A Novel Robust Screening Assay Identifies Pseudomonas Strains as Reliable Antagonists of the Root-Knot Nematode Meloidogyne incognita

Forty-four bacterial strains isolated from greenhouse soil and beetroots were tested for their antagonistic activity against the plant-parasitic root-knot nematode (RKN) Meloidogyne incognita, which causes significant yield losses in a number of important crops worldwide. Through a novel combination of in vitro and on planta screening assays, Pseudomonas spp. 105 and 108 were identified as the most promising bacterial isolates. Both strains were evaluated for their potential to control different RKN population densities and as root protectants against nematode infestation. Regardless of the application method, both strains significantly reduced root galling caused by M. incognita. These two strains were subjected to whole genome sequencing and de novo genome assembly as a basis for phylogenetic and future functional characterization. Phylogenetic analysis revealed that both Pseudomonas strains cluster within the Pseudomonas fluorescens clade among previously characterized RKN antagonists and Pseudomonas-based biocontrol agents of plant diseases.


Introduction
The plant-parasitic nematode Meloidogyne incognita is one of the most common rootknot nematode (RKN) species worldwide [1,2]. This group of obligate, sedentary plant parasites has a wide host range and can cause significant damage to several important agricultural and greenhouse crops, monocots, and dicots [3,4]. M. incognita infections can cause reduced plant growth, stunting, leaf discoloration, and even complete crop loss. The current estimate of the global yield losses caused by plant-parasitic nematodes places the annual losses in the range of billions of dollars [5], higher than those estimated for insects [6][7][8] and accounting for 14% of the total global crop losses [9].
RKN can reproduce both sexually (amphimixis) and asexually (mitotic and meiotic parthenogenesis). During their life cycle, second-stage juveniles (J2) enter the root tip near the elongation zone and migrate intercellularly to establish a permanent feeding site by inducing the formation of giant cells in the vascular cylinder [10]. While depleting plant nutrients and causing root galling, the juveniles molt and develop a swollen, pear-shaped appearance. Mature female nematodes lay their eggs in a gelatinous matrix outside the root system, allowing the J2 to hatch freely in the soil and find new root tips for propagation. Apart from the adult males, only the infective J2 are mobile and can move in the soil after USA) agar plates (15 g/L Agar-Agar, Kobe I (Carl Roth, Karlsruhe, Germany)). Plates were incubated at 26 • C for 24 h; single colonies were picked and subcultured to ensure pure cultures. Bacterial isolates were identified at the genus level with MALDI-TOF by smearing bacterial cell material from a colony on agar plate onto a MALDI targe for the identification of specific biomarker peptides [30] and 16 S rRNA gene sequencing. Forty-four bacterial isolates (Supplementary Table S1) were then tested for their antagonistic potential against M. incognita.

Antagonistic Screening against Meloidogyne incognita
The bacterial isolates were freshly prepared on TSB (Oxoid/Thermo Fisher Scientific, Waltham, MA, USA) agar plates (15 g/L Agar-Agar, Kobe I (Carl Roth)). Each bacterial strain was cultured overnight at 26 • C in a 250 mL Erlenmeyer flask containing 100 mL TSB medium. Liquid cultures were centrifuged at 6000 rcf for 10 min at room temperature, supernatant decanted, and bacterial pellet resuspended in autoclaved tap water to a final optical density (OD 600nm ) of 0.5.
Nematode:bacteria mixtures were prepared by pipetting 6 mL of a water suspension containing 250 M. incognita J2 into a 25 mL plastic cup and adding 4 mL of the freshly prepared bacterial suspension to a final volume of 10 mL. The nematode:bacteria suspensions were incubated for three days at 20 • C in the dark. Thereafter, 20 mL of steamed soil:silver sand mixture (1:3, v/v) was added to the cups containing the nematode:bacteria suspensions. A three-day-old germinated cucumber seedling (Cucumis sativus cv. Sprinter F1) was then planted in each cup and grown in a climate chamber at 24 • C, 60% humidity, and a 16/8 h light/dark cycle and watered as needed (n = 3). After three weeks, cucumber roots were washed free of soil, and root gall formation caused by viable M. incognita J2 was determined according to Zeck [31]. After the initial screening, the same experimental setup was repeated for the pre-selected bacterial strains with nematicidal activity.

