A Genome-Wide Mutant Screen Identifies XopN and XopX as Core Type III Effectors Required for Peach Infection by Xanthomonas arboricola pv. pruni

TL;DR

This study identifies key genes in a bacterium that causes peach disease, revealing new targets for controlling the infection.

Contribution

A genome-wide screen identifies XopN and XopX as essential type III effectors for peach infection by Xanthomonas arboricola pv. pruni.

Findings

XopN and XopX are core type III effectors crucial for symptom induction in peach infection.

Mutants in hrpF, pstS, and other metabolic genes show reduced virulence despite similar bacterial population levels.

Bacterial multiplication and symptom development are not necessarily linked in this pathosystem.

Abstract

Xanthomonas arboricola pv. pruni causes bacterial spot in peaches, a major disease affecting global Prunus production. Despite its economic significance, the virulence mechanisms that enable X. arboricola pv. pruni to colonize peach tissues and induce characteristic necrotic symptoms remain poorly understood. To identify key virulence determinants, a robust and reliable detached-leaf inoculation system was developed, and a genome-wide forward genetic screen of 2400 Tn5 mutants was conducted. A total of 34 mutants with consistently reduced virulence were identified, representing diverse functional categories including secretion systems, nutrient acquisition, primary metabolism, and regulatory pathways. The most prominent findings were the repeated identification of independent mutants in two type III effector genes, xopN and xopX, highlighting these effectors as central and nonredundant contributors to symptom induction. Mutants in the type III secretion system translocon-associated gene hrpF also showed virulence defects. Additional mutants affecting phosphate uptake (pstS), ammonium transport, and vitamin B6 biosynthesis (pdxA, serC) revealed metabolic requirements essential for in planta fitness. Notably, several mutants reached bacterial population levels comparable to the wild-type isolate but produced little or no symptoms, indicating that bacterial multiplication and symptom development are not necessarily linked. This study provides the first comprehensive genome-wide functional screen of X. arboricola pv. pruni virulence and establishes a framework for dissecting infection mechanisms. The essential factors identified here, particularly XopN, XopX, and key metabolic pathways, represent promising targets for future anti-virulence strategies to manage bacterial spot disease. Characterizing the specific functions of each virulence factor identified in this study will be an important focus of future work.

Figures

Full text

1. Introduction

Xanthomonas is a Gram-negative bacterial genus that infects more than 400 plant species, including agriculturally important crops such as rice, wheat, citrus, tomato, pepper, cabbage, peach, and bean [1]. Within this genus, Xanthomonas arboricola comprises seven pathovars, each adapted to specific hosts. Of these, bacterial spot in stone fruits caused by X. arboricola pv. pruni and bacterial blight in hazelnuts caused by X. arboricola pv. corylina have emerged as significant diseases in several European countries [2]. X. arboricola pv. pruni is the causal agent of bacterial spot disease in Prunus spp., including peaches, plums, and almonds [3,4], and is distributed across Asia, the Americas, Europe, and Oceania. In Japan, bacterial spot in peaches is one of the most damaging bacterial diseases affecting orchards [5]. Bacterial spot affects shoots, leaves, and fruits, producing spring and summer cankers [3]. Spring cankers serve as the primary inoculum source, initiating the annual disease cycle [6]. After primary infection, the bacteria multiply epiphytically on young leaves and subsequently enter plant tissues through natural openings such as stomata or wounds, generating secondary inoculum that leads to summer canker formation. In addition, X. arboricola pv. pruni can persist on leaf surfaces through structured bacterial aggregates (biofilms) and can also colonize woody tissues, further supporting secondary inoculum reservoirs.

Despite the economic importance of X. arboricola pv. pruni, its infection mechanisms remain poorly understood. In Xanthomonas species, the type III secretion system (T3SS) is one of the central virulence determinants, enabling the delivery of effector proteins into host cells to suppress defense responses [7]. Hajri et al. [8] investigated the diversity of type III effector (T3E) repertoires in X. arboricola and demonstrated that the T3SS is an essential virulence mechanism in this species. However, studies directly assessing X. arboricola pv. pruni virulence factors through bacterial gene knockout or mutant analyses have not yet been conducted. Moreover, efforts to investigate X. arboricola pv. pruni pathogenicity have been limited by the lack of reliable, high-throughput inoculation methods. Conventional inoculation techniques such as puncture, spray, or syringe inoculation of shoots often suffer from low reproducibility and require long evaluation periods. Suesada et al. [9] successfully evaluated disease severity among several peach cultivars; however, disease assessment following shoot syringe inoculation required nearly three months. In addition, because this method relies on shoot inoculation, a large number of peach trees are required, making it impractical for large-scale mutant screening. Therefore, the development of robust and rapid inoculation methods is crucial for enabling systematic genetic analysis of X. arboricola pv. pruni virulence.

