Technology-induced selection towards the spoilage microbiota of artisan-type cooked ham packed under modified atmosphere

Food Microbiology 27 (2010) 77–84

Contents lists available at ScienceDirect

Food Microbiology journal homepage: www.elsevier.com/locate/fm

Technology-induced selection towards the spoilage microbiota of artisan-type cooked ham packed under modified atmosphere Charalampos Vasilopoulos a, Hannelore De Maere b, Eveline De Mey b, Hubert Paelinck b, Luc De Vuyst a, Fre´de´ric Leroy a, * a

Research Group of Industrial Microbiology and Food Biotechnology (IMDO), Faculty of Sciences and Bio-engineering Sciences, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium Research Group for Technology and Quality of Animal Products, Department of Industrial Engineering, KaHo Sint-Lieven, Gebroeders Desmetstraat 1, B-9000 Ghent, Belgium

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 3 July 2009 Received in revised form 26 August 2009 Accepted 28 August 2009 Available online 4 September 2009

The microbiota associated with a highly-perishable Belgian artisan-type cooked ham was analyzed through plating and (GTG)5-fingerprinting of isolates throughout its processing chain. The raw tumbled meat was characterized by the presence of a versatile microbiota around 4.8 log(cfu g1), consisting of lactic acid bacteria, staphylococci, Brochothrix thermosphacta, Gram-negative bacteria, and yeasts. Pasteurisation of the ham logs reduced bacterial counts below 2 log(cfu g1) and subsequent manipulations selected for leuconostocs and carnobacteria. Also, B. thermosphacta and several Enterobacteriaceae were found at this stage. During storage in an intermediate high-care area for 2 days, a selection towards certain Enterobacteriaceae (Hafnia alvei, Enterobacter spp., and Pantoea agglomerans) and lactic acid bacteria (mainly vagococci and Streptococcus parauberis) was observed. B. thermosphacta, Leuconostoc carnosum and carnobacteria were also detected, but only after allowing bacterial outgrowth by incubating the meat logs at 7  C for four weeks. After a mild post-pasteurisation process and subsequent handling, incubation of the meat logs at 7  C for four weeks led to outgrowth of Enterobacteriaceae (mainly Enterobacter spp. and Serratia spp.). B. thermosphacta, and lactic acid bacteria (Enterococcus faecalis, Leuc. carnosum, and Carnobacterium maltaromaticum) were also found. After slicing and packaging under modified atmosphere, the microbiota of the refrigerated end-product consisted of leuconostocs, carnobacteria, and B. thermosphacta. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Cooked ham Spoilage Lactic acid bacteria Brochothrix thermosphacta Enterobacteriaceae

1. Introduction Microbial growth and metabolism contribute to the limitation of the shelf-life of cooked meat products. The manifestation of the metabolic action is perceived by the consumer as spoilage, resulting from the combined effect of off-flavours, discolouration, and/or slime formation on the surface of the product (Gram et al., 2002; Bruhn et al., 2004; Holley et al., 2004). Modified-atmospherepackaging (MAP), in combination with chilling, is one of the most widespread methods to delay spoilage in cooked meat products. Not only does packaging act as a barrier against contaminants, it also plays a crucial role in the selection of spoilage microorganisms due to its effect on oxygen availability (Labadie, 1999; McMillin, 2008; Nychas et al., 2008). From the large group of microorganisms that initially colonise the raw meat ecosystem, lactic acid bacteria

* Corresponding author. Tel.: þ32 2 6293612; fax: þ32 2 6292720. E-mail address: fl[email protected] (F. Leroy). 0740-0020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.fm.2009.08.008

and Brochothrix thermosphacta mostly prevail in the cooked endproduct, thereby outcompeting most Gram-negative bacteria such as Pseudomonadaceae and Enterobacteriaceae (Borch et al., 1996; Rattanasomboon et al., 1999; Blixt and Borch, 2002; Gram et al., 2002). Lactic acid bacteria that are most commonly encountered are Lactobacillus spp., Carnobacterium spp., Leuconostoc spp., and Enterococcus spp. (Bjo¨rkroth and Korkeala, 1997; Metaxopoulos et al., 2002; Peirson et al., 2003b; Vasilopoulos et al., 2008). Despite the available knowledge on the nature of the microbial spoilage of MAP cooked meat products, little is reported on the source of the spoilage bacteria and the effect of the processing chain on the selection towards the final microbiota (Ma¨kela¨ and Korkeala, 1987; Bjo¨rkroth and Korkeala, 1997; Borch et al., 1988; Dykes et al., 1991). Studies that deal with the processing chain focus mainly on food pathogens rather than on spoilage microorganisms (Berends et al., 1998; Chasseignaux et al., 2002; Aslam et al., 2004; Byrne et al., 2008). It is not fully clear to what extent microorganisms that are present in meat products originate from the meat itself or rather from handling-related operations in the processing

78

C. Vasilopoulos et al. / Food Microbiology 27 (2010) 77–84

line or from the environment (Bjo¨rkroth et al., 1998; Samelis et al., 1998; Metaxopoulos et al., 2003; Sachindra et al., 2005; Chevallier et al., 2006; Vihavainen et al., 2006). Since bacterial communities that are present throughout the processing line can attach to various surfaces and thus serve as contaminants of the final product, it is of particular importance to determine these communities and their contribution to the shelf-life (Gill and McGinnis, 2004; Brightwell et al., 2006; Gounadaki et al., 2008). During processing of brined raw ham logs, cooking plays inarguably the largest role in bacterial selection. The effect of cooking is often measured by monitoring the ‘‘core’’ temperature of the treated meat logs, which determines the application of appropriate temperature and time combinations. However, deviations from proper cooking practices frequently occur (Metaxopoulos et al., 2003). Moreover, sanitation of the processing line and cooking of the product, even intensive, is not always effective against handling-related post-contamination or presence of thermotolerant vegetative bacteria (Franz and von Holy, 1996a; Jessen and Lammert, 2003; Peirson et al., 2003a). The present study deals with the characterisation of spoilageassociated microorganisms during the processing of a Belgian MAP artisan-type cooked ham. The latter product has de facto a shorter shelf-life than more conventional cooked meat products, mainly due to its intrinsic parameters such as lower salt concentrations than usually applied and absence of preservatives. Leuconostoc carnosum, Carnobacterium divergens, and B. thermosphacta have been shown to dominate the microbiota of this sliced MAP endproduct in the cold-chain at 4 and 7  C (Vasilopoulos et al., 2008). Around room temperature (26  C), however, enterococci dominate. Compounds related to the metabolism of the aforementioned bacteria are detected at the end of the shelf-life of this product, including lactic acid, acetic acid, acetoin, 3-methyl butanol, and hydrogen sulphide (Leroy et al., 2009). It is of utmost importance to understand in what way the cooked ham production process can serve as a source of handling-related bacterial contamination, resulting in specific species domination. Therefore, raw meat and intermediate products from different production stages of a Belgian artisan-type cooked ham were studied to monitor meat-associated microorganisms and their evolution and succession during processing. 2. Methods 2.1. Origin of the samples All samples were obtained from a commercial facility for the production of sliced, MAP artisan-type cooked ham, starting from

