Culture-independent detection of low-abundant Clostridioides difficile in environmental DNA via PCR

Abstract

Clostridioides difficile represents a major burden to public health. As a well-known nosocomial pathogen whose occurrence is highly associated with antibiotic treatment, most examined C. difficile strains originated from clinical specimen and were isolated under selective conditions employing antibiotics. This suggests a significant bias among analyzed C. difficile strains, which impedes a holistic view on this pathogen. In order to support extensive isolation of C. difficile strains from environmental samples, we designed a detection PCR that targets the hpdBCA-operon and thereby identifies low abundances of C. difficile in environmental samples. This operon encodes the 4-hydroxyphenylacetate decarboxylase, which catalyzes the production of the antimicrobial compound para-cresol. Amplicon-based analyses of diverse environmental samples demonstrated that the designed PCR is highly specific for C. difficile and successfully detected C. difficile despite its absence in general 16S rRNA gene-based detection strategies. Further analyses revealed the potential of the hpdBCA detection PCR sequence for initial phylogenetic classification, which allows assessment of C. difficile diversity in environmental samples via amplicon sequencing. Our findings furthermore showed that C. difficile strains isolated under antibiotic treatment from environmental samples were originally dominated by other strains according to PCR amplicon results. This provided evidence for selective cultivation of under-represented but antibiotic-resistant isolates. Thereby, we revealed a substantial bias in C. difficile isolation and research.IMPORTANCEClostridioides difficile is a main cause of diarrheic infections after antibiotic treatment with serious morbidity and mortality worldwide. Research on this pathogen and its virulence has focused on bacterial isolation from clinical specimens under antibiotic treatment, which implies a substantial bias in isolated strains. Comprehensive studies, however, require an unbiased strain collection, which is accomplished by isolation of C. difficile from diverse environmental samples and avoidance of antibiotic-based enrichment strategies. Thus, isolation can significantly benefit from our C. difficile-specific detection PCR, which rapidly verifies C. difficile presence in environmental samples and further allows estimation of the C. difficile diversity by using next-generation sequencing.

Keywords: Clostridioides difficile; antibiotic resistant isolates; detection PCR; environmental DNA; hpdBCA operon; low-abundant microbes; nosocomial infection.

Fig 1

Heatmap of whole-genome ANIm analysis (wgANI). ANIm percentage values were visualized in heatmap.2 with corresponding dendrogram. Phylogenetic clades based on MLST determination are color-highlighted, and MLST types are noted for C1 sub-clusters comprising more than three strains. The outlier strain ST963 is marked with * in the dendrogram.

Fig 2

Clade-wise comparison of percentage identities for wgANIm and detANIm. Given are mean, median, minimal and maximal ANI percentages between two clades of C1–C5, outlier ST963 to each clade of C1–C5, or between all cryptic Clostridioides sp. to C. difficile (all five clades combined). Clade affiliation of an isolate was defined by MLST assignment using FastMLST (33).

Fig 3

Heatmap of cresol-PCR sequence ANIm analysis (detANIm). ANIm percentage values were visualized in heatmap.2 with corresponding dendrogram. Phylogenetic clades based on MLST determination are color-highlighted, and MLST types are noted for C1 sub-clusters comprising more than three strains. Outliers are marked as follows in the dendrogram: strain ST963 with *, C1 outliers DSM 29632 and Z31 with § and #, and C2 outlier Cd1 with $.

Fig 4

Heatmap of cresol-ASV distribution and their relative abundances for selected samples. C. difficile isolates are in italic and depicted next to their original environmental sample. MLST assignment (ST/clade) of each ASV as described in Table 2 is additionally specified below each ASV. Percentage values of the relative abundances are noted for values >0%.