The seedlings of N. benthamiana were grown for 5-6 weeks at 25°C in a greenhouse with 18 h light/6 h darkness. The citrus plants were grown at 28°C. CLas-infected sugar orange plants (Citrus reticulata) were used for DNA extraction, as well as a source for graft-inoculation. The epicotyl segments of Wanjincheng (C. sinesis) were used for gene transformations. Pseudomonas syringae pv. tomato DC3000 (DC3000) ΔhopQ1-1 was grown on King’s B (KB) medium.

For CLas detection, total DNA was extracted from midrib sections of sugar orange plants using Biospin Omini Plant Genomic DNA Extraction Kit (BioFlux, Hangzhou, China). The CLas detection was conducted by both PCR and quantitative PCR (qPCR). The PCR was performed using 16S rDNA primer OI1/OI2c with 2X Taq Master Mix (Vazyme, Nanjing, China) in a 20 μL reaction system. The bacterial populations (CLas cells μg^-1 of citrus DNA) were quantified with qPCR assay that was described by Du M. et al. (2021). In the qPCR assay, CLasgyrA of CLas was detected, and 18S rRNA of citrus was used as the internal reference. BlastTaq™ 2X qPCR MasterMix (abm, Canada) was used for qPCR amplification. Primer pairs are listed in Table S1 .

Total RNA was extracted from plant tissues and psyllids using RNAiso Plus (Takara, Japan). In a 20-μL volume, 1μg of total RNA was reverse transcribed with All-In-One 5X RT MasterMix (abm) following the manufacturer’s instructions. Quantitative RT-PCR (qRT-PCR) assays were conducted to analyze gene transcript level. Primer pairs of selected genes are listed in Table S1 . The genes NbACTIN and CsGAPDH were used as endogenous controls. The qRT-PCR analyses were performed in a 10 μL reaction system. The 2^-ΔΔCt method was used for the determination of relative gene transcription (Livak and Schmittgen, 2001).

Primer pairs and vectors used are listed in Table S1 . Fragments were amplified by PCR by using Q5 high-fidelity DNA polymerase and digested with appropriated restriction endonucleases (NEB, Peking, China), followed by ligation into vectors using ClonExpress II One Step Cloning Kit (Vazyme). For phoA assay, the fragment of CLas4425 signal peptide (4425SP) was inserted into pET-mphoA (T7 promoter). For the transient expression assay, the coding sequences were cloned into the binary Potato virus X (PVX) vector pGR107. To construct the overexpression vector, CLas4425 was ligated into pLGN (35S promoter). For the VIGS assay, fragments amplified from N. benthamiana cDNA were cloned into the binary vector pTRV2. All constructs were sequenced by Tsingke (Chongqing, China).

The resulting vector pET-4425SP-mphoA, pET-phoA (positive control), and pET-mphoA (negative control) were transformed into Escherichia coli Rosetta-gami2 (DE3), then the E. coli cells were incubated at 37°C overnight in indicator LB solid medium [90 ng/mL 5-bromo-4-chloro-3-indolyl phosphate (BCIP), 0.1 mM isopropyl-β-D-thiogalactopyranoside (IPTG), 75 μM Na₂HPO_4, and 50 μg/mL Kanamycin]. The blue transformants indicated that fusion proteins were secreted outside the cells, while white colonies demonstrated a lack of phoA activity.

A. tumefaciens GV3101 (pJIC SA_Rep) strains harboring the respective constructs were cultured in LB medium supplemented with the appropriate antibiotics at 28°C, and then harvested and suspended in infiltration buffer [10 mM 2-(N-morpholine)-ethane sulfonic acid (MES), 10 mM MgCl₂, 0.2 mM acetosyringone to pH 5.6] to appropriate concentrations (OD₆₀₀ of 0.2 for INF1, BAX, otherwise OD₆₀₀ of 0.6 was used). After dark incubation for 2 h at 28°C, the A. tumefaciens suspensions were infiltrated into the upper leaves of 6-leaf-old N. benthamiana plants with needless syringes. Cell death was observed at 5-8 dpi. For the gene expression analysis, 0.1 mg of samples was detached at 2, 4, 6, and 8 dpi, respectively. The collected samples were quickly frozen in liquid nitrogen and stored in -80°C.

