In vitro, in planta, and comparative genomic analyses of Pseudomonas syringae pv. syringae strains of pepper (Capsicum annuum var. annuum)

ABSTRACT Pseudomonas syringae pv. syringae (Pss) is an emerging phytopathogen that causes Pseudomonas leaf spot (PLS) disease in pepper plants. Pss can cause serious economic damage to pepper production, yet very little is known about the virulence factors carried by Pss that cause disease in pepper seedlings. In this study, Pss strains isolated from pepper plants showing PLS symptoms in Ohio between 2013 and 2021 (n = 16) showed varying degrees of virulence (Pss populations and disease symptoms on leaves) on 6-week-old pepper seedlings. In vitro studies assessing growth in nutrient-limited conditions, biofilm production, and motility also showed varying degrees of virulence, but in vitro and in planta variation in virulence between Pss strains did not correlate. Comparative whole-genome sequencing studies identified notable virulence genes including 30 biofilm genes, 87 motility genes, and 106 secretion system genes. Additionally, a total of 27 antimicrobial resistance genes were found. A multivariate correlation analysis and Scoary analysis based on variation in gene content (n = 812 variable genes) and single nucleotide polymorphisms within virulence genes identified no significant correlations with disease severity, likely due to our limited sample size. In summary, our study explored the virulence and antimicrobial gene content of Pss in pepper seedlings as a first step toward understanding the virulence and pathogenicity of Pss in pepper seedlings. Further studies with additional pepper Pss strains will facilitate defining genes in Pss that correlate with its virulence in pepper seedlings, which can facilitate the development of effective measures to control Pss in pepper and other related P. syringae pathovars. IMPORTANCE Pseudomonas leaf spot (PLS) caused by Pseudomonas syringae pv. syringae (Pss) causes significant losses to the pepper industry. Highly virulent Pss strains under optimal environmental conditions (cool–moderate temperatures, high moisture) can cause severe necrotic lesions on pepper leaves that consequently can decrease pepper yield if the disease persists. Hence, it is important to understand the virulence mechanisms of Pss to be able to effectively control PLS in peppers. In our study, in vitro, in planta, and whole-genome sequence analyses were conducted to better understand the virulence and pathogenicity characteristics of Pss strains in peppers. Our findings fill a knowledge gap regarding potential virulence and pathogenicity characteristics of Pss in peppers, including virulence and antimicrobial gene content. Our study helps pave a path to further identify the role of specific virulence genes in causing disease in peppers, which can have implications in developing strategies to effectively control PLS in peppers.

P seudomonas syringae pv.syringae (Pss) is classified as a pathovar of the P. syringae complex, which is notorious for causing diseases in a wide and often overlapping range of host plants from Solanaceae and Leguminosae plants to citrus and stone fruit trees (1).Pss is an emerging pathogen that causes Pseudomonas leaf spot (PLS) disease in peppers (Capsicum annuum var.annuum).Peppers are an important annual crop used for fresh market consumption and processed products (2).In 2020, 4.7 million pounds of peppers were produced in the US, valued at $579 million (3).However, PLS can be a serious disease in pepper plants (4,5), which can significantly reduce the yield and quality of crops due to the formation of severe dark necrotic lesions on pepper foliage.Due to a lack of knowledge on the virulence mechanisms and pathogenicity of Pss in pepper, Pss is difficult to control once it establishes in the field.
Studies have shown that Pss, in other plant hosts such as cherry trees, mango trees, and bean plants, has a repertoire of virulence mechanisms that allow it to persist on seeds over winter, survive harsh abiotic stressors, and colonize the apoplast to cause disease in plants (6)(7)(8).First, biofilm formation via alginate biosynthesis gene cluster aids Pss in surviving abiotic stressors (9).Second, flagellar and twitching motility genes help Pss move from the outside to the inside of the leaf (10).Finally, secretion system gene clusters (types I-VI), which encode proteins that weaken and invade the host immune defenses, help Pss establish as a pathogen in the host plant (11).Pss uses host defense and stress response gene products to evade, survive, and cause disease in host cells.Currently, copper-based (copper sulfate or copper hydroxide) or streptomycin-based antimicrobials are used as chemical control methods for Pss in peppers (12,13).However, in recent years, the intensive application of copper-based antimicrobials to reduce PLS during the growing season has resulted in the emergence of copper-resistant strains of Pss, making copper-based antimicrobial agents less effective (8,14).
Understanding the pathogenicity, virulence mechanisms, and behavior of Pss in peppers is needed for the prudent use of antimicrobials to mitigate PLS in peppers.Not only has the application of antimicrobials been an insufficient control method, but it has also given rise to antimicrobial resistance genes in plant agriculture and created resistant phytopathogens.This serves as a precautionary tale to understand the pathogen at hand before exploiting intensive control methods.Given the lack of effective management of PLS and the knowledge gap about virulence factors of Pss expressed in pepper seedlings, there is a need to explore the gene content of Pss in peppers so that more effective and sustainable management techniques can be developed.Pss causing PLS in peppers was reported as early as 1962 in the US (15), yet there is no published research on the virulence genes present in Pss specifically affecting pepper plants.
The purpose of this study was to investigate differences in virulence of Pss strains in vitro (growth rate, motility, and biofilm production), in pepper seedlings (Pss popula tion and disease symptoms on leaves), and through whole-genome sequence analysis (functional genomics and non-synonymous single nucleotide variations).The genomic study gave insight into genomic features as well as the virulence and antimicrobial gene profile of the Pss strains that cause disease in pepper seedings.Overall, this study, for the first time, provides insights into in vitro and in planta characteristics of Pss in pepper as well as the virulence and antimicrobial resistance gene content of Pss that causes disease in pepper.

