Bacterial strains kept frozen with 15% (v/v) glycerol at −80°C were streaked on Luria-Bertani (LB) agar plates and incubated at 30°C for 16 h. Then, a single colony of each strain was inoculated into 5 ml LB liquid medium and shaken at 30°C, 200 rpm for another 16 h. Antibiotics were supplemented where necessary at the following concentrations: kanamycin 50 μg/ml, tetracycline 50 μg/ml, gentamicin 25 μg/ml, streptomycin 50 μg/ml, and gentamicin 25 μg/ml (Table 1).

The spores of Botrytis cinerea were deposited at −80°C in 15% (v/v) glycerol. Ten microliters of fungal spores were dropped on a potato dextrose agar (PDA) plate and incubated at 25°C in the dark. When B. cinerea mycelia covered PDA plates, the fungal culture was used for inoculating B. xylophilus.

Cultures of the PWN, Bursaphelenchus xylophilus, were obtained from Dr. Hongtao Li at the Hebei Academy of Agriculture and Forestry Sciences and Dr. Kai Guo at Zhejiang A&F University. The fungus B. cinerea was purchased from the Shanghai Bioresource Collection Center (SHBCC).

Bursaphelenchus xylophilus was inoculated on a PDA culture of B. cinerea and incubated at 25°C in the dark until the fungal mycelia were completely consumed by B. xylophilus. Nematodes were collected using the modified Baermann funnel technique (Kitazume et al., 2018; Cesarz et al., 2019; Maehara et al., 2020) and washed with a mixture of 0.1% streptomycin sulfate and 0.002% actinone three times to remove surface microbial contaminants (Liu et al., 2016). Then, these nematodes were used for PWN-killing activity test and in planta biocontrol assays.

For microscopic image analysis of PWN morphology, about 10,000 nematodes were decanted into a burette containing 25 ml of 0.3% carboxymethyl cellulose (CMC) solution. After 12 h, the second-stage juveniles were collected from the top of the burette (Qiu et al., 2016, 2019). These worms were fed with B. cinerea. After 48 h, the L4 juveniles were collected in sterile water and washed with a mixture of 0.1% streptomycin sulfate and 0.002% actinone three times. These L4 juveniles were used for the image analysis.

Wild-type AA4, AA4ΔsdaB (derivative of wild-type AA4 that lacks sdaB), and C△sdaB (AA4ΔsdaB complemented with wild-type sdaB gene) kept frozen with 15% (v/v) glycerol at −80°C were streaked on LB agar plates with corresponding antibiotics (Table 1) and incubated at 30°C for 16 h. Then, a single colony of each strain was inoculated into 5 ml LB liquid medium and shaken at 30°C, 200 rpm for another 16 h. One milliliter of the culture of each strain was transferred into 50 ml LB liquid medium and shaken at 30°C, 200 rpm for 2–3 h until Optical Density at 600 nm (OD₆₀₀) value reached 0.6–0.8 (early stationary phase). Cells were harvested by centrifuging at 9,100 rpm (8,000 × g) at 4°C for 20 min (Thermo, Multifuge X1R, Germany) and washed by phosphate-buffered saline (PBS). The cell precipitations were resuspended in 5 ml PBS before sonication. An ultrasound processor (Sonics, VCX130PB, United States) was used to sonicate the resuspended cell suspensions. The probe was immersed into the suspensions and sonicated the cells on ice for 60 min at a power of 130 W (pulse duration: 2 s on, 1 s off). The resulting cell lysates were centrifuged at 15,300 rpm (14,000 × g) for 30 min. The supernatant was collected and used as cell extracts. The concentration of total protein in cell extract was measured by using a biuret protein assay reagents kit (Solarbio, PC0010, China).