raw meat. During production, the raw meat is subjected to a number of technological treatments, finally resulting in the packaged and sliced MAP end-product (Fig. 1). Briefly, raw deboned meat is injected with brine and tumbled. Next, the raw tumbled meat is shaped into ham logs and subjected to a first pasteurisation process (the actual ‘‘cooking’’ process). An oxygen-impermeable pasteurisation bag is used (Krehalon Benelux NV, Turnhout, Belgium). Pasteurisation is done by submerging the ham logs into water according to the facility’s standards (maximum water temperature of 72  C; F value of 200 min). Subsequently, the logs are cooled, unpacked, washed, and left to dry for a period of 2 days in an intermediate room (‘‘high-care’’ area). After that, logs are repacked with pasteurisation bags and subjected to a second, milder pasteurisation treatment (post-pasteurisation; i.e. a brief shower treatment with steamed water). After a storage period of three weeks at 0–2  C, the final post-pasteurised product is sliced and packed under modified atmosphere containing 70% N2 and 30% CO2. The final product is packaged in two multilayer films. The upper film consists of a multilayer of polyethylene terephthalate/ low density polyethylene/ethylene vinyl alcohol/low density polyethylene (PET/LDPE/EVOH/LDPE), having an oxygen permeability of 2.5 cm3/m2/24 h at 23  C and 50% relative humidity (RH). The lower film consists of a peelable barrier of polyethylene terephthalate/ polyethylene/ethylene vinyl alcohol/polyethylene (PET/PE/EVOH/ PE), having an oxygen permeability of 1 cm3/m2/24 h at 20  C and 0% RH. Samples were collected at five different stages (a-e) of the production process to determine microbial counts and, where feasible, microbial composition (Fig. 1). Raw tumbled meat was sampled to study the initial microbial load and composition (stage a). Cooked meat logs were sampled after the pasteurisation step (stage b), and prior to and after the post-pasteurisation treatment (stages c and d). The final cooked ham was sampled after slicing and packaging (stage e). At the sampling stage d, samples were taken at three different places of the ham log (top, side, and bottom) to examine heterogeneity of the final heating treatment. As microbial numbers were generally too low to determine the bacterial composition through culture-dependent molecular methods (see below), spoilage-induced tests were conducted, allowing the bacterial groups present to reach detectable numbers. Spoilage-induced tests consisted of provoking the outgrowth of the dominating psychrotrophic microbiota at each processing step, by storing the meat samples (obtained at stages a to e) at a constant temperature of 4 or 7  C for a period of four weeks. At the end of this spoilage induction period, samples were taken for further analysis. At all stages investigated, ten ham logs were sampled. The whole sampling experiment was performed twice.

Fig. 1. Overview of the technological process for the production of sliced and modified-atmosphere-packaged (MAP) artisan-type cooked ham. Lower-case letters (a–e) indicate sampling points for the determination of the psychrotrophic meat-associated microbiota. Asterisks denote points where an incubation of the meat during 4 weeks at 4 or 7  C was performed. At stage c samples were also obtained prior to incubation at 7  C for 4 weeks as mentioned in the text.

C. Vasilopoulos et al. / Food Microbiology 27 (2010) 77–84

2.2. Microbiological analyses For microbiological analysis, 10 g of meat were taken from the sampling stages a-e (Fig. 1) and were diluted ten times in saline solution (0.85%, [w/v], NaCl), followed by homogenization in a Stomacher 400 (Seward, Worthington, UK) for 1 min. Focus was on the determination of total aerobic bacteria and lactic acid bacteria, since the end-product is known to suffer from lactic-type spoilage (Vasilopoulos et al., 2008; Leroy et al., 2009). Therefore, aliquots (0.1 ml) of the appropriate decimal dilutions in saline were used to prepare spread plates for the enumeration of total aerobic viable bacteria on Plate Count Agar (PCA; VWR International, Darmstadt, Germany). Modified de Man-Rogosa-Sharpe agar (mMRS), i.e. MRS without sodium acetate, was used at stages a, b, and e for the enumeration and isolation of presumptive lactic acid bacteria. Both inoculated PCA and mMRS agar media were incubated at 30  C for 48–72 h. Isolates were picked up randomly from an appropriate dilution, always in a proportion of 10–20% of the total amount of colonies present. Also, isolates were tested for Gram reaction, cell morphology, and catalase production. In total, 719 bacterial colonies were obtained from all the meat samples taken (Table 1). The isolates were transferred to 10-ml tubes containing Brain Heart Infusion broth (BHI; VWR International) and incubated overnight at 30  C; 1.5 ml of the overnight cultures was transferred to a cryovial, containing 25% (v/v) of glycerol, and stored at 80  C until further use.

