Genome characteristics of the optrA-positive Clostridium perfringens strain QHY-2 carrying a novel plasmid type

Abstract

ABSTRACT

Clostridium perfringens is a bacterial species of importance to both public and animal health. The gene optrA is the first gene that confers resistance to the tedizolid, a last-resort antimicrobial agent in human medicine. Herein, we whole-genome sequenced and analyzed one optrA-positive C. perfringens strain QHY-2 from Tibetan sheep in Qinghai province and identified one optrA plasmid pQHY-2. The plasmid shared similar structure with the optrA-positive plasmids p2C45 and p21-D-5b previously identified in C. perfringens, demonstrating the potential horizontal transmission of the optrA plasmids among C. perfringens strains. Annotation of the optrA-positive plasmids showed optrA and erm(A) located on a segment flanked by IS element IS1216E, and fexA, optrA, and erm(A) located on a segment flanked by IS element ISVlu1, which revealed the possible dissemination mechanism. Additionally, a Tn6218-like transposon carrying aac(6′)-aph(2″) and erm(B) was also detected on pQHY-2, demonstrating the transposition of Tn6218 and spread of antibiotic resistance among Clostridium bacteria. Molecular analysis indicated the optrA-positive plasmids belonged to a plasmid type distinct from the pCW3-like plasmids, pCP13-like plasmids, or pIP404-like plasmids. Further structure analysis showed they might be formed by inserting segments into plasmid pCPCPI53k-r1_1, which coexist with two pCW3-like plasmids and one pCP13-like plasmid in C. perfringens strain CPI 53k-r1 isolated from a healthy human in Finland.

IMPORTANCE

Antimicrobial resistance is now a global concern posing threats to food safety and public health. The pCW3-like plasmids can encode several main toxin genes and three antibiotic resistance genes (ARGs), including tetA(P), tetB(P), and erm(B), which used to be considered as the main carrier of ARGs in Clostridium perfringens. In this study, we found the optrA plasmids, which belonged to a novel plasmid type, could also harbor many other ARGs, indicating this type of plasmid might be the potential repository of ARGs in C. perfringens. Additionally, this type of plasmid could coexist with the pCW3-like plasmids and pCP13-like plasmids that encoded toxin genes associated with gastrointestinal diseases, which showed the potential threat to public health.

INTRODUCTION

Antimicrobial resistance (AMR) is currently one of the important threats to global health as well as food safety (1). The oxazolidinones are last-resort antimicrobial agents used for the treatment of severe infections in humans caused by multidrug-resistant (MDR) Gram-positive bacteria. Linezolid, the first fully synthesized oxazolidinone antibiotic, was approved by the US Food and Drug Administration for clinical use in 2000 (2). However, linezolid-resistant Enterococcus and Staphylococcus isolates have been reported since the year 2000 (3). The first reported linezolid resistance gene, cfr, encodes an RNA methyltransferase that modifies the adenine residue at position 2,503 of the 23S rRNA gene and thereby confers combined resistance to phenicols, lincosamides, linezolid, pleuromutilins, and streptogramin A antibiotics (4).

The OptrA protein, an ABC-F protein encoded by the optrA gene, mediates resistance against phenicols and oxazolidinones, by protecting the ribosomes of the bacteria. Compared to cfr gene, optrA can not only mediate resistance against linezolid but also tedizolid (5). Clostridium perfringens is a Gram-positive, anaerobic, rod-shaped bacterium, which is widely disseminated in natural environments (6). As an opportunistic pathogen, C. perfringens may be present in the intestinal tract of humans and animals, without causing disease (7). When the intestinal environment is disturbed, C. perfringens can multiply rapidly and secrete over 20 protein toxins, presenting a huge threat to livestock breeding and human public health and safety (8). C. perfringens can cause various animal gastrointestinal diseases, including necrotizing enteritis and diarrhea, which lead to huge economic losses in livestock breeding each year (9, 10). Moreover, C. perfringens strains can transmit from animals to humans through the food chain. Food poisoning caused by C. perfringens is one of the leading foodborne illnesses globally, being the second most common cause of foodborne illness outbreaks in the USA and South Korea, and the third most common in the UK and France (11).

