Genetic and Nutritional Dynamics of SynCom in Suppressing Apple Fire Blight
Article information
Abstract
Fire blight disease, caused by Erwinia amylovora, occurs in apples and other Rosaceae plants and is known to cause significant economic damage. The pathogen usually infects flowers during the reproductive growth period of plants, colonizes, and penetrates by producing exopolysaccharides in the stigma. A synthetic microbial community (SynCom) is an artificial community of microorganisms designed to enhance host viability. To construct SynCom, we attempted to identify and utilize the microbial characteristics of apple trees that are not infected with the pathogen compared to those that are infected. In our previous study, we composed SynCom with strains expected to reduce the density of fire blight pathogens through microbiome analysis, strain isolation, and continuous replacement culture. We are able to observe the disease control effect of the constructed SynCom. However, no study has been conducted to clearly determine the genetic mechanism underlying this effect of the SynCom. Here, we present that potential secondary metabolite candidates and nutritional competition with the pathogen were confirmed as biochemical mechanisms through whole genome analysis of SynCom strains. Additionally, by co-cultivating SynCom with the pathogen in limited nutrient conditions, such as apple blossom extracts, which are susceptible to the pathogen, we confirmed the potential of SynCom treatment to reduce the pathogen densities. This study demonstrates that genetic selection using metagenomics can effectively identify microorganisms with potential functional capabilities.
Fire blight disease caused by Erwinia amylovora is a destructive disease that significantly reduces the productivity and quality of crops in the Rosaceae family (Bonn and van der Zwet, 2000; Van der Zwet and Keil, 1979). Fire blight pathogen primarily infects through flowers, colonizing the surface of the flower stigmata. Under favorable conditions, E. amylovora forms colonies and increases to concentrations of 106–7 CFU (Bubán and Orosz-Kovács, 2003; Slack et al., 2022). Then, it enters the inside of host plants through flower nectaries. By forming ooze that contains bacterial cells and exopolysaccharides (EPS), E. amylovora participates in intercellular interactions to form multicellular aggregates (Schachterle et al., 2022). EPS performs the function of protecting E. amylovora cells from external environmental stresses and contributes to the passive spread of the pathogen (Ramey et al., 2004). Also, the viscosity of EPS amylovoran physically blocks the movement inside the xylem vessels and causes wilting symptoms in the host plants (Koczan et al., 2009; Sjulin and Beer, 1978). For E. amylovora to exhibit pathogenicity, an increase in cell population must precede. It has been reported that the cell population of E. amylovora on flowers must exceed 104 cells per flower for the pathogen to infect plants (Dagher et al., 2020). During this process, environmental factors such as temperature and nutrient availability are critical in the proliferation of the pathogen in sites like flowers or vectors. The degree of E. amylovora is significantly influenced by nutritional status during transmission by vectors, which plays a crucial role in the spread and transmission of the pathogen (Zeng et al., 2020). Therefore, if E. amylovora is continuously exposed to nutrient limitations at key infection sites like flowers and vectors, this could effectively lead to the control of fire blight disease.
Although the fire blight disease has been studied for many years in search of effective control methods, an effective solution has yet to be established. Among the methods for controlling fire blight disease, biological control has been studied as a replacement for chemical control, particularly due to the emergence of resistance in the pathogen to chemicals and antibiotics (Aktepe and Aysan, 2023; Manulis et al., 2003). The mechanisms of biological control are classified into direct pathogen suppression and indirect mechanisms that do not directly inhibit pathogens. The mode of action of biological agents, such as fungi, bacteria, and phages, includes direct antagonistic effects against pathogens, as well as indirect mechanisms that promote plant growth to enhance disease resistance (Legein et al., 2020). Direct mechanisms include the production of antimicrobial compounds such as lipopeptides, cell wall-degrading enzymes, or volatile compounds, which directly suppress pathogens (Kim et al., 2019; Viaene et al., 2016). Additionally, these mechanisms include parasitism by bacteriophages that invade bacterial hosts and utilize their nutrients (Ramíreza et al., 2020), as well as the action of enzymes that specifically degrade the virulence factors of pathogens (Raymaekers et al., 2020). Indirect mechanisms contribute to plant health not by directly killing pathogens but by enhancing nutrient uptake ability through processes like indole-3-acetic acid production, phosphate solubilization, and nitrogen fixation (Compant et al., 2005; Pathania et al., 2020). Also, they activate signaling pathways involved in plant immunity, such as those regulated by jasmonic acid, salicylic acid, and ethylene, which in turn stimulate the expression of pathogenesis-related proteins and other defense-related genes (Agbor et al., 2021; Majeed et al., 2014).