In Vitro Screening of Selected Bacterial Strains against Meloidogyne incognita Second-Stage Juveniles
The ability of the selected bacterial strains and the negative control Pseudomonas orientalis F9 [32] to inhibit M. incognita J2 was tested in vitro using 24-well plates (n = 3). Bacteria were processed as described above but adjusted to a final OD 600 of 1.0. Each of the 24 wells was filled with 1 mL of a nematode suspension containing approximately 150 J2 and 1 mL of the OD 600 = 1.0 bacterial suspension (OD 600 of 0.5).
For the exposure of the J2 to the bacterial supernatant, the supernatant was filtered through a 0.2 µm syringe filter and 1 mL was added to a suspension containing 150 J2. The well plates were kept at 20 • C in the dark. For each well, the motility of the first 100 J2 individuals was scored under a light microscope as normal motility, affected motility, or immotile (elongated shape) [33].

Testing of Pre-Selected Antagonistic Pseudomonas Strains under Soil Conditions
The nematicidal activity of the bacterial isolates showing promising effects in vitro and on planta, as well as the control P. orientalis F9, was further tested under soil conditions in a greenhouse using cucumber plants as indicator plants. Pots (∅ = 14 cm) filled with 750 mL of steamed soil:silver sand mixture (1:3, v/v) were inoculated with ca. 4000 J2. Three days after nematode inoculation, 25 mL bacterial suspensions adjusted to an OD 600 of 0.5 were prepared as described above and added to the nematode-infested soil, except for the bacteria-free controls (n = 8). The pots were then covered with plastic foil and kept moist. Seven days after the application of the bacteria, 3-week-old cucumber plants were planted in the soil. The treatments were arranged in a randomized block design in the greenhouse and kept at 25 • C/19 • C, 60% humidity, and a 15/9 h light/dark cycle. Plants were watered according to their needs. After 4 weeks, the plants were uprooted and washed with tap water to remove soil for root gall rating [31].
The experimental setup was repeated for the selected bacterial strains (n = 7), now using 4-week-old tomato cv. Moneymaker plantlets as indicator plants and testing the bacterial suspensions at OD 600 = 0.5 and 1.5.

Evaluation of the Biocontrol Efficacy of Top Pseudomonas Isolates against Meloidogyne incognita by Application to Mineral Wool Plant Starting Plugs
Grodan SBS mineral wool plugs (SBS 36/77, Grodan, Roermond, The Netherlands) were used to grow cucumber seedlings. After 14 days of growth, the plugs were soaked overnight with 10 mL of selected bacterial suspensions (OD 600 of 0.5). Bacteria-soaked cucumber plugs were then planted in 14 cm ∅ pots. Three days earlier, the soil:silver sand mixture of these pots had been inoculated with approximately 4000 M. incognita J2. Each treatment (selected Pseudomonas strains and control, P. orientalis F9) was replicated seven times (n = 7). Plants were arranged in a randomized block design in the greenhouse and grown at 25 • C/19 • C, 60% humidity, and a 15/9 h light/dark cycle. After four weeks, plants were uprooted and washed with tap water to remove substrates for root gall rating [31].

Evaluation of the Biocontrol Efficacy of Selected Pseudomonas Strains against Different Population Densities of Meloidogyne incognita
The top Pseudomonas strains were tested at low (2000 J2) and high (8000 J2) M. incognita soil concentrations (n = 8 for each treatment and nematode concentration, including 8 untreated control plants). Pots (∅ = 14 cm) were prepared with steamed soil:silver sand mixture (1:3 v/v) inoculated with J2. Bacterial suspensions were prepared as described above, but resuspended to a final OD 600 of 1.5. Tomato cv. Moneymaker seedlings grown in transplanting trays for 4 weeks were planted in infected pots and arranged in a randomized block design. Tomato growth was measured weekly. Forty days after planting, root galling was evaluated according to Zeck [31].

Genomic DNA Extraction, Sequencing, De Novo Assembly, and Annotation
Total DNA was isolated from bacterial cells grown in TSB overnight according to [34]. The quality and quantity of the extracted DNA was assessed on a 0.8% (w/v) agarose gel, followed by the Qubit dsDNA GR assay (Life Technologies, Thermo Fisher Scientific, Waltham, MA, USA).

Phylogenetic Analysis
A phylogenetic tree based on 70 Pseudomonas strains, including the two de novo assembled Pseudomonas strains 105 and 108, was calculated based on 16 core genes (rpsJ, glnK, rplK, infA, rpsE, rplV, rplP, rpsG, rpsC, rpsR, rpsL, mreB, ihfA, rpsK, rplN, rpsU) identified by Page et al. [48]. Multi-sequence alignment of all genes was performed with the ClustalW algorithm [49] using MEGA-CC [50], and a phylogenetic tree was inferred using the Maximum Likelihood method with MEGA-CC and 100 bootstrapping iterations. The 68 strains were selected based on a literature search of available Pseudomonas genomes with putative activity against nematodes, a previous study of Pseudomonas isolates with biocontrol potential that placed them in the context of known Pseudomonas subclades [51], and the five closest genomes of the two isolates reported here (based on sequence identity as estimated by Mash [52] of all publicly available strains on NCBI). All sequences were downloaded from the NCBI database.

