Novel insights into the genetic basis and transcriptomic elucidation of virulence attenuation in Pseudomonas syringae pv. actinidiae variants

ABSTRACT

Pseudomonas syringae pv. actinidiae (Psa) is the causal agent of kiwifruit bacterial canker, but the factors affecting its pathogenicity in natural settings remain poorly explored. In this study, we isolated two Psa strains, G126 and G282, from infected kiwifruit orchards in Guizhou Province of China. Both isolates, categorized as Psa-biovar 3, were confirmed through Psa3-specific primers and phylogenomic analysis. Pathogenicity assays on kiwifruit cultivar “Hongyang” leaves and branches showed significantly reduced numbers of necrotic spots and reduced lesion sizes upon infection with G282 compared to the G1 positive control strain, while G126 showed a nonpathogenic phenotype. Additionally, both strains failed to induce a hypersensitive response in nonhost Nicotiana tabacum plants and exhibited significantly reduced promoter activity of the hrpR/S, hrpL, and hrpA genes, which are crucial for the type III secretion system (T3SS). Genomic sequencing revealed that the T3SS of G126 was defective due to a single-nucleotide polymorphism in the hrpR gene, while G282 was entirely deficient in the type VI secretion system (T6SS), which potentially regulates the expression of T3SS genes. Transcriptomic analysis showed widespread alterations in key aspects of the secretion system, protein transport, and signal transduction, further supporting the phenotype characteristics exhibited by the strains. This study enhances our understanding of the genetic basis of nonpathogenic and partially pathogenic Psa isolates, highlighting the functional interdependencies between T3SS and T6SS.

Introduction

Pseudomonas syringae pv. actinidiae (Psa) is the causal agent of bacterial canker in kiwifruit, a disease characterized by necrotic spots and lesions on leaves and twigs [1]. Infection by Psa compromises the host’s defense system, leading to significant agricultural losses. Central to Psa pathogenicity is the type III secretion system (T3SS), which enables the bacterium to deliver effector proteins into host cells. Like other type III secretion system (T3SS)-mediated infection-causing agents, the T3SS in Psa has been extensively studied as a model organism [2–5]. The T3SS is encoded by the hrp/hrc gene clusters and is a conserved virulence determinant among various plant pathogens, including species within Erwinia, Pseudomonas, and Ralstonia [6–9].

Regulation of the T3SS is a complex, multi-tiered process. At the heart of this regulatory network is the alternative sigma factor HrpL, a member of the extracytoplasmic function family sigma factor, which controls the expression of T3SS genes. HrpL-mediated regulation and coordination with other T3SS-encoding genes precisely control the expression and timely induction of associated genes in the pathway [10,11]. This gene, along with its alternate sigma factor RpoN (σ⁵⁴), interacts with a conserved “hrp-box” located in its corresponding gene promoter to regulate T3SS function [2,12–16]. To date, natural evolution or synthetic intervention-mediated interruption of the hrpL gene results in impaired hrpL transcriptional activity, leading to loss of pathogenicity in Psa [11]. In addition, enhancer binding proteins HrpR and HrpS function as positive regulatory factors for the hrpL promoter, but their mechanism of action has not been established. Some studies indicated that DDE (Asp-Asp-Glu) transposon insertion disrupted the key virulence regulator hrpR gene, and loss of Italian Psa3 recognition in tobacco and eggplant was due to hrpR and hrpS disruptions by ISPsy31 and ISPsy36, respectively [17–19]. However, Psa3 isolated from cultivated Actinida arguta in New Zealand was linked with recombination between IS630 transposases and conserved mobile elements, especially the Tn 6212-encoded LysR regulator, which has global effects on chromosomal gene expression [18,20]. In addition to genetic regulation, pathogenic microorganisms undergo virulence disparity due to spontaneous mutations [21,22]. Moreover, the factors driving virulence reduction remain unclear and may differ across various pathogens [23,24]. The loss of pathogenicity genes carried on mobile genetic elements, such as the tumor-inducing (Ti) plasmids in Agrobacterium tumefaciens, has been associated with the emergence of nonpathogenic strains during evolution [20,25]. A similar example has been documented in the nonpathogenic Psa isolate C17 [26].

Nevertheless, some cases of virulence loss in naturally occurring phytopathogens remain poorly documented and require further investigation [21]. The present study investigated the genetic cause of the complete and partial loss of pathogenicity of Psa strains G126 and G282, respectively. In addition, we revealed the response of both strains to genetic modifications impacting virulence-related pathways, specifically the T3SS and type VI secretion system (T6SS). Transcriptomic analyses helped explore genetic profiles and enriched KEGG pathways, elucidating molecular mechanisms underlying these pathogenicity disruptions. Our findings underscore the central role of the T3SS in the pathogenicity of Psa3, not only as a determinant of virulence but also as a critical factor in the pathogen’s adaptation to diverse environmental conditions. Our study reports the nature of two strains coexisting with reduced virulence together with virulent strains in mixed populations. The emergence of nonpathogenic variants, such as G126, sheds light on the evolutionary trajectories of syringae under selective pressures, balancing resource allocation between pathogenicity and survival.

Materials and methods

Bacterial strains and culture conditions

Pseudomonas syringae pv. actinidiae (Psa) strains were obtained from kiwifruit orchards located in Guizhou Province, China, as part of a field survey. The Psa3 strain G1 (GenBank accession ID: JAESNH01), used in this study, was derived from diseased trunk tissue of kiwifruit plants [19,27]. All these strains were present in symptomatic plants as a mixed population. Additionally, the nonpathogenic Psa3 strain G126 (GenBank accession ID: JBLLLG0) and the weakly pathogenic strain G282 (accession ID: JBLLLF0) were isolated from symptomatic leaf tissues of kiwifruit. For recombinant plasmid vector cloning and plasmid extraction, Escherichia coli (E. coli) DH5α was employed. To assess transcriptional expression in Psa3 strains, the expression plasmid pDSKGFPuv [28] was utilized. This same vector was also applied in the development of expression and luciferase reporter strains. All bacterial strains were first revived and cultured in Luria – Bertani (LB) medium. P. syringae strains were cultured in hrp derepressing (HDM) medium at 25 °C to assess transcriptional activity, while E. coli strains were grown in LB medium at 37 °C. Details of the constructed strains, along with their associated plasmids and primers, are provided in Table 1.