Table 1 Distribution of microorganisms recovered from an artisan-type cooked ham processing line that were identified up to genus or species level after isolation on PCA or mMRS agar. Processing stagea

Total number of isolates

Major bacterial groups detectedb

a

109

Brochothrix thermosphacta (XVb) Carnobacterium maltaromaticum (XIIa, XIIIb) Enterobacter sp. (IVb) Klebsiella sp. (VIIIa, VIb) Lactobacillus plantarum (IIIa) Lactobacillus sakei (IIa) Leuconostoc carnosum (XVIa, XVIIa, XIXb) Leuconostoc mesenteroides (VIIa) Pediococcus sp. (Va) Staphylococcus epidermidis (XIVb) Staphylococcus saprophyticus (Ia) Staphylococcus sp. (IVa)

b

189

B. thermosphacta (Xb0 ) Carnobacterium divergens (IIIb0 ) C. maltaromaticum (Ia0 ) Enterobacter sp. (XIVb0 ) Hafnia alvei (XIIb0 , XIIIb0 ) Lactobacillus curvatus (IVa0 , Va0 ) L. sakei (VIIa0 , XIa0 ) Lactococcus lactis (Xa0 ) Leuc. carnosum (VIa0 , VIIa0 , Ib0 ) Leuc. mesenteroides (IIa0 , IIIa0 , IIb0 ) Serratia proteamaculans (IVb0 , VIIb0 , IXb0 )

c

205

B. thermosphacta (VIIIb0 ) C. divergens (IIb0 ) C. maltaromaticum (Ib0 ) Enterobacter amnigenus (VIIb) Enterococcus faecalis (XIb) H. alvei (Ib, Xb, VIIb0 ) L. lactis (XVb0 ) Leuc. carnosum (XIIIb0 ) Pantoea agglomerans (Vb, VIb, IVb0 ) Streptococcus parauberis (IIb) Vagococcus carniphilus or V. fluvialis (XIIb)

d

150

E. faecalis (VIIb0 ) B. thermosphacta (IXb0 ) C. maltaromaticum (VIIIb0 ) E. amnigenus or Enterobacter sp. (IVb0 , XIb0 , XIIb0 , XIVb0 ) H. alvei (XVb0 ) Kluyvera ascorbata (VIb0 ) Kluyvera intermedia (XIIIb0 ) Leuc. carnosum (XVIIIb0 ) Serratia liquefaciens (Ib0 ) S. liquefaciens or S. plymuthica (IIb0 0 ) S. proteamaculans (IIIb0 )

e

66

2.3. DNA extraction and (GTG)5-PCR fingerprinting of bacterial isolates Cells obtained from cultures grown overnight in BHI were used for DNA analyses. Bacterial DNA was extracted as previously described, based on an enzymatic lysis of the cells and subsequent use of phenol-chloroform-isoamyl alcohol (Gevers et al., 2001). The obtained DNA served as a template for further PCR analyses. Using the oligonucleotide primer (GTG)5 (50 -GTGGTGGTGGTG GTG-30 ), PCR-fingerprinting was applied to discriminate between the collected bacterial isolates. A T3000 Thermocycler (Biometra GmbH, Goettingen, Germany) was used for PCR amplification. The procedure previously described by Gevers et al. (2001) was followed. Image analysis of the obtained (GTG)5 patterns was carried out using BioNumerics Version 5.1 software (Applied Maths, Sint-Martens-Latem, Belgium). The approach was based on the Pearson product-moment coefficient. Band identification was performed prior to similarity matrix calculation. Dendrograms were created with the Unweighted Pair Group Method with Arithmetic Average (UPGMA) clustering algorithm. From each cluster obtained through numerical analysis of the (GTG)5 fingerprints, assignment of each group of bacteria was performed by comparing the (GTG)5 patterns with an in-house library, consisting of over 200 distinct fingerprints originating from bacterial isolates from food origin. In addition, sequencing of the 16S rRNA gene of a representative isolate from the assigned clusters was performed for verification of the cluster identity. Fingerprints that were not present in the library but found in abundance throughout the experiment were subjected to 16S rRNA gene sequencing too. Therefore, DNA was amplified using the primers pA (50 -GAGTTTGATCCTGGCTCAG-30 ) and pH (50 -AAGGAGGTGATCCAGCCGCA-30 ) according to Edwards et al. (1989). The PCR products of approximately 1.5 kb were checked by electrophoresis, purified using the QIAquick PCR Purification Kit (Qiagen N.V., Venlo, The Netherlands), and sequenced in a commercial facility (VIB, Antwerp, Belgium). The sequencing results were evaluated using BLAST analysis of the NCBI database.

79

B. thermosphacta (IIIb40 ) C. divergens (IIIb70 ) C. maltaromaticum (Ib40 ) Leuc. carnosum (IIb40 , Ib70 ) Leuc. mesenteroides (IIb70 )

a

Processing stage as referred to in Fig. 1. Roman numbers denote the cluster number as referred to the text (originally according to the corresponding figures that are not shown). Suffixes ‘‘a’’ and ‘‘b’’ denote mMRS agar and PCA isolates, respectively; the prime symbol denotes isolates that were obtained from products previously stored at 7  C for 4 weeks. For stage ‘‘e’’, suffixes ‘‘4’’ and ‘‘7’’ were added to denote the incubation temperature. Identification was carried out by comparing (GTG)5-PCR patterns with an in-house library and confirmation of the cluster assignment by sequencing of the 16S rRNA gene. b

3. Results 3.1. Raw meat The total initial microbial load on PCA of raw tumbled meat prior to formation and pasteurisation of the ham logs (samples from stage a) was equal to 4.83  1.25 log(cfu g1). The microbial

80

C. Vasilopoulos et al. / Food Microbiology 27 (2010) 77–84

population enumerated on mMRS agar equalled 3.22  1.08 log (cfu g1). (GTG)5-PCR fingerprinting revealed a large diversity among the microbial populations present on the raw meat (Fig. 2 as a representative figure for clustering of (GTG)5-PCR fingerprints; Table 1). The dendrograms obtained after image analysis of all 109 isolates obtained from mMRS agar and PCA (50 and 59 isolates, respectively) yielded 36 different fingerprint clusters. Of this total, 17 different clusters were obtained from mMRS agar (Fig. 2a) and 19 different clusters were obtained from PCA (Fig. 2b). Only two clusters were shared between mMRS agar and PCA isolates (clusters XIIa and XVIIa, and clusters XIIIb and XIXb, respectively), assigned to Carnobacterium maltaromaticum and Leuc. carnosum, respectively. From the mMRS isolates-based DNA fingerprints, the largest group of lactic acid bacteria to colonise the raw tumbled meat was attributed to Leuc. carnosum. The corresponding fingerprints were grouped in clusters XVIIa (18 isolates) and XVIa (two isolates). Other lactic acid bacteria isolated from mMRS agar were assigned to C. maltaromaticum (four isolates, cluster XIIa), Leuconostoc mesenteroides (two isolates, cluster VIIa), Pediococcus sp. (one isolate, cluster Va), Lactobacillus sakei (one isolate, cluster IIa), and Lactobacillus plantarum (one isolate, cluster IIIa). Cluster Ia (three isolates) was assigned to Staphylococcus saprophyticus. Sequencing of the 16S rRNA gene of the only isolate of cluster IVa did not differentiate further than species level (Staphylococcus sp.).

