Comparative genomics of the Liberibacter genus reveals widespread diversity in genomic content and positive selection history

Abstract

'Candidatus Liberibacter' is a group of bacterial species that are obligate intracellular plant pathogens and cause Huanglongbing disease of citrus trees and Zebra Chip in potatoes. Here, we examined the extent of intra- and interspecific genetic diversity across the genus using comparative genomics. Our approach examined a wide set of Liberibacter genome sequences including five pathogenic species and one species not known to cause disease. By performing comparative genomics analyses, we sought to understand the evolutionary history of this genus and to identify genes or genome regions that may affect pathogenicity. With a set of 52 genomes, we performed comparative genomics, measured genome rearrangement, and completed statistical tests of positive selection. We explored markers of genetic diversity across the genus, such as average nucleotide identity across the whole genome. These analyses revealed the highest intraspecific diversity amongst the 'Ca. Liberibacter solanacearum' species, which also has the largest plant host range. We identified sets of core and accessory genes across the genus and within each species and measured the ratio of nonsynonymous to synonymous mutations (dN/dS) across genes. We identified ten genes with evidence of a history of positive selection in the Liberibacter genus, including genes in the Tad complex, which have been previously implicated as being highly divergent in the 'Ca. L. capsica' species based on high values of dN.

Keywords: bacteria; evolution; pan-genome; plant pathogen; positive selection.

Figure 1

The maximum likelihood phylogeny of the 52 Liberibacter accessions and 1 outgroup species included in this study. The phylogeny was constructed using the nucleotide core gene alignment as the input. Each isolate is labeled at the tips and is colored according to the species designation.

Figure 2

Gene annotation for the set of core and accessory genes identified across the Liberibacter genus. The distribution of functional categories are faceted for the set of 436 core genes and 2,639 accessory genes. The y-axis denotes both the COG category letter designation and the category description.

Figure 3

Liberibacter genome relatedness based on average nucleotide identity (ANI). ANI, measured in percent identity, was calculated between all possible pairs of accessions, resulting in 1,378 pairwise ANI calculations. The dendrograms represent the Liberibacter phylogeny based on the core gene alignment.

Figure 4

The distribution of functional categories for core and accessory genes separated by species. For each COG functional category, the number of genes annotated with that particular COG function in each species are depicted and separated by color. The COG definitions per letter are the same as in Figure 2.

Figure 5

The abundances of genes involved in four different KEGG metabolism pathways amongst the different Liberibacter species: (A) amino acid, (B) carbohydrate, (C) energy, and (D) nucleotide. Abundance refers to the total gene count, and each boxplot represents the total number of genes annotated for that KEGG function in each genome analyzed in this study. Significance was assessed using ANOVA with post hoc Tukey HSD test, and the stars note the level of significance such that *p < 0.5, **p < 0.01, ***p < 0.001.

Figure 6

A comparison of a NJ tree based on distances due to accessory gene presence/absence patterns (left) to the maximum likelihood tree based on the core gene alignment (right; Figure 1). As in Figure 1, the isolates are labeled at the tips of the trees, with clades colored according to species. Lines connect the same isolate between the two trees, with angled lines representing topological discordance between phylogenies.

Figure 7

The distribution of estimated values of ω, or dN/dS, under the One-Ratio model (M0) which estimated a single value of ω across an entire gene tree for the set of (A) 436 core genes and (B) 2,639 accessory genes.

Figure 8

A summary figure of the overarching comparative genomics outcomes from this study. (A) Our pipeline consisted of investigating whole genomes of Liberibacter from six different species. (B) Using whole genomes, we found that Lso had the lowest ANI within species, highlighting its high genetic diversity. (C) Pan-genome analysis identified a set of 436 core genes and 1,698 accessory genes. We investigated the phylogenetic relationships by building a phylogenetic tree using the core gene nucleotide alignment and comparing it to a tree built from the presence absence patterns of accessory genes. The comparison revealed different topologies between the two trees as the Laf species clustered more closely with Las based on core gene sequences but was farther from Las based on accessory gene content. (D) We estimated dN/dS across core and accessory genes to investigate for signatures of selection. We found a pattern of purifying selection across core genes, but variable dynamics across the set of accessory genes with several inferred to have experienced positive selection.