Table 1.

Strains, vectors, and primers used in this study.

Strains	Description	Reference (s)
Pseudomonas syringaepv. actinidiae
G1	Psa biovar 3 strain	[19]
G126	Psa biovar 3 strain	This study
G282	Psa biovar 3 strain	This study
G126 hrpRSC-T	Complimentary strain of G126	This study
Escherichia coli
DH5α	For cloning of recombinant plasmids
Plasmids
pDSK-GFPuv	Vector to construct luciferase reporter strains	[28]
pK18mobSacB	Vector for creating knockout strain	[29]
pDSK-PHrpRS:Nluc	Luciferase reporter vector of hrpRS gene promoter	This study
pDSK-PHrpL:Nluc	Luciferase reporter vector of hrpL gene promoter	This study
pDSK-HrpA:Nluc	Luciferase reporter vector of hrpA accumulation	This study
G1-PHrpRS:Nluc	Luciferase reporter vector of hrpRS gene promoter of G1	[19]
G1-PHrpL:Nluc	Luciferase reporter vector of hrpL gene promoter of G1	[19]
G1-HrpA:Nluc	Luciferase reporter vector of hrpA gene promoter of G1	[19]
G126-PHrpRS:Nluc	Luciferase reporter vector of hrpRS gene promoter of G126	This study
G126-PHrpL:Nluc	Luciferase reporter vector of hrpL gene promoter of G126	This study
G1126-HrpA:Nluc	Luciferase reporter vector of hrpA gene promoter of G126	This study
G282-PHrpRS:Nluc	Luciferase reporter vector of hrpRS gene promoter of G282	This study
G282-PHrpL:Nluc	Luciferase reporter vector of hrpL gene promoter of G282	This study
G282-HrpA:Nluc	Luciferase reporter vector of hrpA gene promoter of G282	This study
Primers	Sequences (5”−3”)
HrpR/S-C-F3	CCGGTACCTCGGCTTGGTATGATGATATTACT	This study
HrpR/S-C-R2	GCTCTAGAGGTTCCAGCGTCTTTGCA
M13F	GTTTTCCCAGTCACGAC	Lab stock
M13R	CAGGAAACAGCTATGAC
P0F	CTGCAACAGGCGACGGCGAGGC	[30]
P6R	CATAGGCTTCTGGTTTTCTTCCTGATCC
hrpL-F	AGCCGGGTTATGTTCGC	This study
hrpL-R	TTGAGTCGAGGATCACAATCT
hrpANLuc-F	GGTGGCTCTGGTGGCGGTGGCTCTGGTGGTATGGTCTTCACACTCGAAGATTTC	This study
hrpANLuc-R	TCAGAAATAATTTACGCCAGAATGCGTTCG

Plant material used

The kiwifruit cultivar Actinidia chinensis var. chinensis ’Hongyang’ was taxonomically identified by Professor Xia Liu of College of Landscape Architecture and Life Science, Chongqing University. A voucher specimen (voucher number: CUAS-Hymh01) has been deposited in the herbarium of the Chongqing University of Arts and Sciences Herbarium (acronym: HY1) for future reference. For pathogenicity assays, wound-based infections were performed on one-year-old branches (current-season growth) harvested from three-year-old plants of the kiwifruit cultivar Actinidia chinensis var. chinensis “Hongyang,” following protocols adapted from Zhao et al. [19]. These branches, typically 0.5–1 cm in diameter, were selected for their consistent susceptibility to Psa infection. Concurrently, young leaves from the same cultivar were used for pathogenicity evaluation via the leaf disc inoculation method. To investigate HR induction, experiments were performed on four-leaf-stage tobacco plants (Nicotiana tabacum).

Pathogenicity assessment and recording of bacterial counts

Pathogenicity testing was conducted using established protocols [19,31]. For wound inoculation assays, bacterial strains G1, G126, and G282 were grown under agitation (200 rpm) at 25°C for 16 h. Cells were pelleted via centrifugation (5,000 rpm, 5 min), washed three times with sterile water, and resuspended to a suitable optical density (OD₆₀₀nm = 0.2). The trimmed twigs (40 cm) of Hongyang were surface-sterilized with 0.5% sodium hypochlorite, rinsed with double-distilled water, and air-dried under sterile conditions. Based on the twig length, four suitable incisions (2 mm width × 1 mm depth) were made per shoot using an aseptic blade. Bacterial suspension (10 µL) was applied for treatment, and sterile water for controls to each wound. Twigs were placed in sterile trays containing moist filter paper, while sterile conditions were maintained by wrapping them with plastic sheets. Culture conditions were established as temperature 16°C with a total of 16 h with an interval of 8 h light and dark cycle and 95% humidity level. Lesion development was documented and imaged after 10 days.

For leaf disc assays, young kiwifruit leaves were sterilized (0.5% sodium hypochlorite, 3 min), rinsed, and cut into uniform 1 × 1 cm discs. Bacterial suspensions were introduced via vacuum infiltration (0.1 MPa, 15-s pulses × 3), followed by three sterile water rinses. Discs were placed on 0.5% water agar plates and sampled at 0-, 2-, and 7-days post-inoculation to assess in vivo bacterial proliferation.

Bacterial quantification involved homogenizing three leaf discs per replicate in 1 mL sterile water. Homogenates were serially diluted (10-fold increments), and 0.1 mL aliquots were spread onto LB agar plates using a sterile spreader. After 48-h incubation at 25°C, colony-forming units (CFUs) were enumerated.

The number of bacteria in the original sample was calculated using the following formula:

$$\mathrm{CFU}/\mathrm{mL}=\frac{\mathrm{Number}\mathrm{of}\mathrm{colonies}\mathrm{counted}}{\mathrm{Volume}\mathrm{of}\mathrm{sample}\mathrm{plated}\left(\mathrm{mL}\right)}\times \mathrm{Dilution}\mathrm{factor}$$

where CFU, an abbreviation for colony – forming units, represents the quantity of viable bacteria.