Several studies have attempted to identify virulence related genes in plant pathogenic bacteria using forward genetic approaches. In particular, Tn5 transposon mutagenesis has been widely applied to generate large mutant libraries that are screened based on visible reductions in disease symptoms, allowing direct identification of genes required for pathogenicity. Such symptom-based virulence screens have been conducted in several plant pathogenic bacteria including Pseudomonas [10,11,12,13,14], Xanthomonas [15,16], and Pectobacterium [17]. These studies revealed numerous virulence-related genes, including those involved in secretion systems, regulatory cascades, and metabolic pathways, highlighting the power of phenotype-driven screening to uncover diverse factors essential for disease development. Further, randomly barcoded transposon insertion-site sequencing (RB-TnSeq) has been conducted in several plant pathogenic bacteria. In Xanthomonas campestris pv. campestris, recent RB-TnSeq analyses identified genes required for early colonization of hydathodes and demonstrated that nutrient acquisition and metabolic adaptation, not classical virulence genes, dominate fitness during initial host entry [18]. However, these population-based approaches tend to overlook essential virulence determinants such as T3SS and its effectors because defects in these genes can be functionally complemented by neighboring cells. From a disease-control perspective, even if bacterial populations increase within the host, they pose little problem to growers as long as no visible disease symptoms develop. Thus, identifying the bacterial genes specifically required for symptom induction holds significant potential for developing targeted control strategies.

Despite the severe impact of bacterial spot caused by X. arboricola pv. pruni, its molecular virulence mechanisms remain poorly understood. Notably, no systematic forward genetic screen using individually isolated mutants has been conducted to comprehensively identify virulence determinants in this pathogen. Here, inoculation assays using detached leaves that allow for rapid and reproducible assessment of X. arboricola pv. pruni virulence are described. Using these high-throughput methods, a forward genetic screen was performed to identify virulence-associated genes in X. arboricola pv. pruni. This screen revealed several candidate virulence factors, including components of the T3SS, transporters, amino acid metabolic pathways, and regulatory systems. Notably, the results demonstrate that the T3Es XopN and XopX are essential for X. arboricola pv. pruni virulence. These findings provide an important foundation for future studies on X. arboricola pv. pruni infection biology and contribute to broader insights into plant–pathogen interactions in Prunus species and related pathosystems.

2. Materials and Methods

2.1. Bacterial Strains, Plasmids, and Growth Conditions

The bacterial isolates and plasmids used in this study are described in Supplementary Table S1. X. arboricola pv. pruni isolate PXA19-15-4, provided by the Okayama Prefectural Technology Center for Agriculture, Forestry, and Fisheries, Research Institute for Agriculture, was used as the pathogenic strain to inoculate peach leaves. X. arboricola pv. pruni was grown on YPGA medium (yeast extract 7 g/L, peptone 7 g/L, glucose 7 g/L, and agar 15 g/L) or mannitol–glutamate (MG) medium at 27 °C. Escherichia coli strains were grown on LB medium at 37 °C. The bacterial suspension was prepared in sterile distilled water, with cell density adjusted to an OD_600_ of 0.1 corresponding to approximately 2 × 10^8^ CFU/mL.

2.2. Plant Materials

Peach (Prunus persica) cv. Shimizu Hakuto, which is susceptible to X. arboricola pv. pruni [8], was used for virulence assays. One-year-old peach plants propagated from cutting were grown in a greenhouse under natural environmental conditions and protected from rainfall. Newly expanded leaves from uniform, healthy plants were selected for inoculation.

