes. QS‐dependent genes
   were identified by comparing transcriptomes of Pseudomonas syringae
   11528 wild‐type strain with those of the ΔpsyI mutant. Gene ontology
   (GO) enrichment terms of QS‐dependent genes linked to bacterial
   motility (a); Enrichment pathways of QS‐dependent genes (b). Each bar
   represents the percentage of genes (red, up‐regulated; blue,
   down‐regulated) belonging to the GO term or pathway shown in the y‐axis

Figure 2.

   Figure 2
   [126]Open in a new tab

   Expression profiles of genes linked to flagellar assembly. Each column
   of the heat map represents the Log[2] RPKM of each gene in Pseudomonas
   syringae 11528 wild‐type strain (left) and the ΔpsyI mutant (right)
   with a green‐black‐red scheme. Red, high expression; green, low
   expression

3.3. Phenotypic analyses of P. syringae 11528 wild type and ΔpsyI mutant

3.3.1. Bacterial motility

   When cells of P. syringae 11528 wild type and ΔpsyI mutant were
   inoculated on low‐agar plates at 27°C for 36 hr, the ΔpsyI mutant
   showed irregular dendritic colony pattern, which is a typical
   characteristic of P. syringae AHL‐mutant swarming behavior (Quiñones
   et al., [127]2005; Taguchi et al., [128]2006), while it was not
   observed on the plates of the wild type (Figure [129]3a). Moreover, the
   ΔpsyI mutant cells spreaded rapidly away from the point of inoculation
   but the wild‐type cells exhibited limited swarming motility. We
   measured colony diameter to indicate swarming motility. Colony
   diameters of the ΔpsyI mutant was ~2.2‐fold larger than that of the
   wild type (Figure [130]3b). Results imply that swarming motility is
   repressed by AHL‐mediated QS system in P. syringae 11528.

Figure 3.

   Figure 3
   [131]Open in a new tab

   Quorum sensing (QS)‐dependent swarming motility. Swarming motility
   phenotype (a) and swarming distance (b) of Pseudomonas syringae 11528
   wild‐type strain (top) and the ΔpsyI mutant (bottom) on semisolid
   King's B (KB) plates. Sterile filter discs placed on KB semisolid
   plates were inoculated with 1 × 10^7 cells and plates were incubated at
   27°C for 36 hr

3.3.2. Colonization observation

   We also investigated the plant‐colonizing ability of P. syringae 11528
   using fluorescence‐labeling technology combined with CLSM. GFP‐labeled
   P. syringae 11528 wild‐type and ΔpsyI mutant strains were constructed
   and then spray‐inoculated on tobacco leaves with the same inoculum
   concentration. CLSM observation showed that the majority of both
   strains assembled in the glandular trichomes at both 1 and 3 days
   postinoculation (Figure [132]4). Moreover, we compared the epiphytic
   populations of viable cells recovered from inoculated tobacco leaves.
   In contrast to the wild type, the observed population sizes of the
   ΔpsyI mutant were larger, especially at 3 days after inoculation
   (Figure [133]5). These findings imply that the ΔpsyI mutant exhibits a
   larger epiphytic population in inoculated leaves that results and more
   robust plant colonization ability when compared with the wild type.

Figure 4.

   Figure 4
   [134]Open in a new tab

   Colonization of GFP‐labeled Pseudomonas syringae 11528 strains on
   tobacco leaves. Tobacco leaves were spray‐inoculated with P. syringae
   pv. tabaci 11528 wild‐type strain (top) or the ΔpsyI mutant (bottom) at
   the concentrations of 10^6 CFU/ml and were observed in the glandular
   trichomes at 1 and 3 days after inoculation by confocal laser scanning
   microscopy

Figure 5.