The nematocidal activity of the 43 efficient PWN-killing bacterial strains of AA4ΔsdaB and CΔsdaB was evaluated by the fluorescent staining-based PWN-killing activity test, while Bacillus pumilus YLT40 and Paenibacillus polymyxa M-1 were utilized for confirming the reliability of the assay. The cell density of bacterial culture suspension was adjusted to OD₆₀₀ = 1.0 for each strain used in the test. Each well of the 96-well microplates was added with 40 μl of bacterial culture suspension and 20 μl of worm suspension containing 40–50 nematodes, while 40 μl of LB liquid medium or PBS and 20 μl of worm suspension was employed as negative control. Five wells were used for each treatment designated with the code of the strain listed in Supplementary Tables 1, 2. The inoculated microplates were incubated at 25°C in the dark for 24 h. Then, a plate washer (Biobase, BK-9622, China) was employed to change the liquid cultures with sterile water. 4’6-diamidino-2-phenylindole (DAPI, Solarbio, C0065, China) was used to stain the PWNs, and the final concentration of DAPI in each well was 5 μg/ml. After being stained for 1 h, the nematodes were observed by a confocal laser scanning microscope (Zeiss, LSM 800, Germany) using a excitation laser of 353 nm and collecting emission of 465 nm, where the dead worms displayed bright fluorescence while the live ones were dim.

To test the nematocidal activity of the cell extract of wild-type AA4, AA4ΔsdaB, and CΔsdaB, the extract containing total proteins of 5 mg/ml was mixed with 20 μl of worm suspension containing 40–50 nematodes in one well of a 96-well microplate. The remaining steps are the same as those for the assays with bacterial cell suspensions.

The mortality and corrected mortality of B. xylophilus were calculated as follows (Faria et al., 2013; Guo et al., 2016; Liu et al., 2019):

Two-tailed t-test (GrapPad Prism 8) was used for the statistical analysis. All these experiments above were repeated five times.

Z′factor, a combination of signal interval and variation, is a parameter used to evaluate the reliability of results obtained from an experiment for guiding a larger-scale experimentation. It is calculated as:

where σ_p and σ_n represent the standard deviations of positive control and negative control, respectively, and μ_p and μ_n represent the means of positive control and negative control, respectively. Z′factor was calculated by using the software Excel (Microsoft). The theoretical value of Z′factor is 1, and a Z′factor value between 0.5 and 1 indicates the outcomes of a given experiment is highly reliable, while a test of low reliability usually possesses a Z′factor value of less than 0.5 (Zhang X. D. et al., 2020).

Identification of the efficient bacterial PWN-killing strains were carried out by using 16S rRNA gene sequencing analysis. Genomic DNA was extracted from an overnight LB culture of each strain by utilizing a bacterial genomic DNA extraction kit (Tiangen, DP302-02, China) following the manufacturer’s protocol. The 16S rRNA gene was amplified in 25-μl PCR reactions using the universal primers 27F and 1492R (Table 2) (Monciardini et al., 2002; Niu et al., 2013; López and Alippi, 2019). The reaction mixture contained 1 μl of template DNA, 12.5 μl of master mix (Takara, RR350Q, Japan), 1 μl of each of the forward and reverse primers, and 9.5 μl of sterile deionized water. Amplifications were performed using a T100 Thermal Cycler (Bio-Rad, 621BR47532, United States) with the following cycle conditions: initial denaturation at 95°C for 5 min; 30 cycles at 94°C for 30 s, 58°C for 30 s, and 72°C for 1.5 min; and a final extension at 72°C for 10 min. Amplicons were purified using a PCR purification kit (Omega, D6492-02, China) and sequenced using 27F primer at Tsingke Biotechnology Co., Ltd. DNA sequences were inspected for base-caller errors and were trimmed by removing any ambiguous trailing or leading bases using the software GAP4 in the STADEN Package¹. The sequences were compared with those of the reference organisms by Basic Local Alignment Search Tool (BLAST) at the National Center for Biotechnology Information (NCBI) website² and by Sequence Match at the Ribosomal Database Project (RDP) website³. The 16S rRNA gene sequences of efficient nematocidal strains were deposited in the GenBank database. The accession numbers were listed in Supplementary Table 1.