Hypersensitive response assays

The HR induction ability of G126, G126-hrpRS^C-T, G282 was compared to G1 positive control strain. All the bacterial strains were cultured in LB to a suitable optical density and pelleted to adjust OD600 nm = 0.2 and suspended in 10 mM MgCl2. Already established four-leaf staged tobacco (N. tabacum) plant was used. Tobacco leaves were blunt syringe infiltrated with serially diluted fivefold bacterial suspensions of the mentioned strains, reaching a penetration of 1 cm, and incubated for 28 h. Later, trypan blue (0.6 mg/mL) staining was performed at 95 °C for 5 min and washed with chloral hydrate (2.5 mg/mL) consecutively after 2 h until complete decolorization, and leaves were then snapped.

Construction of luciferase reporter strains and luciferase assay

Gene-specific primers (Table 1) were designed using Primer Premier 5.0 to amplify target sequences, with genomic DNA from Psa strain G1 serving as the template. The open reading frames (ORFs) of hrpR/S were amplified via PCR using primer pair HrpR/S-C-F3/R2 (Table 1). Amplification products were resolved by 1% agarose gel electrophoresis and validated by restriction digestion with EcoRI/PstI. Concurrently, the plasmid vector pDSKGFPuv was linearized with EcoRI/PstI, and the hrpR/S fragments were ligated into the vector using T4 DNA ligase (16°C, 4 h). The recombinant plasmid was electroporated into strain G126, and transformants were selected on King’s B (KB) agar plates containing 50 μg/mL kanamycin (25°C, 48 h). Putative positive colonies were verified by PCR amplification with primers M13F/R (Table 1), followed by Sanger sequencing to confirm vector assembly and generate the complemented strain G126-pHrpR/S. Using analogous cloning strategies, hrpA and hrpL reporter constructs were generated in strains G126 and G282 with gene-specific primers (Table 1).

Later, competent cells of tested strains were prepared after culturing them in LB broth and washed with sterile water. The successful vector constructs of luciferase reporters of Psa G1, G126, and G282 were electroporated and selected on LB agar plates containing suitable kanamycin (50 µg/ml) resistance. Positive transformants were confirmed by PCR and cultured to OD_600 nm = 0.4. Later, liquid KB media as T3SS inhibiting and HDM for T3SS induction was inoculated with 50 µL inoculum of each strain to measure the luciferase activity. Luciferase expression was detected with a chemiluminescence detector (GloMax NAVIGATOR, Promega, Madison, WI, USA).

Genomic DNA extraction, sequencing and genomic analysis

For genomic DNA isolation, fresh colonies were streaked from stock and revived each strain onto LB agar plates and incubated at 25°C for 48 h. Strains G1, G126, and G282 were cultured in LB broth at 25°C under agitation (200 rpm) for 16 h. Cells were harvested by centrifugation (5,000 rpm, 5 min), subjected to three successive sterile water washes, and resuspended in 200 μL sterile water. Bacterial cells were pelleted, and genomic DNA was extracted using a commercial bacterial genomic DNA extraction kit. DNA quality and purity were assessed via capillary electrophoresis (Bioanalyzer 2100, Agilent Technologies) and nanodrop spectrophotometry. The extracted DNA was used for constructing a 250 bp small-fragment library, followed by Illumina sequencing with a data output of 1 Gb. Quality control and filtering of the raw reads were performed via Fastp software . The clean reads were de novo assembled with the Unicycler software [32]. Assembled genomes were annotated using Bakta to capture comprehensive gene predictions and functional annotations and identify virulence factors, including T3SS effectors and T6SS components. Assembled genomes were aligned to the (P. syringae pv. actinidiae Shaanxi_M228; GenBank accession ID: CP032631.1) via Bowtie 2 software [33,34]. The clean reads were de novo assembled with the Unicycler software [32]. Assembled genomes were annotated using Bakta [35]. A total of 102 Psa reference genomes were selected for constructing the phylogenetic tree with Parsnp software [36]. The nonpathogenic G126 was closely related to the virulent reference strain G48 [37], M256 (WGS accession: MDXH01) [19], and P220 (alternative strain name P115, Genome accession no.: CP032870, CP032871) [38]. The low-virulent strain G282 was closely related to the virulent reference strain YXH1 (CP136506, CP136507) [39] and G35 (WGS accession: JAESNF01) [37]. We then compare the gene contexts within each groups (group1: G126, G48, and M256; group 2: G282, YXH1, and G35) with Roary pipeline [40]. Type 3 effectors were identified by the Blast search against the PsyTEC effector protein database of Pseudomonas [41,42] and machine-learning-based prediction with Effectidor II software [43]. The distribution of effector genes in Psa3 genome was determined by Blast search with E values under 10⁻⁵ and query coverage exceeding 60%. The effector protein was considered present if the best Blast hit with over 98% similarity and 99% query coverage, and it was considered absent if no hits were found. If the alignment cannot meet the criteria of “present,” the putative effector was manually checked for potential frameshift and truncation, and recorded as an “incomplete” effector. The hrp/hrc clusters were aligned with the Mauve version 2.4.0 [44].

For T6SS cluster comparisons, indicated strains were compared to the reference using Mauve or MUMmer to confirm known variations. For SNP detection, the indicated strain’s T3SS cluster was compared. The alignment results were visualized via the Integrative Genomics Viewer (IGV) Genome Browser [45].

RNA extraction and transcriptomic sequencing

Strains G1, G126, and G282 were grown as mentioned earlier in T3SS-inducing condition for 6 h. The total RNA was extracted using the commercial RNA Mini Kit (Thermo Scientific) and stabilized using RNA Protect Bacteria Reagent (QIAGEN). RNA quality was assayed on a “Bioanalyzer 2100 system” (Agilent Technologies). Commercially available kits were used to prepare the first and second strand cDNA, and residual activity of RNA was resolved using RNase-H. After all the processes, second-strand cDNA was generated by the cDNA Synthesis Kit (Invitrogen, A48570). Illumina adapters were ligated to synthesize the sequencing libraries to size-selected cDNA fragments (370–420 bp) by the AMPure XP system (Beckman Coulter). Paired-end sequencing (150 bp reads) was performed on an Illumina NovaSeq platform by a commercial sequencing provider (Novogene Co., Ltd., Beijing, China), prior to library quality validation using the Bioanalyzer 2100 system. The obtained raw reads were processed by Trimmomatic v0.36 to remove low-quality bases and adapters. Clean reads were aligned to reference genome (M228) using Bowtie2 v2.2.3. Gene expression quantification (read counts per gene) was performed with HTSeq v0.6.1, and normalized expression values (FPKM: fragments per kilobase of transcript per million mapped reads) were calculated. Differential gene expression analysis was conducted using DESeq2 (v1.18.0) with significance thresholds set at an adjusted p-value < 0.05. Gene Ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses of differentially expressed genes (DEGs) were performed using GOSeq (R package) and KOBAS v2.0 [46], respectively.

