Intra-species diversity of Clostridium perfringens: A diverse genetic repertoire reveals its pathogenic potential

Abstract

Clostridium perfringens is the causative agent of many enterotoxic diseases in humans and animals, and it is present in diverse environments (soil, food, sewage, and water). Multilocus Sequence Typing (MLST) and Whole Genome Sequencing (WGS) have provided a general approach about genetic diversity of C. perfringens; however, those studies are limited to specific locations and often include a reduced number of genomes. In this study, 372 C. perfringens genomes from multiple locations and sources were used to assess the genetic diversity and phylogenetic relatedness of this pathogen. In silico MLST was used for typing the isolates, and the resulting sequence types (ST) were assigned to clonal complexes (CC) based on allelic profiles that differ from its founder by up to double-locus variants. A pangenome analysis was conducted, and a core genome-based phylogenetic tree was created to define phylogenetic groups. Additionally, key virulence factors, toxinotypes, and antibiotic resistance genes were identified using ABRicate against Virulence Factor Database (VFDB), TOXiper, and Resfinder, respectively. The majority of the C. perfringens genomes found in publicly available databases were derived from food (n = 85) and bird (n = 85) isolates. A total of 195 STs, some of them shared between sources such as food and human, horses and dogs, and environment and birds, were grouped in 25 CC and distributed along five phylogenetic groups. Fifty-three percent of the genomes were allocated to toxinotype A, followed by F (32%) and G (7%). The most frequently found virulence factors based on > 70% coverage and 99.95% identity were plc (100%), nanH (99%), ccp (99%), and colA (98%), which encode an alpha-toxin, a sialidase, an alpha-clostripain, and a collagenase, respectively, while tetA (39.5%) and tetB (36.2%), which mediate tetracycline resistance determinants, were the most common antibiotic resistance genes detected. The analyses conducted here showed a better view of the presence of this pathogen across several host species. They also confirm that the genetic diversity of C. perfringens is based on a large number of virulence factors that vary among phylogroups, and antibiotic resistance markers, especially to tetracyclines, aminoglycosides, and macrolides. Those characteristics highlight the importance of C. perfringens as a one of the most common causes of foodborne illness.

Keywords: Clostridium perfringens; genomic epidemiology; intra-species diversity; multilocus sequence typing; toxinotypes.

Figure 1

Origin and toxinotype profile of 372 Clostridium perfringens genomes. (A) Distribution of isolates by country of origin. (B) Distribution of isolates by source. (C) Distribution of isolates by toxin type. (D) Toxinotype diversity among hosts. All figures were created with R software.

Figure 2

Multilocus sequence typing (MLST)-based phylogeny and goeBURST full minimum spanning tree of 195 C. perfringens MLST profiles among 322 genomes. (A) MLST-based phylogeny tree obtained with fastMLST. The outer ring shows the origin of the isolates as indicated in the legend. Clusters are represented by different colors on the inside of the ring. (B) MLST-based minimum spanning tree obtained with goeBURST. Sequence types (STs) are displayed as circles. Founder STs were defined as the STs with the greatest number of single-locus variants. Circle size indicates the number of isolates in every particular ST, with each color representing a different source type. Lines represent closely related isolates and line length illustrates STs that vary by one, two, or more alleles in their MLST profile. Clonal complexes (CC), defined as closely related STs that differ up to two of the eight loci used in the MLST scheme from a common founder (DLV, double locus variant), are designated by dashing lines.

Figure 3

Phylogenetic grouping of C. perfringens using two different approaches. (A) Phylogenetic tree based on the core genome of 372 genomes. Five main phylogroups are highlighted in different colors. Source type and toxinotype are shown on the right panels. (B) Five phylogenetic networks based on the Neighbor-Joining (NJ) algorithm are shown with different colors.

Figure 4

Virulence-associated and antimicrobial resistance (AMR) genes in 372 C. perfringens genomes. The core genome-based phylogenetic tree showed phylogroups highlighted in different colors. Matrix of absence and presence of genes shows (A) virulence genes (B) antimicrobial resistance (AMR) genes. The minimum coverage threshold needed for detection of these genes was 70% and the percentage of identity was 99.95%. The resulting trees were visualized and edited using iTOL v.5 (Letunic and Bork, 2021).