   Figure 5
   [135]Open in a new tab

   Epiphytic population of Pseudomonas syringae 11528 strains. Tobacco
   leaves were spray‐inoculated with P. syringae 11528 wild‐type strain or
   the ΔpsyI mutant at the concentrations of 10^6 CFU/ml and were observed
   at 1 and 3 days after inoculation by confocal laser scanning
   microscopy. Magnification: ×1000 (Bar=20 μm), ×400 (Bar=50 μm)

3.3.3. Pathogenicity

   To explore the effect of AHL‐dependent regulation in P. syringae 11528
   on plant infection, pathogenicity tests were conducted via plant
   inoculation. The wild‐type and ΔpsyI mutant cells were infiltrated into
   tobacco leaves and the sizes of necrotic lesion were measured at
   various time points. Although necrotic leaf tissues were initially
   observed in tobacco leaves inoculated with either the ΔpsyI mutant or
   the wild type at 5 days postinoculation, disease symptoms incited by
   the ΔpsyI mutant were more pronounced than that by the wild type
   (Figure [136]6a). After 7 days of infection, lesion sizes were
   significantly larger when tobacco plants were treated with ΔpsyI mutant
   than the wild type (Figure [137]6b). These data indicate that
   AHL‐mediated QS system affects the interactions between tobacco‐P.
   syringae 11528 and represses disease development.

Figure 6.

   Figure 6
   [138]Open in a new tab

   Pathogenicity tests of Pseudomonas syringae 11528 strains. Leaf
   symptoms (a) and lesion sizes (b) on tobacco leaves that were induced
   by P. syringae 11528 wild‐type strain or the ΔpsyI mutant. Tobacco
   plants were infiltrated with P. syringae 11528 cells at concentrations
   of 10^8 CFU/ml during T phase. Leaf symptoms and lesions sizes were
   assayed at 3, 5, 7, and 9 days postinoculation under moist conditions
   at 25°C

4. DISCUSSION

   P. syringae 11528 is known to be a phytopathogen that can cause
   wild‐fire disease in soybeans and tobacco plants. Its QS system
   possesses psyI and psyR (Elasri et al., [139]2001), and the psyI is
   responsible for the AHL biosynthesis. To explore the AHL‐dependent QS
   regulation on gene expression in P. syringae 11528, we compared the
   transcriptomes of P. syringae 11528 ΔpsyI mutant with those of the wild
   type at three given time points. A total of 1118 QS‐regulated genes
   were found to be differentially expressed in the transition from
   exponential to stationary (T) phase, while dozens of QS‐regulated genes
   were identified in both L and E phases (Table  [140]3). These data
   demonstrate that the AHL‐mediated QS system may play different roles in
   transcriptional regulation during P. syringae 11528 growth. The
   clearest effect of QS regulation on gene expression was observed during
   T phase. Similar to previous studies, the onset of many QS‐dependent
   processes occurred during the late logarithmic to early stationary
   phase (Gao et al., [141]2015; Schuster, Lostroh, Ogi, & Greenberg,
   [142]2003). Thus, the regulatory function of the AHL‐mediated QS system
   is associated with the process of bacterial growth along with the
   specificity and timing (Schuster et al., [143]2003). This is consistent
   with the cell density‐dependent nature of QS, which can regulate the
   gene expression in response to the growth of bacterial cells (Keller &
   Surette, [144]2006). Recently, many studies have carried out
   transcriptome analysis of plant pathogens to identify QS‐regulated gene
   expression patterns and have revealed a large number of genes that are
   subject to QS regulation, such as 6.2% of coding sequences in the P.
   aeruginosa genome (Chugani et al., [145]2012), 19.6% of coding
   sequences in B. glumae BGR1 genome (Kim et al., [146]2013), 21.4% of
   coding sequences in B. gladioli BSR3 genome (Kim, Park, Choi, Kim, &
   Seo, [147]2014), and 11.5% of coding sequences in B. glumae PG1 genome
   (Gao et al., [148]2015). In those previous reports, each study focused
   on an individual organism under different growth conditions, so the
   amount of genes regulated by QS varied. In P. syringae 11528 genome,
   6,057 potential protein‐coding genes were identified by Studholme
   et al. ([149]2009); therefore, we inferred that ~18.5% of those genes
   were differentially expressed in an AHL‐mediated QS regulatory manner.
   This finding implies that QS system is a significant regulator on gene
   expression in P. syringae 11528.

