Isolation of Clostridium Perfringens type C from a Neonatal Yangtze Finless Porpoise (Neophocaena Asiaeorientalis Asiaeorientalis)

Jia Li Institute of Hydrobiology Chinese Academy of Sciences Richard William McLaughlin Gateway Technical College Yingli Liu Institute of Hydrobiology Chinese Academy of Sciences Junying Zhou Wuhan University Zhongnan Hospital Xueying Hu Huazhong Agricultural University Xiaoling Wan Institute of Hydrobiology Chinese Academy of Sciences Haixia Xie Institute of Hydrobiology Chinese Academy of Sciences Yujiang Hao Institute of Hydrobiology Chinese Academy of Sciences Jinsong Zheng (  zhengjinsong@ihb.ac.cn ) Institute of Hydrobiology Chinese Academy of Sciences https://orcid.org/0000-0002-6541-1594


Introduction
The Yangtze nless porpoise (Neophocaena asiaeorientalis asiaeorientalis) is a freshwater subspecies of the narrow-ridged nless porpoise. The porpoise is endemic to the middle and lower reaches of the Yangtze River, as well as Poyang and Dongting Lakes in China (Committee on Taxonomy 2012; Gao and Zhou 1995). The population size of this animal has declined rapidly over several decades. Some of the reasons include illegal shing, over shing, pollution, boat transportation, dams and various other human activities in the Yangtze which are detrimental to the porpoise (Wang 2009;Zhao et al. 2008;Mei et al. 2012). As a result, by the end of 2017, the population has declined to approximately 1012 individuals (Huang et al. 2020). The Yangtze nless porpoise was recently upgraded to the First Order of Key Protected Animals in China. It is also currently listed in the International Union for Conservation of Nature Red Data Book as a critically endangered species (Wang et al. 2013).
Clostridium perfringens is an anaerobic bacterium which is capable of endospore formation. In both humans and animals, the bacterium can cause numerous diseases, in part as a result of its ability to secrete several different toxins (Guran and Oksuztepe 2013). C. perfringens is divided into ve different toxinotypes, which are type A, B, C, D, and E (Siqueira et al. 2012;Milton et al. 2017). This is based on the four major toxins, alpha, beta, epsilon, and iota which different strains of the bacterium can produce. All type C strains must produce both the alpha (CPA) and beta toxins (CPB) (Uzal et al. 2010). In addition, there are some type C isolates that also make enterotoxin (CPE) or beta2 toxin (CPB2) (Sayeed et al. 2008;Uzal et al. 2010). It has been suggested that the beta2 toxin could play a role in caprine enterotoxemia and could cause mortality in kids (Garcia et al. 2013;Dray 2004). Type C can frequently be the cause of enteritis and enterocolitis in neonatal animals, such as sheep, goats, and pigs (Songer 1996;Songer and Uzal 2005;Diab et al. 2012). There are also reports of disease caused by C. perfringens in cattle (Milton et al. 2017;Rahman et al. 2012). However, studies on aquatic mammal neonates are very rare.
A male Yangtze nless porpoise was born on May 19, 2018 in the Wuhan Baiji Dolphinarium, Hubei, China. This neonate died on June 14. During this period, the newborn baby was completely arti cially feeding. After birth, the head of the calf collided once with the wall of the pool, causing a wound on the right side of the head. Symptoms of obvious decreased appetite appeared before death. In this study, we examined the blowhole, lung, stomach and fecal samples from this neonatal Yangtze nless porpoise to assess the presence of pathogenic bacteria that could have contributed to its death. As a result, a total of three Clostridium perfringens type C strains was identi ed in the fecal samples and then further characterized for bio lm formation and antibiotic susceptibility.