Plasmids are important for the virulence of Clostridial species (12). Based on the type of replication initiator, the plasmids of C. perfringens are categorized into three broad families, consisting of the pCW3-like plasmids, pCP13-like plasmids, and pIP404-like plasmids. The size of pCW3-like plasmids varies from ~45 kb to ~135 kb, while the size of pCP13-like plasmids varies from ~36 kb to ~58 kb (13). However, the size of pIP404-like plasmids is smaller (~10 kb), and the plasmid structure is relatively simple (14). It is reported that the pCW3-like plasmids can harbor various toxin genes including β-toxin coding gene cpb, ɛ-toxin coding gene etx, and C. perfringens enterotoxin coding gene cpe (15, 16). Moreover, some antibiotic resistance genes (ARGs) [tetA(P), tetB(P), and erm(B)] can also locate on the pCW3-like plasmids (17, 18), which may lead to the co-spread of toxin genes and ARGs. The transfer of pCW3-like plasmids is based on the Tcp locus, composed of origin of transfer (oriT), tcpA to tcpJ, tcpM, and tcpK (19). Compared with pCW3-like plasmids, less toxin genes (cons.cpb2, becA, and becB) are known to locate on the pCP13-like plasmids, and no ARGs have been detected on them (20). The transfer of pCP13-like plasmids is based on the Pcp locus, an ~2.7-kb sequence segment consisting of pcpA to pcpT (21).

This study aimed to provide insights into the genetic characteristics of the optrA gene and the optrA plasmids in C. perfringens, as well as the genomic content of C. perfringens that harbor these plasmids. Herein, we whole-genome sequenced and analyzed the optr-positive C. perfringens strain QHY-2 isolated from Tibetan sheep in Qinghai province, compared it with other strains, including those harboring similar plasmids, and further investigated the structure and transferability of the optrA plasmids.

RESULTS

WGS analysis

After polymerase chain reaction (PCR) assays, agarose gel electrophoresis, and sequence alignment, QHY-2 was found to be positive for both phenicols-resistant gene fexA and oxazolidinones/phenicols-resistant gene optrA. Besides florfenicol, QHY-2 also showed strong resistance to chloramphenicol (MIC >64 µg/mL) and linezolid (MIC = 32 µg/mL). Whole-genome sequencing (WGS; NCBI BioSample: SAMN33306092) resulted in a total length of 3,528,159 bp, consisting of the chromosome sequence (3,361,707 bp) and an optrA-positive plasmid (166,452 bp, designated as pQHY-2). Eight types of IS elements were detected in the genome, where ISCpe5, ISCpe2, IS1470, and ISCpe4 were detected on the chromosome, while ISVlu1, IS256, IS1216E, and ISCldi2 were detected on plasmid pQHY-2. Nine ARGs were also detected in QHY-2, in which optrA, fexA, aminoglycoside-resistant genes aac(6′)-aph(2″), and lincosamide/macrolide-resistant genes erm(Q), erm(B), and erm(A) located on plasmid pQHY-2, while aminoglycoside-resistant genes ant (6)-Ib and tetracycline-resistant genes tet (22) and tetA(P) were identified on the chromosome. In addition, 10 toxin genes, including phospholipase C coding gene plc, alpha-clostripain coding gene cloSI, perfringolysin O coding gene pfoA, collagenase coding gene colA, hyaluronidase coding genes nagH, nagI, and nagK, and exo-alpha-sialidase coding genes nanI and nanJ, were also detected on the chromosome.