   For the gene expression levels of those 1118 QS‐regulated genes, the
   most pronounced regulation was found for genes associated with
   bacterial flagella and chemotaxis (Table  [150]4), which was further
   confirmed by subsequent GO and pathway enrichment analysis
   (Figure [151]1). The bacterial chemotaxis and flagellar assembly
   pathways were enriched, among which flagellar assembly was completely
   down‐regulated, and this was consistent with the expression profiles of
   all genes involved in flagellar assembly in P. syringae 11528 wild‐type
   and ΔpsyI mutant strains (Figure [152]2). Bacterial flagella and
   chemotaxis play important roles in bacterial motility and plant
   colonization (Alavi et al., [153]2013; de Weert et al., [154]2002),
   processes that allow bacterial cells to seek out better survival
   environments (Aizawa, [155]2001). Swarming motility tests confirmed in
   part the RNA‐seq data, showing that the motility of the ΔpsyI mutant
   was significantly enhanced (Figure [156]3). In the observation of plant
   colonization, both P. syringae 11528 wild‐type and ΔpsyI mutant cells
   were mostly viewed in the glandular trichomes (Figure [157]4). One
   possible explanation for this observation is that the distribution of
   cells on leaf surfaces is approximately associated with the spatial
   heterogeneity of nutrients available on leaf surfaces (Leveau & Lindow,
   [158]2001), and glandular trichomes harbor an abundance of nutrients
   responsible for bacterial survival (Maggi et al., [159]2010). Similar
   observations have been obtained in other previous reports (Taiz &
   Zeiger, [160]1998). Motility is considered to a virulence factor and is
   vital for plant–pathogen interactions, because it helps pathogen to
   locate resources and promotes pathogen invasion into plant tissues
   (Quiñones et al., [161]2005). Thus, the ∆psyI mutant with more motility
   possessed a greater ability to access to nutritious sites and a larger
   population in glandular trichomes at 1 days after inoculation when
   compared to the wild type (Figure [162]5). A rapid multiplication of
   inoculum would occur on tobacco leaves under moist conditions (Quiñones
   et al., [163]2005). At 3 days after inoculation, a more significant
   difference of epiphytic population was observed between the ∆psyI
   mutant and the wild type (Figure [164]5). A direct correlation between
   robust motility and a large epiphytic population was obtained in the
   work of others (Haefele & Lindow, [165]1987; Lindow, Andersen, &
   Beattie, [166]1993), further indicating that motility contributes to
   epiphytic fitness of phytopathogen on plants. Therefore, the
   AHL‐dependent regulation in P. syringae 11528 does not affect
   distribution site but repress motility as well as epiphytic growth of
   this pathogen on plants, resulting in the negative influence of QS
   system on early plant colonization.

   More significant wild‐fire disease symptoms and larger lesion sizes
   were observed in tobacco leaves treated with P. syringae 11528 ΔpsyI
   mutant compared with those resulting from treatment with the wild type
   (Figure [167]6). There may be two possible explanations for an increase
   in the severity of disease incited by the ΔpsyI mutant. During
   tobacco‐P. syringae 11528 interactions, P. syringae 11528 acts both as
   a saprophyte that lives epiphytically on plant surfaces and as a
   pathogen that resides within the leaf apoplast (Hockett, Burch, &
   Lindow, [168]2013). These two states are closely related. The early
   epiphytic phase is critical for the later course of disease as it
   provides inoculum for subsequent infection (Hockett et al., [169]2013;
   Quiñones et al., [170]2005). A larger epiphytic population of the ΔpsyI
   mutant was observed, presumably resulting in a higher inoculum for
   later plant infections when compared to the wild type. As an
   alternative explanation, in addition to genes involved in motility,
   many genes were negatively regulated in an AHL‐dependent manner,
   including genes encoding pilus, EPS, secretion systems, and the
   two‐component system (Table  [171]4). Those genes were responsible for
   bacterial virulence. For example, pilus and EPS contribute to adherence
   to surfaces and are linked to biofilm formation (Alavi et al.,
   [172]2013); secretion systems, such as type II, type III, and type VI
   secretion systems, contribute to bacterial virulence (Bernard, Brunet,
   Gueguen, & Cascales, [173]2010; Preston et al., [174]2005; Studholme
   et al., [175]2009), in which the type III secretion system and its TTEs
   are required for pathogenesis in plants and play critical roles in
   plant–pathogen interactions (Studholme et al., [176]2009; Yang, Lee,
   Cha, & Baik, [177]2011); and the two‐component system has been reported
   to regulate the production of EPS as well as motility and is required
   for virulence in P. syringae (Marutani et al., [178]2008; Willis,
   Holmstadt, & Kinscherf, [179]2001). Summarily, the expression of
   various virulence factors was supppressed by AHL‐dependent QS,
   resulting in increased virulence of the ΔpsyI mutant. Thus, due to a
   larger epiphytic population and being more virulent, the ΔpsyI mutant
   caused more serious wild‐fire disease in inoculated tobacco leaves. To
   initiate pathogenesis, plant pathogenic bacteria must first enter plant
   tissues and colonize on plants (Melotto et al. [180]2006). Plant
   infection was a later stage of plant–pathogen interactions than plant
   colonization, since wild‐fire disease lesions incited by the ΔpsyI
   mutant were obvious only 5 days or more after inoculation. Hence,
   AHL‐dependent QS regulation in P. syringae 11528 may be involved in
   both early and late stages of disease development. Similar reports
   revealed that an AhlI‐AhlR QS system in P. syringae B728a regulated not
   only early water soaking but also late tissue maceration in bean plants
   (Quiñones et al., [181]2005), but AHL‐mediated QS in E. carotovora
   subsp. atroseptica (Pectobacterium atrosepticum) affected late plant
   infection rather than initial plant colonization (Smadja et al.,
   [182]2004).