Bacterial fingerprints that were scattered over the clusters VIa, IXa, Xa, XIIa, and XVa were Gram-positive bacteria, yielding a positive (clusters VIa, IXa, XIIIa, and XVa) or negative (cluster Xa) catalase reaction. They were not detected anywhere else in the processing line. The isolates belonging to the remaining fingerprints were characterized as yeasts (7 isolates, cluster XIVa) and catalasepositive, Gram-negative bacteria (one isolate, cluster VIIIa; one isolate, cluster XIa). The isolate of cluster VIIIa was identified to species level as Klebsiella sp. B. thermosphacta was the most common isolate on PCA (14 isolates, cluster XVb), representing one fourth of the total isolates. Other groups were Leuc. carnosum (ten isolates, cluster XIXb), Enterobacter sp. (five isolates, cluster IVb), Klebsiella sp. (four groups, cluster VIb; different fingerprint than cluster VIIIa), C. maltaromaticum (three isolates, cluster XIIIb), and Staphylococcus epidermidis (two isolates, cluster XIVb). The 21 remaining isolates were scattered over 13 different clusters of bacterial DNA fingerprints. They were all catalase-positive, Gram-negative rods and were only found at this stage of the production process. 3.2. Pasteurisation process Immediately after pasteurisation (samples from stage b; Fig. 1), a decrease of viable microorganisms on PCA was observed to values below the detection limit of the method [ < 2 log(cfu g1)] for all ten samples investigated. After incubation of these ham logs at 7  C

Fig. 2. Representative dendrograms derived from (GTG)5-PCR fingerprints on all bacterial isolates from a) mMRS agar and b) PCA originating from raw tumbled meat (stage a in Fig. 1). The assignment of each group of bacteria was based on comparing (GTG)5-PCR patterns with an in-house library and confirmed by sequencing of the 16S rRNA gene of representative isolates (Table 1).

C. Vasilopoulos et al. / Food Microbiology 27 (2010) 77–84

for a period of four weeks, the surviving cells developed to population counts of 7.17  0.54 log(cfu g1) on PCA, with minimum and maximum values of 6.21 and 8.00 log(cfu g1), respectively. On mMRS agar, microbial populations were estimated after incubation at 7  C at 6.65  1.15 log(cfu g1), with minimum and maximum values of 4.08 and 7.54 log(cfu g1), respectively. From a total of 189 isolates originating from the incubated, pasteurised logs, 80 were obtained from mMRS agar and 109 from PCA (Table 1). (GTG)5-PCR fingerprinting of mMRS agar isolates resulted in ten distinct profiles (Table 1). Almost half of the mMRS agar isolates were assigned to Leuc. carnosum, consisting of two distinctive patterns (28 and 10 isolates in groups VIa0 and VIIa0 , respectively). Whereas the DNA fingerprints of group VIa0 were similar to previously obtained fingerprints for Leuc. carnosum from the final MAP product (Vasilopoulos et al., 2008), the fingerprint for group VIIa0 was different. It differed also from the DNA fingerprints obtained from the isolates derived from the raw tumbled meat. Two other major groups were assigned to Lb. sakei, demonstrating at least two different DNA fingerprints belonging to clusters VIIIa0 (eight isolates) and IXa0 (nine isolates), and Lactococcus lactis (14 isolates, cluster Xa0 ). In addition, C. maltaromaticum (four isolates, cluster Ia0 ), two distinct types of Leuc. mesenteroides (three isolates, cluster IIa0 ; one isolate, cluster IIIa0 ), and two distinct types of Lactobacillus curvatus (two isolates, cluster IVa0 ; one isolate, cluster Va0 ) were found. From the 109 PCA isolates, the major clusters were assigned to leuconostocs and carnobacteria, i.e. Leuc. carnosum (12 isolates, cluster Ib0 ), Leuc. mesenteroides (nine isolates, cluster IIb0 ), C. divergens (18 isolates, cluster IIIb0 ), and C. maltaromaticum (ten isolates, cluster Vb0 ). Cluster Xb0 (three isolates) was assigned to B. thermosphacta, whereas the clusters XVIIb0 (three isolates) and XVIIIb0 (four isolates) were characterized as Gram-positive, catalase-positive bacteria. The remaining isolates were Gram-negative rods and several of them were found to be Enterobacteriaceae. Serratia proteamaculans formed two different (GTG)5-PCR fingerprints, resulting in the clusters IVb0 (five isolates) and VIIIb0 (seven isolates), whereas cluster IXb0 (five isolates) shared 99% similarity with the 16S rRNA gene sequence of S. proteamaculans. Three isolates (clusters XIIb0 and XIIIb0 ) were assigned to Hafnia alvei. The Enterobacter sp. previously encountered in the raw tumbled meat was also found at this stage (one isolate, cluster XIVb0 ). 3.3. High-care area prior to post-pasteurisation Sampling of ten ham logs that were kept for a maximum period of 2 days in a cold room (2  C) in the high-care area (stage c, Fig. 1), revealed a contamination level of 1.90  0.54 log(cfu g1) on PCA, with a minimum and maximum value of 1.00 and 3.63 log(cfu g1), respectively. The 90 isolates obtained directly from the ham logs on PCA were assigned mainly to members of Enterobacteriaceae (Table 1). H. alvei was most dominant, counting for more than half of the isolates (54 isolates, clusters Ib and IXb). Other clusters were assigned to Enterobacter amnigenus (13 isolates, cluster VIIb) and Pantoea agglomerans (eight isolates, clusters Vb and VIb). Only two clusters were assigned to Gram-positive bacteria, namely Streptococcus parauberis (one isolate, cluster IIb) and Vagococcus carniphilus/fluvialis (two isolates, cluster XIIb). The remaining twelve isolates formed eight distinct clusters, consisting of non-determined Gram-negative, catalase-positive rods. When the ham logs were incubated at 7  C for four weeks, a final microbial population of 8.06  0.79 log(cfu g1) was obtained on PCA and shifts in diversity were observed (Table 1). The largest cluster (31 isolates, cluster VIIIb0 ) was related to B. thermosphacta. From the Enterobacteriaceae mentioned above, H. alvei (15 isolates, cluster VIIb0 ) and P. agglomerans (two isolates, cluster IVb0 ) were recovered.