Toxin typing indicated most of the C. perfringens isolates recovered from food animals in China belonged to C. perfringens type A, and many toxin genes, including the cpe gene linked to human foodborne gastroenteritis, were detected in them. In the phylogenetic tree, QHY-2 clustered together with another Tibetan sheep C. perfringens type A strain, QHY-18, from Qinghai province and four goat C. perfringens type D strains (21-D-1, 21-D-2, 21-D-3, and 21-D-4) from Shaanxi province (Fig. 1). Interestingly, the another optrA-positive C. perfringens type D strain (21-D-5) isolated from the same source as the strains 21-D-1, 21-D-2, 21-D-3, and 21-D-4, did not cluster within this clade, showing a certain genetic diversity of C. perfringens from the same source. The C. perfringens isolates from pigs appeared to be distinct from the other isolates in the cluster heatmap based on toxin gene profiles, particularly clade I, which was mainly composed of the C. perfringens isolates from pigs (Fig. S1). Moreover, 94.1% (16/17) of the beta-2 toxin coding gene con.cpb2 was detected in the C. perfringens isolates from pigs (Fig. S2), suggesting that the porcine species might be an important reservoir of C. perfringens isolates containing this virulence marker.

Fig 1

Phylogenetic tree based on the SNPs of the 91 C. perfringens strains from food animals in China. The branch in which QHY-2 locates on is marked with green shadow. The three optrA positive C. perfringens strains were marked with red stars. C. perfringens type A, D, and G strains were marked with green, red, and blue squares, respectively. SNP, single-nucleotide polymorphism.

Structure of the optrA plasmids

The optrA gene on plasmid pQHY-2 (Fig. 2a) shared a 99.9% sequence identity with the optrA gene initially detected on Enterococcus faecalis plasmid pE349 (GenBank accession number KP399637) (23). The only polymorphism occurred at base 1,387 (from A to C), which leaded to the change of amino acid at position 463 from threonine to proline. Moreover, the respective segments containing optrA and fexA were both flanked by ISVlu1 on plasmids pE394 and pQHY-2, indicating the possible horizontal transfer of optrA and fexA from E. faecalis to C. perfringens, vice versa. Sequence alignments between pQHY-2 and the other two optrA-positive plasmids (p2C45 and p21-D-5b) previously reported in C. perfringens showed pQHY-2, p2C45, and p21-D-5b shared similar plasmid structure and high sequence identity (Fig. 2b), demonstrating the possible transmission of the optrA plasmids among C. perfringens isolates (24, 25). Some genetic elements that were homologous to parts of the clade 3 pathogenicity locus (PaLoc) insertion in C. difficile were assigned the designation Tn6218, and Tn6218 transposons were mainly detected in C. difficile strains (26, 27). In this study, one Tn6218-like transposon that shared high sequence identities with the Tn6218 transposon in C. difficile strain Ox42 (BioSample: SAMEA2052297) and the Tn6218-like transposon in C. perfringens strain 21-D-5 (BioSample: SAMN26351020) was detected on plasmid pQHY-2. Additionally, ARGs aac(6′)-aph(2″) and erm(B) were detected on the transposons (Fig. 2c), indicating the dissemination of Tn6218 and associated antibiotic resistance among Clostridium bacteria.

Fig 2

Character of the optrA-positive plasmids. (a) Structure of plasmid pQHY-2. (b) Sequence alignments among plasmids p2C45, pQHY-2, and p21-D-5b. Arrows indicate the directions of transcription of the genes. Tn6218-like transposons on plasmids pQHY-2 and p21-D-5b were marked with pink shadow. (c) Sequence analysis of Tn6218 transposon in C. difficile Ox42 and C. perfringens pQHY-2 and p21-D-5b. Arrows indicate the directions of transcription of the genes.