H₂O₂ accumulation was detected using DAB staining. The N. benthamiana leaves were detached at 3 dpi and incubated in DAB solution (1 mg/mL DAB, 1:2000 Tween-20, and 10 mM Na₂HPO₄), then vacuumed until the leaves were infiltrated with DAB solution (Du et al., 2022). The sample leaves were placed on a shaker at 60 rpm overnight, then distained in 95% ethanol and photographed.

For programmed cell death (PCD) observation, infiltrated leaves were incubated in trypan blue staining solution (0.067% w/v trypam blue, 11.11% w/v phenol, 1 volume of glycerol, lactic acid and distilled water, respectively, and 6 volume of 95% ethanol), boiled for 2 min, then transferred into chloral hydrate (1.25 g/mL). The sample leaves were places on a shaker at 60 rpm until they were fully destained (Shen et al., 2022).

The mature form of CLas4425 fused with red fluorescent protein (RFP) was cloned into PVX. For its entrance into the nuclei, the nuclear localization signal (NLS) was added at N-terminus of CLas4425:RFP. Leaves of N. benthamiana were collected at 3 dpi. Subcellular localization of CLas4425 was determined using an FV3000 confocal equipped with a UV light source (Olympus, Tokyo, Japan). H2B:green fluorescent protein (GFP) and Pm : GFP were used as markers for the nucleus and cytoplasmic membrane, respectively.

The TRV-based gene silencing system was conducted to silence BAK1, SOBIR1, and BIK1 in N. benthamiana. A. tumefaciens GV3101 strains carrying respective pTRV2 constructs were mixed with pTRV1 in equal ratios to a final OD₆₀₀ of 0.2. pTRV2:GUS served as a control, and pTRV2:PDS was used to visualize the silencing process (Liu et al., 2002). A. tumefaciens suspensions were infiltrated into cotyledons of 4-leaf-old seedlings. The silencing efficiency was validated with qRT-PCR of RNA from the leaves in the locations corresponding to the albino leaves.

Protein was extracted from infiltrated leaves using the Plant Protein Extraction Kit (Solarbio, Peking, China). Total protein was separated by 12% SDS-PAGE, and the protein samples were transferred to a polyvinlidene difluoride (PVDF) membrane. Anti-HA-tag served as primary monoclonal antibody at a dilution of 1:2000. HRP-conjugated goat anti-mouse IgG was used as secondary antibody at a dilution of 1:5000.

Epicotyl segments of Wanjincheng seedlings were used for A. tumefaciens-mediated transformation. The details of performance were as previously described (Peng et al., 2015). Transgenic shoots were determined by GUS staining before being micrografted onto Wanjingcheng seedlings in vitro. The resulting plantlets were further grafted onto C. limon in a greenhouse. The expression levels of CLas4425 in transgenic plants were validated with qRT-PCR.

To analyze the effect of CLas4425 on N. benthamiana immunity, N. benthamiana plants were infiltrated with CLas4425, NLS : CLas4425, and GFP, respectively one day before DC3000 ΔhopQ1-1 inoculation. DC3000 was prepared with OD_{600 =} 0.0001 (10⁴ cfu/mL) and injected into six N. benthamiana plants for each treatment. Leaves were detached at 3 dpi and used for symptom observation and bacterial growth analysis.

To determine the role of CLas4425in HLB pathogenicity, five independent CLas4425-overexpressed lines were used via a grafting method. WT Wanjingcheng plants of the same age were used as controls. Branches from each of these plants were inoculated by stems of CLas-infected sugar orange plants, with the presence of CLas confirmed by PCR. Leaf samples were collected monthly, and midribs from three mature leaves per plant were used for DNA extraction and the determination of CLas populations.