81

Other clusters were ascribed to lactic acid bacteria, i.e. Leuc. carnosum (13 isolates, cluster XIIIb0 ), C. maltaromaticum (19 isolates, cluster Ib0 ), C. divergens (13 isolates, cluster IIb0 ), Enterococcus faecalis (two isolates, cluster XIb0 ), Lc. lactis (three isolates, cluster XVb0 ), and Leuconostoc sp. (one isolate, cluster XIVb0 ). 3.4. Post-pasteurisation process Following the mild post-pasteurisation treatment (samples from stage d, Fig. 1), all samples investigated (ten ham logs) possessed a total viable count level on PCA below 100 cfu g1. After incubating the logs at 7  C for four weeks, a final density of 7.85  0.63 log(cfu g1) was obtained on PCA. Samples were taken from the surface (below, top, and side of the incubated hams), but no differences were observed. In total, 150 isolates were picked up from PCA. A large diversity was observed (Table 1), represented by 18 different DNA fingerprints. Enterobacteriaceae were dominant and consisted mainly of Serratia spp., i.e., Serratia liquefaciens (19 isolates, cluster Ib0 ), S. proteamaculans (29 isolates, cluster IIIb0 ), and S. liquefaciens/plymuthica (two isolates, cluster IIb0 ). In addition, Kluyvera ascorbata (18 isolates, cluster VIb0 ), Kluyvera intermedia (four isolates, cluster XIIIb0 ), and H. alvei (six isolates, cluster XVb0 ) were found. Despite the distinct fingerprints, clusters IVb0 , XIb0 , XIIb0 , and XIVb0 (total of 41 isolates) were all assigned to Enterobacter sp./amnigenus. Six more isolates (cluster Vb0 ) were identified as ‘‘Enterobacteriaceae bacterium’’ according to the NCBI database (accession number AY633489). In addition to the Enterobacteriaceae, B. thermosphacta was also found (three isolates, cluster IXb0 ), as well as several lactic acid bacteria, i.e., E. faecalis (ten isolates, cluster VIIb0 ), Leuc. carnosum (three isolates, cluster XVIIIb0 ), and C. maltaromaticum (two isolates, cluster VIIIb0 ). 3.5. End-product After slicing of the final product (samples from stage e, Fig. 1), samples were stored under MAP until the end of the shelf-life [ > 6 log(cfu g1), i.e. according to industrial practice], at 4 and 7  C for four and three weeks, respectively. After storage at 4  C, 26 bacterial isolates were picked up from mMRS agar and PCA (Table 1). The DNA fingerprints obtained corresponded to Leuc. carnosum, C. maltaromaticum, and B. thermosphacta. All 12 isolates from mMRS agar were assigned to Leuc. carnosum. From the 14 PCA isolates, seven were assigned to C. maltaromaticum (group Ib40 ), four to B. thermosphacta (group IIIb40 ), and three to Leuc. carnosum (group IIb40 ). After storage at 7  C, 40 isolates were picked up from PCA and mMRS agar. The DNA fingerprints obtained were ascribed to Leuc. carnosum, C. divergens, and Leuc. mesenteroides (Table 1). From the 19 mMRS agar isolates, one isolate was assigned to Leuc. mesenteroides, whereas the other isolates were all Leuc. carnosum. The 21 isolates from PCA were ascribed to C. divergens (ten isolates, group IIIb70 ), Leuc. carnosum (nine isolates, group Ib70 ), and Leuc. mesenteroides (two isolates, group IIb70 ). No Br. thermosphacta was recuperated after incubation of the product at 7  C. All Leuc. carnosum isolates obtained gave (GTG)5-PCR fingerprints identical to a Leuc. carnosum type previously shown to dominate this type of sliced MAP artisan-type cooked ham when incubated at 4–7  C (Vasilopoulos et al., 2008). 4. Discussion The present study focused on the determination of major microbial contaminants throughout the production chain of a MAP artisan-type cooked ham that is subjected to major product losses due to its spoilage-sensitive character. The microbial diversity of

