Comparative Genomic Analysis of the Lettuce Bacterial Leaf Spot Pathogen, Xanthomonas hortorum pv. vitians, to Investigate Race Specificity

Abstract

Bacterial leaf spot (BLS) of lettuce caused by Xanthomonas hortorum pv. vitians (Xhv) was first described over 100 years ago and remains a significant threat to lettuce cultivation today. This study investigated the genetic relatedness of the Xhv strains and the possible genetic sources of this race-specific pathogenicity. Whole genome sequences of eighteen Xhv strains representing the three races, along with eight related Xanthomonas strains, were included in the analysis. A maximum likelihood phylogeny based on concatenated whole genome SNPs confirmed previous results describing two major lineages of Xhv strains. Gene clusters encoding secretion systems, secondary metabolites, and bacteriocins were assessed to identify putative virulence factors that distinguish the Xhv races. Genome sequences were mined for effector genes, which have been shown to be involved in race specificity in other systems. Two effectors identified in this study, xopAQ and the novel variant xopAF2, were revealed as possible mediators of a gene-for-gene interaction between Xhv race 1 and 3 strains and wild lettuce Lactuca serriola ARM-09-161-10-1. Transposase sequence identified downstream of xopAF2 and prophage sequence found nearby within Xhv race 1 and 3 insertion sequences suggest that this gene may have been acquired through phage-mediated gene transfer. No other factors were identified from these analyses that distinguish the Xhv races.

Keywords: Xanthomonas; bacterial plant pathogens; comparative genomics; effectors; plant-microbe interactions.

FIGURE 1

Bioinformatics workflow for comparing whole genome sequences.

FIGURE 2

Genome assembly completeness assessment for Xhv and related strains. The completeness of the genome assemblies produced in this study was evaluated using the benchmarking universal single copy orthologs method (BUSCO; Manni et al., 2021). The proportions of complete and single-copy orthologs in these assemblies were compared to those of previously published Xanthomonas assemblies retrieved from NCBI’s genome database, including X. campestris pv. campestris (ATCC 33913^T; GenBank assembly accession GCA_000007145.1), X. populi (CFBP 1817^T; GCA_002940065.1), X. arboricola pv. juglandis (CFBP 2528^T; GCA_001013475.1), X. hortorum pv. pelargonii (CFBP 2533^PT; GCA_012922215.1), X. hortorum pv. taraxaci (CFBP 410^PT; GCA_012922225.1), X. hortorum pv. cynarae (CFBP 4188^PT; GCA_002939985.1), X. hortorum pv. carotae (CFBP 7900; GCA_000505565.1), X. hortorum pv. gardneri (CFBP 8163^PT; GCA_012922265.1), X. hortorum pv. vitians (CFBP 8686^PT; GCA_012922135.1), X. campestris pv. coriandri (ICMP 5725^PT; GCA_019201305.1), X. fragariae (PD 885^T; GCA_900183975.1), and X. hortorum pv. hederae (WHRI 7744^T; GCA_003064105.1).

FIGURE 3

Phylogenetic tree constructed from whole genome SNP data. Maximum-likelihood phylogeny of Xanthomonas hortorum pv. vitians strains of race 1 (blue), race 2 (red), and race 3 (green) and related Xanthomonas type and pathotype strains (black) used in this study. SNP data was extracted from trimmed Illumina sequence reads and core aligned using the snippy-multi pipeline (Seemann, 2015) and the tree was built from sequence alignment using CLC Genomics Workbench. This phylogeny was constructed using the neighbor-joining clustering method and the general time-reversible substitution method with rate variation (4 categories) and estimated topology. Bootstrap values are shown above branches and branch length represents the expected number of nucleotide substitutions per site. Branches shorter than 0.0100 are shown as having a length of 0.0100.

FIGURE 4

Phylogenetic trees constructed from (A) XopAQ amino acid alignment and (B) XopAF amino acid alignment. Maximum-likelihood phylogenies were constructed for the two genes present in Xhv race 1 and race 3 strains (blue and green) but missing from Xhv race 2 strains. Nucleotide sequences for these genes were extracted from the whole genome sequence assemblies and converted to amino acid sequences using OrthoFinder. Alignments and trimming were completed using MEGA11, as well as the phylogeny building using the Jones-Taylor-Thornton (JTT) substitution model, uniform substitution rates among sites, and nearest-neighbor-interchange (NNI) method for tree inference. Bootstrap values are shown beside branches and branch length represents the expected number of amino acid substitutions per site. Asterisks indicate published sequences used as references for proteins HopAF1 and XopAF (Astua-Monge et al., 2000; Petnicki-Ocwieja et al., 2002) and XopAQ, HopAQ1, Rip6, and Rip11 (Guttman et al., 2002; Mukaihara et al., 2010; Potnis et al., 2011).

FIGURE 5

Predicted Secretion system gene clusters for Xhv and related Xanthomonas strains. Presence of each labeled gene in the majority of Xhv and related strains is indicated by a black arrow, and absence from all strains is indicated by a white arrow. Asterisks, carets, and crosses indicate exceptions listed here. (A) T2SSs xps and xcs—exceptions include X. hortorum pv. taraxaci and X. hortorum pv. gardneri, for which the entire xps gene cluster is predicted, and Xhv race 1 strain BP5179, for which the genes xpsI, xpsJ, and xpsD were truncated; (B) T3SS—exceptions include X. hortorum pv. hederae, which had a large insertion between hpa1 and hrcC and lacked hrpE, X. campestris pv. coriandri, which lacked hpa1 and hrpW, and X. hortorum pv. pelargonii, which lacked hrpE and hrpD6; (C) T4SSA—exceptions include Xhv race 1 strains BS0541 and BS3044, which encoded the additional gene virB6, X. hortorum pv. gardneri, which encoded the additional virB6 gene and lacked virB11, X. hortorum from radicchio, which only encoded for virD4, and X. campestris pv. coriandri, which only encoded for virB4, virB8, virB9, and virD4; and (D) T6SS—which was only predicted for the X. campestris pv. coriandri strain and not found in any other strain tested.