Genomic Insights into Virulence Factors and Multi-Drug Resistance in Clostridium perfringens IRMC2505A

Clostridium perfringens is a spore-forming, Gram-positive anaerobic pathogen that causes several disorders in humans and animals. A multidrug-resistant Clostridium strain was isolated from the fecal sample of a patient who was clinically suspected of gastrointestinal infection and had a recent history of antibiotic exposure and diarrhea. The strain was identified by 16s rRNA sequencing as Clostridium perfringens. The strain’s pathogenesis was analyzed through its complete genome, specifically antimicrobial resistance-related genes. The Clostridium perfringens IRMC2505A genome contains 19 (Alr, Ddl, dxr, EF-G, EF-Tu, folA, Dfr, folP, gyrA, gyrB, Iso-tRNA, kasA, MurA, rho, rpoB, rpoC, S10p, and S12p) antibiotic-susceptible genetic species according to the k-mer-based detection of antimicrobial resistance genes. Genome mapping using CARD and VFDB databases revealed significant (p-value = 1 × 10−26) genes with aligned reads against antibiotic-resistant genes or virulence factors, including phospholipase C, perfringolysin O, collagenase, hyaluronidase, alpha-clostripain, exo-alpha-sialidase, and sialidase activity. In conclusion, this is the first report on C. perfringens from Saudi Arabia that conducted whole genome sequencing of IRMC2505A and confirmed the strain as an MDR bacterium with several virulence factors. Developing control strategies requires a detailed understanding of the epidemiology of C. perfringens, its virulence factors, and regional antimicrobial resistance patterns.


Introduction
Clostridium perfringens (C. perfringens) is a spore-forming, Gram-positive bacterium. C. perfringens was first discovered in 1981 at the Johns Hopkins Hospital. It is a causative agent of gastroenteritis, myonecrosis, and enterotoxemia. Initially, it was known as Bacillus aerogenes capsulatus. Before the final name, Clostridium perfringens, this pathogen was called Bacillus welchii [1][2][3][4][5][6][7][8][9][10]. C. perfringens can produce spores that are tolerant to environmental stress and have a wide variety of toxins involved in the pathogenesis of diseases. Spores of

Results
The assembled genome of IRMC2505A had 108 contigs with a total length of 4,322,225 bp and an average G + C content of 28.28% (Table S1). This genome belongs to the Bacteria superkingdom, and its taxonomy is cellular organisms > Bacteria > Terrabacteria group > Firmicutes > Clostridia > Eubacterial > Clostridiaceae > Clostridium > Clostridium perfringens. The Clostridium perfringens genome has 4159 protein-coding sequences (CDS), 80 transfer RNA (tRNA) genes, and 7 ribosomal RNA (rRNA) genes (Table S1). Figure 1 shows a circular graphic of the distribution of Clostridium perfringens IRMC2505A genome annotations. It also presents an overview of the genome's specific biological processes and structural complexes. The IRMC2505A genome is close to Clostridium perfringens  (Tables 1 and 2), functional categorization ( Figure 1 and Figure S1), and the number of specialty genes. We also performed a phylogenetic tree and genome comparison between Clostridium perfringens IRMC2505A and Clostridium perfringens ATCC 13124 195103.10 ( Figure 2 and Figure S2). However, the genome comparison between IRMC2505A and ATCC 13124 195103.10 revealed dissimilar arrangements ( Figure 2B). Protein analysis revealed the presence of 2748 proteins with functional assignments and 1411 hypothetical proteins (Table S1). The proteins with functional assignments included 870 proteins with EC numbers and 692 proteins with GO assignments. The Clostridium perfringens IRMC2505A genome has 3608 genus-specific protein families (PLFams) and 3702 cross-genus protein families (PGFams).
Toxins 2023, 15, x FOR PEER REVIEW 3 of 17 Figure 1 shows a circular graphic of the distribution of Clostridium perfringens IRMC2505A genome annotations. It also presents an overview of the genome's specific biological processes and structural complexes. The IRMC2505A genome is close to Clostridium perfringens based on type (Table 1 and Table 2), functional categorization ( Figure 1 and Figure S1), and the number of specialty genes. We also performed a phylogenetic tree and genome comparison between Clostridium perfringens IRMC2505A and Clostridium perfringens ATCC 13124 195103.10 ( Figure 2 and Figure S2). However, the genome comparison between IRMC2505A and ATCC 13124 195103.10 revealed dissimilar arrangements ( Figure 2B). Protein analysis revealed the presence of 2748 proteins with functional assignments and 1411 hypothetical proteins (Table S1). The proteins with functional assignments included 870 proteins with EC numbers and 692 proteins with GO assignments. The Clostridium perfringens IRMC2505A genome has 3608 genus-specific protein families (PLFams) and 3702 cross-genus protein families (PGFams).