The biological control effects of wild-type AA4, AA4ΔsdaB, and CΔsdaB were examined by using both 3-year-old and 1-month-old Scots pine (Pinus sylvestris) tree seedlings grown under greenhouse conditions. The 3-year-old seedlings were purchased from Longsheng nursery at Harbin. Ten milliliters of bacterial suspension (OD₆₀₀ = 1.0) were inoculated on the 3-year-old seedlings by spraying. Then, a 2–4 cm silt was made on the surface of bark located at approximately 10–15 cm above the soil. One hundred microliters of PWN suspension containing about 5,000 worms were injected into the trunk. A small piece of sterilized cotton was fixed on the trunk to cover the silt (Xue et al., 2019). The plants were placed in a greenhouse under the following conditions: 16 h of light (day) and 8 h of dark (night), 25°C, and a relative humidity of 70%. Plants were watered periodically and maintained in the greenhouse. Thirty days after inoculation, the symptoms present on the seedlings were recorded with a camera (Canon, 80D, Japan). The severities of PWD were evaluated based on the disease ranks described in earlier studies (Qiu et al., 2016, 2019) as follows: In rank 0, all needles are green. In rank 1, less than a quarter of needles turn yellow. In rank 2, 25–75% of needles turn yellow. In rank 3, more than 75% of needles turn yellow and less than 50% of needles get wilted. In rank 4, more than 50% of the needles get wilted. Five plants were used for each of the two treatments designated as PWN (inoculation of seedlings with PWN alone) and PWN + AA4 (inoculation of seedlings with PWN jointly with wild-type AA4), respectively.

For the experimentations with 1-month-old Pinus sylvestris seedlings, the seeds were surface-sterilized by immersing in 5% KMnO₄ for 60 min and placing in a growth chamber at 30°C in the light for 48 h. After germination, the seeds with roots were sown in the plant growth substrates prepared by mixing soil and vermiculite in a ratio of 3:1. Thirty days after emergence, 5 ml of bacterial suspensions (OD₆₀₀ = 1.0) or cell extracts (containing total proteins of 5 mg/ml) of wild-type AA4, AA4ΔsdaB, and CΔsdaB were sprayed on the needles of Scots pine seedlings, respectively. In the meantime, 100 μl of PWN suspension containing approximately 3,000 worms were injected into the needles by using a syringe. A small piece of sterile cotton was fixed on the needles to cover the pinhole. The plants were incubated in greenhouse under the same condition on which the 3-year-old seedlings grew. The number of dead seedlings and symptoms were recorded on the ninth day after inoculation. Ten plants were used for each of the 11 treatments designated as PWN (inoculation of seedlings with PWN alone), PWN + ΔsdaB (inoculation of seedlings with PWN jointly with AA4ΔsdaB cell suspension), PWN + CΔsdaB (inoculation of seedlings with PWN jointly with AA4CΔsdaB cell suspension), PWN + WT (inoculation of seedlings with PWN jointly with wild-type AA4 cell suspension), PWN + ΔsdaBCE (inoculation of seedlings with PWN jointly with AA4ΔsdaB cell extract), PWN + CΔsdaBCE (inoculation of seedlings with PWN jointly with AA4CΔsdaB cell extract), PWN + WTCE (inoculation of seedlings with PWN jointly with wild-type AA4 cell extract), PWN + ΔsdaBS (inoculation of seedlings with PWN jointly with AA4ΔsdaB culture supernatant), PWN + CΔsdaBS (inoculation of seedlings with PWN jointly with AA4CΔsdaB culture supernatant), PWN + WTS (inoculation of seedlings with PWN jointly with wild-type AA4 culture supernatant), and DW (treatment with distilled water), respectively.

The severities of PWD were evaluated based on the disease ranks shown in Supplementary Figure 3. The disease severity indices were calculated as follows:

where Xi is the number of pine trees in each disease rank, Ai is the disease rank, X is the total number of pine trees, and Amax is the maximum rank (Qiu et al., 2019; Xue et al., 2019). Two-tailed t-test (GrapPad Prism 8) was used for the statistical analysis. These experiments were repeated three times.

The cell extract of E. ludwigii AA4 was prepared as above. Forty microliters of cell extract containing total proteins of 5 mg/ml and 20 μl of worm suspension containing 40–50 L4 juveniles of PWNs were added in a single well of the 96-well microplates. At the meantime, the combination of 40 μl of PBS and 20 μl of worm suspension was employed as negative control. The microplates were stored at 25°C in the dark for 24 h. Then, the worms were picked and put into 10 μl of PBS by injector pinhead. After being washed three times, the worms were transferred to a slide. Ten microliters of PBS were then dropped on the samples. The morphological changes of B. xylophilus were observed and recorded by a light microscope (Olympus, BX43, Japan).