2.3. Generation of X. arboricola pv. pruni Mutants

Mutant construction was achieved by inserting gene-disrupted constructs into the X. arboricola pv. pruni wild-type isolate. Briefly, fragments of each target gene were amplified by PCR with primers listed in Supplementary Table S2. PCR was performed using KOD One PCR Master Mix (TOYOBO, Osaka, Japan) with an initial denaturation at 98 °C for 2 min, followed by 30 cycles of 98 °C for 10 s, 55–65 °C for 5 s, and 68 °C for 5 s, and a final extension at 68 °C for 5 min. The resulting PCR products were inserted into the pARO191 plasmid using an In-Fusion Snap Assembly Master Mix (Takara Bio, Kusatsu, Japan) at the BamHI site. The constructions were introduced into E. coli S17-1 and transferred into X. arboricola pv. pruni by conjugation, with mutants selected on YPGA plates containing 30 µg/mL kanamycin and 100 µg/mL rifampicin. The resulting strains were then used for subsequent inoculation tests on detached peach leaves.

2.4. Bacterial Inoculation

Detached peach leaves were surface-sterilized with 70% ethanol followed by 5% NaClO and rinsed thoroughly with sterile distilled water. For dip inoculation, leaves were submerged in an X. arboricola pv. pruni suspension containing 0.02% Tween 20 to mimic natural infection. For syringe inoculation, bacterial suspensions of defined concentrations were syringe-inoculated into detached leaves using a 1 mL blunt syringe. After inoculation, leaves were placed in trays with petioles immersed in sterile water and covered with plastic wrap to maintain high humidity (85–95% RH). Inoculated leaves were incubated at 22–24 °C under a 12 h light/12 h dark photoperiod (200 μE m^−2^ s^−1^).

To quantify bacterial populations in dip-inoculated leaves, infected leaves were surface-sterilized with 5% H_2_O_2_, washed, and homogenized. The samples were serially diluted and plated on YPGA medium. To assess bacterial growth in syringe-inoculated leaves, leaf disks were harvested using a 3.5 mm-diameter cork-borer from syringe-inoculated leaf zones. The bacterial CFUs were counted and normalized as CFU per g or cm^2^ of leaf tissue, respectively. The bacterial populations were evaluated in at least three independent experiments.

2.5. Generation of a X. arboricola pv. pruni Genomic Tn5 Mutant Library

The transposon was introduced into X. arboricola pv. pruni by conjugation with E. coli S17-1 which possessed pBSLC1 [19]. The transposon insertion region was integrated into X. arboricola pv. pruni chromosome randomly. Replica plates for all transconjugants were made and used for the inoculation assay.

2.6. Forward Genetic Screening

The workflow of forward genetic screening was described previously [11]. For the first screening, the bacterial suspension was adjusted to approximately 2 × 10^5^ CFU/mL. For the second screening, the virulence-reduced mutants obtained from the first screening were adjusted to 2 × 10^5^ CFU/mL. Leaves were inoculated by syringe inoculation of at least ten parts on at least two different leaves (Supplementary Figure S1). Symptoms were evaluated based on the disease index (Supplementary Figure S1) and classified into four categories based on the ratio of diseased area to the inoculated area. For each experiment, WT was always included on the same leaf, and at least 10 inoculation sites were evaluated across different leaves. The virulence score was calculated using the following formula: virulence score = 3 − (mean disease index of the WT − mean disease index of the mutant). In the first screening, mutant strains with a virulence score of less than 1 were selected as virulence-reduced mutants. In the second screening, mutant strains with a virulence score less than 1.5 were selected as virulence-reduced mutants.

2.7. Insertion Site Identification by Plasmid Rescue

Genomic DNA of the mutants that showed reduced virulence on peaches were purified by using a Nucleospin Microbial DNA Kit (Takara Bio) and digested with HindIII, XhoI, SphI, XbaI, or KpnI (Takara Bio). The resultant DNA was ligated with Ligation high Ver. 2 (TOYOBO), then introduced into E. coli DH5α competent cells. Plasmid DNA was purified from the transformants, and transposon-insertion sites were identified by sequencing with the M13 forward primer. X. arboricola pv. pruni Xap78 (NCBI GenBank: GCA_051126325.1) genome was used for annotation.

2.8. Growth Curve Assay

X. arboricola pv. pruni strains including the WT and mutants were grown at 28 °C for 24 h in YPGA broth. The strain suspensions were adjusted to an OD_600_ of 0.01 with fresh YPGA broth, and the bacterial growth dynamics were measured at OD_600_ for 24 h.