Statistical analysis

GraphPad Prism 9.2.0 (GraphPad Software Inc., USA) was used for graphing and statistical analysis.

Results

Potential Psa3 candidates show reduced and decreased pathogenicity

In our previous study, Pseudomonas syringae pv. actinidiae (Psa) biovar 3 strain G1 was isolated from symptomatic kiwifruit leaf tissues and characterized as highly virulent [11]. From that same composite sample, two distinct strains, G126 and G282, were subsequently isolated, both exhibiting reduced virulence compared to G1 within the Psa biovar 3 population. Two Psa biovar 3-specific primers Psa-3F/R [47] and P0F/P6R were used to confirm the lineage-specific identity of G126 and G282 as Psa biovar 3 (Figure 1(A)). Phylogenomic analysis of whole genome sequences further resolved G126 and G282 within the Psa biovar 3 clade. The cluster of core genes analysis based phylogeny indicated G126, grouped with G48, and M256; G282 grouped with YXH1, and G35; all isolated from Guizhou, China, while G1 clustered with highly virulent clade biovar 3-Peach (Figure 1(B)). To contextualize their evolutionary relationships, phylogenomic analysis of global Psa strains using Parsnp software revealed clustering patterns correlating with geographical origin, and all analyzed strains grouped into their specific clade (biovar-I to VI), underscoring their regional lineage (Figure S1).

Figure 1.

Psa3 primers specific identification of G126 and G282 and their phylogenetic tree representing that both strains shared a biovar 3 clade. (A) gel electrophoresis of G1, G126 and G282 PCR product amplified by 2 Psa3 specific primers. (B) phylogenomic analysis of Psa strains with Parsnp software. Each tip shows the WGS accession, stain name, and geographical location of the reference genome. The gene contexts within each groups (group 1: G126, G48, and M256; group 2: G282, YXH1, and G35) indicate difference in virulence and highly virulent clade is represented in peach color.

To evaluate their reduced pathogenicity, G126 and G282 were infiltrated into leaves of the Hongyang kiwifruit cultivar, with G1 serving as a positive control. At 10 days post-inoculation (dpi), G1-infiltrated leaf discs exhibited extensive brown necrotic lesions, whereas G282 and G126 displayed minimal or no necrosis, respectively (Figure 2(A)). These findings indicate that while both strains exhibit significantly reduced virulence relative to G1, G282 retains slight pathogenicity compared to the nonpathogenic phenotype of G126 in the Hongyang cultivar.

Figure 2.

Determining the cause of loss of virulence in two Psa-biovar 3 isolates, and bacterial counts were recorded. (A) three variable virulence types of Psa-biovar 3 (Psa3) were infiltrated into the sterilized “Hongyang” leaves under vacuum and placed in 0.5% water agar medium in triplicate. Photographs were taken at 5 dpi. G1 showed strong disease development as compared to less pathogenic strain G282 and G126 show loss of pathogenicity. G126 virulence was quite like water treatment. (B) leaves containing G1, G282 and G126 were collected at three different time points (0 days post inoculation (dpi), 2 and 7dpi), and bacterial counts were recorded. At 0 dpi, G1 and G282 population surpassed 2-fold higher than that of G126 which loss virulence. (C) in vitro growth dynamics of Psa-biovar 3 strains G1, G282, and G126 in LB liquid medium. Growth was monitored by optical density (OD_600 nm) over time, with time to lag phase indicated. Data represent means ± SD from three independent replicates, (D-E) the nano- luciferase (NLuc)- based reporter plasmid, pDSK-hrpR/S-NLuc, pDSK-hrpL-NLuc, pDSK-hrpA-NLuc, was constructed and transformed into G1, G126 and G282. For NLuc quantification, the reporter strains were cultured in the hrp-derepressing medium (HDM) for 24 h (OD_600 nm = 0.1) and the NLuc intensity was quantified using a GLOMAX luminometer (Promega). Student’s t test or ANOVA analysis was employed and different letters indicate significant differences at p < 0.05.

Bacterial population counts during pathogenicity assays revealed distinct growth dynamics of the tested strains in plant tissues, correlating with their virulence potential. The analysis of bacterial population dynamics demonstrated significant differences in viable cell recovery at 2- and 7-days post-infiltration (dpi). The population of G282 was significantly reduced compared to G1 (p < 0.05) but higher than that of G126 (Figure 2(B)), indicating partial proliferative capacity in G282 despite its attenuated virulence relative to the control strain. In contrast, G126 exhibited markedly slower growth, with bacterial counts consistently lower than both G1 and G282 (Figure 2 B-C). The in-planta reduced bacterial population dynamics in G126 aligns with its severely impaired virulence, while the slow growth in rich media independent of T3SS functional activity is related to a potential metabolic mechanism substantially affected due to virulence impairment, particularly the T3SS pathway.

To assess the transcription expression of key virulence-related genes, including hrpA, hrpR/S, and hrpL, which are crucial for the T3SS cascade. Nanoluciferase (NLuc)-based reporter plasmids containing the corresponding promoters from G1, G126, and G282 were introduced into their respective strains. After culturing them in hrp-derepressing medium for 6 h, the NLuc-based transcriptional activity was measured. Statistical analysis revealed that hrpL and hrpR/S promoter activity in G282 was significantly reduced compared to G1 but higher than in G126 (Figure 2(D)), suggesting attenuated, though not abolished, pathogenicity in G282 relative to the severely impaired G126. Similarly, a comparative analysis of hrpA promoter activity in G282-pDSK-hrpA and G126-pDSK-hrpA showed strong repression compared to the G1 control (Figure 2(E)). Collectively, the diminished activity of hrpR/S and hrpL promoters in each strain implies disruption of the HrpR/S-HrpL regulatory cascade, a key hierarchical pathway governing pathogenicity, which likely contributes to the strain’s reduced virulence.

