Vital Surveillances: Molecular Epidemiology of Type F Clostridium perfringens Among Diarrheal Patients and Virulence-Resistance Dynamics — 11 Provinces, China, 2024

From January 2 to May 28, 2024, we conducted a cross-sectional study to determine C. perfringens prevalence among inpatients at 11 provincial hospitals across China. The study included hospitals in Shandong (n=230), Guangxi (n=100), Henan (n=196), Gansu (n=190), Shaanxi (n=243), Fujian (n=238), Hunan (n=104), Guangdong (n=177), Jilin (n=200), Jiangxi (n=300), and Zhejiang (n=350) PLADs. Participating departments included Gastroenterology and Neurology. A total of 2,068 fecal or rectal swab samples were collected using ESwab^TM collection kits (Copan, Brescia, Italy). For C. perfringens isolation, we processed either a small fecal sample or 0.2 mL of transport medium with 50% ethanol, followed by centrifugation and plating on TSC agar for anaerobic incubation. Suspected colonies underwent further purification on blood agar and definitive identification using MALDI-TOF MS (Bruker Daltonik GmbH, Bremen, Germany).

Antimicrobial susceptibility testing of C. perfringens isolates was performed using the Etest method following Clinical & Laboratory Standards Institute (CLSI) guidelines (M100-S29:2019). Nine antimicrobial agents were evaluated: metronidazole, penicillin, amoxicillin, tetracycline, ciprofloxacin, cefoxitin, linezolid, clindamycin, and erythromycin. For erythromycin and ciprofloxacin testing, we applied breakpoints equivalent to clindamycin and fluoroquinolones, respectively, due to the absence of specific CLSI guidelines for C. perfringens. C. perfringens ATCC 13124TM served as the quality control strain.

Genomic DNA extraction was performed using PureLink Genomic DNA Mini Kit (Invitrogen, Carlsbad, CA, USA). DNA libraries were indexed using TruSeq DNA PCR-free Sample Preparation Kit (Illumina, Inc., San Diego, CA) and sequenced on the Illumina HiSeq X Ten System, generating 300-bp paired-end reads with minimum 150-fold coverage per isolate. Raw reads underwent trimming and assembly using SPAdes v3.11.1, followed by targeted analysis of AMR and virulence genes using ABRicate against relevant databases, employing thresholds of >90% identity and >75% coverage.

Type F C. perfringens isolates were retrieved from the National Center for Biotechnology Information (NCBI) database on August 27, 2024, using specific search criteria: “Toxin_genotypes: cpe & plc” and “species_taxid:1502.” 91 Type F C. perfringens isolates, with their sources, countries of origin, and accession numbers are documented in Table 1 (13). Single-nucleotide polymorphisms (SNPs) were identified through sequence alignment using Snippy v4.6.0 (https://github.com/tseemann/snippy) (14), which generated a core genome alignment profile. Pairwise SNP distances were calculated using Snp-dists v0.6. We constructed a phylogenetic tree based on core-genome SNPs using Parsnp within the Harvest suite, with midpoint rooting and visualization enhanced through iTOL v6.25 (15).

Clinical data were extracted from the hospital information system. We employed the Wilcoxon test to analyze differences in antimicrobial resistance and virulence genes, while Pearson chi-square and Fisher’s exact tests were used to evaluate statistical significance (P<0.05) in gene frequencies and resistance phenotypes.

All supporting data for this study are included in this article and its Supplementary Information. The genome assemblies of C. perfringens have been deposited in NCBI under BioProject accession number PRJNA1154412. Additional data are available from the corresponding authors upon reasonable request.

Among 2,068 non-duplicated stool specimens collected from patients across 11 provincial tertiary hospitals in 2024, 17 Type F C. perfringens isolates were identified, yielding a prevalence rate of 2.38% [95% (confidence interval) CI: 1.95%, 2.91%]. These isolates were distributed across 6 PLADs, with isolation rates varying from 0.9% in Shandong to 2.0% in Henan, Jilin, and Guangxi PLADs.

The demographic distribution of Type F C. perfringens cases closely mirrored the overall study population, with cases showing a mean age of 54.00±24.00 years and a gender distribution of 52.9% female versus 47.1% male (compared to the overall study population: age 37.51±11.98 years, 45.48% female versus 54.52% male). Notably, 64.7% of Type F C. perfringens isolates were recovered from patients presenting with diarrhea in gastroenterology departments.

Antimicrobial susceptibility testing revealed that most Type F C. perfringens isolates demonstrated susceptibility to linezolid, cefoxitin, and amoxicillin. However, 47.1% of isolates exhibited high resistance (>32 µg/mL) to metronidazole, and all strains showed intermediate resistance to erythromycin (Figure 1 and Supplementary Table S1).

Figure 1. 

Distribution of antimicrobial resistance patterns among 17 F toxinotype C. perfringens isolates against 9 antimicrobial agents across 8 distinct categories.

Phylogenetic analysis (Figure 2) revealed distinct clonal clusters of Type F C. perfringens isolates, each characterized by specific virulence and resistance determinants. Whole-genome sequencing identified tetracycline resistance genes [tet(A) and tet(B)], which corresponded with observed phenotypic resistance patterns. Among the analyzed isolates, we identified 12 distinct virulence factors: colA, nagH, nagI, nagJ, nagK, nagL, nanH, nanI, nanJ, pfoA, plc, and cpe. All isolates harbored the essential virulence determinants cpe (encoding enterotoxin CPE) and plc (encoding α-toxin). Notably, 18.0% (3/17) of isolates lacked nagI, nagJ, nagK, nagL, nanI, nanJ, and pfoA genes. The clustering patterns suggested the emergence of distinct clonal lineages, with an apparent inverse relationship between resistance gene carriage and virulence factor repertoire, indicating a potential fitness trade-off between resistance and virulence mechanisms.