   A high sequence identity (85%) is present between the PsyI in P.
   syringae 11528 and LuxI‐homologs (AhlI) in P. syringae B728a (Quiñones
   et al., [183]2004). Similar AHL‐mediated QS regulation on traits
   relevant to plant–pathogen interactions exists in these two pathovars,
   such as negative regulation of swarming motility and the formation of
   disease lesions, while there is an obvious divergence in regulation of
   EPS production, a down‐regulation in P. syringae 11528 but an
   up‐regulation in P. syringae B728a (Quiñones et al., [184]2005).
   Additionally, two virulence traits, iron transport as well as toxin,
   were postively controlled by QS system in P. syringae 11528, while QS
   regulation is also required for the production of exoenzymes involved
   in tissue maceration in P. syringae B728a (Quiñones et al., [185]2005).
   Moreover, Taguchi et al. ([186]2006) have demonstrated that a psyI
   deletion mutant of another P. syringae pathovars, P. syringae 6605,
   exhibited enhanced swarming motility and EPS production and decreased
   siderophore production, biofilm formation, and virulence against
   tobacco. It seems likely that different microbes possess distinct QS
   systems and lead to significant variations in the regulation of gene
   expression, or there may be other regulator(s) which control or
   coordinate QS system to regulate bacterial activities. In P. syringae
   B728a, AefR and GacA are separate regulators, both of which act as
   activators of AhlI‐dependent QS system via independent pathways
   (Quiñones et al., [187]2004). As proposed by Cha and collaborators,
   GacA regulon and iron regulon differently affect the AHL production,
   and both of them coordinately regulate virulence traits related to the
   pathogenesis of P. syringae 11528, and there may be an interaction
   between them (Cha, Lee, Lee, Oh, & Baik, [188]2012). Therefore,
   additional studies are required to further identify other regultors and
   characterize the interaction between them and PsyI‐dependent QS system
   in P. syringae 11528.

   In conclusion, our transcriptional analysis revealed that the most
   significant regulation of AHL‐mediated QS system on gene expression
   occurred during T phase in P. syringae 11528. A total of 1118
   QS‐regulated genes were identified, including numerous genes involved
   in pathogenicity on plants. Moreover, phenotypic assays revealed that
   QS‐dependent traits are involved in motility, epiphytic growth, and
   disease course of plants. These findings suggest that AHL‐regulated
   traits in P. syringae 11528 may be involved in both early plant
   colonization and late plant infection during plant–pathogen
   interactions. This study provides insights into the effects of
   QS‐dependent regulons on the processes of epiphytic growth and
   virulence of P. syringae11528 and extends our understanding of
   AHL‐mediated QS regulation, including positive and negative regulation,
   on plant–pathogen interactions.

CONFLICT OF INTEREST

   None declared.

Supporting information


   [189]Click here for additional data file.^ (291.5KB, pdf)

ACKNOWLEDGMENTS