Protective role of native root-associated bacterial consortium against root-knot nematode infection in susceptible plants

Shikai La; Jiafan Li; Si Ma; Xingqun Liu; Lihong Gao; Yongqiang Tian

doi:10.1038/s41467-024-51073-7

. 2024 Aug 7;15:6723. doi: 10.1038/s41467-024-51073-7

Protective role of native root-associated bacterial consortium against root-knot nematode infection in susceptible plants

Shikai La ^1,^2,^#, Jiafan Li ^1,^#, Si Ma ¹, Xingqun Liu ¹, Lihong Gao ^1,^✉, Yongqiang Tian ^1,^✉

PMCID: PMC11306399 PMID: 39112511

Abstract

Root-knot nematodes (RKNs) are a global menace to agricultural crop production. The role of root-associated microbes (RAMs) in plant protection against RKN infection remains unclear. Here we observe that cucumber (highly susceptible to Meloidogyne incognita) exhibits a consistently lower susceptibility to M. incognita in the presence of native RAMs in three distinct soils. Nematode infection alters the assembly of bacterial RAMs along the life cycle of M. incognita. Particularly, the loss of bacterial diversity of RAMs exacerbates plant susceptibility to M. incognita. A diverse range of native bacterial strains isolated from M. incognita-infected roots has nematode-antagonistic activity. Increasing the number of native bacterial strains causes decreasing nematode infection, which is lowest when six or more bacterial strains are present. Multiple simplified synthetic communities consisting of six bacterial strains show pronounced inhibitory effects on M. incognita infection in plants. These inhibitory effects are underpinned via multiple mechanisms including direct inhibition of infection, secretion of anti-nematode substances, and regulation of plant defense responses. This study highlights the role of native bacterial RAMs in plant resistance against RKNs and provides a useful insight into the development of a sustainable way to protect susceptible plants.

Subject terms: Plant physiology, Microbial ecology, Microbiome

Root-knot nematodes (RKNs) are global soil-borne parasites that damage most plants. Here, the authors show that native root-associated bacterial consortium exhibit general suppression of RKN infections by offering protective functions.

Introduction

Root-knot nematodes (RKNs; Heteroderidae) are the most severe soil-borne parasites damaging almost all economically important crops in the world¹. It was estimated that RKNs parasitize the majority of vascular plants and infect nearly every species of flowering plants, causing tremendous yield losses estimated to billions of dollars². Among RKNs, Meloidogyne incognita represents a widespread species and is considered to be one of the most damaging crop parasites worldwide^3,4. M. incognita is extremely difficult to control because it is an apomictic species with a wide host range (making it difficult to use crop rotation as a management measure) and characterized by a sedentary lifestyle, high reproductive rate and short generation time^1,2. Traditionally, nematicides and soil fumigants were used to control M. incognita infection during crop production⁵. However, these agents generally have unfavorable toxicities to humans and nontarget soil organisms⁶, emphasizing the need for alternative strategies.

Theoretically, the use of nematode-resistant cultivars is the most efficient and safe strategy for M. incognita control. There are important RKN-resistant genes such as the Mi gene of tomato. However, most of the current cultivated crops lack dominant resistance genes to M. incognita and show a high degree of susceptibility under field conditions^1,3. Notwithstanding this, owing to their sessile nature, plants have evolved various types of strategies (e.g. adjusting growth rate and morphology, integrating multiple signals, and eliciting innate immunity) to counteract the adverse effects of (a)biotic stresses in their environment⁷. More importantly, plants do not fight alone when dealing with adverse conditions because they are constantly exposed to the microbes that inhabit the soil surrounding the roots⁸. It appears that root-associated microbes (RAMs) co-evolve with their plant hosts and are capable of promoting plant health through conferring plant tolerance to stresses⁹, enhancing plant resistance to parasites¹⁰, and advancing the root architecture¹¹. As a result, plants rely greatly on their RAMs to mitigate various kinds of environmental stresses.

In recent years, emerging evidence suggests that biotic stresses can induce plants to re-assemble their native RAMs in order to maximize both nutrition and defense when plants are challenged by harsh environment^8,12. Even more inspiring, several recent studies have revealed the importance of native RAMs in the rhizosphere, which provides a first line of defense against stresses in the soil^13,14. The protective effect of rhizosphere RAMs became well known primarily due to the discovery of disease-suppressive soils, in which a susceptible plant host suffers little or no disease even though the parasites persist in the soil¹⁵. This disease suppression is largely attributed to the existence of beneficial microbes in the rhizosphere upon pathogen challenge⁸. It is noteworthy that the biotic stress-induced re-assembly of native RAMs in plants is not exclusive to the disease-suppressive soils, but may exist widely in plant-soil systems^12,16,17. The recent theoretical framework of host-microbe co-evolution also infers that although the presence and significance of RAMs may vary by plant species and environmental conditions, the existence of beneficial RAMs is probably conserved throughout the plant kingdom¹⁸. Theoretically, therefore, RAMs may act as protectors against a wide range of parasites including RKNs. To date, however, little information is available regarding the role played by RAMs during the plant-RKN interaction.

As a type of soil-borne parasite, RKN will inevitably interact with RAMs during its attack on plant host roots. Since the existence of beneficial microbes is likely a conserved strategy for plants^12,16–18, we hypothesize that native RAMs may exhibit general suppression of RKNs to maximize defense against infection, irrespective of soil types. Moreover, because of the tremendous diversity of RAMs nurtured by plant roots^19,20, we also hypothesize that the RAM consortium, rather than individual members, provides enduring resistance against RKN infection. Identifying these two hypotheses allowed us to uncover the ecological significance of RAMs in plant protection against RKN infection. To test these hypotheses, we focused on cucumber (Cucumis sativus L.), a globally important crop having no dominant resistance genes to M. incognita and lacking RKN-resistant varieties. We examined whether native RAMs are universally beneficial for plants to resist M. incognita infection and whether native bacterial RAMs have functional redundancy in promoting plant resistance^21,22. Moreover, we evaluated the potential of native root-associated bacterial strains to inhibit M. incognita infection and constructed various simplified synthetic communities (SynComs) based on a design that randomly selected a gradient of increasing strain number²³. Our results show that native root-associated bacterial consortium is effective for controlling root-knot nematode disease of cucumber, thus expanding the range of potential options for controlling parasites of susceptible plants.

Results

Plants show low susceptibility to M. incognita in the presence of native microbiomes

We first collected twelve soils with various properties from different areas of China (Supplementary Fig. 1a, b) and observed that plants exhibited a noticeable decrease in the number of galls under unsterilized conditions compared to γ-sterilized conditions when artificially infected with M. incognita (Supplementary Fig. 1c), indicating the ubiquity of reduced susceptibility of plants to M. incognita in the presence of native microbiomes. To discern whether the observed RKN suppression is general (attributable to the global microbial load) or specific (attributable to a few functional microbial groups), we conducted a soil transplantation experiment and observed that only half of the twelve soils exhibited specific RKN suppression (Supplementary Fig. 1d). Subsequently, to further test whether native microbiomes have beneficial effects on plant resistance against M. incognita, cucumber plants were grown in unsterilized or γ-sterilized soils and further infected artificially with M. incognita (Fig. 1a and Supplementary 1d, e) by using three soils having different geographical origins (i.e. site 1-BJ, site 2 SD and site 3-SX; Fig. 1b and Supplementary Table 1). There were significant differences between these three soils in microbial compositions with respect to the relative abundance of different phyla, α-diversity (observed species) and β-diversity visualized using principal coordinates analysis (PCoA) (Fig. 1c). Despite this, in all three soils, plants exhibited weaker gall symptoms (smaller and fewer galls) in roots (Fig. 1d) and a significant reduction in the numbers of galls and egg masses (reduction: galls 33.2% ~ 46.7%, egg masses 32.7% ~ 49.6%; Fig. 1e and Supplementary Fig. 2) under unsterilized conditions than under γ-sterilized conditions when challenged by M. incognita, emphasizing the universal role of native microbiome in plant protection.

To characterize M. incognita-induced variation in RAMs, we conducted a detailed analysis of microbial communities in the rhizosphere (RS) and root endosphere (RE) of plants infected or not infected with M. incognita (Supplementary Fig. 3). Analysis of microbiome composition revealed that there were significant differences between nematode treatments (infected vs. non-infected) in RS bacteria in all three soils and in RE bacteria in two (i.e. site 1-BJ and site 2-SD) of the three soils (all p < 0.05; Fig. 1f). In contrast, no significant differences were observed for fungi in most situations (except for RS fungi in the site 3-SX). These results suggest that nematode infection modified bacterial rather than fungal RAMs. Moreover, Linear discriminant analysis of Effect Size (LEfSe) demonstrated that in all three soils, several bacterial taxa were consistently enriched in either RS or RE under conditions of nematode infection (Fig. 1g). Together, these results confirmed the M. incognita-induced re-assembly of native bacterial RAMs, which occurred commonly in soils. In addition, M. incognita infestation did not lead to significant changes in soil properties under unsterilized conditions (Supplementary Table 2), suggesting that the change of plant rhizosphere secretions, rather than soil conditions, might play a positive feedback role in the process of bacterial assembly after infection with M. incognita.

M. incognita infection alters root bacterial microbiome assembly at the temporal scale

To explore patterns of temporal turnover of bacterial RAMs under M. incognita infection, we investigated root bacterial microbiomes along the life cycle of M. incognita in non-infected and infected plants (Fig. 2a, b). As expected, plants showed low susceptibility to M. incognita in the presence of RAMs at the temporal scale (Supplementary Fig. 4). Although bacterial microbiomes exhibited a complex and variable response to nematode infection during the process of M. incognita development in roots (Fig. 2c), its composition (visualized using PCoA) was generally influenced by the interaction of nematode treatments (infected vs. non-infected) and sampling times (reflecting the life cycle of M. incognita) (Fig. 2d). The first axis of PCoA displayed M. incognita-mediated temporal shifts in bacterial community composition: while non-infected and infected plants followed similar trajectories, infected plants started diverging from non-infected plants as soon as M. incognita invasion occurred (particularly in RE at 3 dai) and showed an apparent deviation in both RS and RE at 35 dai (Fig. 2e). Particularly, the LEfSe analysis indicated that the number of enriched bacterial taxa (40) exceeded the number of depleted bacterial taxa (25) in root-associated compartments following M. incognita infection in at least one temporal comparison (Fig. 2f). To decipher the role played by plant rhizosphere secretions in the process of bacterial microbiome assembly, we also characterized M. incognita infection-induced metabolic variations in the rhizosphere (at 14 dai) and observed the distinct metabolic patterns between nematode treatments (infected vs. non-infected; Fig. 2g–i). Furthermore, the linkage analysis revealed strong co-occurrence relationships between the changes in rhizosphere metabolites and bacterial OTUs after M. incognita infection (Fig. 2j). Together, these results indicate that M. incognita infection-responsive bacterial taxa follow diverse trajectories along the life cycle of M. incognita in root-associated compartments.

Fig. 2 — a The life cycle of *M. incognita* in cucumber roots. Samples were collected on 3, 7, 14, 21, 28, and 35 days after inoculation (dai) with *M. incognita*. *M. incognita*-infected roots were stained with acid fuchsin to visually inspect the nematode and egg masses. Scale bar, 200 μm. The experiment was repeated four times with similar results. b Schematic representation of the design of the temporal-scale experiment. c Stacked bar plot showing the composition of bacterial RAMs. d Principal coordinate analyzes (PCoA) based on Bray-Curtis distance showing the distinction of bacterial RAMs of plants. Statistical significance was determined using the permutational multivariate analysis of variance (PERMANOVA). e β-diversity (expressed by the first axis of PCoA) patterns in the bacterial RAMs. The trend lines represent the mean values for each treatment throughout the experiment (mean ± SEM). f *M. incognita* infection-responsive bacterial taxa. The linear discriminant analysis (LDA) scores reflecting the *M. incognita* shaped bacterial taxa at the genus level (the absolute LDA score > 3.0) at each time point. Number of differentially abundant OTUs detected in *M. incognita*-infected roots as compared to non-infected roots was also shown. g PLS-DA score plots displaying differences in the composition of rhizosphere metabolites. h Volcano plot of differentially rhizosphere metabolites. i Heat map of differentially rhizosphere metabolites. Colors indicate the Z score values of rhizosphere metabolite abundance (Benjamini–Hochberg test, two-sided). j The linkage analysis reveals the co-occurrence relationships between the changes in rhizosphere metabolites and bacterial OTUs after *M. incognita* infection. The solid lines indicate a significant correlation (p < 0.05) and color depth indicates the level of correlation. r, Person correlation coefficient. Source data are provided as a Source Data file.

Bacterial diversity loss of native microbiome exacerbates plant susceptibility to M. incognita

To examine whether functional redundancy exists among RAMs in protecting plants against M. incognita, we applied the dilution-to-extinction approach to create a broad diversity gradient in the native microbiome (Fig. 3a). As expected, distinct assembly processes of bacteria occurred in both RS and RE after dilution and re-introduction (Fig. 3b and Supplementary Fig. 5). However, a dramatic loss of bacterial α-diversity (observed OTUs) with increasing dilution level was observed in RS (p < 0.001; Fig. 3c) but not in RE (p = 0.42; Supplementary Fig. 5i). Furthermore, significant differences in bacterial microbiome composition were observed among the dilution levels in both RS and RE (both p < 0.05; Fig. 3d and Supplementary Fig. 5j), and the dilution level had a strong correlation with bacterial β-diversity (expressed by PCoA 1) in RS (p < 0.001; Fig. 3e) but not in RE (p = 0.19; Supplementary Fig. 5k). Remarkably, the severity of root gall symptoms exhibited an increasing trend with an increase in the dilution level under conditions of M. incognita infection (Fig. 3f–h and Supplementary Fig. 5c). Notably the bacterial diversity was very closely and positively associated with its predicted functionality (expressed by the Fric index, p < 0.0001; Fig. 3i and Supplementary Fig. 6), indicating that loss of diversity implies loss of the potential function of the bacterial community. Obviously, the diversity and functionality of the RS bacterial community were tightly and negatively linked to plant susceptibility to M. incognita (reflected by the numbers of galls and egg masses per gram root, a lower number representing a lower susceptibility, all p < 0.001; Fig. 3j). Together, since bacterial diversity loss of native microbiome exacerbated plant susceptibility to M. incognita, M. incognita suppression was attributed mainly to the rhizosphere bacterial consortium (rather than individual members), which employed a synergistic regulation (but not individual effect) strategy in protecting plants against M. incognita.