Materials And Methods
Sample collection and isolation of Clostridium perfringens Five fecal samples were collected directly from the anus of the neonatal porpoise on days 2, 3, 7, 12 and 23 during arti cial feeding. All fecal samples were frozen at -20°C until used. Another fecal (rectal) sample was removed by dissection on day 27. Postmortem was performed within three hours of its death. Amies Media swabs (Copan, Italy) were used to obtain lung and blowhole samples. A sterile soft plastic tube with a diameter of 5 mm was inserted into the forestomach to draw gastric uid which was then placed in a sterile 1.5 ml tube. All sampling procedures used in this study were approved by the Regulations of the People's Republic of China for the Implementation of Wild Aquatic Animal Protection (promulgated in 1993), adhering to all ethical guidelines and legal requirements in China.
Lung and blowhole samples were inoculated directly onto Blood agar plates (Dijing LS0109, Guangzhou, China) and incubated aerobically at 37°C for 24-48 h. Bacteria from the fecal material and the gastric uid were inoculated onto both blood and MacConkey agar plates (DijingLS1009, Guangzhou, China).
The plates were incubated anaerobically at 37°C for 24-48 h. To obtain a pure culture, bacteria were subcultured twice onto Blood agar plates. No growth was observed on the MacConkey agar plates.
Bacteria identi cation based on 16S rRNA gene sequencing In order to release the genomic DNA, representative single colonies were randomly selected and added to 10 µL of Lyse and Go PCR Reagent (Thermo Scienti c, USA), following the manufacturer's recommendations. The 16S rRNA gene was then ampli ed by PCR using primers 27F and 1492R (Lane 1991), which ampli ed nearly the entire 16S rRNA gene. The PCR products were puri ed and then sequenced using the BigDye terminator cycle sequencing ready reaction kit (Applied Biosystem, USA) on an ABI 3730 automated DNA sequencer. Sequences were assembled using DNAMAN 8.0. For bacterial identi cation, sequences were compared to the 16S rRNA gene of existing organisms using the NCBI GenBank Nucleotide Database (Madden 2002).
Detection of toxin genes using PCR ampli cation All primers (Table 1) and PCR cycling conditions were found in the study by van Asten et al. 2009. All reactions were done as singlets using PowerTaq PCR buffer (Tianyi Huiyuan, Beijing, China). A 10 µl aliquot was subject to electrophoresis through a 2.0% (wt/vol) agarose gel which contained ethidium bromide. PCR products were visualized under UV light.
Bio lm formation and quanti cation C. Perfringens strains that had been identi ed by 16S rRNA gene sequencing were inoculated into Reinforced Medium for Clostridia (RCM; Hopebio, Qingdao, China) and incubated at 37°C for 24 h under static conditions. Then the bacteria were then subcultured at a 1:200 dilution into Dulbecco's modi ed Eagle Medium (DMEM; Invitrogen) in a 24-well tissue culture plate embedded with cover slips. The plate was then incubated at 35°C under 5% (vol/vol) CO 2 atmosphere conditions without agitation.
Synchronously, Edwardsiella tarda PPD130/91 (Ling et al. 2020) and its derived strains were also grown without agitation in tryptic soy broth (TSB; BD Biosciences) at 28°C. For the introduction of type III secretion system (T3SS) proteins, E. tarda strains were subcultured in DMEM at 25°C under 5% (vol/vol) CO 2 as a control.
At 24 hours post subculture, the culture supernatant was removed carefully, and the cover slips were gently rinsed three times with prewarmed PBS to remove unattached bacteria. The bacteria that attached to the cover slips were xed in 4% paraformaldehyde and stained with 0.1% crystal violet for 30 min.
Bio lms stained with crystal violet were photographed before being solubilized with a 1% SDS buffer (200 per well) to evaluate the OD 595 using a microplate reader (SpectraMax, M5).
The data was analyzed statistically with SPSS 18.0 software (SPSS Inc., Chicago, IL, USA) and P < 0.001 was considered statistically signi cant.
Antibiotic susceptibility testing Single colonies of C. Perfringens from a pure culture plate were selected and then suspended in 0.9% NaCl to a

Histopathological examinations
Pathological tissues from the lung, liver, kidney, stomach and intestine were sampled and xed in 10% neutral-buffered formalin. Subsequently embedded in para n wax, cut into 5-7 μm sections and stained with hematoxylin and eosin (HE) for histopathological examination.