G126 was functionally impaired in hrpR due to an SNP, rendering it nonvirulent

The pathogenicity defect observed in G126 was not limited to kiwifruit leaves. To further assess this defect, both the G126 strain and the G1 strain were tested on Hongyang kiwifruit branches. After 10 days, G126 caused no necrotic spots, whereas the G1 strain presented visible lesions. Furthermore, the measured length of the necrotic spots induced by G126 was negligible, indicating a significant reduction (100%) in disease development compared with that of the G1 positive control (Figure 3(A-D)). This lack of disease symptoms in G126 strongly suggests that it is impaired in pathogenicity. As Psa3 relies on a functional T3SS to secrete effectors and induce a hypersensitive response (HR), we tested the ability of Psa3-G126 to induce an HR in Nicotiana tabacum leaves via bacterial infiltration and trypan blue staining. In contrast to the highly virulent G1 strain, G126 was unable to induce HR at any of the tested concentrations, further supporting the notion of a nonfunctional T3SS in G126 (Figure 3(C)).

Figure 3.

Determining the cause of reduction in pathogenicity of Psa3 strain G126 (A) G1 and G126 were vacuum infiltrated in sterilized Kiwi branches, and 10 µL of bacterial solution was dropped onto the wound, with sterile water as a control. Branches were placed at 16 °C for 20 days after inoculation, photographs were taken at 10 dpi, and (B) lesion density was measured, and applied the students-t test to find the significance, (C) the bacterial suspension of strains G1, G126 and 10 mM MgCl₂ (OD_600 nm 0.2) was gradient dilution 5 times and infiltrated into tobacco leaves with a blunt-ended plastic syringe. After 28 hpi, the necrotic area was observed by trypan blue staining which indicate the loss of HR in G126 as compared to G1. (D) the strain G126 was sequenced and aligned to positive control G1, and single nucleotide polymorphism (SNP) was found at the hrpR gene at the position of 10^th amino acid “R (arginine)” where C was mutated to T, (E), G126 and its plasmid borne complimentary strain with recovered base mutation of T to C. G1 as positive control and G126 wild and complimentary strain with base “C” insertion in HrpR/S strain were inoculated on tobacco leave for HR test, G126 hrpR/S recovered HR expression. Student’s t test or ANOVA analysis was employed and different letters indicate significant differences at p < 0.05. (F-G) G126 and its plasmid borne complimentary strain (G126^hrpR/S) with recovered base mutation of T to C. G1 as positive control and G126 wild and complimentary strain with base “C” insertion in HrpR/S strain were tested on kiwifruit leaves indicating expression of hrpRS in G126 restored virulence. Student’s t test or ANOVA analysis was employed and different letters indicate significant differences at p < 0.05.

To investigate the molecular basis for the loss of hrpL and hrpA expression in G126, genomic reads obtained via Illumina paired-end sequencing and mapped to the reference as well as genomic reads of the G1 strain mapped, yielding an average read depth of 137.23. After comparing G126 with G1 control, no genetic variations were detected in the genes associated with hrpL expression or its regulatory pathway. However, a single-base “T” was identified in the promoter region of hrpR gene in G126 as a single nucleotide polymorphism (SNP) event as compared to G1 containing base “C,” which affected the 10^th amino acid R (arginine) (Figure 3(D)). To confirm the functional impact of this SNP, we cloned the native hrpR/S genes from G1 into the expression plasmid pDSK-GFPuv, resulting in the construction of the strain G126-hrpR/S^C-T. We then tested the ability of plasmid-borne hrpR/S from G1 to restore HR induction in tobacco leaves and virulence in kiwifruit leaves. Reintroducing the one base to the hrpR/S^C-Tto G126-hrpR/S^C-TG1 hrpR/S in G126 restored HR induction and virulence, demonstrating that the T insertion in hrpR disrupted the ability of G126 to induce HR and virulence (Figure 3(E-G)). These results indicate that the pathogenicity defect in G126 is due to a functional deficiency in hrpR, particularly in the region of the 10th amino acid, and that the restoration of the original amino acid sequence rescues HR induction and pathogenicity.

Pangenomic analysis of Psa 3 genomes reveals genetic variations linked to strain specific virulence-associated loci

To elucidate the genetic determinants underlying virulence attenuation, we conducted a pangenome analysis of Psa biovar 3 strains, including G1, G126, G282, and related global isolates. Roary-based pangenome matrices classified G126 and G282 into two distinct clades (Group 1: G126, G48, M256; Group 2: G282, YXH1, G35) based on gene presence-absence patterns (Figure S2, Tables S1-S2). Comparative analysis revealed 11 unique accessory genes and 29 gene losses in G126 relative to Group 1 strains (G48, M256), and 7 unique genes versus 153 absent loci compared to Group 2 (YXH1, G35) (Figure S2A-B). Notably, YXH1 harbored 586 strain-specific genes, predominantly transposases, which were identifiable in its complete genome assembly but absent in draft-quality genomes of other strains (Figure S2B). Focusing on type III secretion system (T3SS) components and effectors (T3Es), comparative genomics identified key differences between strains. All strains lacked functional hopAB1b (Cluster 1) and a conserved effector locus (CEL) containing hopAA1d (Figure S3A). Strikingly, G282 uniquely lacked hopBN1a, a T3E retained in all other strains. Five critical genetic variations were identified. For instance, SNPs in noncoding regions flanking P220 (putative regulatory sequences). A 16 bp deletion in hopAA1–1 (P220), truncating this effector. An additional SNP in a P220-associated noncoding region. A nonsense SNP in hrpR (G126), resulting in a truncated HrpR protein critical for T3SS regulation (Figure 3(D), S3B). A 280 bp deletion between hrpV and hrcU (G126), disrupting operon architecture in the T3SS gene cluster. These variations, particularly the truncated HrpR and disrupted hrpV-hrcU locus in G126, correlate with its severe T3SS dysfunction and nonpathogenic phenotype. The partial retention of T3SS activity in G282 aligns with its residual virulence, underscoring the hierarchical dependence of pathogenicity on intact regulatory and structural components.