Synthetic microbial community (SynCom) is a synthetic microbial community selected to mimic the functional interactions of microorganisms found in natural ecosystems (Shayanthan et al., 2022). We proposed using SynCom as a biological method for controlling E. amylovora. Our previous study organized a SynCom based on three mechanisms to suppress fire blight disease (Lee et al., 2025). The first is an antibacterial group (A) that directly inhibits E. amylovora. The second is a network group (N) consisting of taxa that negatively correlates with E. amylovora density in the apple endosphere. The third is a pathway group (P), which includes taxa rich in metabolic pathways that are abundantly expressed in healthy apples compared to diseased ones. We observed that using SynCom weakened pathogenicity in various tissues such as apple fruits, rose flowers, and apple plants (Lee et al., 2025). Except for three strains in the antibacterial and pathway groups that showed direct pathogen inhibition, the mechanisms underlying the disease suppression effects of the other strains have yet to be determined.
To elucidate the mechanisms of the SynCom, we conducted genetic and nutritional analyses of these strains. In this study, we present a molecular and biochemical mechanism of action of SynCom in suppressing fire blight disease through genomic analysis and the analysis of nutritional competitive relationships.
Materials and Methods
Whole genome sequencing of SynCom strains
SynCom consists of nine strains in total, including five keystone taxa (N group) within the microbiota network. Two strains were selected based on differences in metabolic pathways (P group), and two were antibacterial strains (A group) (Lee et al., 2025). This study analyzed the genomes of the five keystone taxa and the two pathway-related members. Table 1 shows the strain information from the network and pathway groups in SynCom. All strains were cultured at 28°C for 2–4 days on R2A agar (0.5 g casamino acid hydrolysate, 0.5 g yeast extract, 0.5 g proteose peptone, 0.5 g dextrose, 0.5 g soluble starch, 0.3 g dipotassium phosphate, 0.05 g magnesium sulfate, 0.3 g sodium pyruvate and 15 g agar per L, pH 7.0–7.4) or broth media supplemented with 0.5% mannitol. Genomic DNA was extracted using the Zymo, Quick-DNA HMW magbead kit (Zymo Research, Irvine, CA, USA). PacBio sequencing was conducted for Labrys okinawensis DSM 18385, Terriglobus aquaticus TAH1, Kitasatospora papulosa AF6313, Pseudomonas lundensis AB23 by CJ Bioscience, Inc (Seoul, Korea) and Macrogen, Inc. (Seoul, Korea). Additionally, National Center for Biotechnology Information (NCBI) downloaded sequences for strains obtained from the Korean Agricultural Culture Collection only for those with confirmed complete genomes. The strains for which sequences were downloaded from NCBI include Labrys miyagiensis NBRC 101365 (GenBank accession no. NZ_BSPC00000000), Novosphingobium mathurense SM117 (GenBank accession no. FVZE00000000), and Novosphingobium endophyticum EGI 60015 (GenBank accession number BMHK00000000). The whole genomes of strains in the antibacterial group (Streptomyces recifensis SN1E1, Paenibacillus polymyxa AF2927) were sequenced (Lee et al., 2024). The genomes of closely related strains within the same genus as the SynCom strains were downloaded from NCBI for accurate identification. The average nucleotide identity (ANI) analysis used pyANI v0.3 to determine taxonomic boundaries for SynCom strains in the same genus. The ANI was calculated using fastANI (Jain et al., 2018).