Fig. 3 — a Schematic representation of the design of the microbe gradual-removal experiment. Six levels of dilution (L1: 10⁰; L2: 10^-1; L3: 10^-2; L4: 10^-4; L5: 10^-6; L6: 10^-8) of the soil suspension were used as inocula to create a gradient of microbial diversity in soils. b Chord diagram showing changes in the relative abundance of bacterial RAMs in the rhizosphere at the phylum level. The chord diagram can depict the relationship between serial dilution treatments by displaying the relative abundance of bacterial microbes at the phylum level. c Changes in rhizosphere bacterial α-diversity with increasing simplification of microcosms. d Principal coordinate analyzes (PCoA) based on Bray-Curtis distance showing the distinction of rhizosphere bacterial RAMs of *M. incognita*-infected plants. Statistical significance was determined using the permutational multivariate analysis of variance (PERMANOVA). e Changes in rhizosphere bacterial β-diversity with increasing simplification of microcosms. f Symptoms of the root-knot disease in roots. g, h Changes in gall and egg mass numbers per gram of plant roots with increasing simplification of microcosms (mean ± SEM). Different letters indicate significant differences by one-way ANOVA with post hoc Tukey HSD test (p < 0.05). N = 9 biologically independent samples. i Functionality in relation to the diversity of bacterial RAMs. The Fric index was calculated using the dbFD function in the ‘FD’ R package. j Gall and egg mass numbers per gram root in relation to the diversity and functionality of bacterial RAMs. Values are expressed as a ratio of the most complete microbial consortium. In c, e, i, and j, the curve represents the linear regressions and the shaded area indicates 95% regression confidence interval. Source data are provided as a Source Data file.

A diverse range of native bacterial strains from M. incognita-infected roots have nematode-antagonistic activity

To investigate the M. incognita-suppressing ability of native root-associated bacteria, we established a taxonomically diverse bacterial culture collection from M. incognita-infected roots (Fig. 4a). After removal of clonal duplicates, we obtained a total of 212 unique bacterial strains, which covered 32 bacterial genera, indicating a high diversity of the isolated strains (Fig. 4b). We then performed in vitro detection of the juvenile mortality of M. incognita to screen for bacterial strains capable of suppressing M. incognita. Specifically, we observed that 45 different bacterial strains caused a certain level of mortality in M. incognita juveniles, accounting for 21.2% of the total unique bacterial strains isolated (Fig. 4c). These functional strains were taxonomically assigned to 27 bacterial genera, covering 84.4% of bacterial genera in the current study (Fig. 4d). Notably, there were 21 functional strains (i.e. Ecc, Rp, Pj, Pk, Pp, As, Sn, Cw, Fa, Spu, Ct, Mp, Ba, Sh, Mj, Nn, Psi, Bp, Sm, Bt and Fn) that strongly suppressed survival of M. incognita (corrected nematode mortality > 90%) (Fig. 4c). Moreover, a total of 22 functional strains had no adverse effects on plant performance (if the ratio of plant performance between the functional strain inoculation and the control (not inoculated) > 1.0, it would be recognized as a specific strain having no adverse effect on plant growth; Fig. 4e and Supplementary Fig. 7). Based on the results of nematode mortality detection and plant performance evaluation, we further selected 14 high-performance functional strains (HFSs) that have pronounced M. incognita-antagonistic activity without inhibiting plant performance (Fig. 4f). A pot experiment was further conducted to verify the M. incognita-antagonistic functions of these 14 HFSs in cucumber roots under M. incognita infection. When compared with the control (no bacterial strain), all 14 HFSs effectively reduced the numbers of galls or egg masses per gram root (Fig. 4f, g), demonstrating their ability to reduce plant susceptibility to M. incognita. Collectively, these results emphasized that a diverse range of native root-associated bacterial strains have antagonistic activity against M. incognita infection.

Fig. 4 — a Schematic representation of the design for evaluating the *M. incognita*-suppressing ability of native bacterial strains. Bacterial strains were isolated from the rhizosphere and root endosphere of *M. incognita*-infected cucumber. b Taxonomic classification of the *M. incognita*-induced strains. A total of 212 bacterial strains were randomly isolated, repeatedly purified, and phylogenetically identified based on full-length 16 S rRNA gene sequencing. c In vitro detection of the second-stage juvenile (J2) mortality of *M. incognita* (expresse as corrected *M. incognita* J2 mortality) caused by antagonistic strains. Fourty-five of these 212 bacterial strains were verified to cause a certain level of mortality in *M. incognita* J2s. d Proportion of functional strains taxonomically assigned to bacterial genera to the total functional strains. The red bars show strong suppression of *M. incognita* J2. e Ratios of plant performance of functional strain inoculation to the control (not inoculated). f, g In vivo verification of *M. incognita* infection. Gall (f) and egg mass (g) numbers per gram of plant roots are shown (mean ± SEM). The statistical significance of p values was calculated via a two-tailed, unpaired Student’s t-test. A pot experiment was conducted to verify the *M. incognita*-antagonistic functions of fourteen high-performance functional strains (HFSs; corrected nematode mortality > 90%, and ratio of plant performance compared to control > 1.0) in cucumber roots under *M. incognita* infection. N = 9 biologically independent samples. Source data are provided as a Source Data file.

Multiple SynComs have the capacity to inhibit M. incognita infection in plants

To assess how strain diversity could control M. incognita infection of roots, we conducted an experiment using a design that randomly selected strains from a pool along a gradient of strain number in order to test if the pattern of M. incognita infection shows a consistent directional trend along the number of HFSs (Fig. 5a). The M. incognita infection rate exhibited an inverse S-shaped decreasing trend with increase in the number of HFSs (Fig. 5a), suggesting that maximum additive/synergistic effects were reached with a relatively low number of the HFSs. It was noted that the infection rate reached relatively low levels with 6 HFSs and then maintained relatively constant levels for more HFSs, implying that 6 HFSs were enough to construct a highly simplified SynCom capable of significantly reducing M. incognita infection. Thus, we further randomly combined HFSs (instead of using a targeted approach) and constructed 10 different combinations (SynComs A-J) of 6 HFSs (Fig. 5b). Remarkably, all these SynComs led to a significant reduction (49.1%) in the M. incognita infection rate (all p < 0.05; Fig. 5b). In particular, SynCom C, which consisted of Pseudomonas knackmussii (Pk), Pseudomonas protegens (Pp), Sphingobacterium puteale (Spu), Microbacterium paraoxydans (Mp), Bacillus proteolyticus (Bp) and Stenotrophomonas maltophilia (Sm), led to the strongest inhibition of M. incognita infection (p < 0.0001; Fig. 5b). The six HFSs of the SynCom C were taxonomically assigned to five bacterial genera including Pseudomonas, Sphingobacterium, Microbacterium, Bacillus and Stenotrophomonas, all of which were consistently observed in M. incognita-infected roots in three different soils (Supplementary Fig. 8a, b). A dramatic loss of Pseudomonas, Microbacterium and Stenotrophomonas with increasing dilution level was observed in RS in the dilution to extinction experiment (Supplementary Fig. 8d). Additionally, the relative abundance of Pseudomonas, Microbacterium and Stenotrophomonas was tightly and negatively linked to plant susceptibility to M. incognita (Supplementary Fig. 8d, e), indicating that the loss of these selected microbes corresponds to a significant increase in M. incognita infection and proliferation. However, not all Pseudomonas or Microbacterium strains detected in the microbiome analysis exhibited a correlation with nematode suppression (Supplementary Fig. 8a). In addition, unexpectedly, the OTU sequences enriched in response to nematode infection from the microbiome analysis did not match well the full-length 16 S rRNA sequences of the isolates used to compose the SynComs (Supplementary Fig. 8b, c). An in vitro biofilm formation assay showed that these six HFSs of SynCom C acted synergistically in biofilm formation, as SynCom C formed more biofilm than the separate strains (Fig. 5c). Moreover, these six HFSs were observed to generally attract each other on plates (Fig. 5d). Taken together, these observations indicated that these six HFSs of SynCom C were synergistically assembled as a consortium against M. incognita infection of roots.

Fig. 5 — a Effects of strain diversity on *M. incognita* infection in roots. Values are expressed as a ratio of the average of individual strains (number of strain, 1). The curve represents the logistic regression and the shaded area indicates 95% regression confidence interval. b Efficiency of ten different SynComs (A-J) in inhibiting *M. incognita* J2 infection. c Biofilm formation by single strain or their combination (SynCom C). Different letters indicate significant differences by one-way ANOVA with post hoc Tukey HSD test (p < 0.05). d Attraction between colonies grown at increasing proximity. e Metabolite profiles of single strains. f–i The effects of the overall metabolites of bacterial strains on *M. incognita* egg hatching (f), J2 mortality (g), the expression of *M. incognita flp-1* and *flp-18* genes (h) and *M. incognita* chemotaxis movement towards roots (i). LB, liquid broth. j–m The effects of the specific metabolites (0.5 mM) of bacterial strains on *M. incognita* egg hatching (j), J2 mortality (k), the expression of *M. incognita flp-1* and *flp-18* genes (l) and *M. incognita* chemotaxis movement towards roots (m). L-phenylalanine served as a control. In b and f–m, data are presented as mean ± SEM and statistical significance of p values was calculated via a two-tailed, unpaired Student’s t-test. Biologically independent samples are N = 9 (b), N = 4 (f, g, i, j, k and m), N = 5 (h) and N = 3 (l). In c, j, and k, the lower, upper, and center lines in boxplot represent the 25th, 75th, and 50th percentiles, respectively, and whiskers extend no further than ±1.5 times the inter-quartile range. Source data are provided as a Source Data file.

To further characterize the functions of the separate strains of SynCom C, we examined plant growth promotion (PGP)-related characteristics including PGP functional genes, biochemical and physiological properties, and metabolite profiling. Generally, each of the six HFSs of SynCom C showed more than one PGP-related characteristic (such as phosphate solubilization, siderophore secretion) and carried at least two PGP functional genes (Supplementary Fig. 9). In addition, Bp, Sm, Pp and Mp had the capacity to utilize most carbon sources including sugars, amino acids, hexose acids, carboxylic acids, esters and fatty acids (Supplementary Fig. 10), which are commonly referred to as root metabolites. Pk had the ability to utilize several amino acids and some hexose acids, while Mp was capable of utilizing almost all types of sugars. Furthermore, each of the six HFSs of SynCom C could accumulate one or more specific metabolites (e.g. norvaline and 3-oxoadipic acid for Mp, L-erythrulose for Pk, 2-ethyl-2-hydroxybutyric acid for Pp, 2-isopropylmalic acid, uridine, N-[2-(5-methoxy-3H-indol-3yl)ethyl]acetamide and styrene for Bp, beta-L-fucose for Sm, 3-methyl-2-oxovaleric acid for Spu; Fig. 5e). It may be feasible to enhance the inhibitory effect on RKNs. Notably, the overall metabolites of all bacterial strains, which were obtained from the supernatant of a liquid bacterial culture with an OD₆₀₀ of 0.4, strongly inhibited M. incognita egg hatching and the survival of infective J2s (inhibition: egg hatching 96.7% ~ 99.2%, infective J2 survival 96.6% ~ 100%; Fig. 5f, g). Moreover, the expression of M. incognita flp-1 and flp-18 genes, which are critical for the chemotaxis and parasitism of M. incognita towards cucumber roots, was effectively suppressed by the overall metabolites of all bacterial strains (suppression: flp-1 85.6% ~ 97.2%, flp-18 84.1% ~ 97.1%; Fig. 5h). As a result, the chemotaxis movement of M. incognita towards cucumber roots was significantly inhibited in the presence of bacterial strains’ metabolites (Fig. 5i). Several specific metabolites of bacterial strains (e.g. norvaline, betaine, L-erythrulose, isonicotinic acid, uridine, L-pipecolic acid, pantothenic acid, and beta-L-fucose), which were selected based on their high abundance in the secretions of bacterial strains, exhibited inhibitory effects on egg hatching or the survival of infective J2s at a concentration of 0.5 mM (inhibition: egg hatching 6.1% ~ 59.7%, infective J2 survival 6.7% ~ 21.6%; Fig. 5j, k). Moreover, the expression of M. incognita flp-1 or flp-18 genes, as well as the chemotaxis movement of M. incognita towards cucumber roots, was also inhibited by several specific metabolites (suppression: flp-1 15.2% ~ 48.6%, flp-18 12.2% ~ 57.0%; Fig. 5l, m). Together, these in vitro findings (Fig. 5f–h, j–l), combined with in vivo observations (Fig. 5i, m), suggest that the suppression of RKNs by cucumber root-associated bacterial strains could involve multiple mechanisms, such as the secretion of anti-nematode substances, inhibition of egg hatching, induction of infective J2 mortality, hindrance of chemotaxis, and suppression of migratory activities.

The protective role of SynCom C against M. incognita infection is associated with the regulation of plant defense responses

To verify whether SynCom C triggers plant immune responses under M. incognita infection, we conducted a two-level factorial experiment (SynCom C inoculation × M. incognita infection; Fig. 6a). Inoculation of SynCom C substantially diminished the serious damage caused by M. incognita to both shoots and roots (Mi+SynCom C vs. Mi) (Fig. 6a, b). Moreover, plants showed a significant reduction in the number of galls, egg masses and females in the presence of SynCom C (Mi+SynCom C vs. Mi), indicating the inhibiting effects of SynCom C on M. incognita development in roots (all p < 0.05; Fig. 6c–e). We then analyzed metabolites of the SynCom C (Fig. 6f) and observed that the overall metabolites significantly inhibited M. incognita egg hatching and infective J2 survival (inhibition: egg hatching 99.5%, infective J2 survival 100%), suppressed the expression of M. incognita flp-1 and flp-18 genes (suppression: flp-1 99.5%, flp-18 99.7%), and thereby disrupted the movement of M. incognita towards cucumber roots (Fig. 6g–j). Furthermore, we performed RNA sequencing (RNA-seq) to evaluate the transcriptional responses of roots to SynCom C treatments. The distinct transcriptional patterns were identified between all treatments (Fig. 6k), leading to 724 and 95 differentially expressed genes (DEGs) respectively in M. incognita (Mi vs. mock) and SynCom C (Mi+SynCom C vs. Mi) treatments (Fig. 6l). Closer examination of DEGs revealed dozens of genes associated with defense protein, cell wall metabolism, transcription factor, proteolysis, antioxidant defense system and phenylpropanoid pathway (Fig. 6m). Particularly, several DEGs enriched under M. incognita infection (Mi vs. mock) were associated with plant defense responses (Supplementary Table 3) and were suppressed by inoculation of SynCom C (suppression: 13.6% ~ 49.5%; Mi+SynCom C vs. Mi) (Fig. 6m). These candidate genes were further verified by qRT-PCR (Fig. 6n). Moreover, gene ontology (GO) term enrichment analysis of ‘core’ enriched DEGs (Mi+SynCom C vs. Mi) revealed an over-representation of terms involved in plant immunity (e.g. MAPK signaling pathway; Supplementary Fig. 11). We also characterized SynCom C-induced metabolic variations in roots, and observed the distinct metabolic patterns between treatments (Fig. 6o). There were 34 and 26 differential metabolites (DMs) in M. incognita (Mi vs. mock) and SynCom C (Mi+SynCom C vs. Mi) treatments, respectively (Fig. 6p). Closer examination of DMs revealed more than thirty of metabolites belonging to lipids, flavanones and lignans (Fig. 6q), most of which were reported to be related to plant defense responses (Supplementary Table 4). In addition, dozens of DMs enriched under M. incognita infection (Mi vs. mock) were suppressed by inoculation of SynCom C (suppression: 10.6% ~ 96.2%; Mi+SynCom C vs. Mi) (Fig. 6q). Together, SynCom C rescued plants from M. incognita infection partly through regulation of stress-responsive genes and metabolic pathways (Fig. 6r).