The sdaB gene knock-out mutant of E. ludwigii AA4 was created by using the Red recombinase system (Datsenko and Wanner, 2000; Dahyot et al., 2020; Guérin et al., 2020). Primers were designed according to the information obtained from the whole genome sequence of strain AA4 (Niu et al., 2017). The upstream border sequence of the AA4 sdaB gene was amplified from AA4 chromosomal DNA using primers sdaB-H1up and sdaB-H1down (Table 2), while the downstream border sequence of sdaB was amplified with primers sdaB-H2up and sdaB-H2down (Table 2). A kanamycin resistance cassette, flanked by 23-bp-long sequences homologous with sdaB gene, was amplified from plasmid pKD4 DNA with primers Kup and Kdown (Table 2). Another fragment flanked by 22-bp-long sequences homologous with sdaB gene was amplified from plasmid pUC19 by using primers pUC19up and pUC19down (Table 2). The purified PCR products were assembled by Gibson DNA assembly technology, using a SoSoo Cloning Kit (Tsingke, T-TSV-S1, China) following the manufacturer’s instructions and then cloned into Escherichia coli DH5α. Clones containing the resulting recombinant pU19ΔsdaB vector were selected on LB agar supplemented with kanamycin. The recombinant plasmid pUC19ΔsdaB was isolated with a plasmid extraction kit (Omega, D6943-02, China) following the manufacturer’s instructions and used as template DNA for PCR amplification with the primers sdaB-H1up and sdaB-H2down (Table 2). After purification, the PCR products were introduced into the competent cells of AA4 carrying a Red helper plasmid pKD46 by electroporation (Bio-Rad, 411BR11661, United States) at 2.5 KV. The resulting transformants were selected on LB agar supplemented with kanamycin after incubation for 24 h at 30°C. Homologous recombination was confirmed by PCR and sequencing. Then, the elimination of pKD46 was performed by culturing the correct transformant by shaking at 37–42°C, 200 rpm overnight. The knock-out mutant can only grow at 30°C on the LB agar supplemented with kanamycin but not streptomycin.

To construct a complement strain of AA4ΔsdaB, the complete open reading frame of sdaB gene was amplified by PCR with primers sdaB-DF and sdaB-UR (Table 2) from the genomic DNA of strain AA4 and linked with the fragment flanked by 22-bp-long sequences homologous with sdaB gene and amplified from plasmid pBBR1 using primers sdaB-PBBR1-DF and sdaB-PBBR1-UR (Table 2) by the Gibson DNA assembly technology. The complement plasmid pBBR1-sdaB was introduced into the competent cells of AA4ΔsdaB by electroporation at 2.5 KV, and the resulting transformants were selected on LB agar supplemented with kanamycin and tetracycline after incubation at 30°C for 24 h. The complementation was then confirmed by PCR and sequencing.

To create a random transposon library, E. ludwigii AA4 was transformed with pDL1093 via conjugation at 30°C and outgrown at 42°C in the presence of kanamycin to induce transposition of miniTn10 (Duncan et al., 2018).

For the sake of constructing fluorescence protein-labeled AA4, plasmid pGFP78, obtained from Dr. Qi Wang at China Agricultural University, harboring the gene encoding for green fluorescent protein (GFP), was introduced into the competent cells of wild-type AA4, AA4ΔsdaB, or CΔsdaB by electroporation at 2.5 KV. The resulting transformants were selected on LB agar supplemented with tetracycline, or kanamycin and tetracycline, or a combination of kanamycin, tetracycline, and gentamicin where necessary after incubation at 30°C for 24 h. The correctness of the transformants were then confirmed by PCR, sequencing, and fluorescence microscopy.

The cell extract of E. ludwigii AA4 was prepared as above, except washing and resuspending in the extraction buffer (50 mM K₂HPO₄, 2.5 mM serine protease inhibitors) (Velayudhan et al., 2004). Three aliquots of 1 ml of cell extract containing total proteins of 10 mg/ml was incubated with 500 μl of 300 mM L-serine at 37°C for 6 min. Then, the concentration of pyruvate was determined by using a pyruvate assay kit (Solarbio, BC2205, China). The L-serine hydrolase activity was calculated as:

where c is the concentration of pyruvate in μg/ml, v is the total reaction volume in ml, M is the relative molecular mass of pyruvate, T is the reaction time in minutes, and m is the weight of cell extract in mg.

Two-tailed t-test (GrapPad Prism 8) was used for the statistical analysis. This experiment was repeated five times.