Phylogenetic analysis of the optrA plasmids

A previous study indicated that the optrA plasmid p2C45 could not be conjugated (25). In our study, both conjugation and electrotrans-formation experiments using QHY-2 and 21-D-5 as donor and C. perfringens isolates with low MICs (<8 µg/mL) to florfenicol showed unsuccessful transfer of pQHY-2 and p21-D-5b. To investigate the molecular mechanisms of these optrA-positive plasmids, we performed sequence alignment on the replication initiator coding genes and phylogenetic analysis. These optrA-positive plasmids all harbored two replication initiator coding genes, designated as repA and repB (Fig. 2c). Genes repA and repB on the optrA plasmids shared 100% sequence identity, respectively, indicating they belonged to the same plasmid type. However, repA and repB were not detectable on the pCW3-like plasmids, pCP13-like plasmids, or the smaller pIP404-like plasmids (Fig. S3). Then, a total of 50 C. perfringens plasmids (Table 1) consisting of the 3 optrA plasmids, 22 pCW3-like plasmids, and 25 pCP13-like plasmids were selected to construct a phylogenetic tree, using MAFFT and PHYLIP. Various toxin genes (cpe, etx, netB, etc.) and some ARGs including tetA(P), tetB(P), and erm(B) were detected on the pCW3-like plasmids, while three toxin genes, including cons.cpb2 and the binary enterotoxin BEC coding genes becA and becB (28), were detectable on the pCP13-like plasmids. Compared to the pCW3-like plasmids and pCP13-like plasmids, the optrA plasmids harbored more ARGs, but no toxin genes, and neither the Tcp conjugal transfer locus nor Pcp conjugal transfer locus was detected on them. Moreover, the phylogenetic tree showed the optrA-positive plasmids clustered together on their own distinct clade, apart from the pCW3-like plasmids and pCP13-like plasmids (Fig. 3), indicating the optrA-positive plasmids belonged to a novel type of plasmid that differed from the smaller pIP404-like plasmids and the transferable pCW3-like plasmids and pCP13-like plasmids.

Fig 3

Phylogenetic tree of the C. perfringens plasmids. The pCW3-like plasmids and pCP13-like plasmids were marked with blue and orange shadows, respectively. The branch where optrA-positive plasmids (p2C45, p21-D-5, and pQHY-2) locate on was marked with red line. Toxin genes and ARGs were marked with red and green squares, respectively.

TABLE 1

Information of the plasmids used to construct the phylogenetic tree

Type	Plasmids	Accession	Size (bp)	Toxin genes	ARGs
pCW3-like plasmids	pCW3	DQ366035.1	47,263	–	tetA(P) tetB(P)
	p21-D-5a	GCA_022808975.1	121,548	cpe etx aty.cpb2	–
	pJIR4150	LN835295.1	89,692	tpeL	–
	pJFP55F	CP013041.1	72,549	–	–
	pCPT84-1	MK285057.1	72,052	iap ibp	–
	pJIR3537	CP025504.1	48,779	–	tetA(P) tetB(P)
	pD13122_cpe	MG456815.1	47,758	cpe aty.cpb2	–
	pLLY_N11_3	CP023413.1	72,060	aty.cpb2	tetA(P) tetB(P)
	pCBA7123-01	AP017631.1	46,640	aty.cpb2	tetA(P) tetB(P)
	pFORC3	CP009558.1	56,577	aty.cpb2	tetA(P) tetB(P)
	pCPC216_1	CP075974.1	71,757	cpe	–
	pJIR3535	CP025502.1	81,826	netB	–
	pDel1-1	CP019577.1	82,596	netB	–
	pCPPB-1	AB604032.1	67,479	cpe iap ibp	–
	pCPF4969	AB236336.1	70,480	cpe	–
	pJIR3844	CP025503.1	69,606	aty.cpb2	–
	pCpb2-CP1	JQ655732.1	65,875	aty.cpb2	erm(B)
	pCPAAD1903_1	CP075993.1	49,937	cpe aty.cpb2	–
	pCPAAD1527_2	CP076002.1	75,262	cpe aty.cpb2	–
	pJFP838D	CP013039.1	48,597	cpe aty.cpb2	–
	pCPF5603	AB236337.1	75,268	cpe aty.cpb2	–
	pCP8533etx	AB444205.1	64,753	etx aty.cpb2	–
optrA-positive plasmids	p2C45	NZ_JAAQTM010000004.1	148,618	–	optrA erm(A) fexA lnu(P)
	p21-D-5b	GCA_022808975.1	224,659	–	erm(Q) erm(B) optrA erm(A) fexA lnu(P) aac(6′)-aph(2″)
	pQHY-2	CP118266.1	166,452	–	erm(Q) erm(B) optrA erm(A) fexA aac(6′)-aph(2″)
pCP13-like plasmids	pCP310-85	CP075969.1	57,459	cons.cpb2	–
	pCPT1	MK285059.1	51,991	cons.cpb2	–
	pCPCPM77b_1	CP075909.1	53,676	cons.cpb2	–
	pCPCPI18-1_2	CP075983.1	53,677	cons.cpb2	–
	pCPC269_2	CP075972.1	54,958	cons.cpb2	–
	pCPAAD1900a_2	CP075999.1	54,311	cons.cpb2	–
	pCP13	AP003515.1	54,310	cons.cpb2	–
	pCP-OS1	AP013033.1	54,535	becA becB	–
	pCP-TS1	AP013034.1	54,478	becA becB	–
	pJUM001-1	AP026871.1	49,289	–	–
	pCPAAD1527_1	CP076001.1	58,584	–	–
	pCPCPI63K-r5_2	CP075929.1	53,677	–	–
	pCP318-46	CP075967.1	56,620	–	–
	pCPNY83906550-1	MK285071.1	61,240	–	–
	pMFDS1012647.1	CP106927.1	61,476	–	–
	pCPCPLi6-1_2	CP075914.1	53,677	cons.cpb2	–
	pCPCPLi3-1_2	CP075919.1	58,584	–	–
	pCPC746_1	CP075951.1	61,154	–	–
	pCP497-86_1	CP075962.1	56,257	cons.cpb2	–
	pCPCPI103K-3_2	CP075990.1	58,584	–	–
	pCPCPI75-1_2	CP075924.1	62,433	–	–
	pCPCPI53k-r1_4	CP075938.1	63,715	–	–
	p1	MK275619.1	58,796	cons.cpb2	–
	pBCNF5603	AB189671.1	36,695	–	–
	pCP1293-86_1	CP076005.1	59,188	–	–