Fig. 6 — a Symptoms of the root-knot disease in cucumber plants. Mock, untreated control; Mi, *M. incognita* infection; SynCom C, SynCom C inoculation. b Fresh weight of cucumber plants. c, d Gall and egg mass numbers per gram of plant roots. e Number of nematodes in *M. incognita*-infected plants. f–j The effects of the SynCom C’s metabolites (f) on *M. incognita* egg hatching (g), J2 mortality (h), the expression of *M. incognita flp-1* and *flp-18* genes (i) and *M. incognita* chemotaxis movement towards roots (j). k Principal component analysis of genes expressed in roots. l Venn diagram of differentially expressed genes (DEGs). m Heat map of DEGs. Red text indicates DEGs that were associated with plant defense responses. n The expression of candidate DEGs verified by qRT-PCR. o Partial least squares discriminant analysis of metabolites in roots. p Venn diagram of differential metabolites (DMs). q Heat map of DMs. r A schematic diagram illustrating the regulation of *M. incognita* stress-induced transcriptional and metabolic responses by SynCom C. Transcriptional and metabolic responses are associated with plant susceptibility to *M. incognita*, and SynCom C alters these responses partly through suppression of stress-responsive genes and metabolic pathways. In b–e, g–j, and n, data are presented as mean ± SEM. In b-d, different letters indicate significant differences by one-way ANOVA with post hoc Tukey HSD test (p < 0.05). In g–j, and n, the statistical significance of p values was calculated via a two-tailed, unpaired Student’s t-test. Biologically independent samples are N = 9 (b-e), N = 4 (g, h and j), N = 5 (i) and N = 3 (n). Source data are provided as a Source Data file.

Discussion

Despite the emerging evidence of the biological significance of RAMs for the maintenance of plant health^8,10,24, very few studies have attempted to clarify the role of RAMs in the protection of plants when challenged by the parasitic nematode, a global menace to crop production. Our experiments, on M. incognita-susceptible cucumber plants grown in three native soils, have revealed the universal role of RAMs in plant protection against nematode infection, thus helping us to understand how plants rely on native RAMs to maximize defense against infection. Regardless of soil types, plants seem to exhibit a consistently lower susceptibility to M. incognita under conditions of native soil microbiome (Supplementary Fig. 1c and Fig. 1d, e). This phenomenon emphasizes the biological importance of the healthy microbiome concept, that is, plants rely greatly on their microbiome (the collective communities of microorganisms) for tolerance to a variety of environmental stress factors¹⁷. Furthermore, it highlights that plant microbiome provides additional ecological functions (e.g. enhancing plant resistance and improving plant adaptability) that help plants deal with stresses in their stationary lifestyle¹⁶.

Notwithstanding substantial differences in microbial compositions among soils, plants re-assembled microbes in either RS or RE when challenged by M. incognita (Fig. 1g). Similar studies have demonstrated that Meloidogyne infection influences the assembly of the root microbiome^25,26. This experimental finding is in agreement with recent theoretical and experimental studies on the plant microbiome assembly patterns driven by microbial pathogens, such as the fungal root pathogen Rhizoctonia solani causing damping-off symptoms in sugar beet¹⁹, the bacterial and fungal root pathogens (e.g. Curtobacterium flaccumfaciens and Fusarium oxysporum) causing sudden-wilt disease in Nicotiana attenuate¹⁰ and the fungal leaf pathogen Hyaloperonospora arabidopsidis causing downy mildew in Arabidopsis thaliana⁸. It is noteworthy that M. incognita is genetically and morphologically different from the microbial pathogen, implying that M. incognita and the microbial pathogen have substantial differences in pathogenic mechanisms²⁷. Interestingly, however, our study shows that plants exhibited specific alterations in the assembly of their root bacterial microbiome under M. incognita attack, similar to that under microbial pathogen infection^8,10,19. Moreover, a consistent health promotion in roots was observed under soil conditions with native microbiomes (Supplementary Fig. 1c and Fig. 1d, e). Similarly, the suppression of RKN infectivity by microbiomes was also observed under organic farming practices²⁶. Thus, the protective role of native root-associated bacterial consortium is likely a highly conserved regulatory strategy that allows plants to adapt to various stresses including pathogen infection and abiotic stresses such as drought and salinity^14,28.

The RKN is a severe soil-borne parasite damaging almost all economically important crops in the world¹. Not withstanding the fact that the use of RKN-resistant cultivars is an efficient and safe strategy for RKN control, the lack of dominant RKN-resistance genes is still an underappreciated major challenge for plant protection under field conditions^1,3. Despite that, it is feasible to enhance the ability of plants to assemble native RAMs that provide benefits to their plant host. This could be partly supported by the experimental results of Mendes et al.¹³, who observed that crop trait enhancement could be realized by breeding a variety that can recruit specific beneficial bacterial genera capable of improving plant resistance against soil-borne parasites. Indeed, the plant is a powerful driving force for the selection and evolution of native microbes²⁹. It is becoming increasingly clear that plants should be regarded as holobionts (rather than standalone entities), in which a plant is deeply intertwined with its associated microbiome^17,30. In this study, we found that the M. incognita-induced changes in plant rhizosphere secretions were closely associated with the process of bacterial assembly (Fig. 2j). Moreover, three bacterial genera (i.e. Pseudomonas, Microbacterium and Stenotrophomonas) used in the SynCom C (Fig. 5b) showed a significant correlation with nematode suppression in the dilution to extinction experiment (Supplementary Fig. 8d, e). Similarly, these bacterial genera have also been reported to possess the ability to suppress plant diseases caused by microbial pathogens^19,31,32. Thus, it seems that plants show a broadly similar response to various pathogen challenges, i.e., assembly of beneficial RAMs. However, despite the fact that the depletion of Pseudomonas and Microbacterium correlates with a notable rise in M. incognita infection and proliferation, not all Pseudomonas or Microbacterium strains detected in the microbiome analysis showed a correlation with nematode suppression (Supplementary Fig. 8a). Additionally, no member of the SynComs was commonly found across all samples in the microbiome analysis (Supplementary Fig. 8c). More importantly, the soil transplantation experiment showed that only half of the twelve soils exhibited specific suppression attributable to a few functional microbial groups (Supplementary Fig. 1d). Taken together, these findings confirmed our first hypothesis that native RAMs exhibit general suppression of RKN infections by providing protective functions, but not specific suppression.

Unexpectedly, our temporal-scale experiment revealed a complex interaction between bacterial RAMs and nematode infection, along the life cycle of M. incognita in roots (Fig. 2). This could be explained by the fact that plant roots nurture a tremendous diversity of RAMs^19,20, which interact with each other in a complex manner^24,29. Indeed, nematode infection resulted in alteration of a diverse range of bacterial RAMs, which were changeable across the disease developmental stages (Fig. 2f). Coincidently, our results of the microbe gradual-removal experiment also emphasized the critical importance of bacterial diversity of RAMs for enhancing plant resistance against nematode infection (Fig. 3). Since the diversity of RAMs is generally much higher in RS than in RE^29,33, the rhizosphere microbiome supports more functions for plants to cope with stresses in situ^16,34. Because of the sessile nature of plants, the higher diversity of rhizosphere microbiome is essential for plants to counteract the adverse effects of (a)biotic stresses in their environment^7,35. Clearly, the results of our study highlight the significance of bacterial consortium with high diversity (Fig. 3). Theoretically, success in establishing a diverse microbiome is critical for plants to assemble a complex microbial network capable of counteracting various kinds of stressors^34,36,37. Furthermore, the strong correlation existing between the microbial diversity and functionality of RS and plant performance (Fig. 3i, j) confirmed the significance of the synergistic interactions among rhizosphere RAMs in plant protection.

Although the role of microbiome diversity in maintaining plant health and disease suppression has been suggested, the functional role of potential beneficial taxa still needs to be verified individually and collectively by using culture-based methods^8,19,24. Thus, we established a taxonomically diverse bacterial culture collection from M. incognita-infected roots (Fig. 4a, b). Coincidentally, a diverse range of native bacterial strains isolated from M. incognita-infected roots have nematode-antagonistic activity (Fig. 4c). In particular, several strains (e.g. the members of the genera Bacillus, Chryseobacterium, Pseudomonas, Acinetobacter and Enterobacter; Fig. 4d) have been previously demonstrated to support plant disease resistance and promote plant growth^24,38,39. The high diversity of nematode-antagonistic bacterial strains highlighted the importance of bacterial consortium in supporting plant disease resistance because a more diverse microbiome provides a more complex network carrying a greater number of taxa that support various functions³⁴. To further verify that the bacterial diversity exerted a positive effect on the inhibition of M. incognita infection, we used an innovative approach applying a design that randomly selected strains from a pool along a gradient of strain number (Fig. 5a)²³. Our results showed that increasing strain number resulted in more efficient plant protection against nematode infection. This could be explained by the observed differential effects of bacterial strains on nematode mortality and plant performance (Fig. 4). Notably, M. incognita infection was lowest when six or more bacterial strains were present (Fig. 5a), emphasizing that a certain number of strains (consortium rather than individual members) were needed to maintain a relatively high level of disease resistance. This can be supported by the fact that multiple SynComs consisting of six strains had pronounced inhibitory effects on M. incognita infection in plants (Fig. 5b). Thus, multispecies synergistic interaction, but not individual effect, was the underlying mechanism employed by M. incognita-induced RAMs to reduce plant susceptibility to nematode infection. Indeed, the support of plant disease resistance by multiple species of a SynCom could involve multiple mechanisms, such as direct inhibition (Fig. 4c), biofilm formation in the rhizosphere (Fig. 5c), secretion of anti-nematode substances (Fig. 5e–m), expression of PGP genes (Supplementary Fig. 9), and regulation of plant defense responses (Fig. 6 and Supplementary Fig. 11).

Particularly, in SynCom C, each strain produced specific metabolites that are capable of inhibiting egg hatching, inducing mortality in infective J2s, suppressing the expression of M. incognita flp-1 and flp-18 genes which regulate the chemotaxis of M. incognita towards cucumber roots and control parasitism, and disrupting the movement of M. incognita to cucumber roots (Figs. 5e–m and 6f–j). These actions could hinder infective J2 penetration, prevent giant cell formation, and limit the reproduction of RKNs in the roots (Fig. 6a–e). Primarily, stress-response genes play a crucial role in plants to recognize and respond to infection of various pathogens/parasites (including RKNs)^1,2,17,24. The lower expression of stress-response genes (Fig. 6k–n, Supplementary Table 3) suggests that M. incognita infection in the roots was alleviated in the presence of SynCom C. Moreover, SynCom C was suggested to induce suppression of stress-response metabolic pathways in the roots (Fig. 6o–r, Supplementary Table 4), which might lead to an unfavorable root environment for RKNs and thereby contribute to the suppression of RKNs. The suppression of specific metabolic pathways in the roots could limit the availability of essential nutrients required by RKNs for sustainable feeding site establishment and effective reproduction, leading to a reduction in their infection and proliferation^1,11,17,18. Specifically, the lipid metabolic pathway, which is crucial for RKN parasitism reliant on lipid metabolites⁴⁰, was sufficiently suppressed by SynCom C (Fig. 6q). Although our current study highlighted the role of native bacterial RAMs in preventing nematode infections, there are limitations that need to be addressed in future studies. It needs to broaden the focus to include more plant and nematode species, investigate the mechanisms by which rhizosphere bacteria can trigger reactions in plants without entering roots, and account for the dynamic nature of bacterial populations in soil that depends on environmental conditions, seasons or crop types. Furthermore, to improve the ecological validity of the study, it may be beneficial to explore a more targeted, soil- and dilution-specific isolation approach, which involves isolating microbes separately for each soil sample and corresponding dilutions, and constructing SynComs that are customized to the specific soil conditions.

In summary, our study demonstrates that native bacterial RAMs exhibit general suppression of RKN infections by offering protective functions. Although complex interactions exist between RKN infection and RAMs, the diversity and functionality of native bacterial RAMs are critical for plants to resist RKN infection. A diverse range of native bacterial strains isolated from RKN-infected roots have nematode-antagonistic activity, and thereby can be used to construct SynComs having pronounced inhibitory effects on RKN infection. These findings emphasize the critical importance of native root-associated bacterial consortium in plant protection and provide a useful insight into the development of efficient tools to protect susceptible plants in a sustainable way.

Methods

Plants, soils, nematodes, and experimental design

Cucumber (Cucumis sativus cv. ‘Zhongnong 26’), a crop that is highly susceptible to M. incognita and is subject to serious economic damage caused by RKNs during cultivation worldwide, was used in this study. The variety ‘Zhongnong 26’ is an F1 hybrid cucumber that was developed by crossing the inbred line 01316 as the maternal parent and 04348 as the paternal parent. It was challenging for us to utilize pure lines or clone varieties due to the requirement of a large quantity of seeds for our research. To minimize genetic variation among plants, the following measures were implemented in all subsequent experiments: using seeds from the same original source, selecting seedlings with similar appearances and growth statuses, maintaining a consistent controlled environment for seedling growth, and randomly assigning seedlings when multiple treatment groups were involved. The soils used were collected from twelve different sites having various geographical origins (Supplementary Fig. 1a, b), including site 1-BJ (Beijing; 40°01′41.1″N, 116°17′04.2″E), site 2-SD (Shandong; 36°47′47.7″N, 117°17′44.9″E) and site 3-SX (Shanxi; 37°06′40.8″N, 111°58′39.6″E) in China, by using the soil collecting method described in Niu et al.⁴¹. All soil samples were collected in a responsible manner and in accordance with relevant permits and local laws. The site 1-BJ, site 2-SD and site 3-SX soils were selected based on their comparable textures but contrasting physicochemical and microbial properties (for more information see Supplementary Table 1). Root-knot nematode M. incognita was multiplied in vivo on tomato plants to collect egg masses of M. incognita as described in Warmerdam et al.³. The eggs were incubated and hatched to obtain the infective second-stage juveniles (J2s) of M. incognita, which were collected for plant inoculation. Before inoculation, the hatched J2s were surface-sterilized as described in Warmerdam et al.³.