Impact of T6SS deletion on pathogenicity in G282

To further investigate the role of T6SS in the pathogenicity of G282, we conducted a kiwifruit branch assay. Similar to the leaf disc assay, G282 induced fewer necrotic spots on the branches compared to the Psa3-G1 positive control strain. The length of the necrotic lesions was significantly reduced by 40% in G282, providing strong evidence of its diminished pathogenicity relative to G1 (Figure 4(A-B)). These findings suggest that G282 has a reduced ability to cause disease in both leaves and branches of the Hongyang kiwifruit cultivar. Additionally, no hypersensitive response (HR) was observed in non-host tobacco leaves inoculated with G282, in contrast to the HR induced by the G1 strain (Figure 4(C)).

Figure 4.

Pathogenicity, HR and comparative genomic analysis of Psa3 isolate G282. Both strain were vacuum infiltrated in sterilized Kiwi branches, and 10 µL of bacterial solution was dropped onto the wound, with sterile water as a control and placed at 16 °C, photographs were taken at 10 dpi, and (B) lesion density was measured, and applied the students-t test to find the significance, (C) the bacterial suspension of strains G1 and G282 in 10 mM MgCl₂ (OD_600 nm 0.2) was gradient dilution 5 times and infiltrated into tobacco leaves with a blunt-ended plastic syringe. After 28 hpi, the necrotic area was observed by trypan blue staining which indicate the loss of HR in G282 as compared to G1 (D) the strain G282 was aligned using mauve alignment with G1 and Psa3 genome M228 to indicate the deletion of T6SS cluster in G282. (E) transcriptome gene expression of type 6 secretion system of G126, and G282 compared to control G1 strain. Student’s t test or ANOVA analysis was employed and different letters indicate significant differences at p < 0.05.

Previous studies have demonstrated that deletion of the Type VI Secretion System (T6SS) gene cluster in Pseudomonas syringae pv. actinidiae (Psa) biovar 3 attenuates the expression of Type III Secretion System (T3SS)-related genes, resulting in reduced virulence [39,48]. Notably, the attenuated T3SS activity and hypovirulent phenotype of G282 are somewhat similar to each other. To investigate the molecular basis for the diminished expression of hrpL and hrpA in G282, we performed whole-genome sequencing and aligned the de novo assembled genome of G282 to reference strains P220, G126, and YXH1 using Mauve (v2.4.0 [44];). Comparative genomic analysis revealed the complete absence of the T6SS cluster in G282, whereas this locus was intact in other strains (Figure S4).

Despite the absence of mutations in core T3SS regulatory genes (hrpR, hrpS, hrpA, hrpL; Figure S3B), Mauve-based alignment against the reference genome M228 identified a large-scale deletion encompassing the entire T6SS locus in G282 (Figure 4(D); Figure S4). Transcriptional profiling further confirmed the complete lack of T6SS-associated gene expression (tssM, tssG, tssH) in G282, contrasting with robust expression in G1 and G126 (Figure 4(E)). These findings suggest that the natural deletion of the T6SS cluster in G282 disrupts cross-regulation between secretion systems, leading to suppressed T3SS activity and attenuated pathogenicity.

Transcriptomic insights into G126 and G282

In addition to providing genetic evidence of a nonfunctional T3SS in G126 and a partial effect on the T3SS pathway in G282 due to omission or deletion of the whole T6SS cluster, we performed transcriptomic analyses of both isolates and compared them with the G1 positive control. Among all the genes transcribed, 525 and 604 were upregulated, and 399 and 579 genes were downregulated in both strains as shown in volcano plot (Figure 5A,B). A comparative analysis revealed several key aspects; most importantly, the secretion system, protein transport and signal transduction were affected due to SNP event in hrpR (Figure 5C,D). Gene Ontology (GO) analysis revealed that the major GO pathways included protein transport, secretion, peptide transport, amide transport, peptide secretion, protein secretion, and secretion by cells, which are essential biological processes affected in G126 (Figure 5(C)). Moreover, genes related to the outer membrane were enriched in cellular components and intracellular nonmembrane-bound organelles, and the structural constituent of the ribosome, an indicator of molecular function related to transcription, was influenced in G126 (Figure 5(C)). In contrast, the GO pathways of locomotion, chemotaxis, and signal transduction as biological processes and signal transduction activity genes as molecular functions were highly enriched and influenced in G282 (Figure 5(D)). This indicate that both strains with variable pathogenicity showed unique pathways affected during infection.

Figure 5.

Transcriptomic insights into gene expression and key biological processes, cellular components and molecular functions predicted via transcriptomic data of G126 and G282 analysed by GO, KEGG analysis. (A-B) volcano plot of differentially expressed genes in G126 and G282 compared to G1 strain. (C-D). GO analysis of G126 and G282 compared to control G1 strain. Encircled lines indicate the important molecular functions associated genes affected in G126 and G282.

We then analyzed the highly enriched KEGG pathways upregulated in G126 compared with those in G1, with significant adjusted p values. It indicates that flagellar assembly and bacterial chemotaxis associated with motility and pathogenesis, and essential for T3SS functionality, were influenced due to SNP variation, rendering the strain nonpathogenic. Additionally, certain pathways associated with metabolism in diverse environments and lysine degradation reflect metabolic adaptations due to hrpR gene deficiency. The results of the citrate cycle (TCA cycle), amino sugar and nucleotide sugar metabolism, and glycolysis/gluconeogenesis indicate shifts in central metabolic pathways, likely due to changes in energy demands from disrupted T3SS activity. Tyrosine metabolism and the biosynthesis of secondary metabolite pathways may be indirectly affected, suggesting metabolic rewiring under stress. These findings suggest that T3SS disruption impacts the expression of genes critical for motility (Figure 6(A)).

Figure 6.

Number of genes associated to KEGG pathways upregulated and downregulated in G126 and G282 predicted based on transcriptomic data using KEGG analysis of gene expression reads. (A-B) indicated the upregulated genes in both strains, (C-D) indicates the downregulated pathways that are statistically significant. (E) the calculated gene expressions (log FC) of G126 and G282 after transcriptome analysis of T3SS genes.