Comparison of features of SynCom strain genomes
Antibiotic gene prediction in the SynCom strains
The functional genes were annotated and classified using rapid annotation using subsystem technology (https://rast.nmpdr.org/). To identify the biosynthetic gene clusters of secondary metabolites, genomes were submitted to the antiSMASH bacterial version (antiSMASH 6.0). The genomes of SynCom strains were annotated using the rapid prokaryotic genome annotation (Prokka) 1.14.6 tool for prokaryotic gene prediction. The strains in the pathway group were selected based on their contributions to a total of four pathways in MetaCYC. R software identified the EC numbers included in each pathway (version 4.3.3, The R Foundation for Statistical Computing, Vienna, Austria).
Nutrient utilization assay in vitro
Phenotype microarray screening was conducted for each strain on PM1 and PM2 plates by Biolog Inc. (Hayward, CA, USA). For Gram-negative bacteria, strains cultured in R2A supplemented with 0.5% mannitol were subjected to cell down and washed twice with double-distilled water (ddH2O). The pellet was diluted in 16 mL of IF-0a to create a cell suspension with an OD600 0.38. Subsequently, 31.25 mL of fresh IF-0a was mixed with 0.45 mL of Redox Dye Mix A, and the 7.5 mL diluted cell suspension was added to make a final OD600 0.071. The resulting mixture was dispensed into PM1 and PM2A plates, each with 100 μL of cell suspension using a multipipette. The Gram-positive bacterium was cultured on MS (20 g mannitol, 20 g soya flour, 20 g agar per L) media to ensure spore formation. Spores were harvested with sterile water to create a cell suspension, which was then diluted in IF-0a GN/GP to adjust OD600 0.6. Ten milliliters of IF-0a GN/GP, 0.12 mL of Dye mix G, 1 mL of PM additive solution, and 0.88 mL of cell suspension were mixed. The PM additive solution was produced according to Khatri et al. (2013). Afterward, 100 μL each was dispensed into plate wells using a multi-pipette. The culture plates were incubated at 37°C, and the optical density at 590 nm was measured at 24 h intervals up to 72 h.
Apple flower exudate nutrient competition assay
Apple flowers (cv. Fuji) were disinfected with 0.5 g in 70% ethanol for 30 s and then washed twice with ddH2O. The washed flowers were dried and then sonicated in 5 mL of ddH2O for 30 min. After sonication, the solution was filtered through a 0.45 μm filter, and to remove any residual debris, the supernatant was separated by centrifugation at 1,077 ×g for 5 min. The resulting flower exudate was mixed with the SynCom and E. amylovora TS3128 (Kang et al., 2021) to examine the E. amylovora growth. Rifampicin resistance E. amylovora TS3128 was used for this assay. SynCom was prepared by mixing strains adjusted to OD600 0.5 concentration in equal amounts for each group (A, N, and P groups). When more than two SynCom groups were treated, the mixture of each group was mixed in equal amounts for each group. The E. amylovora suspension was adjusted to OD600 0.1, and the flower extract and SynCom mixture, containing the pathogen in a 1:1:1 ratio, were cultured at 28°C in a shaking incubator. Subsequently, the cell density of E. amylovora was determined at 0, 24, 48, and 72 h on 1/5 TSA (6 g tryptic soy broth, 20 g agar per L) with 50 μg/mL rifampicin. Linear regression analysis was performed using the simple linear regression function [Im()], and R2 values were used to evaluate the goodness-of-fit according to standard methodologies.