For all experiments (Expts. 1–6) mentioned below, cucumber seeds were surface-sterilized, checked for contamination, and germinated as described in Li et al.¹⁴. The germinated seeds were sown in plastic pots (927 cm³ per pot) filled with unsterilized or γ-sterilized soils using sterile tweezers at a density of one seed per pot, and then were grown in a germ-free phytotron (equipped with UV sterilizing lamps used to provide a sterile environment for the cultivation of plants) under conditions with a mean temperature of 26 °C/18 °C (day/night, 14/10 h), a relative humidity of 60–80%, a mean CO₂ concentration of 380 µmol mol^-1 and an average daily photosynthetically active radiation (PAR) of approximately 300 μmol m^-2 s^-1 during the day. For soil sterilization, γ-irradiation at 60 kGy was applied because it can remove microorganisms from soils (no microbial DNA and live microbes were observed) without disrupting soil physiochemical properties¹⁴. The efficiency of γ-irradiation in soil sterilization was verified by quantitative PCR assays for quantifying bacterial 16 S rRNA and fungal ITS genes and gel electrophoresis for checking PCR products using the DNA extracted from soils as a template (Supplementary Fig. 1e, f). For nematode inoculation, 400 infective J2s of M. incognita were inoculated for each 10-d-old cucumber seedling by injecting a suspension containing infective J2s.

Six experiments were conducted. In Expt. 1, which aimed to investigate the ubiquity of reduced susceptibility of plants to M. incognita in the presence of native microbiomes, twelve different soils with various properties were collected from different areas of China (Supplementary Fig. 1a, b) and subjected to soil sterilization treatments including (i) unsterilized soils and (ii) γ-sterilized soils. There was a total of 24 experimental treatments (12 soils × 2 soil sterilization treatments). Each treatment had three independent replicates with 9 seedlings per replicate plot (length × width × height: 33 × 33 × 10 cm). All plants were artificially infected M. incognita. Furthermore, to discern whether the observed M. incognita suppression was general or specific, these twelve soils were also subjected to soil transplantation treatments including (i) γ-sterilized soils and (ii) γ-sterilized soils mixed with 10% (w/w) unsterilized soils. There was a total of 24 experimental treatments (12 soils × 2 soil transplantation treatments). Each treatment had three independent replicates with 9 seedlings per replicate plot (length × width × height: 33 × 33 × 10 cm). All plants were artificially infected M. incognita. The nematode infection in roots was determined 21 and 35 days after M. incognita inoculation.

In Expt. 2, which aimed to examine whether native RAMs are universally beneficial for plants to resist M. incognita infection, regardless of the differences in microbial composition among soils, a two-level factorial experiment (soil sterilization × M. incognita infection) was conducted for each of three soils (i.e. site 1-BJ, site 2-SD, and site 3-SX). The soil sterilization treatments included (i) unsterilized soils and (ii) γ-sterilized soils, while the M. incognita infection treatments included (i) uninfected plants and (ii) M. incognita-infected plants, resulting in a total of 12 experimental treatments (3 soils × 2 soil sterilization treatments × 2 M. incognita infection treatments). Each treatment had four independent replicates with 18 seedlings per replicate plot (length × width × height: 66 × 33 × 10 cm). The physiological properties of seedlings, nematode infection in roots, and microbial communities (including bacteria and fungi) in the bulk soil (BS) were determined 35 days after M. incognita inoculation. In addition, the microbial communities in the rhizosphere (RS) and root endosphere (RE) were also measured to examine whether plants can recruit specific RAMs under M. incognita infection. Since the results of Expt. 2 showed that M. incognita infection induced variation in the bacterial microbiome rather than the fungal microbiome, further experiments focused on the bacterial microbiome.

In Exp. 3, which aimed to uncover the secrets of native microbial interactions occurring during the process of M. incognita development after inoculation, a temporal-scale experiment was conducted under soil conditions with/without M. incognita infection treatments. To eliminate the influence of soil types, these three soils (i.e. site 1-BJ, site 2-SD, and site 3-SX) were mixed evenly (1:1:1, v/v/v) before imposing the treatments. The treatments considered were (i) unsterilized soils plus uninfected plants, (ii) unsterilized soils plus M. incognita-infected plants, (iii) γ-sterilized soils plus uninfected plants, and (iv) γ-sterilized soils plus M. incognita-infected plants. Each treatment had three independent replicates with 18 seedlings per replicate. Based on the disease developmental stages (i.e. the life cycle of M. incognita in roots), plant and microbial samples collected at 3, 7, 14, 20, 28 and 35 days after infection (dai) were recognized as invasion by second-stage juveniles (J2s, initiation of feeding sites), parasitic J2s (formation of feeding sites), feeding third-stage juveniles (J3s, expansion of feeding sites), fourth-stage juveniles (J4s, further expansion of feeding sites), adult females (maintenance of feeding sites) and females with egg masses, respectively. This sampling strategy allowed us to measure the composition and variation of the bacterial microbiome in RS and RE throughout the process of M. incognita development after inoculation.

In Exp. 4, which aimed to examine whether RAM diversity exhibits a correlation with M. incognita suppression, the dilution-to-extinction approach was applied to conduct a microbe gradual-removal experiment by inoculating the γ-sterilized soil microcosms with 1200 g soil per microcosm (length × width × height: 9.63 × 9.63 × 10 cm) with suspensions of the same soil that was not sterilized^21,22. The soil used here was the same as that used in Exp. 3. Six levels of dilution (10⁰, 10^-1, 10^-2, 10^-4, 10^-6, and 10^-8) of the soil suspension were used as inocula to create a gradient of microbial diversity in soils²¹. An initial soil suspension (10⁰) was prepared by thoroughly mixing 200 g soil (equivalent dry mass) with 600 ml sterile distilled water, followed by serial dilution. For each dilution level, the γ-sterilized soils were mixed with 20% (w/w) the soil suspension. To eliminate difference in inoculum biomass caused by serial dilution and to allow bacterial colonization, all γ-sterilized soils inoculated with suspensions were pre-incubated at 20°C for 30 days^21,22. On day 30 after pre-incubation, the bacterial and fungal abundances and diversity in the re-inoculated soils were quantified using methods mentioned in the following sections (Sampling, DNA extraction, microbial quantification and amplicon sequencing, and Bioinformatics analysis of 16 S rRNA and ITS gene profiling) and showed that the incubated soils had globally similar abundances (Supplementary Fig. 5a, b) but different microbial diversities (Supplementary Fig. 5d–g). These incubated soils were used to grow cucumber plants. For each dilution level, there were three independent replicates with 36 seedlings per replicate, and half of the cucumber plants were inoculated with infective J2s of M. incognita (uninoculated plants served as control). On day 35 after inoculation, the physiological properties of seedlings, nematode infection in roots, and bacteria in RS and RE were measured to evaluate plant performance.

In Exp. 5, which aimed to evaluate the potential of native root-associated bacterial strains to inhibit M. incognita infection, bacterial strains were isolated from the RS and RE components of M. incognita-infected roots by colony picking combined with 16 S rRNA gene profiling. The RS and the smashed root (RE) samples of M. incognita-infected cucumber, grown in a mixture of three soils (i.e. site 1-BJ, site 2-SD and site 3-SX) with equal volumes of each soil, were collected for the isolation of bacterial strains (for more details see the section Bacterial isolation, identification and assessment of anti-M. incognita activity). Here, we performed random isolation and screening of root-associated bacterial strains from M. incognita-infected roots. This strategy was chosen as it allows for the inclusion of a wider range of native root-associated bacteria, thereby facilitating a comprehensive understanding of the diversity and ecological functions of these bacteria in M. incognita-infected roots. Although the enrichment and depletion of specific bacterial groups induced by M. incognita infection provide guidance and opportunities for subsequent isolation, characterization and activity testing of microbes, we chose not to conduct targeted screening of these enriched bacterial groups in order to avoid the potential selection bias that could lead to an incomplete representation of the microbial population and the loss of rare species. The antagonistic activity of a bacterial strain towards M. incognita was assessed by the in vitro detection of the juvenile mortality of M. incognita combined with the in vivo verification of nematode infection in roots in sterilized soils (for more details see the section Bacterial isolation, identification and assessment of anti-M. incognita activity).

In Exp. 6, which aimed to determine strain diversity effects on M. incognita infection and thereby to construct simplified SynComs capable of inhibiting M. incognita infection, a design that randomly selected a gradient of increasing strain number (not strain identity or particular strain combinations) was applied to test if M. incognita infection shows a consistent directional trend along the number of bacterial strains²³. This design allowed us to efficiently determine the minimum number of bacterial strains required for constructing an efficient and simplified SynCom (for more details see the section Construction of simplified SynComs). The functional bacterial strains selected for this design were those having pronounced antagonistic activity against M. incognita infection (corrected nematode mortality >90%) without inhibiting plant performance (the ratio of plant performance between the functional strain inoculation and the control > 1.0; for more details see the section Bacterial isolation, identification and assessment of anti-M. incognita activity). After confirming the minimum number (i.e. six) of bacterial strains required for a simplified SynCom, we constructed 10 different simplified SynComs (A-J), all of which contained six functional bacterial strains, and further screened out the best one (SynCom C) that had the highest efficiency in terms of M. incognita infection inhibition. Synergistic interactions between bacterial strains in the SynCom C were evaluated by quantification of biofilm formation and attraction between colonies. Furthermore, a two-level factorial experiment (SynCom C inoculation × M. incognita infection) was conducted for the SynCom C. The SynCom C inoculation treatments included (i) uninoculated plants and (ii) SynCom C-inoculated plants, while the M. incognita infection treatments included (i) uninfected plants and (ii) M. incognita-infected plants, resulting in a total of 4 experimental treatments. This experiment was conducted in both sterilized and unsterilized soils. Each treatment had three independent replicates with 18 seedlings for each replicate. For the SynCom inoculation, individual bacterial strains were cultured in liquid medium, adjusted to an approximate OD₆₀₀ of 1.0 and then pooled together with equal volumes of each strain. Cells of bacterial strains were collected, washed and resuspended in PBS solution (with a final density of 1.0 × 10⁸ CFU ml^-1). The inoculation was performed by pouring the bacterial suspension to the root zone using a sterile lance tip. Each plant received in total 10⁹ cells. The physiological properties of seedlings and nematode infection in roots were determined 35 days after M. incognita inoculation. Moreover, plant transcriptome and metabolite profiling were measured on day 7 after M. incognita inoculation (for more details see the section RNA extraction, transcriptomic analysis and metabolomic analysis).

Nematode infection assays

Nematode infection was evaluated as described in Warmerdam et al.³. Briefly, whole roots were sampled and stained with acid fuchsin, in order to visually inspect the nematode, gall, and egg masses. Fuchsin-stained roots were observed by using an Olympus CX31 microscope (Olympus, Japan), to distinguish the disease developmental stages (i.e. the life cycle of M. incognita in roots) and record the number of galls, juveniles, and egg masses. The numbers of galls and egg masses per gram root were calculated to eliminate the influence of plant morphology on nematode infection⁴².

Measurements of soil properties and plant parameters

Soil texture, pH, electrical conductivity (EC), organic matter, available nitrogen (N), available phosphorus (P), available potassium (K), and nutrient elements N, P, K, Ca, Mg, S, B, Cu, Fe, Mn, Zn, Mo in soils were determined using methods as described in Tian et al.⁴³. Plant biomass was measured by drying plant materials at 75 °C to constant weight. Plant growth parameters (i.e. plant weight, root length, root volume, root surface area, root tips and forks) were measured to evaluate plant performance as described in Li et al.³³. To maintain clean roots after growing in soils, the entire soil-covered root system was delicately removed from the cultivation pot and the loose soil was carefully shaken off. Following a deionized water rinse, the complete root system underwent scanning using a specialized scanner system (Expression 4990, Epson, Long Beach, CA).

Sampling, DNA extraction, microbial quantification and amplicon sequencing

The bulk soil (BS), rhizosphere (RS) and root endosphere (RE) compartments were sampled using the method of Niu et al.⁴¹, and then used to extract and purify the genomic DNA using the PowerSoil DNA Isolation Kit (QIAGEN Inc., CA, USA) in accordance with the manufacturer’s instructions. The integrity of the extracted genomic DNA was verified by electrophoresis in 1% agarose gel, and its purity was measured and quantified by spectrophotometry using a Nanodrop ND-1000 spectrophotometer (Thermo Scientific, Wilmington, DE). The bacterial and fungal universal primers used are listed in Supplementary Table 5. To quantify the bacterial and fungal abundances, quantitative PCR (qPCR) was performed using iQ SYBR Green Supermix (BioRad), template DNA (5 ng) and universal primers for bacterial 16 S rRNA and fungal ITS genes. Three technical replicates showed high reproducibility (mean SEM < 0.6% of mean), so only biological replicates (n = 3) were run. The proportions of bacteria and fungi were calculated based on their Cq values. PCR amplification for bacteria and fungi was performed under the conditions as described in Wang and Sugiyama⁴⁴. Each sample was amplified in triplicate (together with water control) in a 20 µl reaction system containing 4 μl 5×FastPfu buffer, 2 μl dNTPs (2.5 mM), 0.8 μl forward primer (5 μM), 0.8 μl reverse primer (5 μM), 0.4 μl FastPfu polymerase, 0.2 μl BSA and 10 ng of template DNA. If no visible amplification was observed from negative control (no template added), triplicate PCR products were combined and purified using the QIAquick PCR purification kit (Qiagen, Valencia, CA, USA). The combined PCR products were used to prepare sample libraries for sequencing according to the MiSeq Reagent Kit Preparation Guide (Illumina, San Diego, CA, USA) as detailed in Caporaso et al.⁴⁵. Finally, sample libraries were subjected to a single sequencing run on the MiSeq platform (Illumina, Inc., San Diego, CA, USA) as per the standard MiSeq sequencing procedures.

Bioinformatics analysis of 16 S rRNA and ITS gene profiling

Sequencing analyzes for 16 S rRNA (bacteria) and ITS (fungi) genes were performed as detailed in Wang and Sugiyama⁴⁴. The low-quality and chimeric sequences were removed to obtain the valid sequences, which were further binned into operational taxonomic units (OTUs; 97% similarity) and classified taxonomically as described in Li et al.¹⁴. The α-diversity (species richness) was mirrored by the observed species (OTU) number, which was calculated using the R package vegan v. 2.5–6⁴⁶. The differences in α-diversity among samples were tested using Wilcoxon rank-sum tests. The β-diversity (variation in community composition) was calculated based on Bray-Curtis distance, and then subjected to principle coordinates analysis (PCoA) to reflect the variation of microbial community composition that is attributable to soil types and nematode infection treatments⁴⁷. The PCoA and permutational multivariate analysis of variance (PERMANOVA) were carried out using the function ‘pcoa’ and ‘adonis’ of the R package vegan v. 2.5–6⁴⁶. The LEfSe was performed based on the pairwise Wilcoxon rank-sum test (p < 0.05) using the Galaxy Module online (http://huttenhower.sph.harvard.edu/galaxy), to identify the significantly changed bacterial taxa (at genus to phylum levels; the absolute LDA score > 3.0) due to nematode infection in both RS and RE⁴⁸. Circos plots were created using the circlize package of R to visualize the similarities and differences of microbiomes among treatments. The potential function of the microbial community was predicted using PICRUSt2 (https://github.com/picrust/picrust2), and the functional richness (Fric) index was calculated using the dbFD function in the ‘FD’ R package⁴⁹. To identify fungal phyla (and at the family level), the sequence data were analyzed using bioinformatics tools to compare the sequences to known fungal taxa in the UNITE database (https://unite.ut.ee/). Subsequently, the sequences were classified based on their genetic similarities and evolutionary relationships using phylogenetic analysis.