Figure 2. 

Phylogenetic relationships and corresponding antimicrobial resistance phenotypes, virulence characteristics, and genotypic profiles of 17 C. perfringens isolates from China.

Abbreviation: PLADs=provincial level administrative divisions.

Analysis of the genetic environment (Figure 3) revealed high sequence homology between the IS1469-cpe-hp-IS1151 gene cluster in our isolates (82.4%, 14/17) and sequences from a diarrheal canine isolate (strain D13122 plasmid pD13122_cpe, accession No. MG456815.1). This homology suggests potential horizontal gene transfer events between animal and human strains. The cpe gene, flanked by mobile genetic elements including transposons and insertion sequences (IS1469, IS1151), was found integrated into the chromosomal DNA of multiple isolates, indicating a mechanism for stable inheritance of virulence factors.

Figure 3. 

Genetic organization of the cpe locus in Clostridium perfringens.

Note: Arrows indicate gene orientation and function: red (toxin gene cpe), blue (mobile genetic elements), and orange (other protein-encoding genes).

Our 17 Type F C. perfringens isolates were analyzed in comparison with 91 Type F C. perfringens strains from the NCBI database, representing 13 countries and diverse sources (Figure 4). Comparative genomic analysis revealed that none of the NCBI database strains exhibited SNP distances less than 100 from our study isolates, suggesting distinct evolutionary trajectories. Notably, isolates 1-296 from Henan Province showed close genetic relatedness (SNPs<100) to isolates 2-30 and 1-23 from Hunan and Shandong, respectively. Within Henan Province, isolates 1-246, 1-225, and 1-249 demonstrated remarkable genetic similarity with less than 5 SNPs difference. Similarly, in Jilin Province, isolates 2-253 and 2-211 exhibited 100% sequence identity with isolates 2-262 and 2-228, respectively. Analysis of antimicrobial resistance genes revealed a relatively low prevalence of resistance determinants, with tet(A) and tet(B) being the most common at 46.7% and 15.2%, respectively.

Figure 4. 

Global phylogenetic analysis of 108 F type Clostridium perfringens isolates based on core genome SNPs.

This study characterized the epidemiological landscape of Type F C. perfringens across PLADs, revealing a 2.38% isolation rate. Although this prevalence appears relatively low, the exclusive detection of Type F C. perfringens in human samples, particularly from gastroenterology departments, emphasizes its clinical significance in human gastrointestinal health. The predominant isolation from gastroenterology departments aligns with established associations between Type F C. perfringens and diarrheal diseases, corroborating previous research linking these strains to gastrointestinal pathology through the cpe gene (16).

The enterotoxin CPE is a crucial virulence determinant in Type F strains (17), with historical studies reporting CPE detection rates of 40%–70% in gastroenteritis outbreaks (18). While research in Japan demonstrated a predominance of plasmid-mediated CPE with downstream IS1151 sequences in food poisoning outbreaks (19), our analysis revealed a different pattern. The majority of our CPE-positive isolates (82.0%, 14/17) carried chromosomally-encoded CPE associated with IS1469, though some isolates harbored plasmid-borne cpe-IS1151 loci, suggesting potential involvement in extraintestinal C. perfringens infections.

Antimicrobial susceptibility profiles revealed a concerning trend: while most Type F C. perfringens isolates maintained susceptibility to common antibiotics, 47.1% exhibited high resistance to metronidazole, a critical first-line treatment for anaerobic infections (20). The universal intermediate resistance to erythromycin among isolates suggests that antibiotic selective pressure in clinical settings may be driving the emergence of resistant strains, potentially compromising future therapeutic options.

Phylogenetic analysis revealed highly virulent strains (harboring both cpe and plc genes) with relatively few resistance elements, particularly tetracycline-associated transposons (IS1469, IScp2). Core SNP analysis demonstrated evidence of clonal transmission within specific geographic regions, particularly among isolates from Henan, Hunan, Shandong, and Jilin provinces, suggesting localized spread patterns among diarrheal patients.

The identification of genetic similarities between human clinical isolates and those from a diarrheal dog in an antibiotic-prevalent setting underscores the significance of horizontal gene transfer in virulence factor dissemination across species. Furthermore, the exclusive detection of Type F C. perfringens in human cases within our study suggests potential human-specific adaptation or a preferential ecological niche for human colonization. This host specificity, combined with the pathogen’s virulence capabilities, emphasizes the importance of investigating its transmission dynamics, particularly regarding its persistence in human populations despite relatively low isolation rates.

This study’s limitations include a relatively modest sample size that may not fully represent the diversity of China’s population, potentially affecting the generalizability of the findings to different regions and demographic groups. Additionally, the potential selection bias introduced by hospital participation could skew the results, as the hospitals involved might differ from others in terms of patient characteristics, treatment protocols, and care quality.

In conclusion, our findings highlight the significant public health implications of virulence and antibiotic resistance patterns in Type F C. perfringens. The predominant detection of this pathogen in human cases emphasizes its clinical relevance and raises important questions about its transmission mechanisms and host adaptation. These observations underscore the necessity for targeted surveillance and preventive strategies to mitigate potential risks in both clinical and community settings.

Ethical approval was given by the Zhejiang University ethics committee (number 2024–0994). Informed patient consent was waived as samples were taken under a hospital surveillance framework for routine sampling. The research conformed to the principles of the Helsinki Declaration.