Characterization of a potentially novel ‘blown pack’ spoilage bacterium isolated from bovine hide

Authors


Correspondence

Declan J. Bolton, Principal Research Officer, Teagasc Food Research Centre, Ashtown, Dublin 15, Ireland. E-mail: declan.bolton@teagasc.ie

Abstract

Aims

To characterize a psychrotrophic bacterium, designated TC1, previously isolated from a cattle hide in Ireland, and to investigate the ability of this strain to cause ‘blown pack’ spoilage (BPS) of vacuum-packaged beef primals.

Methods and Results

TC1 was characterized using a combination of phenotypic, chemotaxonomic and genotypic analyses and was assessed for its ability to spoil vacuum-packaged beef at refrigerated temperatures. TC1 was Gram-positive and formed elliptical subterminal endospores. The strain was able to grow between 0 and 33°C, with optimal growth between 23 and 24°C. TC1 could be differentiated from its phylogenetically closest neighbour (Clostridium lituseburense DSM 797T) by 16S rRNA gene sequencing, pulsed-field gel electrophoresis and cellular fatty acid composition. TC1 spoiled (BPS) beef within 42 days when inoculated in cold-stored (1°C) vacuum-packed beef.

Conclusions

The phenotypic, chemotaxonomic and genotypic characterization indicated that TC1 may represent a potentially novel, cold-tolerant, gas-producing bacterium of considerable economic significance to the beef industry.

Significance and Impact of the Study

This study reports and characterizes an emerging BPS bacterium, which should be considered in future activities designed to minimize the psychrophilic and psychrotrophic spoilage of vacuum-packaged beef.

Introduction

The 16S rRNA gene sequences of microbes recovered in beef abattoirs and their environments have revealed a considerable diversity of strictly anaerobic, psychrotrophic bacteria (Moschonas et al. 2009b, 2011). These bacteria have been isolated from animal faeces, hides, the gastrointestinal track, soil, as well as sewage and water (Broda et al. 2002, 2009; Boerema et al. 2003), indicating their possible distribution within the abattoir with potential cross-contamination of carcasses and derived meat products.

Such psychrotrophic bacteria belonging to the genus Clostridium spp. can proliferate in vacuum-packed meats, causing a particular type of spoilage known as ‘blown pack’ spoilage (BPS). BPS occurs in chilled-stored (i.e. −1·5 to 4°C) vacuum-packaged meat within 2–4 weeks. It involves the production of large volumes of gas (comprised mainly of carbon dioxide and hydrogen and sufficient to produce severe pack distension), a putrid smell and a metallic sheen on the meat. This type of spoilage, first reported in the USA in 1989, subsequently occurred in the UK, New Zealand, Ireland and Brazil (Dainty et al. 1989; Kalchayanand et al. 1989; Broda et al. 2000c; Moschonas et al. 2009a; Silva et al. 2011) and has since been reported in many other countries worldwide. Meat spoiled in this way is not considered to pose a food safety hazard, but has no commercial value, so BPS represents a considerable economic loss to meat processors.

Psychrotrophic, strictly anaerobic clostridia constitute a quite heterogeneous microbial group with many closely related strains (Hippe et al. 1992; Stackebrandt et al. 1999; Spring et al. 2003; Stackebrandt and Hippe 2005). Direct identification of such organisms is therefore difficult, and the assignment/employment of a variety of tests has been widely used for their identification/characterization (Collins et al. 1994; Broda et al. 1999, 2000b,c; Brightwell and Clemens 2012).

During a study in Irish beef abattoirs to identify sources of BPS (Moschonas et al. 2009b), a previously unknown psychrotrophic bacterium, designated TC1, was isolated. The present study describes the phenotypic, chemotaxonomic and genotypic characteristics of TC1 and examines its ability to cause BPS. Overall, the data of the present study are used to define the relationship of the strain with other closely phylogenetically related strains/species and to establish its commercial significance.

Materials and methods

Isolation and sources of bacteria

The isolate, designated TC1, was recovered from a cattle hide in an Irish beef abattoir, using ethanol treatment of the sample in prereduced peptone yeast extract glucose starch (PYGS) broth (Lund et al. 1990) following streaking and anaerobic incubation on Columbia Blood Agar (CBA; Oxoid, Basingstoke, UK) plate(s) containing 5% (v/v) horse blood, as previously described (Moschonas et al. 2009b). For comparative testing, Cl. lituseburense DSM 797T (=ATCC 25759T) and Cl. bifermentans DSMZ 14991T (=ATCC 638T) strains were purchased from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ, Braunschweig, Germany). All strains were cultured anaerobically in prereduced PYGS broth and subcultured onto CBA containing 5% (v/v) defibrinated horse blood.