Results
Genome profile of the strains of SynCom
The network group comprises two strains of Labrys genus, two strains of Novosphingobium genus, and one strain of Terriglobus genus. The L. miyagiensis NBRC 101365 genome had 7,783,961 bp and contained a total of 7,824 coding sequences (CDSs) (Table 1, Supplementary Fig. 1). NBRC 101365 had a GC content of 62.7% and encodes 2 rRNAs and 45 tRNAs. L. okinawensis DSM 18385 genome had 7,021,699 bp and contained 6,881 CDSs (Table 1, Supplementary Fig. 1). The GC content was 62.6%, and DSM 18385 was confirmed to have 6 rRNAs and 54 tRNAs. Both Labrys genus strains had a high number of genes associated with carbohydrate subsystems. Specifically, L. miyagiensis NBRC 101365 contained 187 genes related to central carbohydrate metabolism, while L. okinawensis DSM 18385 had 181, which is comparatively higher than other strains. N. mathurense SM117 had a genome of 4,843,239 bp in size and had a GC content of 63.3% (Table 1, Supplementary Fig. 2). SM117 contained 4,674 CDSs, including 3 rRNA and 59 tRNA. N. mathurense SM117 notably possessed over 217 genes related to membrane transport, a significantly large number. N. endophyticum EGI 60015 had a genome of 4,345,016 bp in size, and the GC content was 64.2% (Table 1, Supplementary Fig. 2). EGI 60015 contained 3 rRNAs, 49 tRNAs, and 4,259 CDSs. N. endophyticum EGI 60015 included 51 genes related to the metabolism of aromatic compounds, representing a considerable number. The genome of T. aquaticus TAH1 was 4,244,699 bp and had a GC content of 62.6% (Table 1, Supplementary Fig. 3). It contained 3 rRNAs, 47 tRNAs, and 3,533 CDSs. T. aquaticus TAH1 was found to have the highest number of genes involved in protein metabolism, particularly those associated with protein biosynthesis.
The pathway group consists of one strain of Kitasatospora genus and one strain of Pseudomonas genus. K. papulosa AF6313 had a genome of 7,704,549 bp, and the GC content was confirmed to be 71.0% (Table 1, Fig. 1). AF6313 contained 6,788 CDSs, 19 rRNAs, and 67 tRNAs. This strain contained the highest number of genes related to fatty acids, lipids, and isoprenoids among the SynCom strains, with a total of 145 genes. P. lundensis AB23 genome was 5,000,167 bp, and a GC content was 58.6% (Table 1, Fig. 2). AB23 contained 4,535 CDS, 18 rRNAs, and 73 tRNAs. P. lundensis AB23 contained 392 genes related to amino acids and derivatives, with the most abundant genes associated with arginine; urea cycle, polyamine, lysine, threonine, methionine, and cysteine, and branched-chain amino acids.
Genome map of Kitasatospora papulosa AF6313 in the pathway group. (A) Circular genome map of K. papulosa AF6313. The outermost circles indicate the genome size and coding sequences on the forward and reverse strands. Inner rings display the distribution of rRNA (red), tRNA (blue), and tmRNA (yellow) genes, followed by plots of GC content and GC skew [(G − C)/(G + C)]. GC content is color-coded (yellow: 71.0–61.2%, green: 76.2–71.0%), and GC skew is shown in orange (−0.3–0 bp/bp) and blue (0.2–0 bp/bp). (B) Average nucleotide identity (ANI) heatmap showing the genomic similarity between K. papulosa AF6313 and related actinomycetes. The heatmap is color-coded according to ANI values, and a hierarchical clustering tree is displayed. (C) Summary table of genome statistics for K. papulosa AF6313.
Circular genome map of Pseudomonas lundensis AB23 in the pathway group. (A) Circular genome map of P. lundensis AB23. The genome size is 5,051,469 bp with a 58.6–44.4% GC content. From the outermost to the innermost ring: scale in megabases (Mb), forward strand genes (blue), reverse strand genes (pink), tRNA (green), rRNA (orange), GC content (black), and GC skew (purple/green, where purple indicates values >0 and green indicates values < 0). (B) Heatmap showing the average nucleotide identity (ANI) values between P. lundensis AB23 and related species. ANI values range from 83.7% to 100%, with the color gradient indicating similarity (blue: low similarity, red: high similarity). The dendrogram on the left clusters the species based on ANI values, demonstrating the close relationship between P. lundensis AB23 and other Pseudomonas lundensis strains (AU1044 and M101), while distinguishing it from other Pseudomonas species.
Potential role of secondary metabolites of SynCom in fire blight disease control
To analyze secondary metabolites produced by the SynCom strains, the antiSMASH tool (version 6.0) was utilized, and cases where similarity exceeded 30% were confirmed (Supplementary Table 1). Each strain in the network group contained one or two types of secondary metabolites-related genes in the genome. Both strains of the Labrys genus contained the same cluster related to nonribosomal peptides. This cluster showed 100% similarity to rhizomide, a compound with high potential for antibacterial activity (Wang et al., 2018). The two strains of the Novosphingobium genus each carried a gene for carotenoid and ectoine biosynthesis. N. aromaticivorans break down aromatic compounds and convert them into carotenoids (Hall et al., 2023). Ectoine is involved in osmoregulation, enabling survival in high-salinity environments (Gan et al., 2013).
