The strains, plasmids, and primers used in this study are listed in Supplementary Tables S1–S3, respectively. All peptides used were synthesized by GenScript and are summarized in Supplementary Table S4. P. syringae and P. fluorescens were cultured in King’s medium B (KB) (King et al., 1954), Agrobacterium tumefaciens was grown in Luria-Bertani (LB) broth at 28°C, and E. coli was grown in LB broth at 37°C. The following antibiotic concentrations were used: ampicillin, 50 mg/mL; rifampicin, 50 mg/mL; spectinomycin 50 mg/mL; and tetracycline 20 mg/mL. N. benthamiana plants were grown in a chamber maintained at 16 h light/8 h dark, 65% humidity, and temperatures of 24°C during the day and 22°C at night.

ROS measurements were performed as previously described (Wei et al., 2013), with minor modifications. Briefly, fresh Pf0-1 and its derivatives were infiltrated into N. benthamiana leaves. The concentrations of bacterial suspensions are indicated in the figure legends. Leaf disks from the infiltrated area were excised 15 h post-inoculation (hpi) (36 hpi for transient expression) and placed into the wells of a 96-well plate pre-supplied with 100 μL of sterile water. Then, 100 μL of 0.5 mM L-012 (Wako, Japan) in 10 mM morpholinepropanesulfonic acid–KOH buffer (pH 7.4) was added to each well. ROS accumulation was measured by monitoring chemiluminescence using a microplate reader (Tecan, Switzerland).

Transient expression was performed as previously described (Wei et al., 2013). Briefly, transformed Agrobacterium strains were resuspended in 10 mM MES (pH 5.5) containing 200 μM acetosyringone to the desired optical density and incubated at 28°C for 3 h. The suspension was then infiltrated into N. benthamiana leaves using a 1 mL sterile syringe (without the needle). The inoculated plants were maintained in a greenhouse for subsequent assays.

The challenged-inoculation assay was performed as previously described (Wei et al., 2015). Briefly, fresh streaks of Pf0-1 and its derivatives, as well as Pst DC3000, were prepared from isolated colonies and grown overnight at 28°C on KB plates. For the HR suppression assay, Pf0-1 and its derivatives were pre-inoculated into N. benthamiana leaves at 2 × 10⁸ CFU/mL. Six hours later, Pst DC3000 was infiltrated into the pre-treated areas at 2 × 10⁷ CFU/mL. Images were captured 2 d after Pst DC3000 infiltration. For the bacterial growth assay, Pf0-1 and its derivatives were inoculated into N. benthamiana leaves at 4 × 10⁷ CFU/mL. Six hours later, the strain Pst DC3000ΔhopQ1–1 was infiltrated into the pre-treated leaves at 5 × 10⁵ CFU/mL. Leaf disks were harvested 4 d post-inoculation (dpi) to assess the bacterial population.

The effector translocation assay was performed as previously described (Wei et al., 2013), with minor modifications. Briefly, bacterial infiltrations were performed using varying inoculum levels, as indicated in the figure legends. Seventy-two hours later, Pf0-1 (pCPP6225 + pCPP3221) was inoculated into N. benthamiana leaves transiently expressing CSPs or FliC derivatives, and leaf disks were collected 6 h after infiltration. For the challenge effector translocation assay, Pf0-1 (pCPP6225 + pCPP3221) or Pf0-1 (pCPP5316) was challenge-inoculated 6 h after pre-inoculation with Pf0-1 and its derivatives, and the leaf disks were harvested 6 hpi. To assess suppression of effector translocation induced by flg22, csp22, and csp15, peptides were infiltrated into N. benthamiana leaves at a concentration of 1 μM each. At 6 hpi, Pst DC3000 (pCPP3221) expressing AvrPto-Cya at 1 × 10⁷ CFU/mL was challenge-inoculated into the plant leaves. Leaf disks were harvested 6 hpi, excised using a 1.0-cm-diameter cork borer, flash-frozen in liquid nitrogen, and ground in 250 μL of 0.1 M HCl. cAMP levels were measured using a Correlate-EIA cAMP immunoassay kit (Enzo, United States), according to the manufacturer’s instructions.