Similarly, the most enriched KEGG pathways in G282 were very similar to G126, reflecting significant alterations in virulence-related motility functions due to T6SS deletion. Additional pathways than G126 were identified in G282, the two-component system that is essential in regulating the T3SS system, which is highly affected, possibly linking T6SS deletion with signaling and response mechanisms. Moreover, important metabolic pathways, including starch and sucrose metabolism and fructose and mannose metabolism, are altered similar to G126 in G282, reflecting metabolic reprogramming in response to T6SS deletion. Amino sugar and nucleotide sugar metabolism was similarly enriched, showing shared metabolic disruptions with G126, but the biofilm-forming pathways were unique in G282 compared to G126 (Figure 6(B)).

In G126, enriched KEGG pathways that were downregulated, including the ribosome, indicated that increased protein synthesis was affected, possibly compensating for hrpR gene deficiency. Moreover, the beta-lactam resistance pathway was indirectly influenced, indicating the downregulation of resistance mechanism activation, possibly due to stress responses. Lysine biosynthesis and peptidoglycan biosynthesis are essential for cell wall integrity and bacterial survival and are potentially linked to a strain’s metabolic adaptations. Therefore, strain g126 grew more slowly than the other strains did (Figure 6(C)). However, G282 revealed a highly downregulated pathway of the bacterial secretion system, likely due to T6SS deletion, which directly impacts secretion-related genes and processes. Additionally, quorum-sensing disruption indicates changes in cell-to-cell communication, a potential downstream effect of T6SS deletion. Fatty acid biosynthesis and metabolism: These pathways highlight shifts in lipid metabolism, which could be related to altered membrane structures or signaling molecules (Figure 6(D)).

In G126, the hrpR gene mutation disrupts arginine functionality, which appears to have downstream effects on flagellar assembly and T3SS-related pathways (e.g. chemotaxis). The enrichment of metabolic pathways (e.g. amino acid and nucleotide sugar metabolism) reflects the strain’s effort to compensate for impaired T3SS function. In G282, T6SS deletion significantly impacts pathways such as bacterial secretion systems, quorum sensing, and flagellar assembly, emphasizing its role in virulence. The simultaneous alteration of central metabolic pathways suggests a broader impact on bacterial fitness and adaptability. The G126 strain is deficient in the hrpR gene because of a single-nucleotide polymorphism (SNP), where a cytosine (C) is substituted with thymine (T), leading to an amino acid change in arginine, and significant downregulation of the hrpR and hrpS genes is observed, indicating impaired transcriptional regulation of the T3SS cascade. Correspondingly, hrpA and hrpL, which are essential for the structural and functional components of the T3SS, were moderately downregulated. These changes suggest that the hrpR mutation disrupts the regulatory network, compromising the functional ability of the T3SS in G126. These reduced expression levels suggest that disruption of the regulatory network drives T3SS activation. Similarly, the expression of hrpA, which encodes the structural pilus protein, and hrpL, a master regulator, is altered, underscoring the cascade’s sensitivity to hrpR mutations. The transcriptomic gene expression data for G126 and G282 closely matched their phenotypic data (Figure 6(E)). These findings suggest a compensatory response or cross-regulation between secretion systems.

In contrast, the G282 strain lacks the T6SS entirely. Both genetic alterations are reflected in the expression profiles of T3SS-associated genes, including hrpA, hrpL, hrpR, and hrpS, which are crucial regulators of the secretion system (Figure 6(E)). T6SS deletion results in a mild downregulation of T3SS gene such as hrpA and hrpL show marginal expression changes in G282 (Figure 6(E)). These observations indicate that the absence of a T6SS has a limited effect on T3SS functionality. The relatively stable expression of the hrp/hrc genes suggests that G282 retains partial or near-normal T3SS function, possibly due to minimal cross-regulation between the T6SS and T3SS. These contrasting effects highlight the critical role of hrpR in maintaining T3SS activity and the limited influence of T6SS deletion on T3SS gene expression. Overall, these RNA sequencing data underscore the critical role of the hrp/hrc gene cluster in the T3SS and highlight how genetic disruptions, such as hrpR mutation or T6SS deletion, impact the molecular machinery responsible for bacterial pathogenesis. Our results highlight the functional interdependencies between secretion systems (T3SSs and T6SSs) and metabolic or motility-related pathways, revealing how genetic mutations lead to widespread physiological changes.

Discussion

The current study investigated the functional and metabolic adaptations of Psa3 strains G126 and G282 in response to genetic modifications impacting virulence-related pathways, specifically the T3SS and T6SS. Using genetic and transcriptomic analyses, we elucidated the molecular mechanisms underlying T3SS regulation in G126 and G282. Though belonging to biovar 3 clade, G126 showed severely reduced pathogenicity and G282 displayed intermediate virulence (Figure 2A-B). In Psa3 pathovars, this is a compelling phenomenon less likely described as a functional deficiency of the T3SS, driven by transposon insertions in hrpR/S genes [11,17,37]. In a current study, T3SS dependent regulatory cascade mediated pathogenesis in G126 and G282 was significantly impacted in in-vitro plant experiment leading to the lowered expression levels of the hrpR/S, hrpL, and hrpA promoters of both strains compared to the virulent strain G1 (Figure 2). HrpR is a transcriptional regulator that modulates HrpS gene activity, subsequently activating the transcriptional activity of other Hrp genes through the action of hrpL [49–51]. hrpL is a σ⁵⁴-dependent transcription regulator of Psa3 and is regulated primarily by the heterodimer of the enhancer-binding proteins HrpR and HrpS proteins [52]. Genomic sequencing and mapping enabled us to identify the SNP in the hrpR gene specifically targeting the 10^th amino acid in the promoter region in G126 strain. Naturally emerged SNP in hrpR resulted in blocking the normal functionality of hrpL regulator and other T3SS cascade genes characterized as defective hrpL, hrpA expression (Figure 3). These findings indicate that hrpR deficiency can block hrpRS to the hrpA pathway in the T3SS cascade that can significantly impair T3SS functionality, aligned to recent study [11].

