CLOSTRIDIUM TERTIUM ISOLATED FROM FEED

Although Clostridium tertium is supposed to be a foodborne pathogen, the data on its detection in foodstuff s is scarce, and there are no reports on its isolation from feed. In this communication paper, the isolation of C. tertium from a sample of soya semolina is described. C. tertium may be important in diff erential diagnosis, when it is to be distinguished from Clostridium perfringens. It is a unique species due to the lack of key characteristics of the genus it belongs to because it grows in the presence of oxygen and does not produce toxins. It has been well-documented as a human pathogen, although its mechanisms of pathogenicity are still unknown. According to sporadic reports in veterinary medicine, it has been identifi ed as a rare causative agent of infections in cattle, pigs, birds and marine mammals.


INTRODUCTION
Bacterium species of the Clostridium genus are endospore-forming, obligate anaerobes (or relatively oxygen-tolerant) widespread not only in solid and liquid environments (soil, sewage, surface waters, marine sediments, etc), but also in animal and human intestines and, eventually, in animal and plant products. Based on its rRNA structure, the genus comprises extremely heterogeneous species, many of which share phylogenetic similarities with some other bacterial genera (Collins et al., 1994). Owing to its capability to produce enteritis and enterotoxaemia in various domestic animals, Clostridium perfringens is the most important clostridium species in veterinary medicine. Animal feed is one of potential sources of infection. According to regulations on the microbiological criteria for animal feed quality, it is considered safe if no Clostridium perfringens and Clostridium botulinum are detected in 50 g of a sample (Regulation on the Quality of Animal Feed, 2010). Th e isolation and identifi cation of C. perfringens should be done in compliance with the EN ISO 7937 standard, which enables the precise identifi cation and enumeration of the target species in food and animal feeding stuff . Th e identifi cation of other members of Clostridium genus is not part of the routine procedure in laboratories for feed analysis in Serbia and is beyond their diagnostic capacity. For the above mentioned reasons, this case report is a result of an aspiration to satisfy the researchers' curiosity, discover the identity of certain Clostridium isolates from feed and to broaden the knowledge about bacterium species (other than C. perfringens) present in animal feedstuff s and feed. Th e isolation and identifi cation of Clostridium tertium is presented in this communication paper. To the best of our knowledge, Clostridium tertium has not yet been detected in animal feed samples, although it is sometimes present in food of animal origin.

A REPORT ON A LABORATORY CASE
Sample: Soya semolina. Isolation: Clostridia were isolated following the instructions given in the EN ISO 7937:2010 Standard. For further confi rmation, fi ve colonies black in colour due to sulphite reduction -grown on TSC (tryptone-sulfi te-cycloserine) agar (Biokar Diagnostics, France) were chosen. Th ey were inoculated into thioglycollate broth and incubated for 24 hours at 37°C (Fig 1.A). Aft er incubation, 5 drops of thioglycollate culture was inoculated into lactose-sulfi te (LS) broth (Biokar Diagnostics, France) for C. perfringens confi rmation. Aft er 24 hours of incubation at 46°C, LS was examined for gas production and the presence of black colour (sediment of iron sulfi te). Th e formation of black colour has been observed, but Durham's tube was fi lled with gas to less than a 1/4 of its volume (Figure 1.B). According to ISO standard, the test in LS medium should be repeated in this case by transferring 5 drops of culture grown in LS broth to another test tube with the same medium, repeating the incubation in the same conditions. As the repeated test once again failed to confi rm the presence of Clostridium perfringens species, the culture which grew in thioglycollate medium was transferred by streaking onto two plates with Columbia blood agar base with the addition of 5% of defi brinated sheep blood. Th e plates were incubated at 37°C, one in aerobic and the other in anaerobic conditions using GasPak EZ (Becton Dickinson and Company, USA). Aft er 24 hours of incubation, the growth was noted only on the plate which was incubated in anaerobic conditions, which led to the conclusion that the species is a strict anaerobe. However, aft er 48 hours, the growth was also observed in the dish incubated in aerobic conditions, which would have led to ambiguity if it had been a Clostridium species. Th e isolate formed little (about 1 mm in diameter), opaque colonies, surrounded by a zone of incomplete (a) haemolysis ( Figure  1.C). It was confi rmed that it was Clostridium genus by a negative test for catalase and the microscopic appearance of the Gram stained smears: Gram-positive rods with rounded ends were found and oval spores located terminally were rarely present in smears made from cultures grown in anaerobic conditions. Based on characteristic black colonies grown on sulfi te cycloserine agar, Gram-stain morphology and negative catalase test results, the isolates were presumptively identifi ed as Clostridium species.
A. B. C. Due to the lack of tests for biochemical and molecular typing of isolates of Clostridium species other than C. perfringens in our laboratory, the isolate was further processed at the Institute of Public Health of Vojvodina in Novi Sad. Th e identifi cation was performed with matrix-assisted laser desorption/ionization time-of-fl ight mass spectrometry (MALDI-TOF MS). MALDI-TOF MS has been adapted to generate protein mass spectra from whole bacteria and other microorganisms. Th ese spectra can be compared to a reference database for rapid and accurate taxonomic classifi cation of unknown organisms at the genus, species, and, in some cases, at strain levels. Th e isolate was identifi ed by MALDI TOF as Clostridium tertium (Fig 2).