Phenotypic and chemotaxonomic characterization of TC1

Well-isolated colonies of the isolate were cultured on CBA and examined in relation to cell morphology, Gram stain, oxidase and catalase reactions. The temperature range for growth of the isolate was determined in prereduced peptone yeast extract (PY) broth (Holdeman et al. 1977), at pH 7·0, with 0·5% (w/v) glucose, with 2% (v/v) exponentially growing culture inoculated into Hungate tubes as described previously (Broda et al. 1999). The culture was incubated at temperatures ranging between −1·5 and 40°C for up to 21 days.

Cellular fatty acid (CFA) content was determined from stationary-phase cells of TC1 grown at 25°C for 72 h on Reinforced Clostridial Agar (Oxoid) (pH = 6·8 ± 0·2) under anaerobic conditions. Harvesting, extraction and analysis were performed at BCCM/LMG (Ghent, Belgium) as recommended by the MIDI (Microbial Identification System, Inc., Newark, DE, USA), except that cells were harvested from two plates to obtain sufficient concentration of fatty acids in the extract. The whole-cell fatty acid composition was determined by gas chromatography (Mosca et al. 1998) and comparison with the MIDI BHIBLA (Rev 3·80) peak naming table.

Genotypic characterization of TC1

DNA from TC1, Cl. lituseburense DSM 797T and Cl. bifermentans DSMZ 14991T was prepared according to the protocol of Niemann et al. (1997). The 16S rRNA gene of TC1 was amplified by PCR using the following primers (Hutson et al. 1993): pA (forward) 5′-AGA GTT TGA TCC TGG CTC AG-3′ (approx. positions 8–28 in Escherichia coli numbering) and pH* (reverse) 5′-AAG GAG GTG ATC CAG CCG CA-3′ (approx. positions 1542–1522 in Ecoli numbering). PCR-amplified 16S rRNA was purified using the NucleoFast® 96 PCR Clean-up Kit (Macherey-Nagel, Düren, Germany). Sequencing reactions were performed using the BigDye® TerminatorT Purification Kit (Applied Biosystems, Foster City, CA, USA), using an ABI Prism® 3130XL Genetic Analyser (Applied Biosystems). To ensure highly reliable assembled data, the following forward (F) and reverse (R) primers were used to get a partial overlap of sequences: 16F358, 16F536, 16F926, 16F1112, 16F1241, 16R339, 16F519 and 16F1093 (Sundset et al. 2007). Sequence assembly was performed using the program AutoAssembler (Applied Biosystems). A similarity matrix was created using the software package BioNumerics 4·61 (Applied Maths, Belgium), after discarding unknown bases, using a similarity calculation with a gap penalty of 0%, based on pairwise alignment using an open gap penalty of 100% and a unit gap penalty of 0%. Phylogenetic analysis was performed using the software mega4 (Tamura et al. 2007). A resulting tree was constructed using the neighbour-joining method (Saitou and Nei 1987).

DNA–DNA hybridization between TC1 and its closest related strain Cl. lituseburense DSM 797T was carried out at 31°C according to Ezaki et al. (1989). The DNA similarity percentages reported are the means of a minimum of four hybridizations.

Determination of the G+C content of TC1 was carried out according to Mesbah et al. (1989). The value that is reported is the mean of two independent analyses of the same DNA sample.

Phylogenetic comparison among all examined strains was carried out using pulsed-field gel electrophoresis analysis (PFGE). PFGE was carried out after overnight digestion with SmaI, at 37°C, as described by Hung and Bandziulis (1990). The restriction digests were separated by agarose gel electrophoresis in a pulsed electric field (voltage 5·3 V cm−1, angle 120°, 24 h with switch time 1–15 s, linear). Molecular weight markers (DNA ladder 100 bp; Promega, Southampton, UK) were included for normalization of the fingerprints. The DNA fragments were visualized by ethidium bromide staining, photographed under UV illumination, scanned and digitized.

Ability of TC1 to produce gas in cold-stored, vacuum-packed beef

The ability of the isolate to produce gas in vacuum-packed, chilled-stored meats was assessed as follows: portions of beef (10 × 10 × 1 cm each) were inoculated with 0·1 ml of an actively growing culture of TC1 in PYGS, to a final level of 103 CFU cm−2. The inoculated samples were stored for 30 min at room temperature, placed in individual vacuum bags (BB325, Cryovac; SealedAir Ltd, St Neots, UK) and vacuum-packed using Vac Star S220 (Sugiez, Switzerland). In line with industrial practice, vacuum packs were ‘heat shrunk’ by dipping in a water bath, at 90°C for 2–3 s. Inoculated and uninoculated control samples were stored at 1(±1)°C and visually examined every 2 days, for up to 100 days. The experiment was repeated twice with three samples per replicate for each treatment. Upon completion of the experiment, gas in the headspace of the blown packs was analysed as previously described (Moschonas et al. 2009a).