Bacterial isolation, identification and assessment of anti-M. incognita activity

Bacterial isolation was conducted as described in Bai et al.⁵⁰. Briefly, the RS and the smashed root (RE) samples of M. incognita-infected cucumber were resuspended in PBS buffer (0.1 M phosphate buffer, 0.15% Tween 80, pH 7.0) and then homogenized. Homogenates were serially diluted and plated on two solidified bacterial growth media (R2A and TSA), before isolates were picked from plates containing less than 20 colony-forming units. All isolates were purified five times to obtain the purified bacterial strains, which were further identified based on full-length 16 S rRNA gene sequencing using universal primer pair 27 F (5′-AGAGTTTGATCMTGGCTCAG-3′) and 1492 R (5′-CTACGGCTACCTTGTTACGA-3′)²⁴ containing specific barcodes. The unique strains were subsequently deposited in 30% glycerol solutions and stored at − 80 °C for subsequent in vitro detection of the juvenile mortality of M. incognita, in vivo verification of nematode infection in roots, and evaluation of strain effect on plant growth.

The antagonistic activity of bacterial strains towards M. incognita was assessed by the in vitro detection of the juvenile mortality of M. incognita as well as the in vivo verification of nematode infection in roots. The in vitro detection of the juvenile mortality of M. incognita was conducted as described in Wei et al.⁵¹. The sterile 48-well plates were used to carry out an in vitro detection of J2 mortality. Specifically, various bacterial strains were cultured in liquid medium and adjusted to an approximate OD₆₀₀ of 1.0. To obtain bacterial metabolites, the liquid culture of bacterial strains was centrifuged at 10615 × g for 5 min and the resulting supernatant was subsequently filtered through a 0.2 μm membrane. Freshly incubated surface-sterilized J2s were pre-enriched and adjusted to a concentration of 100 J2s per 100 μl. Subsequently, 100 μl of bacterial metabolites and 100 μl of J2 suspension were aspirated into petri dishes, ensuring darkness and maintaining a constant temperature. After an incubation period of approximately 24 h, the mortality rate of the second instar larvae was assessed. Each bacterial strain was replicated nine times. To determine the viability of the J2s, we observed and monitored their mobility. Highly active J2s displayed frequent oscillations of the head and tail, accompanied by body curling. In contrast, J2s with low activity or deceased ones exhibited minimal movement or even complete immobility, along with a rigid body posture. Nematode mortality (M) was calculated according to the equation M (%) = [N₁/(N₁ + N₂)] × 100, where N₁ and N₂ represent the number of dead nematode and surviving nematode, respectively. All mortality data were corrected by the following equation M_A (%) = [(M_t − M_c)/(1 − M_c)] × 100, where M_A is the corrected nematode mortality, and M_t and M_c represent the nematode mortality in the treatment and the control, respectively⁵². For in vivo verification of nematode infection in roots, cells of a bacterial strain were collected, washed and resuspended in PBS solution (with a final density of 1.0 × 10⁸ CFU ml⁻¹) as described in Schmitz et al.⁵³. Inoculation of plants with a bacterial strain was conducted immediately after sowing. The inoculation was performed by pouring the bacterial suspension to the root zone using a sterile lance tip. Each plant received a total of 10⁹ cells, as recommended by Schmitz et al.⁵³. For each bacterial strain, there were 36 seedlings, and half of the cucumber plants were inoculated with infective surface-sterilized J2s of M. incognita (uninoculated plants served as control). On day 35 after inoculation, the physiological properties of seedlings and nematode infection in roots were measured to evaluate plant performance. To comprehensively evaluate plant performance, all plant growth parameters (i.e. plant weight, root length, root volume, root surface area, root tips and forks) were used in a principal component analysis (PCA) as described in Li et al.³³. The first principal component (PC1), which explained the major variation, was recognized as the plant performance index (PPI). It was assumed that a higher PPI meant a better plant performance.

Construction of simplified SynComs

Based on in vitro detection of the juvenile mortality of M. incognita, in vivo verification of nematode infection in roots, and evaluation of strain effect on plant performance, we selected 14 bacterial strains having pronounced antagonistic activity against M. incognita infection without inhibiting plant performance. To determine the minimum number of bacterial strains required for constructing a simplified SynCom capable of efficiently inhibiting M. incognita infection, a design that randomly selected a gradient of increasing bacterial strain number, namely the levels 1, 2, 3, 4, 5, 6, 8, 10, 12, and 14 strains (each replicated 9 times), was applied to test if M. incognita infection shows a consistent directional trend along the number of bacterial strains²³. Since our objective is to determine how an increasing number of bacterial strains influence M. incognita infection, our resolution does not allow statements on specific, individual strain interactions (addressing such interactions would lead to a design encompassing all strain combinations with a huge number of unique treatments and experimental units). Furthermore, we randomly constructed 10 different simplified SynComs (A-J), all of which contained the lowest number of bacterial strains. Based on nematode infection assays, we further screened out the best simplified SynCom (SynCom C) that had the highest efficiency in terms of M. incognita infection inhibition.

Characterization of bacterial strains, and synergistic interactions between bacterial strains

Each bacterial strain in the SynCom C was characterized by the determination of plant growth promotion-related genes (i.e. PGA, hcnBC, acdS, and codA) as described in Fan et al.⁵⁴.

The PGA gene is responsible for the production of Poly-γ-glutamic acid (PGA), a polymer composed of glutamic acid units that can induce the activation of the plant immune system and boost resistance against pathogens. The hcnBC gene encodes the hydrogen cyanide synthase involved in the biosynthesis of hydrogen cyanide (HCN), a chemical deterrent that can inhibit the growth and development of herbivores and pathogens. The acdS gene is involved in the degradation of the plant stress hormone ethylene, which tends to accumulate under various environmental stresses and can contribute to the development of diseases in plants. The codA gene encodes the cytosine deaminase (CODA), which can convert non-toxic compounds into toxic derivatives that affect a range of organisms or biological entities. Moreover, nitrogen fixation, phosphate solubilization, potassium decomposition, siderophore secretion, and phenotypic/physiological profiles (i.e. reducing power, sensitivity to salinity, and carbon source utilization) in the Biolog GEN III MicroPlate^TM test were measured using the methods described by Woźniak et al.⁵⁵. Synergistic interactions between bacterial strains were evaluated by quantification of biofilm formation as well as attraction between colonies as described in Berendsen et al.⁸. An attraction index (AI), defined as the attraction between colonies, was calculated according to the equation AI = 1- [(D1-D2)/D1 + (d1-d2)/d1] /2, where D1 and d1 represent the diameter of isolates B1 and B2 when they were farthest apart, respectively, and D2 and d2 represent the diameter of isolates B1 and B2 when they were nearest apart, respectively. Metabolites of bacterial strains were measured by LC-MS/MS analysis⁵⁶. Briefly, the bacterial liquid sample was mixed with a solution of acetonitrile and methanol (in a 1:1 ratio, v:v) containing 0.02 mg/ml L-2-chlorophenylalanine and subjected to centrifugation to separate the metabolites. The resulting mixture was then dried under nitrogen and re-solubilized with a solution of acetonitrile and water (in a 1:1 ratio, v:v). The prepared samples were subjected to ultrasonication and centrifugation once again before being analyzed using LC-MS/MS on a Thermo UHPLC-Q Exactive HF-X system. Metabolite identification was carried out by comparing the acquired spectra with entries in the HMDB database (accessible at http://www.hmdb.ca/), and the resulting data matrix obtained from the database search was uploaded to the Majorbio cloud platform (https://cloud.majorbio.com) for further data analysis. A full description of the mass spectrometry experiments used to identify metabolites is provided in Supplementary Data 1. All spectral data and associated data used to identify microbial metabolites are provided in Supplementary Data 2 (for bacterial strains) and Supplementary Data 3 (for the SynCom C). The identification of specific metabolites (norvaline, betaine, L-erythrulose, isonicotinic acid, uridine, L-pipecolic acid, pantothenic acid, and beta-L-fucose) was confirmed using authentic standards, based on the retention times and m/z values of the molecular ions along with MS/MS for structural confirmation (Supplementary Table 6).

To uncover the mechanisms underlying nematode suppression by cucumber root-associated bacterial strains, we used several approaches. These included assessments of M. incognita egg hatching rates and mortality in infective J2s, as well as analysis of the expression of M. incognita flp-1 and flp-18 genes. These two genes are critical for the chemotaxis and parasitism of M. incognita towards cucumber roots, as demonstrated in our previous research⁵⁷. Moreover, we examined the movement patterns of M. incognita towards cucumber roots using the dual-choice attraction assay⁵⁷. Both overall metabolites and specific metabolites (e.g. norvaline, betaine, L-erythrulose, isonicotinic acid, uridine, L-pipecolic acid, pantothenic acid, and beta-L-fucose) were used to evaluate their efficacy in nematode suppression. The tested metabolites were selected due to their high abundance in the secretions of bacterial strains. To obtain overall metabolites, the liquid culture (with an OD₆₀₀ of 0.4) of bacterial strains was centrifuged at 10615 × g for 5 min and the resulting supernatant was subsequently filtered through a 0.2 μm membrane. The overall metabolites of Brevibacillus choshinensis, a strain that did not exhibit nematode inhibition during the screening of bacterial strains were used as a control. For selected specific metabolites, a concentration of 0.5 mM, which has been demonstrated to be suitable for detecting the toxicity of bacterial metabolites against plant parasitic nematodes in several previous studies^58–60, was used for subsequent assays. Simultaneously, L-phenylalanine, a metabolite detected in non-effective bacteria, served as a control. According to target quantification using LC-MS/MS, the bacterial strains were capable of producing metabolites at such a level in vitro. Although this concentration level might be higher compared to in situ concentrations in the rhizosphere, it can be reached under in situ conditions when bacterial strains are inoculated into roots. To examine the inhibitory effect of metabolites from bacterial strains or SynCom C on M. incognita egg hatching, egg masses were collected and washed with sterilized water, followed by treatment with 0.05 % (v/v) sodium hypochlorite to separate individual eggs for further analysis. The eggs were then diluted with water to create a suspension solution containing 50 eggs per milliliter and exposed to different concentrations of metabolite solutions to measure the hatching rate on day 7 after exposure⁵⁷. The mortality in infective J2s was detected as mentioned above⁵¹. To analyze the expression of M. incognita flp-1 and flp-18 genes, qRT-PCR was performed using SYBR Green PCR Master Mix (Vazyme) on an ABI 6500 Real-Time PCR System (Applied Biosystems). The primers for qRT-PCR of M. incognita flp-1 gene were forward primer flp1-re-F (5′-GTGCTACAAGTGCCAACAG-3′) and reverse primer flp1-re-R (5′-CCAGCAGCACGTTTAATTCC-3′), and for M. incognita flp-18 gene were forward primer flp18-re-F (5′-GAGATAATTGCAAATGGAGAG-3′) and reverse primer flp18-re-R (5′-CATCGTCTATAAGGAGACCTTG-3′). The M. incognita actin (accession BE225475; forward primer Mi-actin-F 5′-GTTATTCTTTCACCGCAACCG-3′, reverse primer Mi-actin-R 5′-CGTCAGGCAATTCATAGCTC-3′) was employed as the reference gene to normalize the gene expression⁵⁷. In addition, the dual-choice attraction assay was conducted based on the methodology described in Zhang et al.⁵⁷. Briefly, roots treated with bacterial strains’ metabolites or water were placed on opposite sides of a 13 × 13 cm square Petri dish containing sterilized solid agar gel, which was scooped out from the area between two parallel imaginary lines and replaced with 23% Pluronic F-127 gel to create a zone suitable for M. incognita inoculation and movement.

Collection of rhizosphere metabolites

Rhizosphere metabolites of non-infected and infected roots were collected as described in Zhang et al.⁵⁷. Briefly, the roots were harvested from the pots and shaken to remove large soil aggregates before the entire root system was collected in a 50 ml tube containing cold extraction solution. After shaking the tube for 1 min at 4 °C, the rhizosphere solution was separated from root tissues and placed in liquid nitrogen, then stored at -80°C for further analysis. The rhizosphere supernatant was freeze-dried and re-dissolved in a mixture of methanol and water (in a ratio of 1:4, v/v) for metabolomics analysis. Six biological replicates were used for each treatment. A full description of the mass spectrometry experiments used to identify metabolites is provided in Supplementary Data 1. All spectral data and associated data used to identify rhizosphere metabolites are provided in Supplementary Data 4.

RNA extraction, transcriptomic analysis and metabolomic analysis

Total RNA was extracted and purified from root samples using a modified ‘pine-tree’ method, followed by determination of RNA integrity and concentration, according to the method of Chialva et al.⁶¹. RNA-seq library preparation, sequencing, and bioinformatics analysis were performed to obtain transcriptomic profiles as described in Cheng et al.⁶². The original data was filtered using ‘fastp’ to remove reads containing adapters. The paired reads were removed if the N content exceeded 10% of the total bases in any sequencing read, and were discarded if the number of low-quality bases (Q ≤ 20) exceeded 50% of the total bases in the read. Subsequently, the resulting clean read pairs were compared to the cucumber genome sequence (http://cucurbitgenomics.org/ Chinese long CDS V2). The gene alignment was conducted using featureCounts, followed by the calculation of the FPKM (Fragments Per Kilobase of transcript per Million mapped reads) for each gene based on its length. DESeq2 was then employed to analyze differential expression between the two groups, with correction of p-values using the Benjamini & Hochberg method. The adjusted p-value and log2 fold change (log₂FC) were employed as thresholds to determine significant differences. For KEGG (Kyoto encyclopedia of genes and genomes) pathway analysis, the hypergeometric distribution test was performed at the pathway level. The candidate genes associated with plant defense responses were verified by qRT-PCR, which was performed using SYBR Green PCR Master Mix (Vazyme) on an ABI 6500 Real-Time PCR System (Applied Biosystems). The CDS sequences and primers used for qRT-PCR of candidate genes are provided in Supplementary Data 5. The UBI (accession CsaV3_2G036600; forward primer UBI-F 5′-CCTTATTGACCAACCAGTAGT-3′, reverse primer UBI-R 5′-GGACAATGTTGATTTCCTCG-3′) was employed as the reference gene to normalize the gene expression⁶³. To obtain metabolomic profiles, root samples were freeze-dried, crushed and prepared for metabolomic analysis of metabolites using UPLC and ESI-Q TRAP–MS/MS, as described by Zhang et al.⁶⁴. Briefly, the biological samples underwent vacuum freeze-drying and were then ground into a powder. Subsequently, a mixture of the sample powder and internal standard extract was prepared, followed by vortexing and filtration prior to UPLC-MS/MS analysis. A full description of the mass spectrometry experiments used to identify metabolites is provided in Supplementary Data 1. All spectral data and associated data used to identify root metabolites are provided in Supplementary Data 6. Differential metabolites (DMs) for the two-group analysis were determined using VIP ( > 1) and Log₂FC (≥ 1.0) values extracted from OPLS-DA result. Before conducting the OPLS-DA, the data was log-transformed and mean-centered, and a permutation test was performed to prevent overfitting. Data analysis and graphing for both transcriptomic and metabolomic profiles were carried out by a free platform named Metware Cloud (https://cloud.metware.cn). KEGG pathway was plotted by an online tool named OmicStudio (https://www.omicstudio.cn/tool).