Data Analysis
Statistical analyses were performed using the SPSS software, data were tested for homogeneity of variances (Levene's test), and treatments were distinguished by one-way ANOVA with Tukey's Honestly Significant Difference (HSD) post-hoc test (p ≤ 0.05). Results of the in vitro assay using cells and supernatant were expressed as percentages. Significant differences were calculated by comparing data from J2 with normal motility.

Screening of Bacterial Strains with Potential Antagonistic Activity against Root-Knot Nematodes
The antagonistic potential of 44 bacterial strains (Supplementary Table S1) against M. incognita was tested with a newly developed method using a combined "in vitro" onplant cucumber assay ( Figure 1). The screening method was designed to ensure that the J2 were exposed to the bacteria in an aqueous solution and also had the opportunity to infect a host plant. Therefore, in this method, an inhibitory effect of the tested bacteria on the J2 nematodes results in a reduction in or even the absence of root gall formation in the indicator plants. Thus, the effect of the bacteria on J2 plant infectivity can be assessed even if the J2 are visually unaffected. This is an advantage over studies with RKN antagonists, which rely solely on a visual assessment of the viability of the nematodes [53][54][55].
Preliminary characterization revealed that most of the bacterial isolates tested belonged to the genus Pseudomonas (Supplementary Table S1). P. orientalis F9 was included in the study to assess its activity against nematodes, as this strain is known to have antagonistic potential against the fire blight pathogen Erwinia amylovora and also against the oomycete Pythium ultimum [32].
The antagonistic activity of the selected bacteria was determined by the root gall index recorded on cucumber roots caused by M. incognita J2 exposed to the selected bacterial strains and compared to an infection control, in which no bacterial antagonists were added to the J2-infected soil ( Figure 2). The most promising antagonists of M. incognita, i.e., strains 102, 105, 108, 112, 119, and 157, caused a twofold or greater reduction in the galling index.  Pseudomonas strains (102, 105, 108, 112, 119, and 157) were retested for their ability to antagonize RKN (Figure 3). Based on the results of the initial screening, P. orientalis F9 and isolate 113 were added as negative controls for bacterial antagonism since neither strain showed inhibitory effects on RKN-induced galling (Figure 2). With the exception of strain 112, strains 102, 105, 108, 119, and 157 repeatedly showed RKN antagonistic activity, while P. orientalis F9 and isolate 113 again had no significant effect on J2 (Figure 3). The results confirmed the reproducibility of the data obtained in the cucumber assay.
Bacterial cells and sterile filtered supernatants of Pseudomonas strains 102, 105, 108, 112, 119, and 157 were then tested against J2 in a pure in vitro assay to compare the antagonistic effects that occurred with the results obtained from the cucumber experiments. The 7-day in vitro assay showed that, for all strains, bacterial cells added to a J2 suspension significantly inhibited the nematodes after 1 day (Figure 4). However, 4 days (strains F9 and 113) or 7 days (strains F9, 112, and 113) after the application of the cell suspension, the J2 recovered and showed no significant differences compared to the water control. These results were even more pronounced when the supernatant of these strains was tested ( Figure 4). Accordingly, the results support the notion that visual screening at a too early time point may give a misleading indication of RKN control in soil, as was observed with P. orientalis F9 and isolates 112 and 113. These negative controls in the cucumber assay significantly inhibited M. incognita J2 after one day when applied either as a cell suspension or its corresponding supernatant (Figure 4). In the literature, similar RKN in vitro assays using Pseudomonas spp. are typically evaluated after 12, 24, 48, and/or 72 h [56,57], and therefore some important information may be overlooked compared to the newly developed screening assay outlined in Figure 1. incognita second-stage juveniles (J2) after 1, 4, and 7 days. J2 motility was recorded according to normal motility, affected, or immotile nematodes [33]. Error bars represent standard deviations of replicates. * Significantly affected or immotile nematodes in percent (%) relative to the control, calculated using a one-way ANOVA with post-hoc Tukey HSD test (n = 3).