Results

Phenotypic and chemotaxonomic characterization of TC1

TC1 produced circular, β-haemolytic colonies with undulate margins, dull, cloudy and umbonate measuring 3 mm in diameter during anaerobic growth on CBA. Cells in exponential growth phase were Gram-positive, relatively short fat rods (1·0 × 3·0 μm), single or in pairs and nonmotile, forming elliptical subterminal endospores during the late stationary growth phase. The spores caused slight swelling of the maternal cells. Oxidase and catalase reactions were negative. The minimum growth temperature of the isolate was 0°C, and the maximum growth temperature was 33°C (Fig. 1). Optimal growth was observed between 23 and 24°C. The major CFAs of TC1 were C16:0 (16·3%), cis-7-C16:1 (14·0%), C14:0 (11·7%) and cis-9-C16:1 (11·2%). Smaller proportions of C12:0 (5·3%), cis-9-C18:1 (5·2%), iso-C13:0 (3·9%), iso-C11:0 (2·7%) and iso-C12:0 (2·5%) were also detected (Table 1).

Table 1. Comparison of cellular fatty acid compositions of strains TC1 and Clostridium lituseburense DSM 797T (Elsden et al. 1980). Fatty acids occurring at <1·0% in both strains are not listed
Fatty acidTC1

Cllituseburense

DSM 797T

  1. Summed feature 7 contains C17:2 and/or cis-8-C17:1.

  2. Summed feature 10 contains 18:1c11/t9/t6 and/or UN 17·834.

C10:01·5
anteiso-C11:01·5
iso-C11:02·7
C12:05·32·4
iso-C12:02·5
C13:01·11·2
anteiso-C13:02·10·4
iso-C13:03·90·7
C14:011·711·0
C15:01·57·0
iso-C15:01·41·5
C16:016·335·0
iso-C16:01·5
C16:11·0
cis-7-C16:114·02·2
cis-9-C16:111·22·1
C17:01·07·2
iso-C17:02·3
cis-9-C17:11·0
cis-11-C17:11·1
C18:01·611·0
cis-9-C18:15·22·0
cis-11-C18:11·5
Summed feature 72·2
Summed feature 109·0
Figure 1.

Growth rate of TC1 at various temperatures.

Genotypic characterization of TC1

Phylogenetic analysis of the strains, based on their 16S rRNA sequences, showed that TC1 was separated from their closest relatives, that is, Cl. lituseburense (97·3% similarity), Cl. bifermentans (95·8%) and Cl. bartlettii (95·8%). The phylogenetic position of TC1 among members of Cluster XIa of the class Clostridia is shown in the phylogenetic tree in Fig. 2. The results of the DNA–DNA hybridization studies showed that TC1 had 40% DNA homology with the type strain of Cl. lituseburense. The G+C content of TC1 was 29·2 mol%. The extent to which TC1 can be differentiated from its closest neighbours (Cl. lituseburense, Cl. bifermentans) was also demonstrated by comparison of the PFGE banding patterns of the strains (Fig. 3). The PFGE profile of TC1 had more bands in common with Cl. lituseburense than with the other reference strains.

Figure 2.

The position of TC1 within Cluster XIa of the genus Clostridium. Sequences were downloaded from The National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov) and aligned using the online tool ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2/), with a gap opening penalty of 15 and gap extension penalty of 6·66. The evolutionary history was inferred using the neighbour-joining method. The optimal tree with the sum of branch length = 0·45772515 is shown. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches (Felsenstein 1985). The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the maximum composite likelihood method (Tamura et al. 2004) and are in the units of the number of base substitutions per site. All positions containing gaps and missing data were eliminated from the dataset (complete deletion option). There were a total of 1121 positions in the final dataset. Phylogenetic analyses were conducted in mega4 (http://www.megasoftware.net).

Figure 3.

PFGE fingerprints of the strains with the enzyme SmaI. M, molecular weight marker; 1, TC1; 2, Clostridium lituseburense (DSMZ 797T); 3, Cl. bifermentans (DSMZ 14991T).

Ability of TC1 to produce gas in cold-stored, vacuum-packed beef

The isolate produced gas bubbles in the drip of vacuum-packed meats within 28 days at 1°C with subsequent loss of pack integrity (loss of vacuum) taking place on day 42 ± 9. A mixture of CO2 (56·4 ± 1·9%), H2 (15·3 ± 4·3%), N2 (1·2 ± 0·3%), and to a lesser extent, 1-butanol, 1-propanol, isobutanol, butyric acid and 2-methyl-1-butanol were detected in the packs. No gas was produced in uninoculated controls.