In K. papulosa AF6313 of the pathway group, secondary metabolite genes were identified as diverse as those of the strains of the antibacterial group. K. papulosa AF6313 contained antibiotic-related metabolites such as Ni-siderophore, ectoine, T2PKS, PKS-like, terpene, and melanin with antioxidant/antibacterial functions. P. lundensis AB23 had a cluster related to hydrogen cyanide (HCN) and aryl polyene. HCN, known as prussic acid, is a volatile compound commonly produced by soil bacteria, including Pseudomonas spp. (Sehrawat et al., 2022). It was suggested that the strain, particularly those producing HCN, promote plant establishment through increasing phosphate availability (Rijavec and Lapanje, 2016) and exhibits broad-spectrum toxicity against fungi, nematodes, insects, and weeds (Sehrawat et al., 2022). This suggests its potential use as an integrated, sustainable, and eco-friendly approach to help plants combat pathogens.
Overlap in nutrient niche examined through nutrient availability
We investigated the utilization of various nutrients by E. amylovora and the SynCom strains (Fig. 3). E. amylovora showed minimal usage on the PM2A plate, which contains complex, high-molecular-weight carbon sources. In contrast, specific compounds were significantly utilized on the PM1 plate. The compounds with the highest usage by E. amylovora included D-glucose-1-phosphate (OD590 1.90), α-D-glucose (OD590 1.81), sucrose (OD590 1.81), N-acetyl-D-glucosamine (OD590 1.51), D-glucose-6-phosphate (OD590 1.30), L-malic acid (OD590 1.15), D,L-malic acid (OD590 1.07), fumaric acid (OD590 0.94), D-galactose (OD590 0.84), and D-mannitol (OD590 0.83). We then examined whether SynCom strains utilized these compounds similarly.
Carbon source nutrient utilization of synthetic microbial community (SynCom) strains of network and pathway group. SynCom strains were grown in R2A broth media, 0.5% mannitol was added, and subsequently centrifuged at 4,000 ×g for 20 min to pellet the cells. The collected cells were resuspended in 16 mL of IF-0a and diluted to OD600 0.38. Separately, 31.25 mL of fresh IF-0a was mixed with 0.45 mL of Redox Dye Mix A. From the previously prepared cell suspension (OD600 0.38), 7.5 mL was added to this mixture to achieve a final OD600 0.71. The resulting mixture was then aliquoted into the wells of PM1 plates at a volume of 100 μL per well. Incubation was carried out at 37°C, with an optical density of 590 nm recorded every 24 h for a total duration of 72 h.
In the network group, T. aquaticus TAH1 was found to utilize a wide range of carbon sources (Fig. 3). It also demonstrated high utilization rates for key nutrients predominantly used by E. amylovora, including N-acetyl-D-glucosamine (OD590 1.2445), D-galactose (OD590 1.191), D-glucose-6-phosphate (OD590 1.109), D, L-malic acid (OD590 1.0075), α-D-glucose (OD590 1.2085), sucrose (OD590 1.1545), D-glucose-1-phosphate (OD590 1.141), fumaric acid (OD590 0.7875), and L-malic acid (OD590 0.742). In the pathway group, K. papulosa AF6313 utilized N-acetyl-D-glucosamine (OD590 1.633), D-galactose (OD590 1.792), and D-mannitol (OD590 0.609) at levels comparable to those of E. amylovora.