Figure 1.
A circular graphic of the distribution of genome annotations in IRMC2505A. From the outer to inner rings, this includes the contigs, CDS on the forward strand, CDS on the reverse strand, RNA genes, CDS with homology to known antimicrobial resistance genes, CDS with homology to known virulence factors, GC content, and GC skew. The colors of the CDS on the forward and reverse strands indicate the subsystem these genes belong to (see subsystems below). From the outer to inner rings, this includes the contigs, CDS on the forward strand, CDS on the reverse strand, RNA genes, CDS with homology to known antimicrobial resistance genes, CDS with homology to known virulence factors, GC content, and GC skew. The colors of the CDS on the forward and reverse strands indicate the subsystem these genes belong to (see subsystems below).

Antimicrobial Resistance Genes
The CARD (comprehensive antibiotic resistance database), NDARO (national database of antibiotic-resistant organisms), and PATRIC (Pathosystems Resource Integration Center) databases revealed antimicrobial resistance genes (Tables 1 and 2; Supplementary  Tables S2 and S3). Nineteen (Alr, Ddl, dxr, EF-G, EF-Tu, folA, Dfr, folP, gyrA, gyrB, Iso-tRNA, kasA, MurA, rho, rpoB, rpoC, S10p, and S12p) antibiotic targets in susceptible species were identified in the genome of Clostridium perfringens IRMC2505A using k-mer-based detection methods for AMR genes (Table 1). Three genes (GdpD, MprF, and PgsA) involved in protein-altering cell wall charge conferring antibiotic resistance were also identified in the IRMC2505A genome. Only one gene in each category was associated with the resistant mechanism of the isolate Clostridium perfringens: antibiotic inactivation enzyme (NimB), antibiotic target protection protein (TetB(P)), antibiotic target replacement protein (fabV), efflux pump conferring antibiotic resistance (TetA(P)), gene conferring resistance via absence (gidB), and protein involved in antibiotic sequestration (FabK-like). The putative membrane protein (mprF) is similar to the genome of the comprehensive antibiotic resistance database (E-value = 0.0). Tetracycline-resistant MFS efflux pump (TetA(P)) is similar to the genome reported in the NDARO, CARD, and PATRIC databases (E-value = 1 × 10 −234 ) ( Figure 3). We compared regional views of the putative membrane protein, tetracycline-resistant MFS efflux pump, and tetracycline-resistant ribosomal protection type genes in the IRMC2505A genome using the PATRIC cross-genus families (PGfams) method with reference genomes confirming similar subject coverage, query coverage, and identity (Table 2 and Figure 3). Seventeen virulence factors were identified in the Clostridium perfringens IRMC2505A genome (Table 3). More than 50% of the virulence factors are exoenzymes. Using the CARD and VFDB databases, metagenomic read mapping against the template genome revealed significant (p-value = 1 × 10 −26 ) genes with aligned reads against antibiotic-resistant genes or virulence factors, including phospholipase C, perfringolysin O, collagenase, hyaluronidase, alpha-clostripain, exo-alpha-sialidase, and sialidase activity (Table 4).

Discussion
Although many food-borne diseases are surveyed and controlled, and several enteric organisms are monitored through databases, Clostridium perfingens remains one of the only toxicogenic clostridial species identified by local investigations after an outbreak occurs. The bacterium is reported worldwide and has been listed among the leading species of foodborne and non-foodborne gastroenteritis [11]. Although a serious threat is imposed by the rise of multi-drug resistant (MDR)/toxigenic strains, little is known about the prevalence and genetic composition of this pathogen in the Middle East. To our knowledge, the few studies conducted in Saudi Arabia are limited to livestock and raw animal meats. None of them involve whole genome sequencing for species identification or studying antimicrobial resistance and toxic behavioral patterns in clostridial isolates from humans [12,13].
This genomic study of C. perfingens was isolated from a patient in Saudi Arabia's Eastern Province. It would be beneficial to start a surveillance program and elucidate changes in the pathogen's genome. This information is clinically significant since the antimicrobial resistance profile and genome characteristics of these isolates are still understudied in Saudi Arabia. Additionally, recent studies have identified hypervirulent clostridial spores in new environments and genomically linked them to new matrices other than hospital environments due to uncontrolled antimicrobial usage [14,15]. Our findings may also provide insight into the adaptation of these bacteria to new niches.
We believe that a better understanding of C. perfingens resistance mechanisms could decrease the misuse of broad-spectrum antibiotics and AMR genes' potential transfer to other pathogens.