2.9. Statistical Analysis

All data are presented as mean ± standard error (SE). Statistical analyses were carried out using RStudio (version 1.6.0). Differences between the X. arboricola pv. pruni wild-type (WT) isolate and each mutant at the same time point were analyzed using Student’s t-test. Asterisks indicate statistically significant differences (* p < 0.05, ** p < 0.01).

3. Results

3.1. Inoculation Methods

To establish a reliable assay for evaluating X. arboricola pv. pruni virulence in peach leaves, spray inoculation was tested first, which mimics natural infection. Although bacterial populations reached the stationary phase by 5 dpi (Figure 1b), visible necrotic symptoms did not appear until 14 dpi (Figure 1a). This delay suggests that epiphytic multiplication precedes tissue invasion and symptom induction. These observations indicate that dip inoculation is useful for monitoring symptom development and bacterial growth in planta. Syringe inoculation was performed next, which is an inoculation method that delivers bacterial suspensions directly into the leaf apoplast. As with dip inoculation, necrosis developed at 14 dpi (Figure 1c), and bacterial populations increased gradually until 10 dpi (Figure 1d). To determine the inoculum concentration required to induce disease symptoms, detached leaves were syringe-inoculated with bacterial suspensions at 2 × 10^2^, 2 × 10^3^, 2 × 10^4^, 2 × 10^5^, 2 × 10^6^ and 2 × 10^7^ CFU/mL. Necrosis was observed only at inoculum concentrations ≥ 2 × 10^4^ CFU/mL (Supplementary Figure S2a). Notably, increasing the inoculum concentration beyond 2 × 10^5^ CFU/mL did not lead to further increases in bacterial populations (Supplementary Figure S2b,c), indicating that bacterial multiplication reaches a saturation level independent of higher inoculum doses. These results demonstrate that syringe inoculation of detached leaves provides a reproducible quantitative assay for assessing X. arboricola pv. pruni virulence.

To confirm that the two inoculation methods reliably detect differences in virulence, mutants carrying disruptions in the essential T3SS genes hrcC and hrcV were tested. In dip inoculation assays, both mutants produced only faint or no necrotic lesions and exhibited drastically reduced bacterial populations compared with the WT (Figure 2a,b). Further, after syringe inoculation, both hrcC and hrcV mutants showed no symptoms and bacterial populations were also significantly reduced compared with the WT (Figure 2c,d). Together, these findings confirm that both dip and syringe inoculation methods are robust for evaluating X. arboricola pv. pruni virulence. Importantly, because the syringe inoculation allowed rapid, uniform inoculation and provided highly consistent quantitative measurements, syringe inoculation was adopted as the primary method for the forward genetic screen. This approach enabled efficient evaluation of virulence phenotypes across the entire mutant library.

3.2. Isolation and Characterization of Virulence-Reduced X. arboricola pv. pruni Mutants

To identify virulence factors in X. arboricola pv. pruni, a genome-wide forward genetic screen was conducted using 2400 independent Tn5 insertion mutants. Detached peach leaves inoculated with the WT isolate developed characteristic necrotic lesions, whereas 34 mutants consistently showed reduced or no necrosis across repeated assays (Table 1; Figure 3). To determine whether reduced virulence was associated with auxotrophy, growth on MG minimal medium was evaluated. Nine mutants failed to grow under these conditions (Table 1), whereas all mutants grew normally on rich medium (Supplementary Figure S3), indicating that most Tn5 insertions did not cause general growth defects. Next, the Tn5 insertion sites in all 34 mutants were identified, and the disrupted genes were classified into functional categories as described below (Table 1).

3.2.1. Type III Secretion System

Five mutants carried Tn5 insertions in genes associated with the type III secretion system (T3SS) or its effectors. Both xopN and xopX were each disrupted by two independent Tn5 insertions, indicating that the loss of either effector consistently reduces X. arboricola pv. pruni virulence. The insertion site of strain XF23 was located within a PIP-box–like sequence upstream of xopN (TTCGC–N_16_–CTCGC), suggesting that transcriptional regulation of xopN may also be important for virulence. Further, the recovery of multiple independent mutants in the same T3E genes supports the robustness and reproducibility of the screen. A mutant in hrpF was additionally identified, which encodes a putative T3SS translocon protein responsible for forming the pore through which effectors are delivered into the host cell. The virulence reduction in hrpF mutants suggests the importance of T3E translocation for virulence. Consistent with the known importance of T3SS machinery (Figure 2), the screen identified that T3SS and two effectors, XopN and XopX, are essential for X. arboricola pv. pruni virulence.