Inclusion of enzymes in metabolic pathways highly correlated with the incidence of fire blight disease in apples
The complete genomes of strains in the pathway group were annotated using Prokka, and it was confirmed whether the enzymes contributing to four selected pathways, which served as the basis for selecting these strains, contained the corresponding EC numbers. Although not all pathways were fully represented according to EC numbers, it was found that the strains could potentially contribute to the breakdown and metabolism of acetyl-CoA precursors. In the metabolic chain of toluene degradationIII (aerobic) (via p-cresol), K. papulosa AF6313 and P. lundensis AB23 contained enzymes EC 2.8.3.6 and EC 2.3.1.174 at the stage immediately before the acetyl-CoA pathway (Supplementary Fig. 4A). Only K. papulosa AF6313 contributed to the catechol degradation II (meta-cleavage pathway) metabolic pathway, and the strain contained the EC 1.13.11.2, EC 1.2.1.85, and EC 5.3.2.6 enzyme for the step of decomposing catechol into (3E)-2-oxohex-3-enedioate (Supplementary Fig. 4B). In the case of the catechol degradation (meta-cleavage pathway) metabolic pathway, both K. papulosa AF6313 and P. lundensis AB23 strains contributed to the metabolic pathway, but the enzymes contributed by the two strains were different (Supplementary Fig. 4C). K. papulosa AF6313 contained enzymes EC 1.13.11.2 and EC 3.7.1.9 that decompose catechol into 2-oxopent-4-enoate at the beginning of the metabolic cycle. P. lundensis AB23 contains enzymes EC 4.1.3.39 and EC 1.2.1.10 in the step immediately before acetyl-CoA and decomposes 4-hydrozy-2-oxopentanoate. In the superpathway of β-D-glucuronosides degradation pathway, K. papulosa AF6313 contained the EC 3.2.1.31 enzyme that decomposes β-D-glucuronosides and it was confirmed that K. papulosa AF6313 and P. lundensis AB23 contained decomposing enzymes converting 2-dehydro-3-deoxy-D-gluconate to pyruvate just before the acetyl-CoA step (Supplementary Fig. 4D).
Suppression of pathogen growth via nutritional competition
We investigated whether the presence of SynCom affects the density of E. amylovora in the nutrients available in apple flowers (Fig. 4). The antibacterial group, which showed direct antibacterial effects, was observed along with the network and pathway groups to assess their disease suppression effects in an ecosystem with limited nutrients. A linear regression analysis indicated that the concentration of E. amylovora decreased over time in groups treated with antibacterial agents: antibacterial (A), pathway (P), antibacterial+pathway (AP), and antibacterial+network+pathway (ANP). For the control, the slope change over time was not statistically significant, with an R2 of 0.0227 (P > 0.05). The R2 values of the linear regression equations over time for each treatment are as follows: A was 0.7639 (P < 0.001), N was 0.52 (P < 0.001), ANP was 0.506 (P < 0.001), P was 0.354 (P < 0.01), AP was 0.2563 (P < 0.05), NP was 0.0141 (P > 0.05), A+N was 0.0326 (P > 0.05). It can be observed that, except for the NP and AN treatments, other groups showed differences in slope (Fig. 4). This indicated that E. amylovora did not efficiently utilize the nutrients due to SynCom, leading to a reduction in its density. This finding suggested that SynCom actively consumes and competes for nutrients with E. amylovora.
Reduction in the density of Erwinia amylovora during co-cultivation with synthetic microbial community (SynCom) strains in flower extracts. SynCom strains were adjusted to an OD600 0.5 and mixed according to their respective groups. These mixtures were combined with OD600 0.1 E. amylovora and flower extracts in proportion. Then, samples were incubated in a shaking incubator at 28°C. At 0, 24, 48, and 72 h, samples were serially diluted to 10−5 to 10−7 and plated on 1/5 tryptic soy agar media containing rifampicin to determine E. amylovora CFU counts. R2 value was evaluated using a linear model function in the R program. A, antibacterial; P, pathway; N, network.
Discussion
SynCom is an artificial microbial community that performs its function by mixing various strains suitable for the purpose of mimicking the microbial community that exists in the ecosystem (Qiao et al., 2024). By combining functional strains together, SynCom contributes to promoting plant growth and enhancing plant protection against various stresses. We selected the strains based on three mechanisms to compose SynCom to control the fire blight disease in apples (Lee et al., 2024). One of the mechanisms involves the antibacterial group, which selectively isolated strains with direct inhibitory effects against E. amylovora (Kim et al., 2022). In contrast, the network group and the pathway group selected strains based on depth microbiome analysis. The network group consists of strains that are inferred to compete with fire blight pathogen within apple plants, leading to a reduction in pathogen density (Lee et al., 2023b). On the other hand, the pathway group includes strains that are presumed to contribute to metabolic pathways associated with the overall health of apple plants (Lee et al., 2023a). In our previous study, these selected strains exhibited the fire blight disease control effects in various tissues including apple fruits, roses flowers, and apple plants. The network group and pathway group, unlike the antibiotic group, did not exhibit direct antibacterial activity in vitro. However, the mechanisms underlying the observed reduction in pathogenicity in plant hosts identified in the previous study have not been elucidated. Therefore, this study aimed to investigate how such effects are mediated by elucidating the mechanisms in the SynCom.