Statistical analyzes

Statistical analyzes were carried out in R v.4.1.3. Permutational multivariate analysis of variance (PERMANOVA) was performed using the R package vegan v. 2.5-6 (Oksanen et al., 2019) to investigate the effects of soil types and nematode infection on the response of microbial community composition⁴⁶. Effects of soil sterilization, nematode infection, microbial diversity loss and bacterial inoculation on plant performance were analyzed using one-way ANOVA followed by multiple comparisons via Tukey’s test or a two-tailed, unpaired Student’s t-test. Two-way ANOVAs were conducted to test the effects of soil sterilization, nematode infection and their interaction on plant performance, as well as the effects of sampling time, nematode infection and their interaction on microbial diversity. When necessary, data were transformed before conducting statistical analyzes to ensure data normalization and equal variance. Linear models were performed to decipher trends in the changes in α-diversity (observes species richness) and functional richness (Fric index) across the decreasing gradient of microbial diversity (using the ggpmisc package in R), and to explore the relationships between the α-diversity (observed species richness) and root-knot number (using the ggpubr package in R).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplemental information^{(2.4MB, pdf)}

Peer Review File^{(7MB, pdf)}

Supplementary Data 1^{(15.1KB, xlsx)}

Supplementary Data 2^{(656.1KB, xlsx)}

Supplementary Data 3^{(125.4KB, xlsx)}

Supplementary Data 4^{(2.7MB, xlsx)}

Supplementary Data 5^{(13.8KB, xlsx)}

Supplementary Data 6^{(379.8KB, xlsx)}

Reporting Summary^{(2MB, pdf)}

Source data

Source Data^{(2.2MB, xlsx)}

Acknowledgements

The authors appreciate the financial support received from the National Natural Science Foundation of China (Project 32372791 to Y.T.), and the earmarked fund for China Agriculture Research System (CARS-23 to L.G.).

Author contributions

Y.T. and L.G. supervised the study. S.L. and J.L. set up the experiment, collected the samples, and performed the lab work. S.L., J.L., S.M., X.L., and Y.T. analyzed the data. S.L. and J.L. contributed to the isolation and identification of bacterial strains. Y.T. and S.L. wrote the first draft. Y.T. revised the manuscript. All authors commented on the manuscript.

Peer review

Peer review information

Nature Communications thanks Jun Yuan, Youn-Sig Kwak, Desalegn Etalo, Christian Jung and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The raw data of microbial amplicon reads generated in this study have been deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) database under accession codes PRJNA857583 [https://www.ncbi.nlm.nih.gov/sra/SRX16203999], PRJNA857847 [https://www.ncbi.nlm.nih.gov/sra/SRX16141965], and PRJNA857979 [https://www.ncbi.nlm.nih.gov/sra/SRX16141747]. The transcriptomics raw data generated in this study have been deposited in the NCBI SRA database under accession code PRJNA1019930 [https://www.ncbi.nlm.nih.gov/sra/SRX21871799]. The metabolomics raw data of bacterial strains and the SynCom C generated in this study have been deposited in the National Genomics Data Center (NGDC) Open Archive for Miscellaneous Data (OMIX) database under accession code OMIX006899. The metabolomics raw data of the rhizosphere and roots generated in this study have been deposited in the NGDC OMIX database under accession code OMIX006901. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Shikai La, Jiafan Li.

Contributor Information

Lihong Gao, Email: gaolh@cau.edu.cn.

Yongqiang Tian, Email: tianyq1984@cau.edu.cn.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-024-51073-7.

References

1.Rutter, W. B., Franco, J. & Gleason, C. Rooting out the mechanisms of root-knot nematode– plant interactions. Ann. Rev. Phytopathol.60, 3.1–3.24 (2022). 10.1146/annurev-phyto-021621-120943 [DOI] [PubMed] [Google Scholar]
2.Holbein, J. et al. Root endodermal barrier system contributes to defence against plant‐parasitic cyst and root‐knot nematodes. Plant J.100, 221–236 (2019). 10.1111/tpj.14459 [DOI] [PubMed] [Google Scholar]
3.Warmerdam, S. et al. Genome-wide association mapping of the architecture of susceptibility to the root-knot nematode Meloidogyne incognita in Arabidopsis thaliana. N. Phytol.218, 724–737 (2018). 10.1111/nph.15034 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Castagnone‐Sereno, P. et al. Gene copy number variations as signatures of adaptive evolution in the parthenogenetic, plant‐parasitic nematode Meloidogyne incognita. Mol. Ecol.28, 2559–2572 (2019). 10.1111/mec.15095 [DOI] [PubMed] [Google Scholar]
5.Wram, C. L. & Zasada, I. A. Short-term effects of sublethal doses of nematicides on Meloidogyne incognita. Phytopathology109, 1605–1613 (2019). 10.1094/PHYTO-11-18-0420-R [DOI] [PubMed] [Google Scholar]
6.Zasada, I. A. et al. Managing nematodes without methylbromide. Ann. Rev. Phytopathol.48, 311–328 (2010). 10.1146/annurev-phyto-073009-114425 [DOI] [PubMed] [Google Scholar]
7.van Loon, L. C. The intelligent behavior of plants. Trends Plant Sci.21, 286–294 (2016). 10.1016/j.tplants.2015.11.009 [DOI] [PubMed] [Google Scholar]
8.Berendsen, R. L. et al. Disease-induced assemblage of a plant-beneficial bacterial consortium. ISME J.12, 1496–1507 (2018). 10.1038/s41396-018-0093-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Castrillo, G. et al. Root microbiota drive direct integration of phosphate stress and immunity. Nature543, 513–518 (2017). 10.1038/nature21417 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Santhanam, R. et al. Native root-associated bacteria rescue a plant from a sudden-wilt disease that emerged during continuous cropping. Proc. Natl Acad. Sci. USA112, E5013–E5020 (2015). 10.1073/pnas.1505765112 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Verbon, E. H. & Liberman, L. M. Beneficial microbes affect endogenous mechanisms controlling root development. Trends Plant Sci.21, 218–229 (2016). 10.1016/j.tplants.2016.01.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Bakker, P. A., Pieterse, C. M., de Jonge, R. & Berendsen, R. L. The soil-borne legacy. Cell172, 1178–1180 (2018). 10.1016/j.cell.2018.02.024 [DOI] [PubMed] [Google Scholar]
13.Mendes, L. W., Raaijmakers, J. M., de Hollander, M., Mendes, R. & Tsai, S. M. Influence of resistance breeding in common bean on rhizosphere microbiome composition and function. ISME J.12, 212–224 (2018). 10.1038/ismej.2017.158 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Li, H., La, S., Zhang, X., Gao, L. & Tian, Y. Salt-induced recruitment of specific root-associated bacterial consortium capable of enhancing plant adaptability to salt stress. ISME J.15, 2865–2882 (2021). 10.1038/s41396-021-00974-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Schlatter, D., Kinkel, L., Thomashow, L., Weller, D. & Paulitz, T. Disease suppressive soils: new insights from the soil microbiome. Phytopathology107, 1284–1297 (2017). 10.1094/PHYTO-03-17-0111-RVW [DOI] [PubMed] [Google Scholar]
16.Rolfe, S. A., Griffiths, J. & Ton, J. Crying out for help with root exudates: adaptive mechanisms by which stressed plants assemble health-promoting soil microbiomes. Curr. Opin. Microbiol.49, 73–82 (2019). 10.1016/j.mib.2019.10.003 [DOI] [PubMed] [Google Scholar]
17.Trivedi, P., Batista, B. D., Bazany, K. E. & Singh, B. K. Plant–microbiome interactions under a changing world: Responses, consequences and perspectives. N. Phytol.234, 1951–1959 (2022). 10.1111/nph.18016 [DOI] [PubMed] [Google Scholar]
18.Liu, H., Macdonald, C. A., Cook, J., Anderson, I. C. & Singh, B. K. An ecological loop: host microbiomes across multitrophic interactions. Trends Ecol. Evol.34, 1118–1130 (2019). 10.1016/j.tree.2019.07.011 [DOI] [PubMed] [Google Scholar]
19.Mendes, R. et al. Deciphering the rhizosphere microbiome for disease-suppressive bacteria. Science332, 1097–1100 (2011). 10.1126/science.1203980 [DOI] [PubMed] [Google Scholar]
20.Berendsen, R. L., Pieterse, C. M. & Bakker, P. A. The rhizosphere microbiome and plant health. Trends Plant Sci.17, 478–486 (2012). 10.1016/j.tplants.2012.04.001 [DOI] [PubMed] [Google Scholar]
21.Domeignoz-Horta, L. A. et al. Microbial diversity drives carbon use efficiency in a model soil. Nat. Commun.11, 3684 (2020). 10.1038/s41467-020-17502-z [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Yang, G., Roy, J., Veresoglou, S. D. & Rillig, M. C. Soil biodiversity enhances the persistence of legumes under climate change. N. Phytol.229, 2945–2956 (2021). 10.1111/nph.17065 [DOI] [PubMed] [Google Scholar]
23.Rillig, M. C. et al. The role of multiple global change factors in driving soil functions and microbial biodiversity. Science366, 886–890 (2019). 10.1126/science.aay2832 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Zhou, X. et al. Cross-kingdom synthetic microbiota supports tomato suppression of Fusarium wilt disease. Nat. Commun.13, 7890 (2022). 10.1038/s41467-022-35452-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Lamelas, A. et al. The bacterial microbiome of Meloidogyne-based disease complex in coffee and tomato. Front. Plant Sci.11, 510816 (2020). 10.3389/fpls.2020.00136 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Silva, J. C. et al. Organic practices intensify the microbiome assembly and suppress root-knot nematodes. J. Pest Sci.95, 709–721 (2022). 10.1007/s10340-021-01417-9 [DOI] [Google Scholar]
27.Martinson, V. G. et al. Multiple origins of obligate nematode and insect symbionts by a clade of bacteria closely related to plant pathogens. Proc. Natl Acad. Sci. USA117, 31979–31986 (2020). 10.1073/pnas.2000860117 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Williams, A. & de Vries, F. T. Plant root exudation under drought: implications for ecosystem functioning. N. Phytol.225, 1899–1905 (2020). 10.1111/nph.16223 [DOI] [PubMed] [Google Scholar]
29.Xiong, C., Zhu, Y. G., Wang, J. T., Singh, B., Han, L. L., Shen, J. P., Li, P. P., Wang, G. B., Wu, C. F., Ge, A. H., Zhang, L. M. & He, J. Z. Host selection shapes crop microbiome assembly and network complexity. N. Phytol.229, 1091–1104 (2021). 10.1111/nph.16890 [DOI] [PubMed] [Google Scholar]
30.Ravanbakhsh, M., Kowalchuk, G. A. & Jousset, A. Targeted plant hologenome editing for plant trait enhancement. N. Phytol.229, 1067–1077 (2021). 10.1111/nph.16867 [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Deng, X., Zhang, N., Shen, Z., Zhu, C., Liu, H., Xu, Z., Li, R., Shen, Q. & Salles, J. F. Soil microbiome manipulation triggers direct and possible indirect suppression against Ralstonia solanacearum and Fusarium oxysporum. npj. Biofilms Microbi33, 1–10 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Xiong, W., Li, R., Ren, Y., Liu, C., Zhao, Q., Wu, H., Jousset, A. & Shen, Q. Distinct roles for soil fungal and bacterial communities associated with the suppression of vanilla Fusarium wilt disease. Soil Biol. Biochem.107, 198–207 (2017). 10.1016/j.soilbio.2017.01.010 [DOI] [Google Scholar]
33.Li, E., de Jonge, R., Liu, C., Jiang, H., Friman, V. P., Pieterse, C. M., Bakker, J., Jousset, P. A. H. M. & Rapid, A. evolution of bacterial mutualism in the plant rhizosphere. Nat. Commun.12, 3829 (2021). 10.1038/s41467-021-24005-y [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Wagg, C., Schlaeppi, K., Banerjee, S., Kuramae, E. E. & van der Heijden, M. G. A. Fungal-bacterial diversity and microbiome complexity predict ecosystem functioning. Nat. Commun.10, 4841 (2019). 10.1038/s41467-019-12798-y [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Hassani, M., Durán, P. & Hacquard, S. Microbial interactions within the plant holobiont. Microbiome6, 1–17 (2018). 10.1186/s40168-018-0445-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Coyte, K. Z., Rao, C., Rakoff-Nahoum, S. & Foster, K. R. Ecological rules for the assembly of microbiome communities. PLoS Biol.19, e3001116 (2021). 10.1371/journal.pbio.3001116 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Zandalinas, S. I. & Mittler, R. Plant responses to multifactorial stress combination. N. Phytol.234, 1161–1167 (2022). 10.1111/nph.18087 [DOI] [PubMed] [Google Scholar]
38.Macedo-Raygoza, G. M., Valdez-Salas, B., Prado, F. M., Prieto, K. R., Yamaguchi, L. F., Kato, M. J., Canto-Canché, B. B., Carrillo-Beltrán, M., Mascio, P. D., White, J. F. & Beltrán-García, M. J. Enterobacter cloacae, an endophyte that establishes a nutrient-transfer symbiosis with banana plants and protects against the black Sigatoka pathogen. Front. Microbiol.10, 804 (2019). 10.3389/fmicb.2019.00804 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Liu, H., Li, J., Carvalhais, L. C., Percy, C. D., Prakash Verma, J., Schenk, P. M. & Singh, B. K. Evidence for the plant recruitment of beneficial microbes to suppress soil‐borne pathogens. N. Phytol.229, 2873–2885 (2021). 10.1111/nph.17057 [DOI] [PubMed] [Google Scholar]
40.Miyara, S. B., Ionit, I., Buki, P. & Kolomiets, M. Chapter six-the role of lipid signalling in regulating plant-nematode interactions. Adv. Bot. Res.73, 139–166 (2015). 10.1016/bs.abr.2014.12.004 [DOI] [Google Scholar]
41.Niu, B., Paulson, J. N., Zheng, X. & Kolter, R. Simplified and representative bacterial community of maize roots. Proc. Natl Acad. Sci. USA114, E2450–E2459 (2017). 10.1073/pnas.1616148114 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Zhang, X., Peng, H., Zhu, S., Xing, J., Li, X., Zhu, Z., Zheng, J., Wang, L., Wang, B. & Chen, J. et al. Nematode-encoded RALF peptide mimics facilitate parasitism of plants through the FERONIA receptor kinase. Mol. Plant13, 1434–1454 (2020). 10.1016/j.molp.2020.08.014 [DOI] [PubMed] [Google Scholar]
43.Tian, Y., Wang, Q., Zhang, W. & Gao, L. Reducing environmental risk of excessively fertilized soils and improving cucumber growth by Caragana microphylla-straw compost application in long-term continuous cropping systems. Sci. Total Environ.544, 251–261 (2016). 10.1016/j.scitotenv.2015.11.091 [DOI] [PubMed] [Google Scholar]
44.Wang, B. & Sugiyama, S. Phylogenetic signal of host plants in the bacterial and fungal root microbiomes of cultivated angiosperms. Plant J.104, 522–531 (2020). 10.1111/tpj.14943 [DOI] [PubMed] [Google Scholar]
45.Caporaso, J. G., Lauber, C. L., Walters, W. A., Berg-Lyons, D., Huntley, J., Fierer, N., Owens, S. M., Betley, J., Fraser, L. & Bauer, M. et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J.6, 1621–1624 (2012). 10.1038/ismej.2012.8 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Oksanen, J., Blanchet, F., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P., O’Hara, R., Simpson, G., Solymos, P. vegan: Community ecology package. R package v.2.5-6. [WWW document] URL https://CRAN.R-project.org/package=vegan (2019).
47.Javůrková, V. G., Kreisinger, J., Procházka, P., Požgayová, M., Ševčíková, K., Brlík, V., Adamík, P., Heneberg, P. & Porkert, J. Unveiled feather microcosm: feather microbiota of passerine birds is closely associated with host species identity and bacteriocin-producing bacteria. ISME J.13, 2363–2376 (2019). 10.1038/s41396-019-0438-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Segata, N., Izard, J., Waldron, L., Gevers, D., Miropolsky, L., Garrett, W. S. & Huttenhower, C. Metagenomic biomarker discovery and explanation. Genome Biol.12, R60–R60 (2011). 10.1186/gb-2011-12-6-r60 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Laliberté, E., Shipley, B. FD: measuring functional diversity from multiple traits, and other tools for functional ecology. R package version 1.0-11. [WWWdocument] URL https://cran.r-project.org/web/packages/FD/FD.pdf [accessed 30 September 2023].
50.Bai, Y., Müller, D. B., Srinivas, G., Garrido-Oter, R., Potthoff, E. & Rott, M. et al. Functional overlap of the Arabidopsis leaf and root microbiota. Nature528, 364–369 (2015). 10.1038/nature16192 [DOI] [PubMed] [Google Scholar]
51.Wei, L. H., Xue, Q. Y., Wei, B. Q., Wang, Y. M., Li, S. M., Chen, L. F. & Guo, J. H. Screening of antagonistic bacterial strains against Meloidogyne incognita using protease activity. Biocontrol Sci. Tech.20, 739–750 (2010). 10.1080/09583151003714109 [DOI] [Google Scholar]
52.Abbott, W. S. A method of computing the effectiveness of an insecticide. J. Econ. Entomol.18, 265–267 (1925). 10.1093/jee/18.2.265a [DOI] [PubMed] [Google Scholar]
53.Schmitz, L., Yan, Z., Schneijderberg, M., de Roij, M., Pijnenburg, R., Zheng, Q., Franken, C., Dechesne, A., Trindade, L. M., van Velzen, R., Bisseling, T., Geurts, R. & Cheng, X. Synthetic bacterial community derived from a desert rhizosphere confers salt stress resilience to tomato in the presence of a soil microbiome. ISME J.16, 1907–1920 (2022). 10.1038/s41396-022-01238-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Fan, P., Chen, D., He, Y., Zhou, Q., Tian, Y. & Gao, L. Alleviating salt stress in tomato seedlings using Arthrobacter and Bacillus megaterium isolated from the rhizosphere of wild plants grown on saline–alkaline lands. Int. J. Phytoremediat.18, 1113–1121 (2016). 10.1080/15226514.2016.1183583 [DOI] [PubMed] [Google Scholar]
55.Woźniak, M., Gałązka, A., Tyśkiewicz, R. & Jaroszuk-Ściseł, J. Endophytic bacteria potentially promote plant growth by synthesizing different metabolites and their phenotypic/physiological profiles in the Biolog GEN III MicroPlate^TM Test. Int. J. Mol. Sci.20, 5283 (2019). 10.3390/ijms20215283 [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Lin, Z., Li, Y., Zhang, Z., Liu, X., Hsu, C. C., Du, Y., Sang, T., Zhu, C., Wang, Y., Satheesh, V., Pratibha, P., Zhao, Y., Song, C. P., Tao, W. A., Zhu, J. K. & Wang, P. A RAF-SnRK2 kinase cascade mediates early osmotic stress signaling in higher plants. Nat. Commun.11, 613 (2020). 10.1038/s41467-020-14477-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Zhang, X., Song, M., Gao, L. & Tian, Y. Metabolic variations in root tissues and rhizosphere soils of weak host plants potently lead to distinct host status and chemotaxis regulation of Meloidogyne incognita in intercropping. Mol. Plant Pathol.25, e13396 (2024). 10.1111/mpp.13396 [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Kusakabe, A., Wang, C., Xu, Y. M., Molnár, I. & Stock, S. P. Selective toxicity of secondary metabolites from the entomopathogenic bacterium Photorhabdus luminescens sonorensis against selected plant parasitic nematodes of the Tylenchina suborder. Microbiol. Spectrum10, e02577–21 (2022). 10.1128/spectrum.02577-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Kusakabe, A., Molnár, I. & Stock, S. P. Photorhabdus-derived secondary metabolites reduce root infection by Meloidogyne incognita in cowpea. Plant Dis.107, 3383–3388 (2023). [DOI] [PubMed] [Google Scholar]
60.Du, C., Cao, S., Shi, X., Nie, X., Zheng, J., Deng, Y., Ruan, L., Peng, D. & Sun, M. Genetic and biochemical characterization of a gene operon for trans-aconitic acid, a novel nematicide from Bacillus thuringiensis. J. Biol. Chem.292, 3517–3530 (2017). 10.1074/jbc.M116.762666 [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Chialva, M., Salvioli di Fossalunga, A., Daghino, S., Ghignone, S., Bagnaresi, P., Chiapello, M., Novero, M., Spadaro, D., Perotto, S. & Bonfante, P. Native soils with their microbiotas elicit a state of alert in tomato plants. N. Phytol.220, 1296–1308 (2018). 10.1111/nph.15014 [DOI] [PubMed] [Google Scholar]
62.Cheng, Z., Liu, X., Yan, S., Liu, B., Zhong, Y., Song, W., Chen, J., Wang, Z., Che, G., Liu, L., Ying, A., Lv, H., Han, L., Li, M., Zhao, J., Xu, J., Yang, Z., Zhou, Z. & Zhang, X. Pollen tube emergence is mediated by ovary-expressed ALCATRAZ in cucumber. Nat. Commun.14, 258 (2023). 10.1038/s41467-023-35936-z [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Ji, T., Ma, S., Liang, M., Wang, X., Gao, L. & Tian, Y. Reference genes identification for qRT-PCR normalization of gene expression analysis in Cucumis sativus under Meloidogyne incognita infection and Pseudomonas treatment. Front. Plant Sci.13, 1061921 (2022). 10.3389/fpls.2022.1061921 [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Zhang, Y., Zhang, S., Zhang, J., Wei, W., Zhu, T., Qu, H., Liu, Y. & Xu, G. Improving rice eating and cooking quality by enhancing endogenous expression of a nitrogen‐dependent floral regulator. Plant Biotechnol. J.21, 2654–2670 (2023). 10.1111/pbi.14160 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental information^{(2.4MB, pdf)}

