Genomic comparisons confirm Giardia duodenalis sub-assemblage AII as a unique species

Abstract

Giardia duodenalis is a parasitic flagellated protozoan which infects a wide range of mammalian hosts, including humans, and is subdivided into at least eight genetic assemblages commonly thought to represent cryptic species. Molecular studies have shown that G. duodenalis assemblage A, which parasitizes humans and animals, contains several phylogenetically distinct groupings known as sub-assemblages. Molecular studies employing poor phylogenetic-resolution markers routinely recover these sub-assemblages, implying that they represent evolutionarily distinct clades and possibly cryptic species, a hypothesis which is supported by epidemiologic trends. Here, we further tested this hypothesis by using available data from 41 whole genomes to characterize sub-assemblages and coalescent techniques for statistical estimation of species boundaries coupled to functional gene content analysis, thereby assessing the stability and distinctiveness of clades. Our analysis revealed two new sub-assemblage clades as well as novel signatures of gene content geared toward differential host adaptation and population structuring via vertical inheritance rather than recombination or panmixia. We formally propose sub-assemblage AII as a new species, Giardia hominis, while preserving the name Giardia duodenalis for sub-assemblage AI. Additionally, our bioinformatic methods broadly address the challenges of identifying cryptic microbial species to advance our understanding of emerging disease epidemiology, which should be broadly applicable to other lower eukaryotic taxa of interest. Giardia hominis n. sp. Zoobank LSID: urn:lsid: zoobank.org:pub:4298F3E1-E3EF-4977-B9DD-5CC59378C80E.

Keywords: Giardia; assemblage A; comparative genomics; cryptic species; phylogenetics; zoonotic diseases.

Figure 1

Workflow diagram which describes the bioinformatic steps taken to differentiate genomes of G. duodenalis assemblage A strains.

Figure 2

Comparison of genomic relatedness using ANI and multilocus coalescent estimations. Left side: ANI tree constructed from pairwise comparisons of draft genomes. Right side: Coalescent tree estimated by ASTRAL-III using gene trees from all pangenome loci. Filled circles indicate nodes with greater than 95% bootstrap support. Open circles indicate nodes with greater than 95% support in the set of trees estimated by STACEY. Grey dashed lines indicate concordant placement of a genome on opposing trees. Colored dashed lines indicate specific genomes which had differing placement.

Figure 3

Characterization of the assemblage A pangenome. (A, left side) Neighbor-joining tree estimated from SNP distances between genomes. Filled circles indicates nodes with ≥ 95% bootstrap support. (A, right side) phyletic matrix of the assemblage A pangenome. Cells shaded in dark blue indicate a gene is present, and white-shaded cells indicate gene absence. (B) Rarefaction curve estimated from the phyletic matrix shown in (A) with 100 permutations. (C) Dynamics of gene gain/loss between sub-assemblages estimated by GLOOME. Blue bars indicate gene gain events, red bars indicate gene loss, and gray bars indicate net (gain – loss) change.

Figure 4

Functional pathway annotations of unique genes per sub-assemblage. Left: Heatmap of unique genes assigned to KEGG categories. Darker shading indicates greater number of unique gene content annotated to a given category. Right: Stacked bar chart of the percent distribution of EC terms applied to genes unique to a given sub-assemblage (some genes have multiple associated KEGG and/or EC terms, enumerated individually here). White numbers on each bar segment reflect the frequency count of each EC category per sub-assemblage.

Figure 5

Comparison of gene content diversity with genomic relatedness. Each dot represents a pair of genomes. Dots are colored according to the host origin of the genomes.