Results
Isolation and identi cation of C. perfringens Table 2 lists all the bacteria which were isolated from the blowhole, lungs, stomach and fecal samples from the neonatal Yangtze nless porpoise.
Only two fecal samples, collected on days 2 and 27, contained C. perfringens. The colony from the fecal (rectal) sample collected by dissection on day 27 was numbered strain F9#1, and the two colonies obtained from the sample collected on day 2 were numbered strains F9#2 and F9#3.
Pairwise sequence comparisons to sequences deposited in the GenBank nucleotide database using the BLAST algorithm indicated that all the three sequences showed 99% identity to C. perfringens partial 16S rRNA gene, strain 18115 (KP944158) as well as 99% identity to the 16S rRNA gene, partial sequence, of several other C. perfringens strains. These similarity values are above the proposed cutoff value of 98.7% routinely used for the delineation of novel species of bacteria (Stackebrandt and Ebers 2006;Yarza et al. 2014). The 16S rRNA gene sequences from our study have been deposited in the GenBank nucleotide database under the accession number MH569435 to MH569437.
Toxin genotyping of C. perfringens isolates All isolates were classi ed as C. perfringens type C based on the presence of the cpa and cpb genes. As well, all isolates contained the cpb2 gene. No etx, iap and cpe genes were detected (Fig. 1).

Bio lm formation pro le
Previous studies have demonstrated that EseB mediate wild-type E. tarda PPD130/91 is capable of bio lm formation when cultured in DMEM, whereas the mutation strain ΔeseB cannot (Tan et al. 2005;Gao et al. 2015). In this study, we examined C. perfringens strains F9#1, F9#2 and F9#3 for their ability to form a bio lm. As shown in Fig. 2

Histopathological examinations
Histopathological diagnosis showed that there was mild in ammation in the left lung of the neonatal porpoise (Fig S1A), the heart had mild focal degeneration and necrosis (Fig S1B), while other internal organs, such as the liver, kidney and intestine, showed no obvious lesions. Although the lung and heart had some mild lesions, this did not contribute directly to the death of the animal.