In order to investigate the ability of E. ludwigii AA4 to colonize Scots pine needles, 50 ml of LB cultures (OD₆₀₀ = 1.0) of GFP-labeled wild-type AA4, AA4ΔsdaB, and CΔsdaB were sprayed on 21 1-month-old seedlings, respectively. The plants were placed in a greenhouse under the following conditions: 16 h of light (day) and 8 h of dark (night), 25°C, and a relative humidity of 70%. Three aliquots of 50 mg of needles were sampled at the 4th, 48th, 72th, 96th, 120th, and 144th, and 168th hour post inoculation, respectively, for bacterial quantification. The needles were crushed with 1 ml of PBS by using a sterile mortar. The resulting suspensions were diluted and spread on LB agar plates supplemented with tetracycline and incubated at 30°C for 16 h. The colony-forming unit (CFU) numbers were recorded. Two-tailed t-test (GraphPad Prism 8) was used for the statistical analysis. This experiment was repeated five times.

For image analysis, 48 h after inoculation, the needle slices were washed with PBS and transferred to a slide. Ten microliters of PBS were dropped on the samples. The colonization of needles by bacteria was visualized by a Confocal Laser Scanninc Microscopy (CLSM; Zeiss, LSM 800, Germany) using an excitation laser of 480 nm.

To find microorganisms possessing strong nematocidal activity against B. xylophilus for potential biological control of the PWD, we performed a PWN-killing test with 374 bacterial strains from our lab collection. First, we used an efficient PWN-killer Bacillus pumilus YLT40 and a moderate PWN-killing strain Paenibacillus polymyxa M-1 (Niu et al., 2013), identified previously (data not shown), as controls (Figure 1A) to verify if the PWN-Bacterium interaction system employed in this study was reliable for a large-scale screening. The corrected mortality rates of the nematode B. xylophilus treated by B. pumilus YLT40 and P. polymyxa M-1 were more than 85% and less than 25% (Figure 1A), respectively. We calculated a Z′factor score of 0.682 (Figure 1A) which indicated that the screening procedure utilized is sufficiently reliable for selecting powerful PWN-killing strains.

We detected 43 efficient nematocidal bacterial strains exhibiting B. xylophilus mortality rates above 85% (Figure 1A, Supplementary Figure 1, and Supplementary Table 1). Based on 16S rRNA gene sequencing, we found that these bacterial strains belong to 21 species from seven genera, including Bacillus, Pseudomonas, Enterobacter, Chryseobacterium, Stenotrophomonas, Pantoea, and Paenibacillus (Figure 1B), all genera with known nematocidal activity. Among the seven genera, Bacillus harbored 19 nematocidal strains, while Pantoea and Paenibacillus possessed only one PWN-killing strain (Figure 1B). Interestingly, seven Enterobacter strains, including the plant beneficial bacterium E. ludwigii AA4 isolated from maize roots (Niu et al., 2017), exhibited high nematocidal activity against the PWN (Figure 1B). In addition, AA4 did kill almost all of the B. xylophilus clones (96%) in our assay (Figure 1C). In our collection, we found 43 strong PWN-killing strains involving E. ludwigii AA4. Next, we validated the PWN-killing function of strain AA4 in both in vitro and in planta experiments.

To confirm the nematocidal effect of E. ludwigii AA4 against B. xylophilus, we developed a fluorescent dye-exclusion viability test to visualize PWN-killing activity. We used DAPI which differentially stains live and dead PWNs. Dead worms take up DAPI rapidly and show bright fluorescence, while the live ones are relatively non-fluorescent and look dim (Figure 2A).

In our assay, the majority (68%) of PWNs co-cultured with wild-type E. ludwigii AA4 cells in very low density (OD₆₀₀ = 0.063) for 24 h were killed (Figure 2B). The corrected mortality rate of the nematodes treated with wild-type AA4 cells tends to increase with enhanced cell density. The percentage of dead worms dramatically increased from less than 70% to more than 92% when the OD₆₀₀ value of the bacterial culture was raised from 0.063 to 0.5. The corrected mortality rate rose slowly to a maximum of 98.3% when the OD₆₀₀ value of the cell suspension reached 2 (Figure 2B). Our visual test documented that nearly all the nematodes incubated with wild-type AA4 emitted intense fluorescent signals, while almost no fluorescence was detected without bacteria (Figure 2A). These results well corroborated the high mortality rate of B. xylophilus when exposed to wild-type AA4. E. ludwigii AA4 exhibited a remarkable in vitro PWN-killing effect.