Testing of Promising Meloidogyne incognita Bacterial Antagonists under Greenhouse Conditions
Pseudomonas strains 102, 105, 108, 119, and 157 were evaluated for their potential to control RKN under soil conditions. P. orientalis F9 was selected as a negative control for the soil experiment. The bacterial cultures were applied to M. incognita-infested soil and cucumbers were planted seven days after application.
Only cucumber roots grown for four weeks in nematode-infested soil and treated with the 105 or 108 strains showed a significant reduction in root galling compared to cucumbers grown in untreated soil or with other Pseudomonas strains ( Figure 5). The results showed no significant effect on root gall formation for bacterial strains 102, 112, 119, and 157. It may be that the application method used reduced the RKN control performance of the strains tested, as suggested in previous publications [58]. The selected Pseudomonas strains 102, 105, 108, 119, and 157 were further tested as a root bale treatment. Cucumber seedlings grown on mineral wool plugs ( Figure 6) were planted in M. incognita-infested soil. Prior to planting, the mineral wool was inoculated with the selected strains. Figure 6. Selected Pseudomonas strains were tested as root protectants against M. incognita J2 by applying bacterial suspensions to cucumber seedlings grown in mineral wool one day before planting in nematode-infested soil (4000 J2/pot). Cucumber plants were grown for 28 days and root galls were rated according to Zeck's [31] 0-10 scale, where 0 indicates no galls and 10 indicates dead roots.
Pseudomonas strains 102, 105, 108, and 119 showed a significant reduction in root galling compared to control plants ( Figure 7). As in the previous experiments, Pseudomonas strains 105 and 108 showed the strongest reduction in the root galling index. Pseudomonas spp. are known to colonize plant roots [26,57]; therefore, the bacterial treatment of the mineral wool may have promoted the bacterial colonization of the cucumber roots, resulting in protection against M. incognita infection. The ability to control M. incognita J2 using tomato as an indicator plant was tested with the most promising isolates, Pseudomonas strains 105 and 108, and the negative control, P. orientalis F9. Two suspensions with an OD 600 of 0.5 or 1.5 were compared (Table 1). Root gall formation in tomatoes treated with strain 105 or 108 was reduced regardless of the cell concentration, but a significant reduction in root galling was only achieved with the OD 600 = 1.5 bacterial suspension.
Previous studies have shown that Pseudomonas fluorescens strain CHA0 responds specifically to plant species, age, and genotype when tested against M. incognita [59]. Therefore, we hypothesized that Pseudomonas strains 105 and 108 may share a similar trait as CHA0 and act somewhat differently on different crop plants.
To evaluate whether Pseudomonas strains 105 and 108 maintained their control effect at different nematode population densities, their effect was tested in pots inoculated with 2000 or 8000 J2/pot. Both strains showed control of RKN in soil. However, in soil with a higher nematode population density (8000 J2), only strain 105 significantly reduced root gall formation in tomato (Figure 8). Despite the fact that RKN control was significant in our experiments, the control effect was not as strong as, for example, that obtained for Pseudomonas simiae strain MB751. This strain promoted up to an 80% reduction in root galling from 6.3 to 1.2 [60]. However, when compared to the recent study by Zhao et al. [57], the control effect of strains 105 and 108 was stronger than that observed for Pseudomonas protegens strains Sneb1997 or Sneb2001. The beneficial effect of RKN control by Pseudomonas 105 and 108 strains was also observed on tomato plant height development under soils infested with low (2000 J2/pot) and high (8000 J2/pot) population densities of M. incognita (Supplementary Figure S1). The tomato plant height was lower for plants grown in M. incognita-infested soil than for plants grown in M. incognita-infested soil but treated with Pseudomonas 105 or 108 strains. However, plants performed best when the soil was not inoculated with M. incognita. Similar results were observed when the biological nematicide BioAct or the chemical nematicide fluopyram were applied, as neither nematicide was able to restore the same yield and plant growth as M. incognita-free soil [61]. Based on our experimental design, we cannot conclude that Pseudomonas strains 105 or 108 support plant growth and/or seed germination in the absence of nematodes, as reported for P. protegens [57].

Sequencing and De Novo Assembly of the Complete Genome of Pseudomonas 105 and 108
To create an optimal basis for the phylogenetic placement of the strains (see below) and for future functional genomics studies to uncover potential mechanisms of action, the complete genomes of both strains were sequenced and de novo assembled. Motivated by recent studies that have demonstrated that Illumina short-read-based genome assemblies can lack important genes that may underlie antagonistic activity, such as non-ribosomal peptide synthetases, phenazine biosynthesis genes, and type six secretion system effectors [62], a combination of long reads (Oxford Nanopore Technologies) and short reads (Illumina) was used (see Methods). Illumina reads were mainly used for polishing and to assemble potential plasmids. The complete genomes of the strains consisted of one chromosome and one plasmid, respectively, and were subsequently annotated with a local installation of the NCBI's PGAP software (see Methods; Supplementary Table S2). The analysis of the longest repeats indicated that both strains can be classified as difficult to assemble class III genomes [46]; Illumina reads alone would not have allowed us to assemble a complete genome. Finally, an analysis with AntiSmash v6.0.1 [63] predicted several biosynthetic gene clusters of potential relevance (see Supplementary Figure S2A,B), including a lokisin NRP biosynthetic gene cluster, which was shown to exert anti-fungal activity [64]. Moreover, a lokisin derivative was produced in a multispecies bacterial community where the gene and metabolite expression changed depending on the composition [65], underlining the relevance of multispecies consortia.