Discussion
As shown in Table 2 no bacteria were cultured from the gastric uid collected on day 2, while Enterococcus sp. and Clostridium bifermentans were cultured from the gastric uid collected on day 27. Aeromonas hydrophila and Citrobacter freundii were cultured from the lung swabs. Five different types of bacteria were cultured from the blowhole swabs. It is not known if these bacteria permanently resided in the blowhole or if the bacteria were only present in the water and contaminated the tissue. Although there were some mild lesions in the lungs, as shown by histology (Fig. S1A), this did not directly contribute to the death of this animal. Furthermore, aquatic mammals such as porpoises, in the process of choking can develop a lung infection. No obvious evidences of lung or other internal organ damage caused by other Clostridium spp. or pathogenic bacteria, such as Staphylococcus spp., was found at necropsy.
C. perfringens is among the most common cause of foodborne illnesses in humans (Centers for Disease Control and Prevention 2015) and this bacterium can cause disease in several different animals (Diab et al. 2012;Minamoto et al. 2014). In many cases, diseases caused by this bacterium are life threatening or even fatal (Awad et al. 1995). Although adult animals can contract the bacterium and become ill, disease most often occurs in neonates (Timoney et al. 1988). Piglets are highly susceptible to type C infections (Fitzgerald et al. 1988;Johnson et al. 1992). Infections occur in many different neonatal terrestrial mammals, such as calves (Griner and Bracken 1953), lambs (Griner and Johnson 1954), goat kids (Garcia et al. 2012) and foals (Drolet et al. 1990). Sometimes necro-hemorrhagic enteritis results from an infection. This is caused by the beta-toxin being absorbed by the intestine and then going into general circulation (Songer 1996). Death of the animal can result. In marine mammals, such as captive killer whales (Orcinus orca), false killer whales (Pseudorca crassidens), dwarf sperm whales (Kogia sima), and bottlenose dolphins (Tursiops truncatus), C. perfringens is a part of the normal intestinal ora (Walsh et al. 1994). However, the bacterium can still have the ability to cause disease which ranges from mild to severe and in some cases even resulting in death (Sawires and Songer 2006). For example, it caused death of a captive Atlantic bottlenose dolphin (Buck et al. 1987).
In this study, the head of the calf once collided with the wall of the pool after birth, and a wound appeared on the right side of the head. It is assumed that C. perfringens entered the animal through the external wound, and subsequently vegetative cell developed under anaerobic conditions and produced toxins. C. perfringens infections spread rapidly because of the destructive effects of the toxin. It was similar to the death of a captive bottlenose dolphin living in Mystic Marinelife Aquarium (Mystic, Connecticut) in Florida (Buck et al. 1987). Currently, nothing is known about the epidemiology of C. perfringens in the Yangtze nless porpoise. Here, we investigated the presence and toxin diversity of C. perfringens isolated from a neonatal Yangtze nless porpoise that died 27 days after birth.
A total of three C. perfringens strains were isolated in this study ( Table 2). The toxin genes cpa, cpb, and cpb2, were detected in this study, whereas the etx, iap and cpe genes were absent (Fig.  1). The cpa gene can be present in all C. perfringens strains. However, if this gene is not present the bacterium is unable to cause disease (Niilo 1980). All type C strains contain both the cpa and cpb genes. The beta2 toxin, which is encoded by the plasmid-borne cpb2 gene (Gibert et al. 1997), is the cause of most of the clinical symptoms and it is considered the main virulence factor (Sayeed et al. 2008). This toxin has also been reported to cause many enteric diseases in animals (Petit et al. 1999;Garmory et al. 2000). Das et al. 2009 found that C. perfringens containing the cpa gene was present in both healthy piglet controls and diarrheic animals, while isolates carrying the additional cpb2 gene was found only in diarrheic pigs and piglets, but absent in healthy animals. This evidence strongly suggests that the cpb2 gene is implicated in pig enteritis (Das et al. 2009). The toxin produced is a necrotizing and lethal toxin which has cytotoxic effects on intestinal cells (Guran et al. 2014).
A bio lm is a bacterial community which is formed on a surface. A polymeric matrix is developed which allows the bacteria to stick to each other (Hall-Stoodley et al. 2004). Bio lm formation has been shown to have an important role in gastrointestinal infections (Hu et al. 2018). Bio lm formations also helps bacteria to survive and it may play a role in virulence (Charlebois et al. 2014). In a recent study, all isolates from beef, chicken, duck and pork were able to form a robust bio lm. Bio lm formation was also observed in most poultry and swine isolates (83%), as well as in most clinical and commensal isolates (96.7%) (Hu et al. 2018). In this study, one of the three C. perfringens type C strains was able to form a bio lm (Fig. 2). It was reported that the ability of C. perfringens to form a bio lm depends on the environmental conditions (Pantaléon et al. 2014).
Studies have shown signs of acquired antibiotic resistance in C. perfringens indicating that antibiotic resistance is emerging (Lyras et al. 2009;Soge et al. 2009;Charlebois et al. 2002). In a recent study, multidrug resistant (MDR) C. perfringens strains were isolated from dogs in Korea (Chon et al. 2018).
Other studies have reported MDR strains of C. perfringens in other animals (Dutta and Devriese 1981;Marks and Kather 2003;Rood et al. 1978). A recent study in South Korea, in which C. perfringens isolates collected from diseased animals in poultry farms showed susceptibility to ampicillin, penicillin, bacitracin, and erythromycin, but high resistance to apramycin, gentamicin, and streptomycin (Park et al. 2015). A study in Belgium found that C. perfringens resistance to penicillin was very rare (Osman and Elhariri 2013). These ndings are similar to our results. Overuse or the misuse of antibiotics in both human and veterinary medicine, without proper antibiotic susceptibility testing, may be a major factor in increasing antimicrobial resistance (Wellington et al. 2013).

Conclusions
In this study, C. perfringens type C was isolated and characterized from the fecal materials of a neonatal Yangtze nless porpoise that died 27 days after birth. The toxin pro le suggests that the bacterium could have contributed to the death of this porpoise, since some neonatal animals are affected by the toxins produced during a type C infection. In addition, antibiotic resistance poses therapeutic challenges. Future studies will potentially expand our insight into the role C. perfringens plays in the overall health and disease of both adult and neonatal Yangtze nless porpoises.  Figure 1 2.0% agarose gel (wt/vol) electrophoresis of PCR products obtained with six C. perfringens toxin types.