82

C. Vasilopoulos et al. / Food Microbiology 27 (2010) 77–84

the meat product during the main production stages consisted of a variety of microorganisms, mostly including lactic acid bacteria, B. thermosphacta, and Enterobacteriaceae (Vasilopoulos et al., 2008). There has been much debate whether the spoilage-associated bacteria of meat products originate from the meat itself or whether they are just contaminants that reside at the processing area (Borch et al., 1988; Dykes et al., 1991; Franz and von Holy, 1996b; Bjo¨rkroth and Korkeala, 1997; Bjo¨rkroth et al., 1998; Samelis et al., 1998; Metaxopoulos et al., 2002; Sachindra et al., 2005; Vihavainen et al., 2006). In this study, it was shown that spoilageassociated bacteria found at the final MAP product were already isolated at the stage of the raw tumbled meat. Further, it can be suggested that all the manipulations that occur during the manufacturing process contribute to the formation of the so-called ‘‘house microbiota’’. Indeed, the microorganisms are introduced onto surfaces and into the environment of the processing line through contact with the meat and of its handling by the personnel. Also, common daily manufacturing practices and sanitation techniques can select towards different specific groups of bacteria. The meat therefore acts as host for the proliferating fraction of microbes responsible for the spoilage of the final product. Lactic acid bacteria as well as B. thermosphacta are known to play a major role in the spoilage of cooked ham packed under vacuum or modified atmosphere (Franz and von Holy, 1996b; Bjo¨rkroth and Korkeala, 1997; Rattanasomboon et al., 1999; Gram et al., 2002). In the present study, Carnobacterium spp. and Leuc. carnosum were predominantly present in the sliced MAP endproduct, which is in agreement with earlier observations (Dykes et al., 1991; Bjo¨rkroth et al., 1998; Samelis et al., 1998; Vasilopoulos et al., 2008). In addition to the end-product, these lactic acid bacteria were also detected throughout the production process. This included C. maltaromaticum and Leuc. carnosum at all examined stages and C. divergens after the ‘‘high-care’’ area treatments. As the presence of carnobacteria has not yet been confirmed in the gastrointestinal tract of mammals (Leisner et al., 2007), it is hypothesized that they are early contaminants of the processing plants colonising the raw tumbled meat. The meat then acts as the carrier of the microorganisms throughout the processing facility. Within the lactic acid bacteria population of the present study, species variation throughout the production chain was large compared to the limited number of species that were initially present on the raw tumbled meat and eventually dominated the final MAP cooked ham. Nevertheless, some degree of microbial variability can also be detected in the end-product (presence or absence of Leuc. mesenteroides and B. thermosphacta), which may be due to batch variations (Borch et al., 1988) and differences in chill temperature (Vasilopoulos et al., 2008). From the different lactic acid bacteria detected throughout the processing line, several meat-associated ones were identified. V. carniphilus/V. fluvialis and St. parauberis were remarkably present on ham logs prior to post-pasteurisation but were absent in the endproduct. The vagococci, which are lactococci/enterococci-like species, possess a high degree of similarity within their 16S rRNA gene, which hampers their discrimination (based solely on 16S rRNA gene sequencing analysis). V. carniphilus is possibly a meat-borne bacterium as it is frequently isolated from meat-like environments and it is a predominant strain at meat-processing facilities (Shewmaker et al., 2004; Ammor et al., 2005). Antimicrobial activity against a broad spectrum of bacteria, including lactic acid bacteria and Gram-negative bacteria, such as H. alvei belonging to the Enterobacteriaceae, has recently been demonstrated for V. carniphilus (Ammor et al., 2006). Similarly, St. parauberis was detected prior to but not in the end-product. Information existing on the prevalence of the latter species at meat-associated environments is scarce. Up to now, a few isolates from poultry meat and fresh beef

have been described (Koort et al., 2006; Najjari et al., 2008). As it has been previously reported, St. parauberis can indeed survive and grow on MAP meat (Koort et al., 2006). In addition, pediococci, Lb. sakei, and Lc. lactis were found either only in the raw meat (pediococci) or in more than one sample stage (Lc. lactis, Lb. sakei). In a previous study using the same ham type, it has been shown through statistical analysis of (GTG)5-PCR fingerprints of Leuc. carnosum that two different subgroups of this bacterium can dominate the end-product depending on the storage temperature, namely one group that is commonly found at low temperatures (below 12  C) and a second group that dominates the Leuconostoc populations of higher storage temperatures (Vasilopoulos et al., 2008). In this study, the Leuc. carnosum subgroup that is known to prevail at higher temperatures in the end-product (above 12  C) was isolated from the incubated ham logs of the high-care area. They were not recuperated from the raw meat, neither after the pasteurisation process of cooked ham logs, nor during incubation at 4 and 7  C of the final MAP sliced product. This suggests that this subgroup of Leuc. carnosum, probably part of the house microbiota, remains at low levels until it meets favourable environmental conditions. The type that prevails at low temperatures (below 12  C) was omnipresent throughout the production chain of the cooked ham product. Two more types of Leuc. carnosum were occasionally isolated from raw tumbled meat and after the first heating treatment. This variety of different Leuc. carnosum strains indicates the adaptability of this bacterium to a specific niche. Although Enterobacteriaceae were of little importance in the sliced MAP end-product, they were omnipresent throughout its production chain. This was especially the case in the production stages corresponding to ‘‘high-care’’ area treatments and after postpasteurisation. In general, the 16S rRNA gene-limited resolution of Enterobacteriaceae allowed in many cases their identification to species level (Nhung et al., 2007). H. alvei was one of the most dominant contaminants of the processing area found after the pasteurisation, occurring mainly in the drying area of the ‘‘highcare’’ area and onwards. H. alvei is one of the major spoilage enterobacteria found in meat, in particular due to its psychrotolerant character which gives an adaptation advantage over other microbial members (Borch et al., 1996). It has been associated with blowing of packages as well as the production of biogenic amines (mainly putrescine) together with other Enterobacteriaceae members (Durlu-Ozkaya et al., 2001; Gram et al., 2002; Brightwell et al., 2007). In addition to H. alvei, ham logs which were stored in the ‘‘highcare’’ area prior to post-pasteurisation were colonised within two days with Enterobacter spp. and P. agglomerans. Besides the aforementioned species, Serratia spp. and Kluyvera spp. were among the enterobacteria commonly encountered before or after the thermal processes. The presence of these bacteria in meat matrices as well as the potential of synergistic action to provoke spoilage phenomena has been indicated recently (Bruhn et al., 2004; Nychas et al., 2008). It is known that within meat processing factories the levels of Enterobacteriaceae are higher than the ones found at retail level (Stiles and Ng, 1981; Marin et al., 1996). This has been attributed to inadequate hygiene techniques, cross-contamination incidents, and the psychrotrophic traits of these bacteria. Not only can they cause losses prior to distribution due to their proteolytic activity and the formation of off-flavour esters (Garcı´a et al., 2000), they also possess a health-risk potential due to the production of biogenic amines (Shalaby, 1996; Silla Santos, 1996). To face the new challenges of meat products with respect to high quality, nutritional properties, naturalness, and safety (Aymerich et al., 2008; Vandendriessche, 2008) and to plan novel strategies to deal with spoilage phenomena, it is important to

C. Vasilopoulos et al. / Food Microbiology 27 (2010) 77–84

understand the prevalence and behaviour of specific microorganisms under detailed conditions. From the above data it is clear that the meat processing plant is a harbour of different bacteria. It does not necessarily reflect the hygienic conditions of the plant neither the quality of the raw material. It only indicates that care has to be taken of respecting appropriate sanitation conditions and avoiding cross-contamination. In conclusion, variations in conditions throughout the different stages of a Belgian artisan-type cooked ham production chain resulted in a diversity of meat-associated bacteria and an unwanted bacterial presence on the surface of the products. Despite this biodiversity, the MAP end-product is dominated by only a few bacterial groups that are highly competitive and are able to grow out from low contamination levels to large, spoilage-provoking populations. The latter populations outcompete Enterobacteriaceae, resulting mostly in lactic-type spoilage.