3.2.2. Membrane Transporter

Five mutants carried insertions in genes predicted to encode membrane transporters. These included an ammonium transporter (amtB), a phosphate ABC transporter substrate-binding protein (pstS), a fluoride efflux transporter (crcB), and an RND-type periplasmic adaptor protein. InterPro analysis indicated that AmtB and PtsS are associated with nutrient uptake systems, while crcB and the RND adaptor represent efflux-related components. The identification of transporters involved in both nutrient acquisition and efflux suggests that X. arboricola pv. pruni must carefully balance the uptake of essential nutrients with the removal of harmful compounds to successfully colonize peach tissue.

3.2.3. Regulation/Signaling

Several mutants disrupted genes associated with regulatory or signaling functions, and all exhibited reduced virulence. These included a S-adenosyl-L-homocysteine riboswitch (XE37), a response regulator belonging to a two-component regulatory system (XF81), and an EF-hand calcium-binding protein (XY2). Further studies will be required to determine the downstream processes governed by each regulator.

3.2.4. Cofactor/Vitamin Biosynthesis

Three mutants carried insertions in pdxA (XF16 and XV67) and the gene encoded in nicotinate phosphoribosyltransferase (XO84) showed reduced virulence. Two independent mutants disrupted pdxA (XF16 and XV67), which encodes a key enzyme in the deoxyxylose 5-phosphate (DXP)-dependent vitamin B_6_ biosynthesis pathway. We also identified a mutant in serC (XJ36), which functions upstream of PdxA and provides the precursor for the PdxA-catalyzed step. The identification of mutations in both pdxA and serC indicates that the vitamin B_6_ biosynthetic pathway is likely required for X. arboricola pv. pruni virulence.

3.2.5. Carbohydrate Metabolism/Cell Envelope Biosynthesis

Two mutants carried insertions in genes associated with carbohydrate metabolism and cell envelope biosynthesis. One mutation occurred in a nucleoside-diphosphate (NDP) sugar epimerase/dehydratase. Enzymes of this class typically participate in the synthesis of NDP-sugar precursors, which serve as building blocks for bacterial surface glycoconjugates. Another mutant disrupted a glycosyltransferase family 1 protein, a class of enzymes generally involved in transferring sugar moieties to form polysaccharides or glycan structures in the bacterial envelope.

3.2.6. Type II Secretion System

One mutant (XH56) carried a Tn5 insertion in a gene encoding a PilN/GspL domain-containing protein, which is part of the type II secretion system (T2SS). In Gram-negative bacteria, the T2SS exports a wide range of extracellular enzymes and virulence factors, including cell wall-degrading enzymes, proteases, and lipases. The reduced virulence of the XH56 mutant therefore suggests that secreted enzymatic activities may contribute to X. arboricola pv. pruni infection, although the specific substrates and roles of the T2SS in this pathovar remain to be elucidated.

3.2.7. Type IV Secretion/Conjugation System

One mutant (XT34) contained a Tn5 insertion in a TIGR03751 family conjugal transfer lipoprotein (WDL40_17020). This gene is part of a conserved integrative and conjugative element (ICE)-associated type IV secretion/conjugation machinery. ICEs are mobile genomic islands that integrate into the chromosome, commonly at tRNA loci, and mediate horizontal gene transfer through conjugation [20,21,22]. Consistent with this, the disrupted region in XT34 was flanked by two tRNA^Gly^ sequences. Although the specific contribution of this conjugal transfer component to X. arboricola pv. pruni pathogenicity remains unclear, these findings suggest that horizontal gene-transfer-associated elements or their encoded functions may influence virulence.

3.2.8. Amino Acids Biosynthesis

Six mutants contained insertions in genes involved in amino acid metabolism: thrC (threonine biosynthesis), serC (serine biosynthesis), ilvG (branched-chain amino acid synthesis), and glyA (serine–glycine interconversion). All mutants except mtnA failed to grow on MG minimal medium, indicating that they are partially auxotrophic. In contrast, the mtnA mutant, which is defective in the methionine salvage pathway, grew normally on minimal medium, consistent with this pathway being dispensable under our culture conditions. These findings suggest that amino acid synthesis is required for X. arboricola pv. pruni colonization.