We analyzed secondary metabolites directly related to anti-E. amylovora effects based on their genome information. K. papulosa AF6313, selected based on pathway criteria in healthy apples, was found to have weak antibacterial activity against E. amylovora. This is presumed to be because AF6313 contained several secondary metabolites, a characteristic shared with Streptomyces in the antibacterial group Actinomycetes. Strains in the network group include two species from the Labrys genus, two species from the Novosphingobium genus, and one species from the Terriglobus genus. Within the Labrys genus, L. okinawensis was isolated from the root nodules of Entadaphaseoloides plants in Okinawa, Japan, and L. miyagiensis is a strain isolated from grassland soil in Miyagi Prefecture, Japan (Tapia-García et al., 2020). Given its isolation from root nodules, it is likely that Labrys has a plant growth promoting effect through nitrogen fixation ability and has a competitive niche with pathogens for endogenous habitat. EPS is known to play a crucial role in biofilm formation and colonization of microorganisms within the plant endosphere (Afzal et al., 2019; Santoyo et al., 2016). Labrys is characterized by its ability to produce large amounts of EPS when cultured on plates. Additionally, among the two Labrys species in SynCom, L. miyagiensis has demonstrated a particularly strong biofilm formation ability (Lee et al., 2024).
The Novosphingobium genus in the network group, a Gram-negative bacterium, is characterized by its ability to decompose a wide range of aromatic compounds such as phenol, aniline, and nitrobenzene (Liu et al., 2005). In the analysis results of the pathway group, we conducted a comparative study of the metabolic pathways in healthy versus E. amylovora-infected apples, revealing that taxa in healthy apples exhibit a significantly higher expression for the degradation of aromatic compounds pathways. Interestingly, although Novosphingobium taxa in the network group were not selected based on this specific pathway, the bacteria were found to possess a strong capability for aromatic compound degradation within the metabolic pathways identified in the pathway group. Aromatic polymers such as lignin are components of plant cellulosic structures that have a chemically resilient structure (Seo et al., 2009). These structures make it difficult to degrade and thus consider environmental pollutants in agricultural soils. For this reason, the bacterial degradation availability of aromatic compounds helps protect plants from abiotic stress, enhancing their tolerance (Chaudhary et al., 2023; Rodríguez-Valdecantos et al., 2023). Various strains in the genus Novosphingobium have been identified to degrade aromatic compounds, and some of these strains convert these compounds into carotenoids (Linz et al., 2021; Liu et al., 2005; Metz et al., 2024). In our study, the two selected Novosphingobium species were confirmed to possess gene clusters associated with carotenoid production, as identified by AntiSMASH analysis, and clusters related to the degradation of aromatic compounds, as determined through Clusters of Orthologous Groups analysis (Supplementary Table 1). This finding suggested a coherent direction among strains related to pathogen suppression.
The pathway group, as described in our previous study, was selected based on the expressed metabolic pathways associated with the presence or absence of fire blight disease in apple trees. Among the metabolic pathways that differed depending on the presence of fire blight were toluene degradation III (aerobic) (via p-cresol), catechol degradation II (meta-cleavage pathway), catechol degradation I (meta-cleavage pathway), and the superpathway of β-D-glucuronosides degradation. To determine whether the strains with high homology to two amplicon sequence variants (ASVs), which contributed commonly to these four pathways, truly participated in these pathways, we verified using EC numbers. As a result, there were no cases where all EC numbers of the pathways were contributed. In the case of toluene degradation III (aerobic) (via p-cresol), the strains predominantly decomposed components prior to the tricarboxylic acid cycle. For catechol degradation I (meta-cleavage pathway), it is thought that two strains could collaboratively degrade catechol components. This analysis was conducted based on EC numbers using Prokka, which presents limitations in pathway identification, indicating that more detailed analysis is required.