Acknowledgements The authors acknowledge their financial support from the Research Council of the Vrije Universiteit Brussel and the Fund for Scientific Research-Flanders (FWO). This work was particularly supported by an industrial research project of the Flemish Institute for the Encouragement of Scientific and Technological Research in Industry (IWT). FL was supported by a postdoctoral fellowship of the FWO.

References Ammor, S., Rachman, C., Chaillou, S., Prevost, H., Dousset, X., Zagorec, M., Dufour, E., Chevallier, I., 2005. Phenotypic and genotypic identification of lactic acid bacteria isolated from a small-scale facility producing traditional dry sausages. Food Microbiol. 22, 373–382. Ammor, S., Tauveron, G., Dufour, E., Chevallier, I., 2006. Antibacterial activity of lactic acid bacteria against spoilage and pathogenic bacteria isolated from the same meat small-scale facility: 1-Screening and characterization of the antibacterial compounds. Food Control 17, 462–468. Aslam, M., Greer, G.G., Nattress, F.M., Gill, C.O., McMullen, L.M., 2004. Genotypic analysis of Escherichia coli recovered from product and equipment at a beefpacking plant. J. Appl. Microbiol. 97, 78–86. Aymerich, T., Picouet, P.A., Monfort, J.M., 2008. Decontamination technologies for meat products. Meat Sci. 78, 114–129. Berends, B.R., Van Knapen, F., Mossel, D.A.A., Burt, S.A., Snijders, J.M.A., 1998. Salmonella spp. on pork at cutting plants and at the retail level and the influence of particular risk factors. Int. J. Food Microbiol. 44, 207–217. Bjo¨rkroth, K.J., Korkeala, H.J., 1997. Use of rRNA gene restriction patterns to evaluate lactic acid bacterium contamination of vacuum-packaged sliced cooked wholemeat product in a meat processing plant. Appl. Environ. Microbiol. 63, 448–453. Bjo¨rkroth, K.J., Vandamme, P., Korkeala, H.J., 1998. Identification and characterization of Leuconostoc carnosum, associated with production and spoilage of vacuum-packaged, sliced, cooked ham. Appl. Environ. Microbiol. 64, 3313–3319. Blixt, Y., Borch, E., 2002. Comparison of shelf life of vacuum-packed pork and beef. Meat Sci. 60, 371–378. Borch, E., Nerbrink, E., Svensson, P., 1988. Identification of major contamination sources during processing of emulsion sausage. Int. J. Food Microbiol. 7, 317–330. Borch, E., Kant-Muermans, M.L., Blixt, Y., 1996. Bacterial spoilage of meat and cured meat products. Int. J. Food Microbiol. 33, 103–120. Brightwell, G., Boerema, J., Mills, J., Mowat, E., Pulford, D., 2006. Identifying the bacterial community on the surface of Intralox belting in a meat boning room by culture-dependent and culture-independent 16S rDNA sequence analysis. Int. J. Food Microbiol. 109, 47–53. Brightwell, G., Clemens, R., Urlich, S., Boerema, J., 2007. Possible involvement of psychrotolerant Enterobacteriaceae in blown pack spoilage of vacuum-packaged raw meats. Int. J. Food Microbiol. 119, 334–339. Bruhn, J.B., Christensen, A.B., Flodgaard, L.R., Nielsen, K.F., Larsen, T.O., Givskov, M., Gram, L., 2004. Presence of acylated homoserine lactones (AHLs) and AHLproducing bacteria in meat and potential role of AHL in spoilage of meat. Appl. Environ. Microbiol. 70, 4293–4302. Byrne, B., Lyng, J., Dunne, G., Bolton, D.J., 2008. An assessment of the microbial quality of the air within a pork processing plant. Food Control 19, 915–920. Chasseignaux, E., Gerault, P., Ragimbeau, C., Salvat, G., Colin, P., Ermel, G., 2002. Ecology of Listeria monocytogenes in the environment of raw poultry meat and raw pork meat processing plants. FEMS Microbiol. Lett. 210, 271–275.