3.2.9. Central Carbon Metabolism

Two mutants carried Tn5 insertions in genes involved in central carbon metabolism. One mutant disrupted ppsA, which encodes phosphoenolpyruvate synthase, and the other disrupted a gene encoding class I fructose-bisphosphate aldolase. Both mutants showed markedly reduced necrosis and decreased bacterial populations in vivo compared with the WT. Neither mutant grew on MG minimal medium, while both grew normally on rich medium. Because PpsA and fructose-bisphosphate aldolase supply central carbon intermediates that fuel glycolysis, gluconeogenesis, and the pentose phosphate pathway, loss of these enzymes likely restricts metabolic flexibility. Their failure to grow on minimal medium reinforces that robust carbon flux is essential for survival in the nutrient-poor leaf apoplast. These findings suggest that X. arboricola pv. pruni requires efficient central carbon metabolism to adapt and colonize peach tissue.

3.3. Bacterial Multiplication In Vivo

To determine whether reduced symptom development was associated with decreased bacterial proliferation, in vivo bacterial populations of all virulence-reduced mutants (Figure 3) were quantified. For amino acid metabolism mutants, all isolates except XF66 exhibited growth defects on MG minimal medium; however, their phenotypes in vivo were not uniform. Notably, the serC (XJ36) and argininosuccinate synthase (XV51) mutants, despite showing auxotrophy in vitro, multiplied to levels comparable to the WT inside peach leaves (Figure 3a). In contrast, the mtnA mutant (XF66), which showed no growth defect on minimal medium, did not display enhanced fitness in vivo (Figure 3a). These observations indicate that auxotrophy in vitro does not necessarily predict bacterial proliferation inside host tissues, and that the peach apoplast may supply certain amino acids or alternative metabolic routes that partially compensate for specific biosynthetic defects.

As expected, all T3SS-related mutants exhibited significantly lower bacterial populations than the WT (Figure 3b), consistent with their strong impairment in disease development. Moreover, the bacterial populations of the mutants related to regulation/signaling and cofactor/vitamin biosynthesis were also significantly reduced compared with the WT (Figure 3d,e).

Among transporter mutants, the bacterial populations of XG88, XQ90, and XT52 were significantly reduced compared with the WT, whereas those of XG89 and XL76 were not (Figure 3c). Notably, the ammonium transporter mutant XG88 displayed a clear decrease in bacterial population. These results suggest that while several transporters contribute to symptom development, ammonium uptake in particular is required for effective bacterial growth in vivo.

Interestingly, some mutants displayed reduced bacterial populations that coincided with reduced disease symptoms, whereas others developed reduced symptoms despite proliferating to WT levels (Figure 3; Supplementary Figure S4). These observations indicate that bacterial multiplication and symptom progression are not always tightly correlated in X. arboricola pv. pruni, and that some virulence factors affect host responses independently of bacterial growth.

4. Discussion

In this study, a reliable detached-leaf inoculation system was established for X. arboricola pv. pruni and performed a forward genetic screen to identify key virulence factors. The most striking outcome of this screen was the repeated identification of independent mutants in two T3Es genes, xopN and xopX, highlighting these effectors as central and nonredundant determinants of X. arboricola pv. pruni virulence in peach. From the 2400 mutants screened, 34 were identified that consistently showed reduced virulence, representing diverse functional categories, including secretion systems, nutrient uptake, metabolic pathways, and regulatory networks. The identification of multiple independent mutations within key pathways further demonstrates the depth and robustness of the screening strategy.