The peptides were infiltrated into N. benthamiana leaves at a concentration of 100 nM. Three leaf disks were collected from the infiltrated area 6 hpi and ground in liquid nitrogen. Total RNA was isolated using the Plant RNA Kit (R6827-02, OMEGA, United States), following the manufacturer’s instructions. cDNA was synthesized using the FastKing RT kit (KR116-02; TIANGEN, China). Real-time PCR reactions were performed using the SYBR^® Green Premix Pro Taq HS qPCR Kit (AG11701, Accurate Biology, China) and ROX Reference Dye (4 mM) (AG11710, Accurate Biology, China). The expression levels of WRKY22 and ERF1a were quantified using the qPCR primers listed in Supplementary Table S1. NbEF1α was used as a reference gene.

The protein sequences used in this study are available under the following accession numbers: AAO55887.1 (CspA for Pst DC3000), ABA73672.1 (P. fluorescens), AAZ34509.1 (P. savastanoi pv. phaseolicola 1448A), UZD98366.1 (P. corrugata B21-055), UUI32708.1 (P. putida ATCC 12633), AAO57601.1 (CspB for Pst DC3000), ABA72944.1 (P. fluorescens), AAZ36544.1 (P. savastanoi pv. phaseolicola 1448A), UZD96690.1 (P. corrugata B21-055), UUI33322.1 (P. putida ATCC 12633), AAO54799.1 (CspC for Pst DC3000), ABA76132.1 (P. fluorescens), AAZ35758.1 (P. savastanoi pv. phaseolicola 1448A), UZD94220.1 (P. corrugata B21-055), and UUI36241.1 (P. putida ATCC 12633).

Statistical analyses were performed using IBM SPSS Statistics software. Different letters indicate significant differences (p < 0.05 by one-way ANOVA).

To identify the CSPs in Pst DC3000, we performed a BLASTp analysis using the amino acids of csp15 and csp22—two peptides with PAMP activity that contain the RNP-1 epitope of CSP (Felix and Boller, 2003; Wang L. et al., 2016). Three proteins, CspA (AAO55887.1), CspB (AAO57601.1), and CspC (AAO54799.1), were identified by means of this method. Sequence alignment revealed that all three proteins were highly conserved in Pst DC3000, differing by only one amino acid from csp15 and five amino acids from csp22 (Figure 1A). We also performed multiple alignments of the three CSPs with the CSPs of other plant and environment associated-Pseudomonas strains, including the well-studied strain P. fluorescens Pf0-1. The results showed that these three CSPs were well conserved; CspA shares 94% homology with its homologs, CspB shares 95.43%, and CspC shares 96.86%. Collectively, these findings suggest that these CSPs may exhibit similar functions across diverse Pseudomonas strains (Figure 1B).