83

Chevallier, I., Ammor, S., Laguet, A., Labayle, S., Castanet, V., Dufour, E., Talon, R., 2006. Microbial ecology of a small-scale facility producing traditional dry sausage. Food Control 17, 446–453. Durlu-Ozkaya, F., Ayhan, K., Vural, N., 2001. Biogenic amines produced by Enterobacteriaceae isolated from meat products. Meat Sci. 58, 163–166. Dykes, G.A., Eugene Cloete, T., von Holy, A., 1991. Quantification of microbial populations associated with the manufacture of vacuum-packaged, smoked Vienna sausages. Int. J. Food Microbiol. 13, 239–248. Edwards, U., Rogall, T., Blocker, H., Emde, M., Bottger, E.C., 1989. Isolation and direct complete nucleotide determination of entire genes. Characterization of a gene coding for 16S ribosomal RNA. Nucleic Acids Res. 17, 7843–7853. Franz, C.M.A.P., von Holy, A., 1996a. Thermotolerance of meat spoilage lactic acid bacteria and their inactivation in vacuum-packaged Vienna sausages. Int. J. Food Microbiol. 29, 59–73. Franz, C.M.A.P., von Holy, A., 1996b. Bacterial populations associated with pasteurized vacuum-packed Vienna sausages. Food Microbiol. 13, 165–174. Garcı´a, C., Martı´n, A., Timo´n, M.L., Co´rdoba, J.J., 2000. Microbial populations and volatile compounds in the ’bone taint’ spoilage of dry cured ham. Lett. Appl. Microbiol. 30, 61–66. Gevers, D., Huys, G., Swings, J., 2001. Applicability of rep-PCR fingerprinting for identification of Lactobacillus species. FEMS Microbiol. Lett. 205, 31–36. Gill, C.O., McGinnis, J.C., 2004. Microbiological conditions of air knives before and after maintenance at a beef packing plant. Meat Sci. 68, 333–337. Gounadaki, A.S., Skandamis, P.N., Drosinos, E.H., Nychas, G.-J.E., 2008. Microbial ecology of food contact surfaces and products of small-scale facilities producing traditional sausages. Food Microbiol. 25, 313–323. Gram, L., Ravn, L., Rasch, M., Bruhn, J.B., Christensen, A.B., Givskov, M., 2002. Food spoilage-interactions between food spoilage bacteria. Int. J. Food Microbiol. 78, 79–97. Holley, R.A., Peirson, M.D., Lam, J., Tan, K.B., 2004. Microbial profiles of commercial, vacuum-packaged, fresh pork of normal or short storage life. Int. J. Food Microbiol. 97, 53–62. Jessen, B., Lammert, L., 2003. Biofilm and disinfection in meat processing plants. Int. Biodeter. Biodegr 51, 265–269. Koort, J., Coenye, T., Vandamme, P., Bjo¨rkroth, J., 2006. Streptococcus parauberis associated with modified atmosphere packaged broiler meat products and air samples from a poultry meat processing plant. Int. J. Food Microbiol. 106, 318–323. Labadie, J., 1999. Consequences of packaging on bacterial growth. Meat is an ecological niche. Meat Sci. 52, 299–305. Leisner, J.J., Laursen, B.G., Pre´vost, H., Drider, D., Dalgaard, P., 2007. Carnobacterium: positive and negative effects in the environment and in foods. FEMS Microbiol. Rev. 31, 592–613. Leroy, F., Vasilopoulos, C., Van Hemelryck, S., Falony, G., De Vuyst, L., 2009. Volatile analysis of spoiled, artisan-type, modified-atmosphere-packaged cooked ham stored under different temperatures. Food Microbiol. 26, 94–102. Ma¨kela¨, P., Korkeala, H., 1987. Lactobacillus contamination of cooked ring sausages at sausage processing plants. Int. J. Food Microbiol. 5, 323–330. Marin, M.E., Carrascosa, A.V., Cornejo, I., 1996. Characterization of Enterobacteriaceae strains isolated during industrial processing of dry-cured hams. Food Microbiol. 13, 375–381. McMillin, K.W., 2008. Where is MAP going? A review and future potential of modified atmosphere packaging for meat. Meat Sci. 80, 43–65. Metaxopoulos, J., Mataragas, M., Drosinos, E.H., 2002. Microbial interaction in cooked cured meat products under vacuum or modified atmosphere at 4 C. J. Appl. Microbiol. 93, 363–373. Metaxopoulos, J., Kritikos, D., Drosinos, E.H., 2003. Examination of microbiological parameters relevant to the implementation of GHP and HACCP system in Greek meat industry in the production of cooked sausages and cooked cured meat products. Food Control 14, 323–332. Najjari, A., Ouzari, H., Boudabous, A., Zagorec, M., 2008. Method for reliable isolation of Lactobacillus sakei strains originating from Tunisian seafood and meat products. Int. J. Food Microbiol. 121, 342–351. Nhung, P.H., Ohkusu, K., Mishima, N., Noda, M., Shah, M.M., Sun, X., Hayashi, M., Ezaki, T., 2007. Phylogeny and species identification of the family Enterobacteriaceae based on dnaJ sequences. Diagn. Microbiol. Infect. Dis. 58, 153–161. Nychas, G.J.E., Skandamis, P.N., Tassou, C.C., Koutsoumanis, K.P., 2008. Meat spoilage during distribution. Meat Sci. 78, 77–89. Peirson, M.D., Guan, T.Y., Holley, R.A., 2003a. Thermal resistances and lactate and diacetate sensitivities of bacteria causing bologna discolouration. Int. J. Food Microbiol. 86, 223–230. Peirson, M.D., Guan, T.Y., Holley, R.A., 2003b. Aerococci and carnobacteria cause discoloration in cooked cured bologna. Food Microbiol. 20, 149–158. Rattanasomboon, N., Bellara, S.R., Harding, C.L., Fryer, P.J., Thomas, C.R., AlRubeai, M., McFarlane, C.M., 1999. Growth and enumeration of the meat spoilage bacterium Brochothrix thermosphacta. Int. J. Food Microbiol. 51, 145–158. Sachindra, N.M., Sakhare, P.Z., Yashoda, K.P., Narasimha Rao, D., 2005. Microbial profile of buffalo sausage during processing and storage. Food Control 16, 31–35. Samelis, J., Kakouri, A., Georgiadou, K.G., Metaxopoulos, J., 1998. Evaluation of the extent and type of bacterial contamination at different stages of processing of cooked ham. J. Appl. Microbiol. 84, 649–660.

84

C. Vasilopoulos et al. / Food Microbiology 27 (2010) 77–84

Shalaby, A.R., 1996. Significance of biogenic amines to food safety and human health. Food Res. Int. 29, 675–690. Shewmaker, P.L., Steigerwalt, A.G., Morey, R.E., Carvalho, M.S., Elliott, J.A., Joyce, K., Barrett, T.J., Teixeira, L.M., Facklam, R.R., 2004. Vagococcus carniphilus sp. nov., isolated from ground beef. Int. J. Syst. Evol. Microbiol. 54, 1505–1510. Silla Santos, M.H., 1996. Biogenic amines: their importance in foods. Int. J. Food Microbiol. 29, 213–231. Stiles, M.E., Ng, L.K., 1981. Enterobacteriaceae associated with meats and meat handling. Appl. Environ. Microbiol. 41, 867–872.

Vandendriessche, F., 2008. Meat products in the past, today and in the future. Meat Sc 78, 104–113. Vasilopoulos, C., Ravyts, F., De Maere, H., De Mey, E., Paelinck, H., De Vuyst, L., Leroy, F., 2008. Evaluation of the spoilage lactic acid bacteria in modifiedatmosphere-packaged artisan-type cooked ham using culture-dependent and culture-independent approaches. J. Appl. Microbiol. 104, 1341–1353. Vihavainen, E., Lundstro¨m, H.S., Susiluoto, T., Koort, J., Paulin, L., Auvinen, P., Bjo¨rkroth, K.J., 2006. Role of broiler carcasses and processing plant air in contamination of modified-atmosphere-packaged broiler products with psychrotrophic lactic acid bacteria. Appl. Environ. Microbiol. 73, 1136–1145.