^a

–, no characterized toxin genes/ARGs.

Possible formation mechanism of pQHY-2

To investigate the possible mechanisms of plasmid pQHY-2, we performed sequence blast on pQHY-2 and the other C. perfringens plasmids available in the National Center for Biotechnology Information (NCBI) database and fortunately found an ARGs-negative plasmid named pCPCPI53k-r1_1 (CP075935.1), which was identified in C. perfringens strain CPI 53k-r1 isolated from healthy human in Finland (29). Sequence analysis of plasmids pQHY-2 and pCPCPI53k-r1_1 indicated that they shared similar plasmid structure and 78% sequence identity, and both harbored the replication initiator coding genes repA and repB (Fig. S4), demonstrating pCPCPI53k-r1_1 and the optrA-positive plasmids belonged to the same plasmid type. Compared to pQHY-2, there were two segment (segment I and segment II) deletions in pCPCPI53k-r1_1. Segment I (ISVlu1-hp-fexA-IS1216E-optrA-hp-erm(A)-IS1216E-hp-hp-ISVlu1) encoded optrA, fexA, and erm(A), while segment II (hp-hp-hp-erm(Q)-▲Tn6218-hp) encoded erm(Q), erm(B), and aac(6′)-aph(2″). Therefore, we speculate that the optrA plasmids were formed through inserting segment I and segment II into pCPCPI53k-r1_1, based on the transfer ability of IS elements and Tn6218 transposon (Fig. 4).

Fig 4

Speculative formation process of plasmid pQHY-2. Segment deletions on the plasmids were represented by blue shadows.

DISCUSSION

The bacterium C. perfringens is of importance to animal and human health, as well as food safety (30). Frequently found in food system environments, C. perfringens presents a risk to food animal health and may cross-contaminate associated ingredients or food products, with the potential to cause sporadic outbreaks of disease in human populations, including gastroenteric illness (31). An important pathogenic bacterium, C. perfringens possesses several toxin genes and generates a number of dangerous toxins, which lead to huge economic losses in domestic animals and the food industry each year (9, 32). Moreover, C. perfringens associated with dairy farm systems show diverse genotypes and contain various virulence makers (16). It therefore remains important to maintain surveillance and toxin genes profiling of C. perfringens in farms, foods, and environments, to protect humans and animals from C. perfringens infections.