Among the virulence factors identified, disruptions in the T3SS and its associated effectors produced some of the strongest virulence defects (Table 1; Figure 3). This aligns with the well-established role of the T3SS as the primary virulence mechanism in Xanthomonas species [7]. The screen also recovered two independent mutants in xopN, making it one of the most repeatedly identified T3E-related genes. This strongly suggests that XopN is indispensable for X. arboricola pv. pruni virulence in peach. The coding-region mutant (XN94) caused a severe loss of virulence, whereas the regulatory mutant (XF23), which carries a Tn5 insertion within a PIP-box–like motif upstream of xopN, displayed a milder phenotype. In Xanthomonas, HrpX activates hrp and several effector genes via binding to plant-inducible promoter PIP boxes [23]. Thus, the phenotypic gradient between XN94 and XF23 provides functional genetic evidence that both XopN protein function and appropriate HrpX-dependent induction are required for full virulence. XopN is one of the most conserved core effectors in Xanthomonas, and its virulence functions have been documented across multiple pathovars [24,25,26,27,28]. It suppresses pathogen-associated molecular pattern-triggered immunity (PTI), interacts with host transcription factors such as VOZ proteins, interferes with thiamine metabolism, and can even facilitate effector translocation in cooperation with HrpF [27,29,30]. These diverse activities illustrate why disruption of XopN results in such a pronounced phenotype in X. arboricola pv. pruni. These findings highlight XopN as a central virulence determinant with multiple mechanisms for suppressing host defenses. Similarly, two independent xopX mutants (XH42, XI8) were identified, reinforcing that XopX is also essential for X. arboricola pv. pruni virulence (Table 1). XopX suppresses PTI responses, including ROS burst, MAPK activation, and callose deposition, and acts in combination with other core effectors such as XopN, XopQ, and XopZ to suppress immune signaling and contributes to virulence of several Xanthomonas spp. [24,28,31,32]. Additionally, XopX modulates cell death and supports bacterial proliferation and disease symptom development [33], further highlighting its broad immunosuppressive functions. Together, these findings highlight XopN and XopX as central, nonredundant virulence determinants in X. arboricola pv. pruni, despite the species possessing one of the largest and most conserved T3E repertoires in X. arboricola [8]. The repeated recovery of both genes from our forward genetic screen demonstrates that these two effectors form the core functional backbone of the X. arboricola pv. pruni effector arsenal on peaches, emphasizing their importance in host adaptation and disease development.

Interestingly, the hrpF mutant exhibited reduced virulence, but its defects in symptom development and in vivo bacterial multiplication were consistently less severe than those observed in hrcC and hrcV mutants (Figure 2 and Figure 3). This difference is likely attributable to the distinct functions of these genes within the T3SS. HrcC and HrcV are core structural components required for assembling the secretion apparatus, and their disruption completely abolishes effector secretion. By contrast, HrpF acts at downstream stages of the secretion pathway interacting with the pilus protein HrpA, contributing to translocon formation, and modulating hrp gene expression via HrpG [34]. Because HrpF influences secretion efficiency rather than the integrity of the secretion channel itself, loss of HrpF is expected to partially impair, but not eliminate, effector translocation. This mechanistic distinction likely explains the comparatively attenuated phenotype of the hrpF mutant in this study.

Five mutants disrupted transporter-related genes, highlighting the importance of both nutrient uptake and efflux during infection (Table 1; Figure 3). The bacterial populations of the ammonium transporter mutant (XG88) were shown to be significantly reduced compared with the WT (Figure 3), suggesting that ammonium transporter contributes to colonization in vivo. The contribution of ammonium transporters to plant-pathogenic bacterial virulence has not been reported to date. In animal pathogens, ammonium transporters influence several virulence-associated traits, including biofilm architecture in Streptococcus mutans [35], and motility and epithelial colonization in enterohemorrhagic E. coli [36]. Determining how ammonium uptake functions in the plant apoplast will be an important direction for future research.

Additionally, XQ90 (disrupted in crcB) and XT52 (disrupted in RND-type periplasmic adaptor protein) also exhibited reduced bacterial populations compared with the WT (Figure 3). Although CrcB is best known for protecting bacteria from fluoride toxicity, recent work in Pseudomonas putida demonstrated that CrcB contributes to tolerance to a broad range of stresses including acid, oxidative, ionic, and membrane stresses, indicating that its physiological role extends well beyond fluoride detoxification [37]. RND transporter functions to extrude various substrates including antibiotics and host-derived molecules [38]. Recent work in Xanthomonas campestris showed that the plant defense hormone salicylic acid (SA) is sensed by the transcriptional regulator HepR and subsequently expelled by the RND efflux pump HepABCD, and that this SA efflux is required for full virulence on Brassica hosts [39]. Although efflux systems are likely critical for detoxifying host-derived antimicrobials, the specific substrates and in planta roles of these transporters remain unclear and will require further investigation.