We then investigated the disease-suppressing efficacy of E. ludwigii AA4 against PWD on 3-year-old Pinus sylvestris greenhouse-grown seedlings. Thirty days after inoculation, we found that the needles of the seedlings treated with PWNs together with wild-type AA4 were much greener and less wilted than those inoculated with B. xylophilus only (Figure 3A). The disease index of PWD on the P. sylvestris plants treated with PWNs together with wild-type AA4 (19.5) was significantly lower than that treated with B. xylophilus alone (86.9) (Figure 3B), indicating that the PWD severity was substantially reduced by treatment with wild-type AA4. E. ludwigii AA4 was capable of efficiently suppressing PWNs not only under in vitro conditions but also in in planta and was a strong bacterial biocontrol strain against PWD caused by B. xylophilus.

To further understand the PWN-killing effect of E. ludwigii AA4, culture supernatants and cell extracts were prepared for the DAPI-staining based nematocidal activity test. Treatment of B. xylophilus with the culture supernatant of wild-type AA4 did not affect the mortality rate of the nematode (Figure 2A). Wild-type AA4 cells did dramatically enhance fluorescence of the B. xylophilus nematodes (Figure 2A). This result was corroborated by the mortality rate of the PWNs, suggesting that the death rate of worms exposed to wild-type AA4 culture supernatant (less than 20%) was significantly less than that of B. xylophilus incubated with cells of wild-type AA4 (more than 60%) (Figure 2B). By contrast, we detected intense fluorescent signals from the worms treated with wild-type AA4 cell extract (Figure 4A) and found that more than 80% of PWNs were dead after being exposed to the extract (Figure 4A), indicating its strong nematocidal effect. Thus, our data indicates that the cell extract of wild-type AA4 rather than the culture supernatant contains the nematocidal activity.

To elucidate the mode of action of the nematocidal effect of E. ludwigii AA4 against PWN, we choose its cell extract for investigating morphological variations of AA4-treated B. xylophilus. The L4 stage nematodes were incubated with wild-type AA4 cell extract. Afterward, the morphological changes of the worms were observed and recorded using an optical light microscope. Exposure of PWN to cell extract-induced formation of vast amounts of vacuoles filled the internal tissues of nematodes completely (Figures 4Bi,iii). Newly formed small vacuoles accumulate and progressively fuse into giant ones, leading to membrane rupture and death (Maltese and Overmeyer, 2014; Armenta and Dixon, 2020). This process is defined as methuosis (Cai et al., 2013; Chen et al., 2014; Nirmala and Lopus, 2020), a type of non-apoptotic cell death widely described in animals (Jang et al., 2016; Rajasekharan et al., 2017; Song et al., 2021). The death of B. xylophilus caused by AA4 might be attributed to methuosis induced by the damage to cell membranes.

In these vacuolated worms, some vital organs, e.g., intestine, pharynx, and genital glands, disappeared in the mass of bubbles. We also failed to capture a clear structure of stylet at the head of PWN treated with wild-type AA4 cell extract (Figures 4Bi,iii). The reproductive organ of male worms, the copulatory spicule, dispersed after being incubated with cell extract (Figure 4Bv). In contrast, we observed fewer vacuoles in the control nematodes exposed only to PBS. All the critical internal organs were intact and clearly visible in the control worm bodies (Figures 4Bii,iv,vi). Therefore, E. ludwigii AA4 appeared to destroy crucial internal organs of B. xylophilus through causing a series of morphological defects.

To resolve the molecular mechanisms underlying the strong nematocidal effect of E. ludwigii AA4, we characterized genes involved in its PWN-killing efficacy using transposon and PCR-targeted mutagenesis. First, by screening a miniTn10 transposon mutant library of strain AA4, we identified a mutant with greatly reduced nematocidal activity against B. xylophilus. We further analyzed the genomic site where the transposon was located and found a miniTn10 element that was inserted into an open reading frame (ORF) located between kilobase positions 3,956,912 and 3,955,545 (Supplementary Figure 4). This ORF was identified as the sdaB gene encoding the L-serine dehydratase beta subunit, indicating that the sdaB gene might be involved in the PWN-killing effect of AA4.