Historically, the use of antibiotics as growth promoters for food producing animals was widespread globally, and high amounts of antimicrobial agents were used for this purpose, in addition to treating bacterial diseases in livestock farming (33). A recognition of the growing problems surrounding AMR and public health has since led to tighter controls around antimicrobial stewardship (34). AMR in bacteria, especially foodborne pathogenic bacteria, still poses a serious threat to public health security. Among the C. perfringens strains isolated from humans, food animals, and foods worldwide, strong resistance against sulfonamides has been observed, while moderate resistance against tetracycline, penicillin, lincomycin, and clindamycin was also noted. However, resistance was rarely observed against doxycycline, florfenicol, and linezolid (35 - 39).

In this study, we whole-genome sequenced and analyzed the optrA-positive C. perfringens strain QHY-2 isolated from Tibetan sheep in Qinghai province. Ten toxin genes were detected in QHY-2, and diverse toxin genes were observed among the C. perfringens strains from food animals in China. Among them, plc, cloSI, and colA were detected in all C. perfringens isolates, while the other toxin genes, including beta2 toxin coding gene cons.cpb2, C. perfringens enterotoxin coding gene cpe, and epsilon toxin coding gene etx, were only detected in some of the C. perfringens genomes. The complex virulence makers among C. perfringens isolates from food animals indicated standardized feeding management should be strengthened to prevent the transmission and pathogenicity of them. Moreover, the toxin gene repertories of C. perfringens varied with the animal species source, suggesting targeted preventive measures should be undertaken to limit C. perfringens-associated diseases. Besides the toxin genes, various ARGs were also observed among the C. perfringens strains. Among them, tetA(P) was found to have the highest prevalence rate (94.5%, 86/91), followed by tetB(P) (64.8%, 59/91), tet (22) (37.4%, 34/91), ant (6)-Ib (35.2%, 32/91), erm(Q) (35.2%, 32/91), and lnu(P) (29.7%, 27/91) (Fig. S5). The high prevalence of tetracycline-resistant and lincomycin-resistant genes partly explained the mild resistance against tetracycline, lincomycin, and clindamycin. However, no sulfonamides-resistant genes were identified among the C. perfringens isolates. Further studies should be carried out to explain the strong resistance against sulfonamides in C. perfringens.

OptrA protein encoded by optrA gene can mediate resistance against oxazolidinones (linezolid and tedizolid), the last-resort antimicrobial agents used for the treatment of severe MDR Gram-positive bacterial infections (40). After the discovery of optrA gene in Enterococcus faecalis strain E349 in 2015, it has been detected among the Gram-positive Enterococcus and Staphylococcus isolates (41 - 43), and the Gram-negative Campylobacter isolates from humans, hospital environments, foods, and animals in China (22, 44). Prevalence of optrA may pose challenge to the treatment of bacterial infectious diseases and potential co-selection of other ARGs, which can cause huge threats to public health and safety. In this study, we found the optrA genes in C. perfringens shared similar gene environments with the optrA genes in Enterococcus, Staphylococcus, and Campylobacter strains, indicating the spread of optrA gene among bacterial strains. Additionally, the presence of other plasmid-borne resistance genes fexA, erm(Q), aac(6′)-aph(2″), erm(B), and lnu(P), particularly erm(A) and fexA that are located on a segment flanked by IS element ISVlu1 or IS1216E together with optrA, may contribute to the co-selection of optrA. C. perfringens is one of the main causes of gas gangrene and foodborne diseases in humans (45), which can transfer between food animals and humans through animal products. AMR of C. perfringens from food animals can pose serious threats to food safety and public health. The various ARGs in C. perfringens strains from food animals in China suggested that the rational use of antibiotics should be strengthened to prevent the increase and spread of AMR.