Two mutants carrying disruptions in ABC transporter–related genes were also identified. Among them, the pstS mutant (XL76), which affects the periplasmic substrate-binding component of the PstSCAB high-affinity phosphate uptake system, displayed a near-complete loss of virulence. Because the PstSCAB system is specifically required for efficient phosphate acquisition under low-phosphate conditions, this phenotype suggests that X. arboricola pv. pruni experiences phosphate limitation inside peach leaf tissues. Similar results were reported in Xanthomonas citri pv. glycines that deletion of the pstSCAB operon or the PhoB regulator abolished in planta multiplication and disease development due to the phosphate-poor soybean apoplast [40,41]. Together with the findings presented here, these studies suggest that the peach apoplast is likewise phosphate-limited and that X. arboricola pv. pruni requires the high-affinity PstSCAB transporter to secure sufficient phosphate during infection.

Two independent Tn5 mutants of pdxA (XF16 and XV67) were identified as virulence-reduced mutants. The pdxA gene encodes 4-hydroxythreonine-4-phosphate dehydrogenase, a key enzyme in the deoxyxylose 5-phosphate (DXP)-dependent vitamin B_6_ pathway. A similar role was recently demonstrated in Xanthomonas oryzae pv. oryzae, where pdxA or pdxJ mutants were completely auxotrophic for vitamin B_6_ and fully lost virulence in rice [42]. Moreover, a mutant in serC was also identified, which encodes the SerC functioning upstream of PdxA and supplying precursors for vitamin B_6_ biosynthesis. The simultaneous recovery of both serC and pdxA mutants provides strong genetic evidence that interruption at any point in this pathway impairs vitamin B_6_ biosynthesis and consequently reduces X. arboricola pv. pruni pathogenicity. Vitamin B_6_ plays a primary role acting as a cofactor for a large number of essential enzymes [43]. Importantly, plants use a DXP-independent pathway to produce VB_6_, whereas Xanthomonas relies exclusively on the DXP-dependent pathway. This metabolic divergence suggests that the bacterial vitamin B_6_ pathway may serve as a pathogen-specific target for anti-virulence strategies, which aim to suppress virulence without inhibiting bacterial growth and therefore impose lower selective pressure for resistance. However, it remains unclear whether X. arboricola pv. pruni can access vitamin B_6_ or its precursors from host tissues, or whether alternative metabolic routes could compensate for pathway disruption. Further work will be required to address these questions and to evaluate whether targeting vitamin B_6_ biosynthesis is a practical strategy for disease control.

Importantly, the virulence factors identified here represent promising targets for anti-virulence approaches, which aim to suppress pathogen aggressiveness while imposing lower selective pressure than traditional bactericides [44,45]. Core mechanisms uncovered in this work, particularly XopN, XopX, nutrient acquisition such as PstSCAB, and metabolic factors like PdxA, may serve as future molecular targets for pathogen-specific disease management. Anti-virulence concepts have been extensively explored in human pathogens; for instance, Staphylococcus aureus has served as a model organism for anti-virulence therapeutics, with efforts focused on pore-forming toxins, immune evasion factors, and the quorum sensing system [46,47]. A similar paradigm is highly relevant in agriculture, where reducing symptom expression rather than eliminating bacterial viability could provide a sustainable route to disease suppression while minimizing resistance risks. Nevertheless, translating these findings into practical control measures will require further work. First, the precise roles and host targets of each virulence factor identified here need to be characterized in detail. Second, because this screening and phenotyping relied on detached-leaf assays, the contribution of these factors should be validated under whole-plant and field conditions, where environmental fluctuations and seasonal host physiology may influence disease outcomes. Such follow-up studies will be essential to determine whether targeting these symptom-associated virulence mechanisms can deliver reliable protection against bacterial spot in commercial orchards.

5. Conclusions

This study provides the first comprehensive forward genetic screening of virulence in X. arboricola pv. pruni. Multiple virulence factors were identified, including components of the T3SS, transporters, primary metabolic pathways, and regulatory systems. Among these, the repeated identification of mutants in xopN and xopX highlights these two effectors as essential and nonredundant determinants of symptom induction in peach. It was also found that some mutants produced little or no symptoms despite reaching WT-like bacterial populations, indicating that symptom expression and bacterial multiplication do not always correlate in this pathosystem. The recovery of independent mutants across several key pathways underscores the robustness of our screening strategy and provides a solid foundation for future genetic analyses of X. arboricola pv. pruni pathogenicity. Further studies using whole-plant inoculation will be necessary to characterize the specific roles of each virulence factor identified in this study. Overall, this study establishes a genetic framework for understanding X. arboricola pv. pruni infection biology and opens the way for developing sustainable, virulence-targeted strategies to manage bacterial spot disease in peach and other Prunus species.