Another plausible explanation for the emergence of nonpathogenic variants lies in the resource allocation trade-offs faced by plant-pathogenic bacteria. Pathogenicity traits such as T3SS expression impose significant metabolic costs, making their maintenance challenging under certain conditions [53,54]. Similar phenomena have been observed in other pathogens, such as R. solanacearum, where mutations in the quorum-sensing regulator PhcA resulted in nonpathogenic variants with increased metabolism [53]. A further supporting hypothesis was put forward for Psa3, where natural recovery of nonpathogenic variants to a pathogenic form was observed in-planta multiplication in R. solanacearum [55]. The enrichment of pathways such as flagellar assembly and bacterial chemotaxis suggested that the SNP mediated evolution in G126 resulted in compromised motility, a critical function for host colonization and pathogenicity. The decreased expression of T3SS-related genes in G126 corroborates the diminished secretion system functionality, further supported by the downregulation of key genes in these pathways. Some other studies show alike results targeting in a different way. For example, the transposable element ISPsy32 was found to bind to noncoding sequences upstream of the hrpR gene in isolates G25 and M227, disrupting T3SS functionality [37]. Similarly, previous studies in Italy reported HR-negative isolates caused by transposition events involving ISPsy36 [17]. Interestingly, although these transposition events lead to reduced pathogenicity, TE-associated genes presented low transcription levels during infection [56], which aligns with observations that nonpathogenic isolates are detected infrequently during Psa3 isolation. The pangenomic analysis revealed that indicated strains retain conserved virulence loci like the hopAA1d effector (CEL), and their T3SS functionality is differentially impaired through distinct genetic lesions. In G126, a nonsense mutation in hrpR (Figure 3(D)) and deletion between hrpV-hrcU (Fig. S3B) disrupt the HrpRS-L hrp regulatory cascade, completely abolishing T3SS activity despite intact effector genes. In contrast, G282 maintains functional T3SS regulators but shows attenuated virulence due to deletion of the only present T6SS cluster (Figure 4(D)), suggesting this secretion system contributes to T3SS activation, possibly through interbacterial competition or stress response modulation. The universal retention of hopAA1d (CEL) across all strains, including nonpathogenic G126, implies that this effector may serve functions beyond virulence. This explains the absence of CEL effector expression and nonpathogenicity despite genetic conservation. In contrast, G282 retained an intact HrpR/S-L cascade but suffered deletion of T6SS-1, which indirectly suppresses T3SS activity by disrupting redox balance and epiphytic competition [48]. The residual virulence in G282 aligns with the leaky expression of CEL effectors, whereas the loss of accessory effectors (e.g. hopBN1a) appears noncritical. Thus, virulence attenuation in Psa3 can stem from either direct regulatory collapse (G126) or compromised secretion system crosstalk (G282).

Interestingly, the enrichment of metabolic pathways in both strains underscores the critical interplay between secretion systems and core metabolic networks. While G126 had a pronounced effect on ribosomal and cell wall biosynthesis pathways, G282 had stronger perturbations in lipid and fatty acid metabolism, suggesting strain-specific compensatory mechanisms in response to T3SS and T6SS deficiencies, respectively. Disrupting these systems could serve as a potential strategy for controlling bacterial pathogens. For example, targeting key regulators such as hrpR or components of the T6SS machinery may weaken the ability of a pathogen to colonize and infect host plants. Furthermore, the observed metabolic shifts suggest potential vulnerabilities that could be exploited for antimicrobial interventions. Disruption of central metabolic pathways or quorum sensing could synergize with existing strategies to mitigate the pathogenicity of Psa. In addition, we identified enriched KEGG pathways to construct DEGs whose trajectory explained the expression of the T3SS and T6SS in G126 and G282, which demonstrated that T3SS genes were downregulated in the former strain, whereas no expression of the T6SS was evident in the latter strain (Figure S5). We were able to identify only one T6SS cluster that is essentially involved in modulating pathogenicity linked to the T3SS cascade. While few studies have explored these evolutionary trajectories [57], the emergence of Psa3 in China during the 1980s, which coincided with the domestication of kiwifruit, presents a unique case for examining pathogen evolution in response to agricultural intensification [19,26].

In symptomatic and non-symptomatic environments, nonpathogenic Psa3 isolates may interact with virulent strains in ways that influence disease progression. Conversely, T3SS-deficient variants may trigger PAMP-triggered immunity, which could hinder the spread of virulent strains [58,59]. However, in nonhost environments, less aggressive P. syringae variants present a fitness advantage and may persist in agricultural ecosystems [58]. Thus, finding the root cause of those nonpathogenic Psa3 variants may lead to pinpointing an adaptive strategy for designing climate-smart solutions against plant pathogens.

Conclusions

This study elucidates the genetic and molecular mechanisms driving the loss or reduction of virulence in Psa. Through transcriptomic analysis, we determined that the nonpathogenic phenotype of the G126 strain is attributed to an SNP event in the hrpR gene, resulting in an amino acid substitution at the 10th position. This mutation disrupts the functionality of the enhancer-binding protein HrpR, leading to a significant reduction in the expression of key T3SS-associated genes, including hrpR, hrpS, hrpL, and hrpA. Conversely, in the G282 strain, the deletion of the T6SS cluster further highlights the genetic variation contributing to differences in pathogenicity. Furthermore, the observed regulatory complexity of the T3SS highlights the involvement of multiple genetic factors and regulatory networks that govern the expression of this system. Overall, this study provides a comprehensive understanding of the molecular basis of virulence attenuation in Psa3 and offers potential avenues for developing innovative disease management strategies for kiwifruit cultivation.

Funding Statement

National Key Research and Development Program of China [Grant No. 2022YFD1400200], the National Natural Science Foundation of China [Grant No. 31860486, 32001864], and the earmarked fund for the China Agriculture Research System [Grant No. CARS-26-04A].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All the numerical data presented in this study are publicly available at https://doi.org/10.6084/m9.Figureshare.28423463. The whole-genome sequence data of G126 and G282 have been submitted to the National Centre for Biotechnology Information (NCBI) under BioProject accessions PRJNA1205097 and PRJNA1205096, respectively, https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1205096/and https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1205097. The genome assemblies of the isolates G126 and G282 are publicly accessible under accessions JBLLLG000000000 and JBLLLF000000000; (https://www.ncbi.nlm.nih.gov/GenBank/). The RNA-Seq dataset is are publicly accessible at gene expression omnibus (GEO) repository (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE286024) under accession number GSE286024.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21505594.2025.2543983