To determine whether the CSPs of Pst DC3000 modulate plant immunity in vivo, we cloned the CDSs of CspA, CspB, and CspC and transiently expressed them in N. benthamiana leaves under the control of the 35S promoter. First, we assessed whether the CSPs stimulated ROS accumulation. It has been well established that FliC was a major PAMP of Pseudomonas recognized by N. benthamiana, we therefore took advantage of FliC and FliC fused with the PR1a signal peptide (SP-FliC) as controls. As previously reported, FliC and SP-FliC induced weak and strong ROS bursts, respectively (Wei et al., 2013). However, CspA, CspB, and CspC did not stimulate ROS accumulation (Figure 2A). In the HR suppression analysis, Agrobacterium tumefaciens harboring pEarleyGateS101 carrying csp genes were inoculated into N. benthamiana leaves at 2 × 10⁸ CFU/mL and then incubated for 48 h before partially overlapping challenge inoculation with a suspension of 4 × 10⁷ CFU/mL of wild-type Pst DC3000. None of the CSPs inhibited the HR induced by Pst DC3000 (Figure 2B). The activation of PTI inhibits the translocation of effector proteins (Anderson et al., 2014). Therefore, we next investigated whether CSPs suppressed effector translocation by employing Pf0-1 harboring plasmids pCPP3221 expressing AvrPto-Cya and pCPP6225 (a derivative of T3SS⁺ pHIR11 with an unmarked deletion of the genes encoding the effector HopA1 and its chaperone ShcA). N. benthamiana leaves were infiltrated with A. tumefaciens cells and then challenged 72 h later with an overlapping inoculation of Pf0-1 (pCPP6225 + pCPP3221) for the Cya-based translocation assay, in which the translocation of effectors into plant cells can be clearly detected on the basis of the Cya activity as a high production of cAMP (Mukaihara and Tamura, 2009). As shown in Figure 2C, the expression of CspA and CspC, but not CspB, significantly suppressed AvrPto-Cya translocation, as well as SP-FliC, compared to EV and FliC, indicating that CspA and CspC of Pst DC3000 suppress effector translocation when transiently expressed in N. benthamiana leaves. Collectively, these results suggest that the CSPs of Pst DC3000 can suppress effector translocation in N. benthamiana leaves.

To further probe into the function of CSPs in vivo and to circumvent the effects of effectors, we employed P. fluorescens Pf0-1 in subsequent assays. Mutants lacking individual CSPs (ΔcspA, ΔcspB, or ΔcspC) and the triple mutant ΔcspABC were generated and subjected to ROS burst assay. The ΔfliC mutant served as a positive control, while 10 mM MgCl₂ served as a reagent control. As expected, Pf0-1 strongly induced ROS accumulation compared to the mock treatment, whereas ΔfliC induced negligible ROS production (Figure 3A). ROS levels in N. benthamiana leaves treated with ΔcspA, ΔcspB, ΔcspC, or the triple mutant ΔcspABC were comparable to those in leaves treated with Pf0-1, suggesting that deleting the three CSPs could not impair the ROS burst (Figure 3A).

Next, we examined whether the ability of Pf0-1 to suppress HR was associated with CSPs. N. benthamiana plants were inoculated with Pf0-1 and its derivatives at 4 × 10⁷ CFU/mL, incubated for 6 h, and then challenge inoculated with wild-type Pst DC3000 at 2 × 10⁷ CFU/mL. All treatments, including Pf0-1, ΔcspA, ΔcspB, ΔcspC, or the triple mutant ΔcspABC, suppressed the HR induced by Pst DC3000, whereas the control ΔfliC mutant lost this ability, indicating that mutating these three CSPs does not affect the HR suppression ability of Pf0-1 (Figure 3B). Next, we used Pf0-1 (pCPP6225 + pCPP3221) to investigate whether CSPs play a role in suppressing effector translocation. Pf0-1 (pCPP6225 + pCPP3221) was infiltrated into N. benthamiana leaves 6 h after inoculation with Pf0-1 and its derivatives to measure cAMP levels. The results showed that cAMP levels induced by Pf0-1, ΔcspA, ΔcspB, or ΔcspC treatments were analogous, but were significantly lower than the mock treatment. However, ΔcspABC inoculation generated notably higher cAMP levels than Pf0-1, ΔcspA, ΔcspB, or ΔcspC treatments, suggesting that CspA, CspB, and CspC of Pf0-1 were functionally redundant in mitigating the effector injection restriction (Figure 4A).