Characterization of Pss strains
A total of 16 Pss strains were isolated from pepper plants harboring characteristic PLS symptoms.These samples came from three different counties in northern Ohio between 2013 and 2021 (Table S1).The 16 strains were fluorescent on Pseudomonas F agar, oxidase-negative, arginine dihydrolase-negative, and levan-positive on NSA medium, induced a hypersensitive response (HR) 24 h post-infection on tobacco (Nicotiana tabacum) leaves cv."Samsun, " and did not cause soft rot on potato slices (Table S1).These features indicate that the strains belong to the LOPAT group Ia, which is characteristic of P. syringae pathovars (16).Further, PCR assays using two primers (hrpZ and syrB genes) identified the 16 isolates as being Pss strains (Table S1).The hrpZ (encoding for a T3SS hypersensitive response and pathogenicity protein) primer is commonly used as a species-specific primer for Pseudomonas syringae (17), and the syrB (encoding for a protein involved in the biosynthesis of syringomycin toxin) primer has recently been shown to be a pathovar-specific primer for Pss (18).

Pss strain
The disease severity (number of lesions on the surface of two lower leaves for each seedling) was recorded at 3, 7, and 14 days post-infection (dpi).The disease incidence (number of seedlings showing PLS symptoms) was recorded at 3, 7, and 14 dpi.The bacterial load in inoculated seedlings was determined at 0, 3, 7, and 14 dpi using the direct dilution plating method.All Pss strains but SM226-1 caused characteristic PLS symptoms on inoculated pepper leaves 3, 7, and 14 dpi.A gradual increase in disease severity was observed over time (average number of lesions of 73.2 ± 54.3 at 3 dpi, 96.1 ± 68.0 at 7 dpi, and 245.5 ± 122.8 at 14 dpi).At 3 dpi, strains SM51-19 and SM1038-14 caused the highest disease severity (P < 0.05) with mean lesion counts of 169.9 ± 85.2 and 186.0 ± 147.9, respectively.Most strains (n = 9/16) caused moderate disease severity, with the number of lesions ranging from 49.8 ± 39.7 to 94.1 ± 90.1 (P < 0.05).Four strains caused low disease severity on pepper seedlings with the number of lesions ranging from 16.1 ± 12.0 to 41.3 ± 34.7 (Fig. 1A, P < 0.05).Only one Pss strain (SM226-1) did not cause disease symptoms on pepper foliage (P < 0.05).

Growth, motility, and biofilm formation of the Pss strains
Growth, biofilm formation, and motility in nutrient-limited conditions were determined (Fig. 2) to understand how these phenotypes correlate with the observed variation in Pss population and disease severity in pepper seedlings.Different growth patterns were observed between the Pss strains in M9 minimal broth over 48 h at 28°C (Fig. 2A; P < 0.05).SM181-4 and SM914-13 had the significantly highest growth rate whereas SM190-8 and SM04-2028-04 had the lowest growth rates (Fig. 2A, P < 0.05).The strains with the highest growth rates in vitro did not have the highest disease severity in planta (Fig. 1).
All Pss strains produced biofilm that was measured after 72 h at 28°C in nutrient-limit ing M9 minimal broth.However, the strains varied in the quantity of biofilm produced (Fig. 2B).Strain SM226-1 produced significantly (P < 0.05) more biofilm [average optical density (OD) ]570 1.4 ± 0.2] than all other strains.Strains SM1031-4, SM1030-14, and  SM156-18 produced similar amounts of biofilm (average OD 570 1.0 ± 0, 0.8 ± 0.03, and 0.8 ± 0.06, P < 0.05).The remaining strains produced relatively low quantities of biofilm.Similar to the growth rate trends, the strains that produced the highest quantities of biofilm in vitro did not have the highest disease severity in planta.In fact, the highest biofilm-producing strain (SM226-1) caused no disease symptoms in planta at 3 dpi although an average of 6.3 ± 0.4 log CFU/g of the bacterial population was detected (Fig. 1).
Motility (swarming) of the Pss strains was assessed in a semi-solid LB agar (0.3%) medium for 24 h at 28°C.All Pss strains were motile (Fig. 2C); however, the motility of Pss strains did not follow the patterns observed during the growth and biofilm assays.Strains SM109-18, SM156-18, SM51-19, SM155-18, and SM914-13 were significantly more motile [average diameter of the halo (mm) of 7.9 ± 0.02, 8.2 ± 0.01, 8.3 ± 0.06, 8.4 ± 0.04, and 8.7 ± 0.2; P < 0.05] compared to the other remaining 11 Pss strains, which had relatively low motility [average diameter of halo (mm) between 4.7 ± 0.01 and 6.1 ± 0.01; P < 0.05].Except for SM51-19, the strains that were highly motile did not induce the most severe disease in planta.In fact, the moderately motile strains SM226-1 and SM207-3 caused, respectively, no symptoms and high disease severity, indicating that motility measured in vitro could not explain the variation in pathogenicity of Pss strains seen in planta.
Additionally, a linear regression analysis between in vitro data (growth, biofilm, and motility) and in planta data (disease severity) only showed a significant correlation between biofilm and disease severity at 3 dpi (r 2 = 0.37, P < 0.05) and 7 dpi (r 2 = 0.38, P < 0.05).The results of the regression analysis indicated that there was no significant relationship (P > 0.05) between the growth or motility of Pss strains and disease severity on 3 and 7 dpi.The multivariate correlation analysis conducted with a Benjamini-Hochberg (BH) correction showed that none of the genes were significantly  a The mean completeness based on the presence of universal genes that are present in the sequenced genomes that are expected to be in the genomes was 99.90% ± 0.152%.
correlated.Scoary detected a total of 31 genes (n = 1, 3, and 28 for growth, biofilm, and motility, respectively) that were associated with in vitro characteristics.However, after the Benjamini-Hochberg correction, none of the genes were significantly associated (Table S2).

Genome characteristics and phylogenetic relationships of the 16 Pss strains
The average genome assembly size of the 16 Pss genomes was 6.04 ± 0.10 Mbp with an average GC content of 59.12% ± 0.27%.The average N50 contig length was 229.6 ± 104.9 kbp, and the average number of contigs was 147 ± 72 (Table 1).Human contami nant contigs were identified by running Kraken2 v.2.1.2(19) with the Standard Plus Fungi database downloaded from https://benlangmead.github.io/aws-indexes/ on 8 April 2022 and option "-confidence 0.5, " and then, the contaminant contigs were removed from our assemblies using the extract_kraken_reads.py script from KrakenTools v.1.2(20).The average depth of coverage was 57% ± 8.26%.A phylogenetic analysis using kSNP3 was conducted to determine the relatedness of the 16 Pss strains to other P. syringae strains (n = 18) downloaded from NCBI both within and outside the P. syringae pv.syringae phylogenetic group (Fig. 3A; Tables S1 and S3).A total of 936,714 single nucleotide polymorphisms (SNPs) were identified across the 34 genomes, 1,845 of which were core (shared) SNPs.Based on the core SNPs, a phylogenetic tree with all 34 genomes was generated (Fig. 3A).Pss strains (n = 15/16) sequenced in this study formed a clade including a Pseudomonas syringae pv.pisi strain isolated from beans and a Pseudomonas syringae pv.syringae strain from millet with a bootstrap value of 0.927.The SM914-13 strain was the only strain outside of this clade grouped with the other Pss and non-Pss strains from NCBI.

Gene content and pangenome analysis
An analysis comparing the gene content among the 16 Pss strains using Roary v.3.11.2 (21) following annotation with Prokka v.1.14.6 ( 22) identified a total of 8,429 genes, of which 3,831 (45.5%) were core genes (defined as genes present in at least 99% of genomes; in our case, all genomes), 0 were soft core genes (present in 95%-99% of genomes), 2,182 (25.9%) were shell genes (present in 15%-95% of genomes), and 2,416 (28.7%) were cloud genes (present in <15% of genomes) (Fig. 3B).The rarefaction curves for the pangenome showed that the plot came close to plateauing (Fig. 3C).Heaps' law used to determine the openness of the pangenome showed that pangenome was closed with a marginal alpha value of 1.000004.Automatic genome annotation using the Rapid Annotations using Subsystems Technology toolkit (RAST) and SEED (23) predicted an average of 5,223 ± 132 coding sequences among the 16 Pss strains.Within these coding sequences, there was an average of 361.3 ± 4.2 subsystems (subsystems in RAST are used to organize related functional roles of genes into categories that have specific biological functions) (23) among the 16 Pss strains containing essential bacterial genes and genes associated with virulence.The virulence-associated genes were related to motility and chemotaxis, secretion systems, invasion and intracellular resistance, iron acquisition, biofilm, stress response, detoxification, bacteriocins, and antibiotic resistance.

Multidrug resistance genes including copper resistance genes
After the annotation of antibiotic resistance genes using the Pseudomonas genome database, ResFinder, CARD, NCBI, and ARG-ANNOT, a total of 27 antibiotic resistance genes were identified within the core and variable genes (Fig. S3).Most of these (n = 12/27) were associated with copper resistance, including genes associated with copper resistance proteins (copBCD and copG) and copper homeostasis and regulator proteins (cutE and cusSR).Other genes associated with copper resistance included copper ATPase, copper chaperone, multicopper oxidase, and cytochrome c heme (ccmFH) genes.There were several other heavy metal resistance proteins including the magnesium and cobalt resistance gene (corC), mercury resistance gene (merR), two genes associated with cadmium resistance [D-cysteine desulfydrase and Cd (II)/Pb (II) regulator], and a heavy metal resistance gene (hmrR).Several other antibiotic resistance genes (n = 7/27) were not associated with heavy metals, including streptothricin acetyltransfer ase (SAT) for streptothricin resistance (closely related to streptomycin), beta-lactamase C and metal-dependent hydrolase of beta-lactamase genes associated with beta-lac tam resistance, penicillin-binding protein for resistance to penicillin, DNA gyrase AB a Two types of plasmids were found between nine Pss strains.Only two genes found in plasmid KY362372 were associated with the T4SS, which is a known virulence factor.associated with fluoroquinolone resistance, and nfxB gene associated with quinolone resistance.A multidrug resistance efflux gene (cflA) was also identified, which has been associated with resistance to various drugs like chloramphenicol.

Plasmids and prophages
Two types of plasmids (KY362372 and CP006257) containing a total of nine genes were found among nine of the Pss strains (Table 2).These genes included two T4SS genes (virB3 and virD4) located in the KY362372 plasmid.The plasmid KY362372 is one of the eight native Pseudomonas syringae plasmids that are part of the pPT23A family.KY362372 is known to carry genes important for virulence and epiphytic colonization of plant leaf surfaces (24).Both KY362372 and CP006257 plasmids have been identified in other Pss strains isolated from mangoes, pears, and millet (24)(25)(26).No prophages were identified in the Pss genomes by PHAge Search Tool Enhanced Release (PHASTER).

Variable genome of Pss strains
A total of 812 genes were identified in the variable genome among the 16 Pss strains (Fig. 4; Table S4).The variable gene profile of Pss strains causing high to moderate disease severity is distinct from the three strains (SM226-1, SM156-18, and SM914-13) causing low disease severity in pepper seedlings.Some of the variable genes present in Pss strains that cause high to moderate disease severity (lesion counts between 213.8 ± 151.3 and 56.1 ± 36.4) appear to be missing from the three strains (SM226-1, SM156-18, and SM914-13) causing low disease severity (lesion counts of 0, 22.6 ± 17.5, and 26.2 ± 16.8), indicating that variation in the gene content might explain the variation in virulence seen between the strains in planta.A principal component analysis (PCA) based on the variable genomes grouped the Pss strains into five clusters (n = 6, 4, 3, 2, and 1 strain in each cluster; Fig. 5).The four strains with high disease severity SM1042-14R, SM51-19, SM207-3, and SM1038-14 with lesion counts of 136.Motility, biofilm, secretion system, and antimicrobial resistance genes were identified in the pangenome and variable genome Virulence genes associated with motility, biofilm, and secretion system (n = 211; Fig. 6) were identified using ABRicate [against the Virulence Factor Database (VFDB)] and by comparing the functional annotations from RAST against the Pseudomonas genome database, Hop database, BastionHub database, and T3SEdb.The Pss strains contained a total of 87 motility genes that contribute to both flagellar and twitching motility (Fig. S4).Majority of the motility genes were associated with flagellar motility (n = 40/87).
Only six out of 27 antimicrobial resistance genes were part of the variable genome.Most of the variability was found with the copper resistance genes (n = 5/6) including copper resistance proteins (copBCD, copG), copper tolerance, and copper chaperone genes (Fig. 6; Fig. S3).Fewer than half of the Pss strains (n = 6/16) carried the variable copper resistance genes.The only other variable gene was the beta-lactamase C and penicillin-binding protein, which was present in all but one Pss strain; SM914-13 was missing the resistance gene to beta-lactamase and penicillin.

Gene and SNP association analyses did not produce significantly correlated genes
A multivariate correlation analysis and an association analysis with Scoary were performed to associate the presence of variable genes (in silico), non-synonymous SNPs within the variable genes, and pseudogenes (focusing on functional gene copies) with in planta phenotypes for all 16 Pss strains.The multiple correlation analysis and the Scoary run were conducted with Benjamini-Hochberg multiple-testing correction, which resulted in a lack of significant adjusted P-values (P > 0.05) for any genes obtained through the multivariate correlation analysis or Scoary (Tables S4 and S5; Fig. S7).A total of 142 variable genes had correlations with disease severity with uncorrected P-values below 0.05, many of which were positively correlated with secretion systems and biofilm genes (Table S4; Fig. S7).A total of 2,549 SNPs were identified in the virulence genes (motility, biofilm, and secretion system) of the 16 Pss genome assemblies using SM914-13 as the reference genome.After the multivariable correlation analysis, 30 SNPs primarily in the secretion system genes were correlated with disease severity (Table S6; Fig. S8).However, due to the small number of genomes analyzed (n = 16), significant correlations could not be detected in this study.

DISCUSSION
Pss is known to infect numerous plant hosts (from stone fruit trees and legumes to Solanaceae crops) and causes economically important diseases in these crops (1).Peppers are one of the crops that can be infected by Pss and show damaging foliar symptoms if left uncontrolled.However, not much is known about the pathogenicity and virulence factors that Pss carries to cause disease in pepper seedlings.In this study, we investigated (i) the pathogenicity and virulence characteristics of Pss infecting pepper seedlings in planta, (ii) some of the major virulence factors of Pss strains in vitro, and (iii) the genetic content focusing on the core and variable virulence and antimicrobial genes carried by the Pss strains to better characterize and understand the virulence profile of Pss strains that infect pepper seedlings.We observed that the Pss strains differed in virulence (Pss bacterial population and disease symptoms) on pepper seedlings.With optimal disease conditions (28°C, 20%-80% relative humidity), disease symptoms appeared within 3 dpi with most Pss strains (n = 15/16) excluding SM226-1.The disease symptoms by most Pss strains (n = 15/16) increased rapidly (average number of lesions of 73.2 ± 54.3 at 3 dpi, 96.1 ± 68.0 at 7 dpi, and 245.5 ± 122.8 at 14 dpi), and after 14 dpi, diseased pepper seedlings were starting to lose leaves.This rapid and substantial disease progression of Pss strains on pepper seedlings further warrants more research to better understand the virulence of Pss in peppers so that it can be controlled.In vitro growth, biofilm, and motility assays were also conducted to understand the virulence of the Pss strains in vitro, and we found that strains differed in in vitro virulence.The in vitro variation in these phenotypes, however, did not correlate with virulence in planta between Pss strains.
Subsequently, whole-genome sequencing was utilized to study the genetic content of the Pss strains to further understand if genetic variation could influence variation in virulence among the 16 Pss strains.Phylogenetic studies using core SNPs indicated that all strains with the exception of SM914-13 were closely related.Nevertheless, we found a total of 812 genes in the variable genome (Fig. 5).The strains were clustered based on their variable gene profile to assess if strains with similar genetic profiles displayed similar virulence and pathogenicity traits.Some strains (SM207-3 and SM1038-14) with high virulence in planta also had a similar variable gene profile.Similarly, on the other end of the spectrum, two strains (SM226-1 and SM156-18) with the lowest virulence in planta also showed similar variable gene profiles.A multivariate correlation analysis and association analysis with Scoary were conducted to associate variable genes with in planta data to identify virulence genes that could help explain the observed variation in virulence of 16 Pss strains.However, after multiple-testing correction, no genes were significantly correlated, likely due to the small sample size of only 16 strains.In the future, as more Pss strains in pepper plants are isolated and sequenced, multivariate correlation analysis and Scoary need to be conducted with in planta data to identify virulence genes that correlate with Pss's disease severity in pepper seedlings.
The genomic content analysis revealed the presence of several types of virulence genes in the Pss strains.Secretion system genes are known to play an important role in virulence and pathogenicity in phytopathogens like Pss (27).The Pss strains in this study contained a repertoire of secretion system genes (n = 94) that are reported to be associated with virulence in other P. syringae pathovars and some Pss strains in other host plants (27)(28)(29).Type III secretion systems are most commonly reported to be associated with virulence as T3SS is essential for virulence in phytopathogens like Pss.We found a total of 43 T3SS genes, of which 17 were part of the variable genome.The T3SS Hop effector genes were the most prevalent among the Pss strains variable genome.Many of these Hop effector genes are common among other Pseudomonas syringae pathovars including pathovar tabaci, tomato (DC3000), and phaseolicola, which are from different genomospecies that infect a wide range of host plants (30,31).The Hop effector genes have a significant influence on the virulence of Pss strains as they modulate and weaken the host immune defense system pathways (30,32).Type VI secretion system (T6SS) genes were another set of secretion system genes that were present in the Pss genomes.We found a total of 18 T6SS genes, of which 4 were part of the variable genome.These genes were associated with the T6SS lipoprotein gene (tssJ), two PAAR repeat protein genes (rhaS), and a sigma-dependent regulator gene (vasH).Most Pss strains (n = 14/16) carried the tssJ, rhaS, and vasH genes.T6SS is part of a newly discovered secretion system that has been linked to virulence in phytopathogens (33).T6SS delivers toxic proteins into plant host cells and contributes to anti-inflammatory processes, virulence, and interbacterial competition (33).T6SS uses a tubular structure to deliver toxic proteins into the target cells.The tssJ (VasD) gene encodes for a core part of the tubular structure (34) whereas the vasH gene is a σ54-dependent transcriptional regulator of hcp genes that encode for a toxic protein called hemolytic co-regulating protein (Hcp) (33).The T6SS can sense environmental (host plant) or bacterial competition signals to activate the T6SS and release (via tubular structure-TssJ/VasD) toxic proteins like Hcp (regulated by the vasH gene) into the target cell.In other P. syringae pathovars, the deletion of any of these genes (tssJ/vasD or vasH) resulted in failure to assemble the T6SS or failure to release the Hcp protein, which subsequently resulted in reduction in virulence in planta (33)(34)(35).Similarly, these T6SS genes may be required for Pss to cause disease in pepper seedlings.Interestingly, the strains that caused low or no disease severity, SM156-18 and SM226-1, respectively, were missing these T6SS genes, while two of the strains that caused high disease severity, SM1038-14 and SM207-3, carried these T6SS genes.
Motility is an important virulence mechanism in P. syringae.In this study, we found a total of 87 motility genes, but only two motility genes (pilA and flp) associated with flagellar and twitching motility were part of the variable genome.The pilA gene has been reported to be important for surface motility in P. syringae pv.tabaci and subsequently for virulence as pilA mutants were unable to induce HR in host tobacco plants (36).Biofilm is another important virulence mechanism in phytopathogens, which helps them survive on the leaf surface and adhere to and colonize the apoplast and provides protection against antimicrobials (37).We found a total of 30 biofilm genes, but only two biofilm genes were part of the variable genome.These included genes that are part of the alginate biosynthesis gene cluster.Mutations in these genes within the alginate biosynthesis gene cluster in Pss infecting bean plants have been shown to reduce virulence and epiphytic fitness (38).Likewise, alginate biosynthesis genes like algJ could also play a role in the virulence and epiphytic fitness of Pss strains infecting pepper plants.
The whole-genome sequence analysis elucidated important antimicrobial resistance genes present in the 16 Pss strains.Almost half (n = 12/27) of the antimicrobial resistance genes found among the Pss strains were associated with copper resistance.All Pss strains carried genes associated with copper resistance, which is concerning because the use of copper-based antimicrobials is the most common chemical control method to manage phytopathogens like Pss in agriculture (14,39).Our results provide further evidence for the lack of effectiveness of copper-based antimicrobials against phytopath ogens and suggest a need for more strategic and comprehensive control methods to control phytopathogens like Pss in agriculture.The antimicrobial resistance profile of the Pss strains also indicated the presence of SAT for streptothricin resistance (closely related to streptomycin), beta-lactamase C genes associated with beta-lactam resist ance, penicillin-binding protein for resistance to penicillin, DNA gyrase associated with fluoroquinolone resistance, and nfxB gene associated with quinolone resistance, and a multidrug resistance efflux gene (cflA) was also identified, which has been associated with resistance to various drugs including chloramphenicol.The presence of these antimicrobial resistance genes is additionally worrisome as these genes could spread to the environment (soil microbiome), animals, and consumers, creating a One Health issue.
Overall, this study provides useful insights into the pathogenicity and virulence characteristics of Pss strains in vitro, in planta in peppers, and their gene content.
Exploring the gene content of Pss can be beneficial in understanding the potential virulence genes required to cause disease in pepper and in the future can help find drug targets to effectively manage Pss in peppers.In the future, more Pss strains in pepper need to be included to conduct in silico-based comparative multivariate correlation and association analyses to narrow down the list of virulence genes that might be associated with Pss's ability to cause disease in peppers.Further, based on such a list, mutagenesis studies or gene expression studies in planta should be conducted to better facilitate the understanding of Pss genes essential for its virulence in pepper seedlings.Our findings provide an important first step toward understanding the virulence and antimicrobial gene content of Pss in peppers and represent the first study to explore the whole-genome sequences of Pss in peppers.

Plant material
Bell pepper (Capsicum annuum) cultivar "California Wonder" seedlings were used for the in planta pathogenicity studies.California Wonder cultivars were used as they are bred to confer resistance to only tobacco mosaic virus and do not have resistance against Pseudomonas syringae strains.The seeds were sanitized using hot water (51°C) and chlorine (5.25%) treatment as described previously (40).Sanitized seeds were sown individually in 96-cell plug trays containing Baccto Professional Grower Mix (Baccto, Houston, TX) and grown in a greenhouse (with optimal growing conditions for pepper at 24°C-28°C temperature, 20%-80% relative humidity, and 12 h photoperiod) (41).The pepper seedlings were grown until the four-true leaf stage (ca., 6-week-old) prior to testing with 16 Pss strains.Plant experiments were conducted in a growth chamber with controlled temperature (24°C), relative humidity (80%), and photoperiod (12 h).Seedlings were watered daily in the pot.

Collection, isolation, and identification of Pss isolates
Sixteen Pss strains isolated from infected pepper plants in northern Ohio (Wayne, Sandusky, and Seneca counties) from 2013 and 2021 were sourced from the OSU Miller lab collection (Table S1).Isolates were recovered from pepper seedlings showing round dark brown necrotic lesions around 2-3 mm with a yellow halo around them. Isolation and identification of the isolates were performed as previously described (15,42).Standardized microbiological and biomolecular tests such as PCR and LOPAT (levan production, oxidase activity, pectolytic capability, arginine dihydrolase, and tobacco hypersensitivity) were conducted on the isolates suspected to be Pseudomonas to confirm that the strains belonged to LOPAT group Ia (details provided in Supplemental material), indicating that the strains are P. syringae strains (43).Further, species-specific PCR targeting hrpZ and syrB virulence genes (Table S7) was conducted to confirm the identification of Pss strains (18,44,45).

Pss virulence on peppers
Six-week-old pepper seedlings (four-true leaf stage) were inoculated by spraying the foliage (5 cm away from the plant at a 45° angle) with a normalized Pss suspension (1 mL per seedling, 1.0 OD 600 = approximately 1 × 10 9 CFU per plant) with a commer cial hand sprayer (Equate 8 oz plastic spray bottle, Texas, USA).The seedlings were rotated between every two sprays to ensure the entire seedling is sprayed.To ensure each seedling received equivalent Pss inoculum, an initial assessment was conducted where spraying 10 sprays from the 8 oz spray bottle resulted in the release of 1 g of liquid (equivalent to 1 mL) as measured using a weighing balance.This assessment was repeated multiple times (more than 10 times), and consistently, 1 g (1 mL) was obtained each time with 10 sprays.Therefore, we used this approach to spray 1 mL (10 sprays using 8 oz spray bottle) on each seedling when inoculating the seedlings.The Pss strain inoculum was prepared as described above in M9 minimal broth at 28°C at 180 rpm overnight.Inoculated seedlings were incubated in a growth chamber (28°C, 20%-80% relative humidity, 12 h photoperiod).The seedlings (n = 14 per treatment group) were watered daily in the pots.Starter fertilizer [2 lbs/50 gal of 12-48-8 (%) nitrogen-phosphorus-potassium (N-P-K)] was applied to the seedling in the greenhouse.The disease severity (number of lesions on the surface of two lower leaves for each seedling) was recorded at 3, 7, and 14 dpi.The disease incidence (number of seedlings showing PLS symptoms) was recorded at 3, 7, and 14 dpi.The bacterial load in inoculated seedlings was determined at 0, 3, 7, and 14 dpi using the direct dilution plating method (46).Briefly, seedlings (1 cm above the soil) were collected individually in Whirl-Pak bags.The weight of each seedling was recorded prior to adding 2 mL of 1× phosphate buffered saline (PBS).The seedlings were macerated, and the macerate was serially diluted and plated on an NBY agar medium supplemented with copper sulfate (25 µg/ mL).NBY plates were supplemented with a non-lethal dose of copper sulfate to create a semi-selective growing condition to reduce the background growth coming from the phytobiome.Colonies characteristic of Pss were counted after 48-h incubation at 28°C.The experiment was performed twice with 14 biological replicates per strain in each experiment, and plants treated with PBS were used as a negative control.

Bacterial growth assay
The growth of Pss in M9 minimal broth was assessed in nutrient-limiting conditions mimicking the limited nutrient conditions present in planta.Each Pss strain was grown for 12 h in M9 minimal broth at 28°C at 180 rpm until reaching the exponential phase.The OD 600 was adjusted to 0.05 (approximately 1 × 10 7 CFU/mL) in fresh M9 minimal broth.In a sterile, non-treated, flat-bottom 96-well plate (Corning Inc, Corning, NY, USA), 100 µL of the adjusted bacterial culture was transferred per well (n = 3 reps per strain).The plate was incubated at 28°C for 48 h in a Sunrise Tecan kinetic microplate reader (San Jose, CA, USA), and kinetic measurements were conducted at OD 600 every 15 mins.The experiment was conducted twice with M9 minimal broth as a negative control.The turbidimetric data were used to determine the growth rate per hour and doubling time, which was calculated using the growthcurver 0.3.1 package in R 4.3.1.

Biofilm assay
The production of biofilm by the Pss strains was assessed using the crystal violet method (47).Each Pss strain was grown for 12 h in M9 minimal broth at 28°C at 180 rpm, when they reached the exponential growth phase.The OD 600 was adjusted to 0.05 (approximately 1 × 10 7 CFU/mL) in fresh M9 minimal broth.In a 96-well plate, 100 µL of the adjusted bacterial culture was transferred per well (n = 3 reps per strain).The plate was incubated for 72 h at 28°C without shaking; then, the supernatant was carefully removed, and the wells were gently washed by adding 175 µL sterile water to remove any planktonic bacteria.The water was discarded, and the biofilm was stained by adding 175 µL of 0.1% crystal violet solution to each well for 15 mins (47).The plate was washed three times using 200 µL of sterile water as described above to remove excess crystal violet.The plate was dried at room temperature before the stained biofilm was solubilized in 200 µL of 95% ethanol.An absorbance microplate reader was used to measure the OD 570 to quantify the level of biofilm in each well.The experiment was conducted twice.Wells treated with M9 minimal broth were used as a negative control.

Motility assay
A motility assay was conducted on the Pss strains on semi-solid agar (47).Each Pss strain was grown for 12 h in LB broth at 28°C at 180 rpm.The OD 600 was adjusted to 0.05 (approximately 1 × 10 7 CFU/mL) in fresh LB broth.One milliliter of 0.3% semi-solid LB agar supplemented with 0.01% tetrazolium chloride was poured into each well of a 24-well plate.A 0.3% agar concentration was used as swarming motility can be observed at this concentration (48).A 31-gauge, 8-mm sterile syringe needle was used to transfer approx. 1 µL of the adjusted bacterial culture into the semi-solid agar at the center of the well using a stabbing motion (n = 2 replicates per strain).The plate was sealed using parafilm to avoid evaporation of the water in the agar.The plate was incubated at 28°C without shaking for 24 h.Wells that received LB were negative controls.The motility of each of the strains was quantified by measuring the diameter of the halo zone after 24 h of incubation.The experiment was conducted twice.

Genomic DNA extraction
Genomic DNA was extracted using a Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA), following the manufacturer's instructions.Pss strains were grown in fresh NBY agar plates overnight at 28°C.A loopful of bacterial colonies were collec ted from the NBY plates and resuspended in 1-mL sterile water.The bacterial suspen sions were vortexed thoroughly and then centrifuged for 7 mins at 4,500 rpm using a benchtop centrifuge.The supernatant was discarded, and Nuclei Lysis solution was added to lyse the bacterial cells to release genomic content.RNase solution and Protein Precipitation solution were added to extract and remove any RNA and proteins from the bacterial cells.The remaining DNA was washed using 70% ethanol and precipitated with absolute alcohol to remove any remaining protein and salt traces (49).The DNA was resuspended in 100 µL nuclease-free water.DNA quality and quantity were validated using the NanoDrop Microvolume Spectrophotometer (Thermo Scientific, Waltham, MA, USA).DNA samples with a 260/280 ratio of approx.1.8 and a 260/230 ratio of 2.0-2.2 were considered standards for pure DNA samples.

Whole-genome sequencing of Pss pepper isolates
Genomic DNA was extracted as described, and only genomic DNA with high quality and quantity (concentration above 50 ng/µL) was used for sequencing (50).A genomic DNA Clean and Concentrator kit (Zymo Research, Irvine, CA, USA) was used to ensure the quality of extracted DNA.Whole-genome sequencing was performed on the Illumina Miseq platform using V2 chemistry with 2 × 250 paired-end chemistry (51).The Nextera XT DNA Sample Prep Kit (Illumina Inc., San Diego, CA, USA) was used to prepare a genomic DNA library (0.3 ng/µL) for sequencing, following the manufacturer's instruc tions.The libraries were normalized using a bead-based procedure and pooled together at equal volumes.The pooled library was denatured and sequenced using Miseq reagent version 2, as described previously (Illumina Inc.) (51,52).

Phylogenomic analysis
To determine the phylogenetic relationship between the 16 Pss strains from this study and available reference genomes of P. syringae pathovars, kSNP3 (v.3.1) was used.kSNP3 is a reference-free and alignment-free method used to identify sites with SNPs that are shared among the assembled genomes (62).kSNP3 uses maximum likelihood approaches to build a phylogenetic tree (63).Twelve different pathovars of P. syringae (7 Pss and 11 non-Pss strains closely related to Pss) were downloaded from NCBI's GenBank to further validate that the strains collected in this study are part of the pathovar syringae (Fig. 3; Table S3).The phylogenetic tree was rooted using Pseudomonas moraviensis as the outgroup taxa.kSNP3 first identifies core SNPs, i.e., those present in all input genomes (n = 34, 16 Pss strains from this study, 7 Pss strains from NCBI, and 11 P. syringae strains from different pathovars).For visualization of the phylogenetic tree, ggtree was used in R (64).

Gene and SNP association analysis
Annotated Pss genomes were used to separate the core and variable genomes.The core and variable genomes were defined based on the presence of the gene in all strains or absence of the gene in at least one strain, respectively.Hierarchical clustering analyses and principal component analyses were used to identify differences in the variable genome (gene content) between the strains.Further, principal component and multivariate correlation analyses were used to compare the variable genomes in vitro (growth, biofilm, and motility) and in planta (disease severity at 3 and 7 dpi) to identify genes associated with specific phenotypic characteristics.A BH false discovery rate (FDR) correction for multiple testing was applied with an FDR-adjusted P < 0.05 for conducted to test for equal variance between the two trials conducted for each of the in planta experiments.Five tests were conducted to test for equal variance between trials, including the O'Brien test, Brown-Forsythe test, Levene test, Bartlett test, and two-sided F-test (P < 0.05).Randomization of the groups within the growth chamber was conducted for each plant experiment.Bacterial load and disease severity data were log-transformed before performing statistical analyses.

FIG 1
FIG 1 Disease severity and bacterial populations on "California Wonder" pepper seedlings 3 and 7 dpi with 16 Pss strains.Disease severity is represented by the number of lesions on two lower leaves of infected seedlings at 3 dpi (A) and 7 dpi (B) and populations of Pss strains in terms of log CFU/g at 3 dpi (C) and 7 dpi (D).The experiment was conducted twice with a minimum of four replicates for each strain at different time points, and data from the two experiments were combined.The black bars represent the median number of lesions or log (CFU/g).Different letters represent strains that are significantly different (P < 0.05), and the error bars represent the standard deviation.

FIG 2
FIG 2 In vitro (growth, biofilm, and motility) characteristics of 16 Pss strains.(A) The growth rate and doubling time of Pss strains, calculated using the growthcurver 0.3.1 package in R 4.3.1.The OD 600 measurement of Pss strains taken over 24 h in M9 media was used in growthcurver 0.3.1 in R to (Continued on next page)

FIG 2 (
FIG 2 (Continued) calculate the growth rate and doubling time.(B) Quantity of biofilm produced by Pss strains after 72 h of incubation at 28°C in M9 minimal broth.OD 570 was measured to quantify biofilm production.(C) Motility of Pss strains was measured after 24 h incubation on Luria-Bertani (LB) semi-solid agar medium (0.3%) at 28°C.Motility was determined by measuring the diameter (in millimeter) of the halo.Each of the in vitro experiments was conducted twice with three replicates each.The bars represent standard error, and the different letters represent strains that are significantly different (P < 0.05).

FIG 3
FIG 3 Phylogenetic, core, and pangenome analysis of 16 Pss strains.(A) Whole-genome core SNP-based phylogeny of phytopathogenic Pseudomonas strains (n = 34).K-chooser was used to find the optimal k-mer (19-mer), and the core SNPs were calculated by KSNP3.The maximum likelihood phylogeny tree was plotted in R. The branch length of the tree correlates with the SNP distance.Each pathovar as well as host/source of isolation is indicated by a different color.Pepper pathovars colored in blue are strains collected and used in this study.The branch for Pseudomonas moraviensis used as the outgroup species is truncated as indicated by the dashed line.The disease severity of the 16 Pss strains is indicated with a color spectrum, where dark blue indicates low disease severity and yellow indicates high disease severity.The bootstrap support for the branches is indicated with a color spectrum, where dark purple indicates low bootstrap support and light yellow indicates high bootstrap support.(B) Pie chart showing the pangenome composition of the 16 Pss strains.Core genes are present in 100% of the strains, shell genes in 15%-99%, and cloud genes in <15% of strains.(C) Rarefaction curves for core and pangenome of the 16 Pss strains.The lefthand plot shows how the size of the core genome (y-axis) decreases with the addition of more genome assemblies (x-axis), with individual points showing the core genome size for each possible combination of genomes for each value of x, the number of included genomes, and the line showing the fitted line following an exponential decay function.The righthand plot is similar but shows the increase of the pangenome with an increasing number of genomes, and the fitted line follows a power-law distribution.

FIG 4
FIG 4 Hierarchical clustering based on presence/absence of variable genes (n=812) of Pseudomonas syringae pv.syringae (Pss) strains (n=16).On the left-hand side, colored boxes around the strain names indicate the cluster for each strain, and the disease severity score at 7 dpi is shown for each strain.On the right-hand side, each column represents a single gene with blue indicating presence of a gene, and red indicating absence of a gene.

FIG 6
FIG 6Virulence (secretion system, motility, and biofilm) and antimicrobial resistance genes of Pseudomonas syringae pv.syringae (Pss) strains (n=16) in the variable genome.A total of 29 secretion system genes, 2 motility genes, 2 biofilm genes and 6 antimicrobial resistance genes were found to be part of the variable genome.The darker grey boxes indicate more than one copy of the gene.

TABLE 2
Plasmids of Pss strains and the genes located in the plasmids a