We next created a knock-out mutant of the sdaB gene by employing the Red (λ, β, exo)-mediated DNA homologous recombination system. The L-serine hydrolase activity of the sdaB gene knock-out deletion mutant AA4ΔsdaB was significantly lower than that of the wild-type AA4 or the AA4ΔsdaB mutant strain complemented with the sdaB wild-type gene (Figure 5B), which indicated that the interruption of the sdaB gene affected the biosynthesis of L-serine dehydratase. Then, we compared the PWN-killing activities of the wild-type AA4 and AA4ΔsdaB by using the fluorescent staining-based nematocidal activity assay (Figure 2A). Nearly no fluorescence could be detected in the worms treated with AA4ΔsdaB cells or their extracts, which confirmed the loss of nematocidal activity of the miniTn10 insertional mutant. Complementation of the mutant with a recombinant plasmid harboring sdaB partially restored the PWN-killing activity (Figure 2A). These results were corroborated by the nematode mortality curves, which documented that the corrected death rate of the PWNs treated by AA4ΔsdaB was significantly lower than those incubated with wild-type AA4, while the mortality of nematodes exposed to the mutant strain complemented with the sdaB wild-type gene increased remarkably (Figure 2B). Therefore, sdaB gene product has a major impact on the nematocidal activity of E. ludwigii AA4.

Then, we investigated the role of the sdaB gene product on the protective effect of E. ludwigii AA4 on P. sylvestris against PWD. Using a Pine-Nematode-Bacterium tripartite interaction system with 1-month-old P. sylvestris seedlings, a lower disease index was estimated when wild-type AA4 cells or their extracts were added to the system. Under these conditions, much less wilted and fewer chlorotic pine seedlings were observed compared to the control containing B. xylophilus alone (Figure 6B). We found more wilted and chlorotic seedlings when AA4ΔsdaB cells or their extracts were added instead of wild-type AA4 cells or their extracts. The number of diseased plants was reduced when the seedlings were treated with PWNs together with cells or cell extracts of the AA4ΔsdaB mutant strain complemented with the sdaB wild-type gene (Figure 6A). These findings were supported by disease severity analysis, where the disease index of plants treated with PWNs and AA4ΔsdaB mutant cells or cell extracts was significantly higher than that with PWNs treated with the cells or cell extracts of wild-type AA4 or the sdaB mutant complemented with the sdaB wild-type gene (Figure 6B). The indices of the PWN infected seedlings treated with cells or cell extracts of the AA4ΔsdaB mutant complemented with the wild-type sdaB gene and seedlings treated with PWNs plus the wild-type AA4 strain cells or cell extracts showed no significant differences. The culture supernatants of the wild-type AA4, the AA4ΔsdaB mutant strain, or AA4ΔsdaB mutant complemented with the sdaB wild-type gene not only exhibited nearly no nematocidal activity against B. xylophilus in vitro, but also failed to display biocontrol efficacy against PWD (Figure 6). The sdaB gene encoding the beta subunit of L-serine dehydratase, is highly related to the PWN-killing activity of E. ludwigii AA4 and may be a key molecular element controlling the interplay between strain AA4 and PWNs.

Efficient colonization of plant surfaces and tissues is fundamental for successful inhibition of pathogens by biological control agents. To investigate the ability of E. ludwigii AA4 to colonize P. sylvestris, we sprayed the needles with a GFP-labeled transgenic strain of AA4. We observed by CLSM that strain AA4 was capable of adhering to the surface of P. sylvestris needles by forming robust biofilms (Figure 7A). Although there was a dramatical reduction of the abundance of AA4 cells sticking to the needles, from around 1.9 × 10⁷ CFU (colony-forming unit)/g (of pine needle fresh weight) 4 h after inoculation to 3.8 × 10⁴ CFU/g 144 h after inoculation, the colonization rates of AA4 at days 6 and 7 (2.8 × 10⁴ CFU/g) showed no significant difference (Figure 7B). This indicates that E. ludwigii AA4 can colonize the needles of P. sylvestris, presumably a precondition for its disease-controlling efficacy against PWD. AA4ΔsdaB cells were equally able to colonize pine needles compared to the wild-type AA4 (Figure 7A), demonstrating that the sdaB gene function is not needed for colonization. The biomass accumulation rate of both strains was at the same level along our 7-day-long experiment (Figure 7B). Therefore, E. ludwigii AA4 may be an efficient colonizer of pine needles, and that the eradication of sdaB gene did not affect its adherence to needles.