Peer Review File^{(7MB, pdf)}

Supplementary Data 1^{(15.1KB, xlsx)}

Supplementary Data 2^{(656.1KB, xlsx)}

Supplementary Data 3^{(125.4KB, xlsx)}

Supplementary Data 4^{(2.7MB, xlsx)}

Supplementary Data 5^{(13.8KB, xlsx)}

Supplementary Data 6^{(379.8KB, xlsx)}

Reporting Summary^{(2MB, pdf)}

Source Data^{(2.2MB, xlsx)}

Data Availability Statement

[CR1] 1.Rutter, W. B., Franco, J. & Gleason, C. Rooting out the mechanisms of root-knot nematode– plant interactions. Ann. Rev. Phytopathol.60, 3.1–3.24 (2022). 10.1146/annurev-phyto-021621-120943 [DOI] [PubMed] [Google Scholar]

[CR2] 2.Holbein, J. et al. Root endodermal barrier system contributes to defence against plant‐parasitic cyst and root‐knot nematodes. Plant J.100, 221–236 (2019). 10.1111/tpj.14459 [DOI] [PubMed] [Google Scholar]

[CR3] 3.Warmerdam, S. et al. Genome-wide association mapping of the architecture of susceptibility to the root-knot nematode Meloidogyne incognita in Arabidopsis thaliana. N. Phytol.218, 724–737 (2018). 10.1111/nph.15034 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Castagnone‐Sereno, P. et al. Gene copy number variations as signatures of adaptive evolution in the parthenogenetic, plant‐parasitic nematode Meloidogyne incognita. Mol. Ecol.28, 2559–2572 (2019). 10.1111/mec.15095 [DOI] [PubMed] [Google Scholar]

[CR5] 5.Wram, C. L. & Zasada, I. A. Short-term effects of sublethal doses of nematicides on Meloidogyne incognita. Phytopathology109, 1605–1613 (2019). 10.1094/PHYTO-11-18-0420-R [DOI] [PubMed] [Google Scholar]

[CR6] 6.Zasada, I. A. et al. Managing nematodes without methylbromide. Ann. Rev. Phytopathol.48, 311–328 (2010). 10.1146/annurev-phyto-073009-114425 [DOI] [PubMed] [Google Scholar]

[CR7] 7.van Loon, L. C. The intelligent behavior of plants. Trends Plant Sci.21, 286–294 (2016). 10.1016/j.tplants.2015.11.009 [DOI] [PubMed] [Google Scholar]

[CR8] 8.Berendsen, R. L. et al. Disease-induced assemblage of a plant-beneficial bacterial consortium. ISME J.12, 1496–1507 (2018). 10.1038/s41396-018-0093-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR9] 9.Castrillo, G. et al. Root microbiota drive direct integration of phosphate stress and immunity. Nature543, 513–518 (2017). 10.1038/nature21417 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Santhanam, R. et al. Native root-associated bacteria rescue a plant from a sudden-wilt disease that emerged during continuous cropping. Proc. Natl Acad. Sci. USA112, E5013–E5020 (2015). 10.1073/pnas.1505765112 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Verbon, E. H. & Liberman, L. M. Beneficial microbes affect endogenous mechanisms controlling root development. Trends Plant Sci.21, 218–229 (2016). 10.1016/j.tplants.2016.01.013 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Bakker, P. A., Pieterse, C. M., de Jonge, R. & Berendsen, R. L. The soil-borne legacy. Cell172, 1178–1180 (2018). 10.1016/j.cell.2018.02.024 [DOI] [PubMed] [Google Scholar]

[CR13] 13.Mendes, L. W., Raaijmakers, J. M., de Hollander, M., Mendes, R. & Tsai, S. M. Influence of resistance breeding in common bean on rhizosphere microbiome composition and function. ISME J.12, 212–224 (2018). 10.1038/ismej.2017.158 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Li, H., La, S., Zhang, X., Gao, L. & Tian, Y. Salt-induced recruitment of specific root-associated bacterial consortium capable of enhancing plant adaptability to salt stress. ISME J.15, 2865–2882 (2021). 10.1038/s41396-021-00974-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Schlatter, D., Kinkel, L., Thomashow, L., Weller, D. & Paulitz, T. Disease suppressive soils: new insights from the soil microbiome. Phytopathology107, 1284–1297 (2017). 10.1094/PHYTO-03-17-0111-RVW [DOI] [PubMed] [Google Scholar]

[CR16] 16.Rolfe, S. A., Griffiths, J. & Ton, J. Crying out for help with root exudates: adaptive mechanisms by which stressed plants assemble health-promoting soil microbiomes. Curr. Opin. Microbiol.49, 73–82 (2019). 10.1016/j.mib.2019.10.003 [DOI] [PubMed] [Google Scholar]

[CR17] 17.Trivedi, P., Batista, B. D., Bazany, K. E. & Singh, B. K. Plant–microbiome interactions under a changing world: Responses, consequences and perspectives. N. Phytol.234, 1951–1959 (2022). 10.1111/nph.18016 [DOI] [PubMed] [Google Scholar]

[CR18] 18.Liu, H., Macdonald, C. A., Cook, J., Anderson, I. C. & Singh, B. K. An ecological loop: host microbiomes across multitrophic interactions. Trends Ecol. Evol.34, 1118–1130 (2019). 10.1016/j.tree.2019.07.011 [DOI] [PubMed] [Google Scholar]

[CR19] 19.Mendes, R. et al. Deciphering the rhizosphere microbiome for disease-suppressive bacteria. Science332, 1097–1100 (2011). 10.1126/science.1203980 [DOI] [PubMed] [Google Scholar]

[CR20] 20.Berendsen, R. L., Pieterse, C. M. & Bakker, P. A. The rhizosphere microbiome and plant health. Trends Plant Sci.17, 478–486 (2012). 10.1016/j.tplants.2012.04.001 [DOI] [PubMed] [Google Scholar]

[CR21] 21.Domeignoz-Horta, L. A. et al. Microbial diversity drives carbon use efficiency in a model soil. Nat. Commun.11, 3684 (2020). 10.1038/s41467-020-17502-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Yang, G., Roy, J., Veresoglou, S. D. & Rillig, M. C. Soil biodiversity enhances the persistence of legumes under climate change. N. Phytol.229, 2945–2956 (2021). 10.1111/nph.17065 [DOI] [PubMed] [Google Scholar]

[CR23] 23.Rillig, M. C. et al. The role of multiple global change factors in driving soil functions and microbial biodiversity. Science366, 886–890 (2019). 10.1126/science.aay2832 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Zhou, X. et al. Cross-kingdom synthetic microbiota supports tomato suppression of Fusarium wilt disease. Nat. Commun.13, 7890 (2022). 10.1038/s41467-022-35452-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Lamelas, A. et al. The bacterial microbiome of Meloidogyne-based disease complex in coffee and tomato. Front. Plant Sci.11, 510816 (2020). 10.3389/fpls.2020.00136 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Silva, J. C. et al. Organic practices intensify the microbiome assembly and suppress root-knot nematodes. J. Pest Sci.95, 709–721 (2022). 10.1007/s10340-021-01417-9 [DOI] [Google Scholar]