COMMENT
Clostridium tertium is abundant in soil, but is also found in animal and human intestines as well as in the commensal microbiome of the mouth cavity (Vanderhofstadt et al., 2010). It is a rare human pathogen. Reportedly, it was fi rst described and its biochemical properties were studied in isolates from war wounds in the First World War . Th erefore, C. tertium is considered capable of causing bacteraemia . In addition, it was found in persons with various ailments: meningitis, septic arthritis, enterocolitis, peritonitis, posttraumatic brain abscess, pneumonia, and necrotizing fasciitis and gangrene. C. tertium does not produce exotoxins and the mechanism of its virulence is not known (Ferrell and Tell, 2001;Ray et al., 2003;Vanderhofstadt et al., 2010). Moreover, its clinical importance is questionable since it is not entirely clear if it is a real pathogen or only a contaminant (Vanderhofstadt et al., 2010). It is supposed that C. tertium does damage to the gut mucosa when colonizing it (Ferrell and Tell, 2001), meaning it can penetrate into the bloodstream .
In veterinary medicine, C. tertium has been recognized as a causative agent of enteritis in cattle and pigs. AlMashat and Taylor in 1984 isolated similar bacteria from cattle with enteritis and phenotypically identifi ed it as Sporolactobacillus species (Silvera et al., 2003). In artifi cially infected cattle, this bacterium caused mild diarrhoea. Ferrell and Tell (2001) reported an isolation of C. tertium from faeces of Trichoglossus moluccanus that vomited and had blood in faeces. It was assumed that contaminated water was the source of infection and that a diet rich in carbohydrates is a favourable medium for bacterial fermentation. Šeol et al. (2006) were the fi rst to accuse C. tertium of causing abscesses, osteomyelitis and, fi nally, death in a dolphin, which was the fi rst detection of this bacterium in marine animals. Postollec et al. (2012) detected nine Clostridium species in various foodstuff , but not C. tertium (although they reviewed certain data during its previous detection). A long time ago, in 1965, Goudkov and Sharpe fi rst published a paper on C. tertium detected in cheese and milk. Th ey claimed that in spite of unfavourable conditions for C. tertium growth in dairy products, it can spoil certain cheese types. Later (Fernández et al., 2015), this bacterium was listed as one of the three clostridium foodborne pathogens in cheese, along with C. botulinum and C. perfringens. Le Bourhis et al. (2005) even developed and validated PCR primers for Clostridium spp. detection in cheese.
C. tertium was also detected in meat samples (Ersöz and Coşansu, 2018). Search for C. diffi cile with the API20A (System for the identifi cation of anaerobes and with serological Clostridium Diffi cile Test Kit) resulted in C. tertium detection in one out of 101 samples of meat products (beef and chicken) collected from the market (Ersöz and Coşansu, 2018). Th e same agent was successfully recovered from foie gras and was proved to be capable of growing during storage at 8°C (Prevost et al., 2013). It was confi rmed that C. tertium spores can be inactivated in meat by hydrostatic pressure and bacteriocins (Kalchayanand et al., 2003).
C. tertium is resistant to high temperatures, it can grow under various atmospheric conditions, can cause diarrhea and is present in both healthy and diseased humans (Silvera et al., 2003). Since it is considered an intestinal commensal in animals, it remains unclear whether animal feed should be regarded as a potential source of infection.