The pathway group included the genera Kitasatospora and Pseudomonas. The genus Kitasatospora is also one of the genera belonging to Actinobacteria, along with the genus Streptomyces in the antibacterial group (Takahashi, 2017). The strain has a high GC content, produces various secondary metabolites, and is phylogenetically very close to the genus Streptomyces. The genus Pseudomonas is a Gram-negative bacterium belonging to the Pseudomonadaceae family with high metabolic diversity, and its relationship with plants has also been studied extensively (Sah et al., 2021). Except for the antibacterial group and K. papulosa AF6313 in the pathway group, which showed direct antibacterial effects, other strains did not show a direct effect on pathogens in our previous study. However, the effect of suppressing disease inside the plant was observed. We hypothesized that ecologically speaking, it may reduce the colonization of E. amylovora by occupying space and increasing density through the uptake of nutrients available in plants. The Biolog phenotype array was conducted to confirm the hypothesis to ensure that the nutrients were required for both E. amylovora and SynCom. Apple blossoms are the part of apple trees most infected by E. amylovora. Nutrient exudate from apple flowers was used to co-culture E. amylovora and SynCom. We then compared the pathogen density to assess whether the density of the pathogen was reduced or not, suggesting that SynCom influenced controlling pathogen populations under nutrient-limited conditions. The antibacterial group, which contained strains with direct antibacterial properties, exhibited the most significant decrease in pathogen density. However, a reduction in pathogen density was also observed in treatments lacking an antibacterial group.
The stigma exudate of flowers is a crucial medium for sustaining bacteria, yeast, and fungi, sustaining a high microbial population density (Stockwell et al., 1999). Due to the abundance of nutrients and microorganisms, there is always a heightened risk of pathogen infection. To naturally defend against pathogen infections, floral nectars and stigma exudates contain various antimicrobial components such as phenolic compounds, secondary metabolites, and nectarin (González-Teuber et al., 2010; Heil, 2011). This highlights that the exudates, nutrients, and antimicrobial compounds in sites prone to microbial infection can significantly influence pathogen infection and growth. The efficient utilization of carbon sources in flowers is crucial for E. amylovora to overcome host defenses and infect host plants (Raese et al., 1977). E. amylovora prefers specific carbon sources, and the primary nutrients, including sorbitol, sucrose, and galactose, are those identified in this study’s assay. We examined the similarities in nutrient metabolism patterns between strains of groups that do not exhibit direct antibacterial effects and E. amylovora. Among the five strains of the network group, some showed high utilization rates for key nutrients such as sucrose, glucose, and galactose, which are also primarily used by E. amylovora. Studies have shown that E. amylovora selectively metabolizes certain nutrients depending on the infection route (Schachterle et al., 2022). Also, the nutrient availability on the surface of plant tissues is limited compared to PM plates. Based on previous studies showing that strains used to control E. amylovora tend to utilize similar carbon sources, we suggest that competition for these limited nutrients is likely to inhibit E. amylovora infection.
SynCom strains contain secondary metabolites biosynthetic clusters that contribute to plant health or have antibacterial effects. It is assumed that they exhibit direct or indirect pathogen suppression effects through the secretion of these substances. Also, SynCom used carbon sources with some overlap, which is important for density increase during the infection process of E. amylovora. When SynCom was co-cultivated with E. amylovora in extracts of apple flowers, a decrease in E. amylovora density was observed compared to its monoculture. This suggests that they have the potential to suppress pathogens through competition for limited nutrients in the ecosystem. It was concluded that further research is necessary to provide more detailed insights into the nutrients present in apple flowers.
Notes
Conflicts of Interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
This research was supported by an agenda research program by the Rural Development Administration [RS-2020-RD009282] and the National Research Foundation of Korea (NRF) [RS-2025-00516084].
Electronic Supplementary Material
Supplementary materials are available at The Plant Pathology Journal website (http://www.ppjonline.org/).