[CR27] 27.Martinson, V. G. et al. Multiple origins of obligate nematode and insect symbionts by a clade of bacteria closely related to plant pathogens. Proc. Natl Acad. Sci. USA117, 31979–31986 (2020). 10.1073/pnas.2000860117 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Williams, A. & de Vries, F. T. Plant root exudation under drought: implications for ecosystem functioning. N. Phytol.225, 1899–1905 (2020). 10.1111/nph.16223 [DOI] [PubMed] [Google Scholar]

[CR29] 29.Xiong, C., Zhu, Y. G., Wang, J. T., Singh, B., Han, L. L., Shen, J. P., Li, P. P., Wang, G. B., Wu, C. F., Ge, A. H., Zhang, L. M. & He, J. Z. Host selection shapes crop microbiome assembly and network complexity. N. Phytol.229, 1091–1104 (2021). 10.1111/nph.16890 [DOI] [PubMed] [Google Scholar]

[CR30] 30.Ravanbakhsh, M., Kowalchuk, G. A. & Jousset, A. Targeted plant hologenome editing for plant trait enhancement. N. Phytol.229, 1067–1077 (2021). 10.1111/nph.16867 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Deng, X., Zhang, N., Shen, Z., Zhu, C., Liu, H., Xu, Z., Li, R., Shen, Q. & Salles, J. F. Soil microbiome manipulation triggers direct and possible indirect suppression against Ralstonia solanacearum and Fusarium oxysporum. npj. Biofilms Microbi33, 1–10 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Xiong, W., Li, R., Ren, Y., Liu, C., Zhao, Q., Wu, H., Jousset, A. & Shen, Q. Distinct roles for soil fungal and bacterial communities associated with the suppression of vanilla Fusarium wilt disease. Soil Biol. Biochem.107, 198–207 (2017). 10.1016/j.soilbio.2017.01.010 [DOI] [Google Scholar]

[CR33] 33.Li, E., de Jonge, R., Liu, C., Jiang, H., Friman, V. P., Pieterse, C. M., Bakker, J., Jousset, P. A. H. M. & Rapid, A. evolution of bacterial mutualism in the plant rhizosphere. Nat. Commun.12, 3829 (2021). 10.1038/s41467-021-24005-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Wagg, C., Schlaeppi, K., Banerjee, S., Kuramae, E. E. & van der Heijden, M. G. A. Fungal-bacterial diversity and microbiome complexity predict ecosystem functioning. Nat. Commun.10, 4841 (2019). 10.1038/s41467-019-12798-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Hassani, M., Durán, P. & Hacquard, S. Microbial interactions within the plant holobiont. Microbiome6, 1–17 (2018). 10.1186/s40168-018-0445-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Coyte, K. Z., Rao, C., Rakoff-Nahoum, S. & Foster, K. R. Ecological rules for the assembly of microbiome communities. PLoS Biol.19, e3001116 (2021). 10.1371/journal.pbio.3001116 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Zandalinas, S. I. & Mittler, R. Plant responses to multifactorial stress combination. N. Phytol.234, 1161–1167 (2022). 10.1111/nph.18087 [DOI] [PubMed] [Google Scholar]

[CR38] 38.Macedo-Raygoza, G. M., Valdez-Salas, B., Prado, F. M., Prieto, K. R., Yamaguchi, L. F., Kato, M. J., Canto-Canché, B. B., Carrillo-Beltrán, M., Mascio, P. D., White, J. F. & Beltrán-García, M. J. Enterobacter cloacae, an endophyte that establishes a nutrient-transfer symbiosis with banana plants and protects against the black Sigatoka pathogen. Front. Microbiol.10, 804 (2019). 10.3389/fmicb.2019.00804 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Liu, H., Li, J., Carvalhais, L. C., Percy, C. D., Prakash Verma, J., Schenk, P. M. & Singh, B. K. Evidence for the plant recruitment of beneficial microbes to suppress soil‐borne pathogens. N. Phytol.229, 2873–2885 (2021). 10.1111/nph.17057 [DOI] [PubMed] [Google Scholar]

[CR40] 40.Miyara, S. B., Ionit, I., Buki, P. & Kolomiets, M. Chapter six-the role of lipid signalling in regulating plant-nematode interactions. Adv. Bot. Res.73, 139–166 (2015). 10.1016/bs.abr.2014.12.004 [DOI] [Google Scholar]

[CR41] 41.Niu, B., Paulson, J. N., Zheng, X. & Kolter, R. Simplified and representative bacterial community of maize roots. Proc. Natl Acad. Sci. USA114, E2450–E2459 (2017). 10.1073/pnas.1616148114 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Zhang, X., Peng, H., Zhu, S., Xing, J., Li, X., Zhu, Z., Zheng, J., Wang, L., Wang, B. & Chen, J. et al. Nematode-encoded RALF peptide mimics facilitate parasitism of plants through the FERONIA receptor kinase. Mol. Plant13, 1434–1454 (2020). 10.1016/j.molp.2020.08.014 [DOI] [PubMed] [Google Scholar]

[CR43] 43.Tian, Y., Wang, Q., Zhang, W. & Gao, L. Reducing environmental risk of excessively fertilized soils and improving cucumber growth by Caragana microphylla-straw compost application in long-term continuous cropping systems. Sci. Total Environ.544, 251–261 (2016). 10.1016/j.scitotenv.2015.11.091 [DOI] [PubMed] [Google Scholar]

[CR44] 44.Wang, B. & Sugiyama, S. Phylogenetic signal of host plants in the bacterial and fungal root microbiomes of cultivated angiosperms. Plant J.104, 522–531 (2020). 10.1111/tpj.14943 [DOI] [PubMed] [Google Scholar]

[CR45] 45.Caporaso, J. G., Lauber, C. L., Walters, W. A., Berg-Lyons, D., Huntley, J., Fierer, N., Owens, S. M., Betley, J., Fraser, L. & Bauer, M. et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J.6, 1621–1624 (2012). 10.1038/ismej.2012.8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Oksanen, J., Blanchet, F., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P., O’Hara, R., Simpson, G., Solymos, P. vegan: Community ecology package. R package v.2.5-6. [WWW document] URL https://CRAN.R-project.org/package=vegan (2019).

[CR47] 47.Javůrková, V. G., Kreisinger, J., Procházka, P., Požgayová, M., Ševčíková, K., Brlík, V., Adamík, P., Heneberg, P. & Porkert, J. Unveiled feather microcosm: feather microbiota of passerine birds is closely associated with host species identity and bacteriocin-producing bacteria. ISME J.13, 2363–2376 (2019). 10.1038/s41396-019-0438-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Segata, N., Izard, J., Waldron, L., Gevers, D., Miropolsky, L., Garrett, W. S. & Huttenhower, C. Metagenomic biomarker discovery and explanation. Genome Biol.12, R60–R60 (2011). 10.1186/gb-2011-12-6-r60 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Laliberté, E., Shipley, B. FD: measuring functional diversity from multiple traits, and other tools for functional ecology. R package version 1.0-11. [WWWdocument] URL https://cran.r-project.org/web/packages/FD/FD.pdf [accessed 30 September 2023].

[CR50] 50.Bai, Y., Müller, D. B., Srinivas, G., Garrido-Oter, R., Potthoff, E. & Rott, M. et al. Functional overlap of the Arabidopsis leaf and root microbiota. Nature528, 364–369 (2015). 10.1038/nature16192 [DOI] [PubMed] [Google Scholar]

[CR51] 51.Wei, L. H., Xue, Q. Y., Wei, B. Q., Wang, Y. M., Li, S. M., Chen, L. F. & Guo, J. H. Screening of antagonistic bacterial strains against Meloidogyne incognita using protease activity. Biocontrol Sci. Tech.20, 739–750 (2010). 10.1080/09583151003714109 [DOI] [Google Scholar]

[CR52] 52.Abbott, W. S. A method of computing the effectiveness of an insecticide. J. Econ. Entomol.18, 265–267 (1925). 10.1093/jee/18.2.265a [DOI] [PubMed] [Google Scholar]

[CR53] 53.Schmitz, L., Yan, Z., Schneijderberg, M., de Roij, M., Pijnenburg, R., Zheng, Q., Franken, C., Dechesne, A., Trindade, L. M., van Velzen, R., Bisseling, T., Geurts, R. & Cheng, X. Synthetic bacterial community derived from a desert rhizosphere confers salt stress resilience to tomato in the presence of a soil microbiome. ISME J.16, 1907–1920 (2022). 10.1038/s41396-022-01238-3 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Fan, P., Chen, D., He, Y., Zhou, Q., Tian, Y. & Gao, L. Alleviating salt stress in tomato seedlings using Arthrobacter and Bacillus megaterium isolated from the rhizosphere of wild plants grown on saline–alkaline lands. Int. J. Phytoremediat.18, 1113–1121 (2016). 10.1080/15226514.2016.1183583 [DOI] [PubMed] [Google Scholar]

[CR55] 55.Woźniak, M., Gałązka, A., Tyśkiewicz, R. & Jaroszuk-Ściseł, J. Endophytic bacteria potentially promote plant growth by synthesizing different metabolites and their phenotypic/physiological profiles in the Biolog GEN III MicroPlate^TM Test. Int. J. Mol. Sci.20, 5283 (2019). 10.3390/ijms20215283 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Lin, Z., Li, Y., Zhang, Z., Liu, X., Hsu, C. C., Du, Y., Sang, T., Zhu, C., Wang, Y., Satheesh, V., Pratibha, P., Zhao, Y., Song, C. P., Tao, W. A., Zhu, J. K. & Wang, P. A RAF-SnRK2 kinase cascade mediates early osmotic stress signaling in higher plants. Nat. Commun.11, 613 (2020). 10.1038/s41467-020-14477-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Zhang, X., Song, M., Gao, L. & Tian, Y. Metabolic variations in root tissues and rhizosphere soils of weak host plants potently lead to distinct host status and chemotaxis regulation of Meloidogyne incognita in intercropping. Mol. Plant Pathol.25, e13396 (2024). 10.1111/mpp.13396 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Kusakabe, A., Wang, C., Xu, Y. M., Molnár, I. & Stock, S. P. Selective toxicity of secondary metabolites from the entomopathogenic bacterium Photorhabdus luminescens sonorensis against selected plant parasitic nematodes of the Tylenchina suborder. Microbiol. Spectrum10, e02577–21 (2022). 10.1128/spectrum.02577-21 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR59] 59.Kusakabe, A., Molnár, I. & Stock, S. P. Photorhabdus-derived secondary metabolites reduce root infection by Meloidogyne incognita in cowpea. Plant Dis.107, 3383–3388 (2023). [DOI] [PubMed] [Google Scholar]

[CR60] 60.Du, C., Cao, S., Shi, X., Nie, X., Zheng, J., Deng, Y., Ruan, L., Peng, D. & Sun, M. Genetic and biochemical characterization of a gene operon for trans-aconitic acid, a novel nematicide from Bacillus thuringiensis. J. Biol. Chem.292, 3517–3530 (2017). 10.1074/jbc.M116.762666 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Chialva, M., Salvioli di Fossalunga, A., Daghino, S., Ghignone, S., Bagnaresi, P., Chiapello, M., Novero, M., Spadaro, D., Perotto, S. & Bonfante, P. Native soils with their microbiotas elicit a state of alert in tomato plants. N. Phytol.220, 1296–1308 (2018). 10.1111/nph.15014 [DOI] [PubMed] [Google Scholar]

[CR62] 62.Cheng, Z., Liu, X., Yan, S., Liu, B., Zhong, Y., Song, W., Chen, J., Wang, Z., Che, G., Liu, L., Ying, A., Lv, H., Han, L., Li, M., Zhao, J., Xu, J., Yang, Z., Zhou, Z. & Zhang, X. Pollen tube emergence is mediated by ovary-expressed ALCATRAZ in cucumber. Nat. Commun.14, 258 (2023). 10.1038/s41467-023-35936-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Ji, T., Ma, S., Liang, M., Wang, X., Gao, L. & Tian, Y. Reference genes identification for qRT-PCR normalization of gene expression analysis in Cucumis sativus under Meloidogyne incognita infection and Pseudomonas treatment. Front. Plant Sci.13, 1061921 (2022). 10.3389/fpls.2022.1061921 [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Zhang, Y., Zhang, S., Zhang, J., Wei, W., Zhu, T., Qu, H., Liu, Y. & Xu, G. Improving rice eating and cooking quality by enhancing endogenous expression of a nitrogen‐dependent floral regulator. Plant Biotechnol. J.21, 2654–2670 (2023). 10.1111/pbi.14160 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Protective role of native root-associated bacterial consortium against root-knot nematode infection in susceptible plants

Shikai La

Jiafan Li

Si Ma

Xingqun Liu

Lihong Gao

Yongqiang Tian

Abstract

Introduction

Results

Plants show low susceptibility to M. incognita in the presence of native microbiomes

Fig. 1. Efficiency of native microbiome in reducing plant susceptibility to Meloidogyne incognita.

M. incognita infection alters root bacterial microbiome assembly at the temporal scale

Fig. 2. Meloidogyne incognita-induced variation in bacterial root-associated microbes (RAMs) at the temporal scale.

Bacterial diversity loss of native microbiome exacerbates plant susceptibility to M. incognita

Fig. 3. Influence of bacterial diversity loss of root-associated microbes (RAMs) on plant susceptibility to Meloidogyne incognita.

A diverse range of native bacterial strains from M. incognita-infected roots have nematode-antagonistic activity

Fig. 4. The ability of native bacterial strains to suppress Meloidogyne incognita.

Multiple SynComs have the capacity to inhibit M. incognita infection in plants

Fig. 5. Construction of simplified synthetic communities (SynComs) capable of controlling Meloidogyne incognita infection.

The protective role of SynCom C against M. incognita infection is associated with the regulation of plant defense responses

Fig. 6. The capacity of synthetic community (SynCom) C to rescue plants from Meloidogyne incognita.

Discussion

Methods

Plants, soils, nematodes, and experimental design

Nematode infection assays

Measurements of soil properties and plant parameters

Sampling, DNA extraction, microbial quantification and amplicon sequencing

Bioinformatics analysis of 16 S rRNA and ITS gene profiling

Bacterial isolation, identification and assessment of anti-M. incognita activity

Construction of simplified SynComs

Characterization of bacterial strains, and synergistic interactions between bacterial strains

Collection of rhizosphere metabolites

RNA extraction, transcriptomic analysis and metabolomic analysis

Statistical analyzes

Reporting summary

Supplementary information

Source data

Acknowledgements

Author contributions

Peer review

Peer review information

Data availability

Competing interests

Footnotes

Contributor Information

Supplementary information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases