Two plasmid-borne virulence genomic islands of Clavibacter michiganensis are genetically diverse and determine the development of wilt symptoms in host plants

Abstract

Plasmids contribute to the efficient adaptation of bacteria to specific niches in nature. The gram-positive bacterium Clavibacter michiganensis carries two plasmid-borne important virulence genes, celA and pat-1, necessary for wilting in tomato. The 88 C. michiganensis field isolates collected between 2011 and 2020 were examined for phenotypic variation, including virulence in host plants. Four isolates lacking plasmids with celA, pat-1, or both failed to cause wilting, and nine isolates, including these four, failed to cause wilting in Nicotiana benthamiana. Whole genome analyses revealed 11 distinct plasmid types, including 9 newly identified, and 10 bacterial groups with different plasmid compositions, despite having almost identical chromosomes. Comparative genomic analyses revealed significant genetic diversity among the plasmids, while three plasmids containing the genomic island (GI) α with celA or GIβ with pat-1 and three newly identified plasmids carrying both islands shared large blocks of synteny. In addition, GIα is closely associated with mobile genetic elements, suggesting the genetic rearrangement or transfer at this locus. These results suggest that C. michiganensis harbors a wide variety of virulence and nonvirulence plasmids, and that there is genetic rearrangement among plasmids in GI regions, determining bacterial virulence in plants.

Keywords: Clavibacter michiganensis; genetic diversity; genomic islands; plasmids; virulence.

Fig. 1

Identification and virulence of Clavibacter michiganensis field isolates. (a) Maximum likelihood phylogenetic tree of 88 field isolates constructed using MEGA11 after alignment of multiple 16S rRNA sequences. Bootstrap values (%) from 1000 replicates are shown at the internal branch nodes. Newly identified field isolates were grouped into two clusters (Group 1 and Group 2), as listed in Supporting Information Table S1. Leifsonia xyli species was used as an outgroup. The NCBI accession numbers of the respective sequences are given in parentheses. (b) Geographic locations of field isolates. Blue and red nodes represent groups 1 and 2, respectively, based on phylogenetic tree results. (c) Disease severity comparison in tomato plants between field isolates and the type strain (n = 5) post root inoculation with each strain at OD₆₀₀ of 2 (c. 2 × 10⁹ CFU ml⁻¹). Low virulent isolates are marked with a green star symbol. The isolate names are color‐coded, with blue and red colors indicating Groups 1 and 2, respectively. Different letters indicate statistically significant differences analyzed by a nonparametric Kruskal–Wallis test with Dunnett's multiple comparison (P < 0.01). Error bars represent ±SD (standard deviation), and the experiment was repeated three times. The experiment was repeated three times. (d) Wilting symptoms by representative strains in tomato plants, photographed 14 d post‐inoculation. Strains within the red box indicate low virulence strains. Mock, 10 mM MgCl₂.

Fig. 2

Role of celA and pat‐1 in Clavibacter michiganensis field isolates as major virulence determinants for wilting symptoms in tomato plants. (a) Virulence assay of C. michiganensis field isolates lacking celA and pat‐1 and their complemented strains, root‐inoculated with each strain at OD₆₀₀ of 2 (c. 2 × 10⁹ CFU ml⁻¹). The type strain, LMG7333^T, and 10 mM MgCl₂ (Mock) were used as positive and negative controls, respectively. Symptoms were photographed 14 d post‐inoculation. (b) Disease severity (n = 10) of isolates 19249‐1 and 20011‐1 with overexpressed celA and pat‐1. Different letters indicate statistically significant differences analyzed by a nonparametric Kruskal–Wallis test with Dunnett's multiple comparison (P < 0.01). In the boxplots, the boxes represent the interquartile range (25^th to 75^th percentiles), and the 'X' marks indicate the mean values. Whiskers extend to the minimum and maximum values, showing the full data range. Each black dot represents an individual measurement from 10 tomato plants. The dotted outline indicates ±SD (standard deviation). The experiment was repeated three times with similar results.

Fig. 3

Virulence of Clavibacter michiganensis field isolates in Nicotiana benthamiana plants using the root‐dipping method. (a) Differential virulence of 88 C. michiganensis field isolates in N. benthamiana plants. All field isolates were inoculated with each isolate at OD₆₀₀ of 2 (c. 2 × 10⁹ CFU ml⁻¹) using the root‐dipping method on 2‐wk‐old seedlings. Low virulent isolates are marked with a green star symbol. Error bars represent ± SD (standard deviation). (b) Wilting symptoms induced by low virulent isolates among C. michiganensis field isolates, photographed 14 d post‐inoculation. (c) Disease severity quantification (n = 5) in N. benthamiana plants infected with low virulent isolates. The type strain, LMG7333^T, and 10 mM MgCl₂ (Mock) were used as positive and negative controls, respectively. In the boxplots, the boxes represent the interquartile range (25th to 75th percentiles), and the 'X' marks indicate the mean values. Whiskers extend to the minimum and maximum values, showing the full data range. Each black dot represents an individual measurement from 10 tomato plants. The dotted outline indicates ±SD. The experiment was repeated three times with similar results.

Fig. 4

Plasmid distribution in Clavibacter michiganensis field isolates. (a) Heat map of the average nucleotide identity values for chromosome sequences of representative C. michiganensis field isolates and their plasmid composition. Cc, C. capsici; Cs, C. sepedonicus; Ci, C. insidiosus. (b) Distribution of 88 field isolates with each plasmid. (c) Ten groups of 88 C. michiganensis field isolates with different plasmid compositions. Numbers indicate the number of field isolates with indicated plasmids/total number of isolates. (d) Distribution of plasmid combinations in 88 field isolates by year. Plasmids identified in C. michiganensis field isolates were confirmed by whole genome sequencing or PCR assay and then arranged by year.

Fig. 5

Plasmid variation and synteny in Clavibacter michiganensis field isolates. (a) Phylogenomic tree of the plasmids based on orthologous protein sequences present in plasmids. The maximum likelihood (ML) tree was generated using Iq‐Tree v.2.4.0 with 1000 bootstrap replicates. The numbers at the nodes indicate the percentage support based on 1000 bootstrap replicates. A cladogram subtree of pCM7 is shown separately. Cc, C. capsici; Cs, C. sepedonicus; Ci, C. insidiosus. (b) The pyCircos plot representation displaying synteny among plasmids in C. michiganensis field isolates. In both figures, the color corresponds to the individual plasmid groups of C. michiganensis.

Fig. 6

Syntenic relationship among plasmids carrying the mobilization and/or conjugal transfer module. Open reading frames are indicated by arrows. Shared regions with a high degree of sequence similarity are depicted as gray and red synteny blocks, where gray represents collinear regions (same order) and red represents inverted regions (reverse order). The GIα regions with celA and GIβ regions with pat‐1 on plasmids are highlighted by black and blue arrows, respectively. MGEs, mobile genetic elements.

Fig. 7

Genomic islands with virulence genes in plasmids of Clavibacter michiganensis field isolates. Genetic structure and gene content of genomic islands containing celA genes (a) and pat‐1 genes (b) on virulence plasmids. The name of the isolate(s) in parentheses indicates the representative isolate with the plasmid. In addition, the percentages in parentheses indicate the query coverage and identity compared to the GIs of pCM11.

Fig. 8

Major putative virulence gene profiles in 25 Clavibacter michiganensis field isolates. The presence and absence of all identified putative virulence genes are denoted by green and black colors, respectively, while yellow indicates pseudogenes. The virulence profiles of the 25 isolates are color‐coded: red for severe symptoms or strong hypersensitive response (HR), orange for moderate symptoms, and sky blue for no symptoms. Additionally, the plasmid composition of each isolate is depicted in blue.