FliC is a major Pseudomonas PAMP recognized by N. benthamiana. Therefore, we investigated whether the ability of CSPs to suppress effector translocation was related to FliC. We constructed Pf0-1ΔfliC mutants with individual or combined deletions of cspA, cspB, and cspC. Effector translocation assays showed that ΔfliC induced higher cAMP levels than Pf0-1 (Figure 4B). Moreover, cAMP levels induced by Pf0-1ΔfliC, ΔfliCΔcspA, ΔfliCΔcspB, or ΔfliCΔcspC were similar, but significantly lower than those induced by ΔfliCΔcspABC (Figure 4B). These data indicate that CSPs are redundant in suppressing effector translocation independent of FliC.

To compare the relative contributions of FliC and CSPs to effector translocation inhibition, we conducted further assays. N. benthamiana leaves were inoculated with Pf0-1 and its derivatives at 2 × 10⁸ CFU/mL and challenged 6 h later with Pf0-1 (pCPP6225 + pCPP3221) at 4 × 10⁷ CFU/mL. Effector translocation assays showed that Pf0-1ΔfliC induced significantly higher cAMP levels than Pf0-1ΔcspABC, indicating that FliC inhibits effector translocation more effectively than CSPs (Figure 4C). Furthermore, we transformed pCPP5372-AvrPto into Pf0-1 derivatives carrying pCPP6225, infiltrated them into N. benthamiana leaves at 5 × 10⁸ CFU/mL, and kept the plants in the chamber for 3 days. The results showed that ΔfliC and ΔfliCΔcspABC, but not ΔcspABC, caused AvrPto-induced chlorosis, suggesting that mutation of FliC allowed more AvrPto translocation than CSPs (Figure 4D). Additionally, ΔfliCΔcspABC induced significantly higher cAMP levels than ΔfliC and ΔcspABC, further supporting the conclusion that CSPs suppress effector translocation independently of FliC (Figure 4C). We further analyzed the effect of these mutations on HopA1 translocation. Pf0-1 was transformed with pCPP5316 carrying the functional T3SS cluster and HopA1-Cya, and challenge-inoculation effector translocation assays were performed. The results showed that ΔfliC induced significantly higher cAMP levels than Pf0-1, but lower than ΔfliCΔcspABC, suggesting that both fliC and cspABC mutations increase HopA1 translocation (Supplementary Figure S1). We also investigated whether mutations in CSPs alleviated PTI-associated inhibition of challenge-inoculated pathogen growth. Pf0-1 and its derivatives were infiltrated into N. benthamiana leaves at 4 × 10⁷ CFU/mL and incubated for 6 h before challenge inoculation with Pst DC3000ΔhopQ1–1 at 5 × 10⁵ CFU/mL. The results showed that at 4 dpi, no significant difference in pathogen population suppression was observed between the CSP-deleted mutants and Pf0-1 or Pf0-1ΔfliC (Supplementary Figure S2). Collectively, these results suggest that the suppression of effector translocation mediated by CSPs is moderate compared to that mediated by FliC.

To further validate the role of CSPs in restricting effector translocation, we synthesized csp22 and csp15 peptides, which represent the RNP-1 epitope of CSP with PAMP activity as well as the conserved flg22 epitope. We aimed to test whether inoculation with these peptides could induce PTI-associated suppression of effector translocation by assessing their ability to inhibit HR and the cAMP levels produced by Cya fusion. We found that HR was nearly eliminated in N. benthamiana leaves pre-infiltrated with flg22 following challenge inoculation with Pst DC3000, whereas it was partially eliminated in leaves pre-infiltrated with csp22 or csp15 (Figure 5A). Moreover, pre-infiltration with flg22, csp22, or csp15 resulted in significantly lower cAMP levels than the mock treatment, indicating that these peptides effectively suppressed effector translocation in N. benthamiana (Figure 5B). Additionally, inoculation with these peptides upregulated the PTI marker gene—WRKY22—and the ethylene-related marker gene—ERF1a —consistent with previous reports (Felix and Boller, 2003; Saur et al., 2016) (Supplementary Figure S3). Collectively, these findings suggest that csp22 and csp15 trigger PTI-associated suppression of effector translocation.