Phylogenetic Analyses
Phylogenetic analysis confirmed the phylotype and found 100% similarity between our isolate sequences, C. perfingens ATCC 13124 195103.10, and C. perfringens str. 13 195102.6 species. The last species, C. perfringens str. 13, was sequenced for the first time in 2002 by Shimizu et al. It is a natural isolate found in the soil and classified as a type A strain [16]. Genome annotation distribution in IRMC2505A showed that the average genome length, GC content, and rRNA copy number were 4,322,225 bp, 28.28%, and 7, respectively. This anaerobic organism had several genes belonging to metabolism (523), protein processing (238), virulence (100), cellular processes (179), energy (135), and DNA and RNA processing (107,61). A recent study claimed that C. perfringens has the highest genome plasticity (by calculating the openness index (OI)) out of 51 other bacterial species. In other words, this bacterium has a highly open genome to acquire new adaptive genes for survival in hostile external environments. These properties are reflected by its low GC content (around 28%) and high 16S rRNA gene copy number [17].

Antimicrobial Resistance
Many research studies have indicated the emergence of multi-resistant strains of clostridial species and their significant risk to humans. In this study, the antimicrobial resistance profile of C. perfingens showed a multi-drug-resistant isolate involving nineteen resistance genes with different resistance mechanisms. Most of these genes were in the "antibiotic target in susceptible species" category. Many of these identical genes, with their antimicrobial resistance mechanisms, have been identified and described in different gram-negative bacteria where homologous combinations are found [18].
Predicting AMR genes based on NDARO/CARD/PATRIC databases showed tetracycline resistance (% query coverage and % identity > 97%) and defensin-like cationic antimicrobial peptides (CAMPs) (% query coverage = 100% and % identity = 95%). We detected two tetracycline-resistant genes, tetA(P) and tetB(P), which are known to encode proteins that mediate active tetracycline efflux and ribosomal protection tetracycline-resistant mechanisms, respectively [19]. The anti-defensins gene (mprF) reported codes for multiple peptide resistance factor proteins, which are involved in CAMP resistance [20]. Moreover, elfamycin-resistant genes were also detected with high query coverage (96%) but displayed an identity of 83%. These proteins are known as inhibitors of the Elongation Factor-Thermo Unstable (EF-Tu) used in the cell to translate mRNA transcripts into proteins [21].
Antimicrobial agents are widely used in animals as growth promoters and can lead to the emergence of antimicrobial-resistant enteric microflora, including C. perfringens and its spores [22]. Hence, antibiotic-resistant C. perfringens should be investigated in many matrixes (animal, human, food, soil . . . ) [23]. Motamedi et al. conducted the first systematic review in 2021 on C. perfringens as a causative agent of antibiotic-associated diarrhea (AAD) in hospitalized patients [24]. These studies were conducted mainly in Asia, Europe, and Africa and showed 14.9% C. perfringens among patients. However, insufficient data were found in these areas. Consequently, this finding stresses the urgent need for a surveillance program to control AAD and guarantee efficient treatment.
A similar situation was observed in Saudi Arabia, where a recent meta-analysis (2022) on the frequency of antibiotic-resistant Clostridium species found no evidence for the antimicrobial resistance patterns reported in KSA. Most of these limited studies used conventional methods (selective culture on agar media, E-tests, and disk diffusion assays) [25]. These reference methods only isolate and identify the studied bacteria and do not distinguish between different pathogens based on their virulence phenotypes. Limited research studies in KSA investigated the resistance and occurrence of C. perfringens in camel meat. They showed these bacteria's high resistance to tetracycline and ceftiofur drugs and, to a lesser extent, other antimicrobial agents (erythromycin, oxytetracycline, penicillin, metronidazole, clindamycin, and lincomycin) [12,26,27].