Quantification of H₂O₂ production was performed using Hydrogen Peroxide (H₂O₂) Content Assay Kit (Molfarming, Nanjing, China), and the OD was measured using Varioskan Flash Microplate Reader (Thermo Scientific, China). Fresh tissues were collected from fully mature leaves and finely ground with liquid nitrogen, and H₂O₂ accumulation was assayed following the manufacturer’s instructions. The experiment comprised three biological replicates and was repeated three times.

An in silico analysis suggested CLas4425 contained a 19 amino signal peptide in the N-terminal ( Figure 1A ). Accordingly, an alkaline phosphatase (phoA) fusion assay was employed to validate the export of CLas4425 via the Sec translocon (Liu et al., 2019). After overnight incubation, E. coli cells harboring pET-4425SP-mphoA turned blue on indicator LB medium as the positive control, whereas the negative control remained white ( Figure 1B ). Based on these data, CLas4425 was identified as a typical Sec-delivered effector.

To explore the importance of CLas4425 to CLas pathogenicity, we investigated its expression profiles in CLas-infected citrus and psyllids, as well as during the disease process. As a result, CLas4425 was highly expressed (ca. 16-fold) in infected citrus plants compared to that in psyllids ( Figure 1C ). In addition, CLas4425 could be detected at 2 months after CLas colonization, and its expression level stayed the same during disease development ( Figure 1D ). Our findings imply that CLas4425 plays a critical role in planta at early infection stage.

A transient expression system was applied to assess the potential role of CLas4425 in CLas virulence. The coding sequence of CLas4425, without the secretion signal, was cloned into the binary PVX vector pGR107. PVX-GFP and PVX-BAX served as negative and positive control, respectively. Compared to BAX, CLas4425 induced mild cell death symptom ( Figure 2A ). The progression, in which CLas4425-triggered cell death happened, was recorded ( Figures S1A, B ). To avoid the influence of PVX, we sequentially constructed pLGN-CLas4425. Cell death was also observed in CLas4425-infiltrated leaves ( Figure S1C ).

Protein functions are correlated with their subcellular localization. To visualize the distribution of CLas4425 in plant cells, we transiently expressed RFP and CLas4425:RFP in N. benthamiana leaves, and observed their localization pattern with a confocal microscope. Free RFP was found to be distributed in multiple subcellular compartments ( Figures 2B ; S2 ), while CLas4425:RFP was localized in the cytoplasm and cytoplasmic membrane ( Figure 2B ). To reveal the role of subcellular localization involved in CLas4425-triggered cell death, the nuclear localization sequence (NLS) (Liu et al., 2019) was fused in frame to CLas4425 in PVX ( Figure S2 ). Transient expression illustrated that NLS : CLas4425 could not induce cell death, and the protein expression was validated ( Figure 2C ). Therefore, localization in cytoplasm and cytoplasmic membrane is required for CLas4425-activated cell death.

Brassinosteroid insensitive 1-associated kinase 1 (BAK1), a suppressor of BIR1-1 (SOBIR1), and BIK1 participate in several types of PTI signaling pathways (Liebrand et al., 2014; Kadota et al., 2015). Moreover, BIK1 is necessary for bacterial resistance during ETI (Yuan et al., 2021). As a PAMP, INF1-induced cell death relies on BAK1 and SOBIR1 (Heese et al., 2007; Franco-Orozco et al., 2017). Here, we expressed CLas4425 in N. benthamiana silenced for BAK1, SOBIR1, and BIK1. The silencing of these genes was verified using qRT-PCR (p<0.001) ( Figure 2D ). CLas4425-triggered cell death developed in the TRV : BAK1, TRV : SOBIR1, and TRV : GUS treated plants. On the contrary, cell death was abolished in BIK1-silenced N. benthamiana plants. However, INF1-activated cell death was not altered in TRV : BIK1-treated plants. Immunoblot verified that GFP, INF1, and CLas4425 were correctly expressed in N. benthamiana. Conclusively, CLas4425-elicited cell death is dependent on BIK1.