Discussion

The present study reports the characteristics of TC1, an anaerobic bacterium isolated from a cattle hide in Ireland. Classification of TC1 based on its 16S rRNA sequence showed that the strain does not belong to the validly described species of genus Clostridium (sensu stricto, cluster I), but it belongs to a currently nondescribed genus falling in cluster XIa of the class Clostridia with the proposed name ‘Peptostreptococcaceae’ (Wiegel et al. 2006). The study showed that as the isolated strain TC1 is not taxonomically related to the known BPS-causing species belonging to genus Clostridium (cluster I), it is imperative that future studies will re-evaluate the BPS potential of other genera, like the species falling within cluster XIa.

Genotypic characterization showed that the strain had 97·3% similarity of the 16S rRNA gene with Cl. lituseburense DSM 797T. In general, strains sharing 97% (or less) of 16S rRNA gene sequence similarity are not members of the same species (Tindall et al. 2010), indicating that TC1 may belong to Cl. lituseburense (as the closest related strain) or may represent a new and separate species. In addition to the above, DNA–DNA hybridization results between TC1 and Cl. lituseburense DSM 797T showed that they had 40% DNA homology. As a DNA similarity of 70% is generally accepted as limit for species delineation (Wayne et al. 1987), TC1 and Cl. lituseburense may belong to separate species. Overall, the strain TC1 was differentiated from its closest phylogenetic relative, Cl. lituseburense, according to their cellular fatty acid profiles, their 16S rRNA gene sequences and their per cent DNA similarity. Such data support the assumption that TC1 may represent a new species. Our results also showed that TC1 could be distinguished from its closest relatives within the cluster XIa of the class Clostridia by using the PFGE profiles from SmaI digestion. Further taxonomic and classification data of TC1 are deemed necessary to provide an improved understanding of the relationships of TC1 and other related taxa.

The study also showed that in addition to Cl. estertheticum and Cl. gasigenes, which represent two well-known ‘blown pack’ spoilage (BPS) bacteria (Dainty et al. 1989; Kalchayanand et al. 1993; Broda et al. 2000c; Clemens et al. 2010; Silva et al. 2011), strain TC1 may also cause BPS within the commercial holding period of vacuum-packed meats, that is, 42 days. Even though TC1 did not cluster with any of the known BPS bacteria, the combination of gases detected in the packs was similar to the ones comprising the gaseous atmosphere of blown packs caused by the known BPS clostridia (i.e. CO2, H2, N2 and 1-butanol, 1-propanol, isobutanol, butyric acid, 2-methyl-1-butanol) (Dainty et al. 1989; Broda et al. 2000a). Future studies are deemed necessary to characterize other physicochemical properties of the final (BPS) samples, like the odour, texture, colour (in the beef surface, core or fat tissue), the pH and the amount of exudate, as per Silva et al. (2011). Such studies will be used to gain a more holistic insight of the spoilage potential and to facilitate comparison of the spoilage occurring between TC1 (isolated from bovine hide) and other organisms isolated from blown packs of beef.

The isolate was recovered from a bovine hide. As cattle hides are frequently contaminated with faeces (Boerema et al. 2003), it is most likely that direct or indirect contamination with faeces provides the means of meat contamination with such organisms. Similar strains have been isolated from a range of other locations of beef abattoirs and implicated in BPS events in Ireland and New Zealand (Broda et al. 1996) indicating their possible role in a proportion of BPS spoilage events in vacuum-packed, chilled-stored meats.

Overall, this study reports the isolation of a potentially novel, strictly anaerobic, gas-producing bacterium. The phenotypic, chemotaxonomic and genotypic data presented in this paper support the hypothesis that TC1 is a new and emerging psychrotrophic bacterium. Spoilage of vacuum-packaged beef within 28 days at 1°C suggests this strain has the potential to cause major economic losses for the beef industry. Future meat industry blown pack spoilage control strategies, especially those targeting the destruction of spores as part of the plant and equipment sanitation programme, should consider this organism during design, validation, monitoring and ongoing verification. In addition to the above, the ability of the organism of causing other types of spoilage and/or its ability for pathogenicity is yet to be determined.

Acknowledgements

The authors wish to gratefully acknowledge the Food Institutional Research Measure (FIRM), administered by the Department of Agriculture, Fisheries and Food, Ireland, for funding this research. Dr Áine Fox is also acknowledged for her technical assistance.

Ancillary