Toxigenicity
C. perfringens is also known for its toxigenicity. Five toxinotypes (A-G) were historically identified based on the toxin produced by the strain (alpha, beta, epsilon, and iota toxins). Afterward, two new toxinotypes (F and G) were added to the scheme, introducing enterotoxin CPE (but not CPB, ETX, or ITX) and NetB toxin, respectively [28]. Many virulence factors contribute to this pathogen's toxicity. C. perfringens is known to produce a wide range of degradative enzymes, including hydrolytic enzymes, membrane-damaging enzymes, pore-forming toxins, and intracellular toxins (proteases, hyaluronidase, collagenase, endoglycosidases, etc.). The virulence analysis of our bacterial isolate showed a variable array of virulence genes and a 97% similarity to the C. perfringens genome. High query coverage sequences (≥97%) were distributed as follows: Sequences correlated to phospholipase C (plc) encoding the alpha toxin, sialidase genes (nanH, nanI), alpha-clostripain genes (cloSI), hyaluronidase genes (nagJ, nagH), and collagenase genes (colA) encoding the κ-toxin. These genes are chromosomal and located specifically in the conserved region of the C. perfringens chromosome [29]. There has been no WGS study on C. perfringens in Saudi Arabia, including the Arabian Peninsula [30].
In conclusion, this WGS study of C. perfringens showed an MDR bacterium with several virulence factors. To our knowledge, our WGS study is one of the first studies conducted in Saudi Arabia and the Arabian Peninsula on C. perfringens, where national databases are still inexistent. To guarantee a complete understanding of C. perfringens' epidemiology, a full study of this pathogen's virulence factors, and antimicrobial resistance patterns is essential. Future studies will also be beneficial in developing an effective disease control strategy.

Isolation and DNA Extraction of C. perfringens IRMC2505A
This project was submitted for ethical approval to the ethical committee of Imam Abdulrahman Bin Faisal University (IRB-2022-01-398). All procedures were performed in accordance with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
A Clostridium perfringens strain was isolated using cycloserine, cefoxitin, and fructose agar medium from the fecal sample of a patient who was clinically suspected of gastrointestinal infection and had a recent history of antibiotic exposure and diarrhea. A positive sample, tested for both toxins A and B by GeneXpert, was cultured on cycloserine cefoxitin fructose agar selective medium (CCFA) (MOLEQULE-ON, Auckland, New Zealand) under anaerobic conditions at 37 • C for 24 h. The total DNA was extracted using the Gentra Puregene Yeast/Bact. Kit (Qiagen, Hilden, Germany). The purity, quality, and quantity of genomic DNA were measured using Nanodrop 2000 (Thermo Scientific, Waltham, MA, USA) as described in the manufacturer's instructions. The isolate was PCR amplified, and 16S rRNA was sequenced to confirm the isolate.

Virulence Genes Identification Using Whole Genome Sequencing
The whole genome of isolate IRMC2505A was sequenced using the Illumina platform. The paired reads were assembled and annotated using PATRIC (BV-BRC 3.28.5) [31] and the RAST tool kit (RASTtk 1.3.0) [32]. We calculated the number of contigs and average G + C content and identified the taxonomy of the genome. We also predicted the proteins and their functional assignments with the Enzyme Commission (EC) [33], Gene Ontology (GO) [34], pathways [35], protein family types [36], and subsystems of protein complexes [37]. Specialty genes in the genome of IRMC2505A were identified using specific source databases for known transporters [38], virulence factors [39,40], drug targets [41,42], and antibioticresistant genes [43]. We detected k-mer-based antimicrobial resistance (AMR) genes [31]. We phylogenetically analyzed the IRMC2505A genome using 100 genes from the National Center for Biotechnology Information (NCBI) reference and representative genomes by Mash/MinHash (Mash v2.3) and PGFams (PATRIC, BV-BRC 3.28.5) [36] aligned with MUSCLE [44], a matrix analysis using RaxML (version 8) with fast bootstrapping [45,46]. Metagenomic read mapping against the selected template was analyzed using the CARD (2020) and VFDB (2019) databases through k-mer alignment [47].

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/toxins15060359/s1, Figure S1: A: A circular graphical display of the distribution of the genome annotations in IRMC2505A. B: An overview of the subsystems of genes for the genome IRMC2505A; Figure S2: Proteome comparison. List of tracks, from outside to inside: Clostridium perfringens ATCC 13124 (195103.10) and Clostridium perfringens IRMC2505A (1502.2785); Table S1: Assembly Details and annotated features of Clostridium perfringens IRMC2505A; Table S2: List of antimicrobial resistance genes identified in the genome of Clostridium perfringens IRMC2505A; Table S3: Functions and classifications of list of antimicrobial resistance genes in the genome of C. perfringens IRMC2505A; Table S4: Gene families are ranked by alignment score for phylogenetic tree of IRMC2505A genome.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data will be available on reasonable request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.