Phylogenetic Analysis of Pseudomonas Strains 105 and 108
Phylogenetic analysis using 68 Pseudomonas reference strains available in the NCBI database confirmed that both bacterial isolates, 105 and 108, belonged to the genus Pseudomonas. Both isolates clustered within the P. fluorescence subgroup together with previously identified Pseudomonas isolates with potential for nematode biocontrol (Figure 9, Supplementary Table S3). However, strains 105 and 108 were placed in separate subclades of the P. fluorescence group.
Strain 105 was closely grouped with the characterized Pseudomonas brassicacearum (GCA_001017815) and Pseudomonas rhizophila (GCA_003033885), while strain 108 was grouped with characterized P. fluorescence strains (Figure 9). However, with bootstrap support of 0.63, the branching of both clades was considered low. The majority of the knots in the tree were between 0.95 and 1, indicating that the phylogenetic tree was robust.
Based on our experiment, the selected Pseudomonas strains were able to protect cucumber root bales (Figure 8) more efficiently than when applied to soil with a higher OD 600 of 1.5 (Table 1). Therefore, we hypothesize that the protective properties of Pseudomonas strains 105 and 108 are related to root colonization.
Overall, the phylogenetic analysis shows that the Pseudomonas strains used in this study cluster not only among the RKN control strains but also among other biocontrol strains, such as the P. fluorescens strain 8GCA_902497605, a potato-pathogen-inhibiting strain, or other Pseudomonas strains isolated from the rhizosphere and phyllosphere of potato plants, R84 and S49, which have been reported to inhibit the growth of Phytophthora infestans mycelia [51].
Therefore, it may be worthwhile to test Pseudomonas strains 105 and 108 as biocontrol agents against other important plant-pathogenic nematodes and agricultural fungal and oomycete pathogens. The release of the complete de-novo-assembled genome sequence will serve as an important basis to identify the mechanism(s) of action against various plant pathogens in future studies. On a broader scale, the assay system developed here should also allow the testing of mixtures of different strains with biocontrol activity for robust and potentially even synergistic effects.

Conclusions
This study serves as a proof-of-principle for the identification of potentially promising biocontrol strains against RKN, with the help of this newly developed screening method.
The straightforward and reliable combination of screening assays resulted in the isolation of Pseudomonas strains that controlled M. incognita in vitro and in soil.
Further investigation of Pseudomonas strains 105 and 108 demonstrated their potential to control M. incognita under greenhouse conditions with varying RKN population densities. The application of the Pseudomonas strains is versatile, as demonstrated by their efficacy when being used as either soil or root treatments.
Phylogenetic analyses revealed that both Pseudomonas strains clustered in the Pseudomonas fluorescens group alongside previously described plant pathogens and RKN antagonists, but apart from each other.
Future investigations of the selected Pseudomonas candidates for RKN control will need to demonstrate their antagonistic potential in larger greenhouse trials and/or during field applications, investigating the influence of not only biotic but also abiotic factors on the maintenance of their biocontrol potential. Elucidation of the mechanisms of action is now possible given the availability of complete genome sequences.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms11082011/s1, Figure S1: Effect of bacterial strains on the height of tomato plants grown in nematode-infected soil; Table S1: Selected bacterial strains used in a screen to identify antagonistic activity against Meloidogyne incognita; Table S2: Selected genome features of Pseudomonas spp. 105 and 108; Figure S2: Graphical output of the AntiSmash prediction server (v.6.0.1) for the two Pseudomonas isolates; Table S3: List of genome sequences of available for Pseudomonas strains that have been linked to plant-parasitic nematode control. Data Availability Statement: Data are contained within the article and appendices. The complete genome sequences of the two Pseudomonas isolates are available from the NCBI under bioproject PRJNA975707 (P105) and PRJNA975710 (P108) and with the accession numbers CP126693-CP126694 (P105) and CP126691-CP126692 (P108).