Many important toxin genes and ARGs that are crucial to the pathology of C. perfringens are located on plasmids, and so these genetic elements play important roles in the cases of C. perfringens infection. According to the type of replication initiator, the plasmids in C. perfringens can be categorized into pCW3-like plasmids, pCP13-like plasmids, and pIP404-like plasmids (21). In this study, we identified one optrA-positive plasmid pQHY-2 in C. perfringens strain QHY-2 from Tibetan sheep in Qinghai province and further compared it with the other two optrA-positive plasmids p2C45 and p21-D-5 previously identified in C. perfringens strains 2C45 (BioSample: SAMN18155071) and 21-D-5 (BioSample: SAMN28005067) isolated from chicken in Shanxi province and goat in Shaanxi province, respectively. Molecular analysis indicated that the optrA-positive plasmids of C. perfringens belonged to a plasmid type, with different replication initiator coding genes, larger plasmid size, and more ARGs that differed from pCW3-like plasmids, pCP13-like plasmids, or pIP404-like plasmids. Further sequence and structure analysis revealed that the optrA-positive plasmids might be formed by inserting segments into the ARGs-negative plasmid pCPCPI53k-r1_1, which belonged to the same plasmid type with the three optrA-positive plasmids. Additionally, pCPCPI53k-r1_1 exists in the C. perfringens strain CPI 53k-r1, together with two pCW3-like plasmids (pCPCPI53k-r1_2 and pCPCPI53k-r1_3), one pCP13-like plasmid pCPCPI53k-r1_4, and two small plasmids (pCPCPI53k-r1_5 and pCPCPI53k-r1_6) harboring bacteriocin coding gene bcn (Fig. S6), demonstrating the possible coexistence of the optrA-positive plasmids and the pCW3-like plasmids and pCP13-like plasmids in C. perfringens strains.

Although the optrA plasmids have not been successfully transferred to date, the dissemination of similar homologous plasmids across multiple strains from humans and animals in different regions suggests a possible horizontal transfer of the ARGs among C. perfringens strains, especially those harbor other toxin plasmids, which poses a potential threat to food safety and public health. In conclusion, this study indicates that C. perfringens isolates from food animals in China harbored various virulence makers and ARGs and supports the need for effective regulations and measures to control the dissemination of antibiotic resistance of C. perfringens, especially those strains encoding key genetic markers linked to severe disease.

MATERIALS AND METHODS

Source of isolate and detection of florfenicol-resistant genes

Eighty fecal samples were collected from Tibetan sheep grazed in Qinghai province, China, in 2020. Thirty-six strains of C. perfringens were eventually recovered from the samples after isolation and molecular-based identification, and antibiotic sensitivity testing was then performed on the isolates. Interestingly, we found one C. perfringens designated as QHY-2 showed strong resistance against florfenicol (MIC = 32 µg/mL) (46), which was rarely observed in C. perfringens. To investigate the causes of strong resistance against florfenicol in QHY-2, in this study, we screened for the florfenicol-resistant genes floR, fexA, fexB, cfr, and optrA with PCR assays, as described in the previous reports (47 - 49).

Whole-genome sequencing and analysis

Genome DNA of QHY-2 was extracted using TIANamp Bacteria DNA Kit (Tiangen, Beijing, China). WGS of the strain QHY-2 was performed with the Nanopore PromethION platform (Biomarker Technologies, Beijing, China). The sequences were assembled with Canu v1.5. Pilon v1.22 was used to improve the draft genome assemblies by correcting bases. The WGS annotations were designed with the Rapid Annotations using Subsystem Technology annotation pipeline (version 2.0) (http://rast.nmpdr.org/) and Prokka (50). AMR and toxin genes were identified by Abricate and standalone BLAST analysis. Conjugative locus of the plasmids was identified by local blast, Tcp locus, and Pcp locus coding genes were used as the BLAST database (18, 21). Nucleotide sequence visualization and comparison were realized through BRIG and Easyfig. A total of 91 genomes of C. perfringens (Table S1), including QHY-2 and the other 90 C. perfringens isolates recovered from food animals in China were used to construct a phylogenetic tree based on single nucleotide polymorphisms, using REALPHY (https://realphy.unibas.ch/realphy/). A phylogenetic tree was visualized using the online tool iTOL (https://itol.embl.de/).

Transformation experiments

The optrA-positive plasmids were extracted and used to electrotransform into the recipient strain C. perfringens ATCC 13124. Three randomly selected C. perfringens isolates with low MICs (<8 µg/mL) to florfenicol were chosen as recipients for conjugation (25).