DC3000 ΔhopQ1-1 was used to determine the virulence of CLas4425. We inoculated N. benthamiana leaves with DC3000 following the infiltration with either CLas4425 or NLS : CLas4425. qRT-PCR analysis revealed that the PVX accumulation level of PVX-CLas4425 made no significant difference to PVX-GFP in N. benthamiana ( Figure S3A ). N. benthamiana expressing CLas4425 exhibited more severe leaf blight than those infiltrated with NLS : CLas4425 and GFP ( Figure 3A ). Analysis of bacterial growth revealed that CLas4425 promoted DC3000 proliferation in N. benthamiana ( Figure 3B ). Since CLas4425 had a negative effect on N. benthamiana immunity reacting to DC3000 infection, we probed the molecular evidence for its interference with plant innate immunity. The expression of a PTI marker gene, NbWRKY29, and genes representing phytohormones were detected (Pang et al., 2020; Situ et al., 2020). The transcript of NbPR2 declined significantly from 4 dpi ( Figure 3C ), whereas the expression level of other detected genes did not vary dramatically ( Figure S3B ). Considering NbPR2 marked the SA pathway, we examined the expression manners of SA upstreaming genes, namely NbEDS1 and NbNPR1. As expected, their expressions were decreased in plants expressing CLas4425 ( Figure 3D ). Hence, CLas4425 probably impairs N. benthamiana defense via targeting the SA signaling pathway.

To assess the significance of CLas4425 in CLas pathogenicity, we generated CLas4425-transgenic Wanjincheng (4425-OE). The obtained transgenic plants were clarified using PCR, qRT-PCR, and GUS assay ( Figure S4A ), and there was no difference between 4425-OEs and wild-type group (WT) in growth and phenotype ( Figure S4B ). Among seven independent transgenic lines, we selected five of them with relatively higher expression of CLas4425 for further studies. The five 4425-OEs (one-year old) were graft-inoculated using CLas-infected stems. Five WTs of the same age served as a control group and were grafted in the same way. qPCR was applied monthly to monitor bacterial growth. Comparison of bacterial titers demonstrated that CLas growth was faster in 4425-OEs than in WTs at 2 (p<0.001) and 3 months post inoculation (MPI) (p<0.01), whereas no difference was observed after 4 MPI (p>0.05) ( Figure 4A ). Leaves from transgenic plants exhibited a blotchy mottle at 6 MPI, while WTs remained asymptomatic ( Figure 4B ). Previous reports documented the accumulation of starch granules and phloem blockage in infected citrus plants (Wang and Trivedi, 2013). We therefore detected the transcription of genes associated with phloem protein (CsPP2), sugar metabolism (CsAPS1), and starch biosynthesis (CsGBSS1 and CsSS1) (Kim et al., 2009). Consequently, CLas promoted the expression of all the detected genes in both WTs and 4425-OEs while CLas4425 enhanced the expression of CsPP2, CsAPS1, and CsGBSS1 in uninfected citrus plants. Compared to those in WTs infected with CLas, the expression of CsPP2 and CsAPS1 was upregulated in CLas-infected 4425-OEs ( Figure 4C ). To sum up, CLas4425 promotes CLas multiplication, and enhances HLB progression.

To understand how CLas4425 promotes CLas colonization, the H₂O₂ and SA levels in citrus were further investigated. The H₂O₂ content was not significantly different between 4425-OEs and WTs. Of interest, CLas stimulated H₂O₂ accumulation in WTs but not in 4425-OEs ( Figure 4D ). Herein, CLas4425 stimulated a slight increase of H₂O₂ in citrus plants, which is consistent with mild HR induced by CLas4425 in N. benthamiana. Moreover, CLas4425 hindered SA accumulation (p<0.05) ( Figure 4E ). We then detected the expression of SA signaling pathway genes. The expression of CsEDS1, CsNDR1, CsNPR1, CsPR1, CsPR2, and CsPR5 was depressed in 4425-OE lines ( Figures 4F, G ). Therefore, CLas4425 suppresses the plant immune system through interfering with the SA signaling pathway.