Multilocus variable-number of tandem repeat analysis (MLVA) for Clostridium tyrobutyricum strains isolated from cheese production environment

International Journal of Food Microbiology 190 (2014) 61–65

Contents lists available at ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Multilocus variable-number of tandem repeat analysis (MLVA) for Clostridium tyrobutyricum strains isolated from cheese production environment Masaharu Nishihara a, Hajime Takahashi b,⁎, Tomoko Sudo a, Daisuke Kyoi b, Toshio Kawahara b, Yoshihiro Ikeuchi c, Takashi Fujita a, Takashi Kuda b, Bon Kimura b, Shuichi Yanahira a a b c

Institute of Food Hygiene, Quality Assurance Department, Megmilk Snow Brand Co., Ltd., 1-1-2 Minamidai, Kawagoe-shi, Saitama 350-1165, Japan Department of Food Science and Technology, Faculty of Marine Science, Tokyo University of Marine Science and Technology, 4-5-7 Konan, Minato-ku, Tokyo 108-8477, Japan Central Food Analysis Laboratory, Quality Assurance Department, Megmilk Snow Brand Co., Ltd., 1-1-2 Minamidai, Kawagoe-shi, Saitama 350-1165, Japan

a r t i c l e

i n f o

Article history: Received 21 January 2014 Received in revised form 9 August 2014 Accepted 15 August 2014 Available online 23 August 2014 Keywords: Clostridium tyrobutyricum MLVA Cheese

a b s t r a c t Clostridium tyrobutyricum is a gram-positive spore-forming anaerobe that is considered as the main causative agent for late blowing in cheese due to butyric acid fermentation. In this study, multilocus variable-number of tandem repeat (VNTR) analysis (MLVA) for C. tyrobutyricum was developed to identify the source of contamination by C. tyrobutyricum spores in the cheese production environment. For each contig constructed from the results of a whole genome draft sequence of C. tyrobutyricum JCM11008T based on next-generation sequencing, VNTR loci that were effective for typing were searched using the Tandem Repeat Finder program. Five VNTR loci were amplified by polymerase chain reaction (PCR) to determine their number of repeats by sequencing, and MLVA was conducted. 25 strains of C. tyrobutyricum isolated from the environment, raw milk, and silage were classified into 18 MLVA types (DI = 0.963). Of the C. tyrobutyricum strains isolated from raw milk, natural cheese, and blown processed cheese, strains with identical MLVA type were detected, which suggested that these strains might have shifted from natural cheese to blown processed cheese. MLVA could be an effective tool for monitoring contamination of natural cheese with C. tyrobutyricum in the processed cheese production environment because of its high discriminability, thereby allowing the analyst to trace the source of contamination. © 2014 Elsevier B.V. All rights reserved.

1 . Introduction Clostridium tyrobutyricum is a gram-positive spore-forming anaerobe that is considered as the main causative organism responsible for late blowing in cheese due to butyric acid fermentation (Bergeres and Sivela, 1990; Klijn et al., 1995; Le Bourhis et al., 2007). Late blowing is caused by the outgrowth of C. tyrobutyricum spores during the ripening period of cheese. The characteristics of late blowing include the production of acetic acid, butyric acid, a large quantity of gas associated with production of carbon dioxide, and hydrogen, as well as the development of a bad odor. Processed cheese is prepared from crushed hard or semi-hard cheese such as Cheddar and Gouda and by adding ingredients, including whey powder, milk powder, cream, butterfat, emulsifier, salt, seasonings and water, heating at 80 °C or higher and thorough mixing. Processed cheese has been largely produced in the United States, Russia and Japan. Especially in Russia and Japan, processed cheese dominates the large part of the total cheese consumption (Sorensen, 2005). Spoilage of processed cheese due to butyric acid fermentation also often occurs because C. tyrobutyricum spores can survive the conditions of cheese ⁎ Corresponding author. Tel./fax: +81 3 5463 0603. E-mail address: [email protected] (H. Takahashi).

http://dx.doi.org/10.1016/j.ijfoodmicro.2014.08.022 0168-1605/© 2014 Elsevier B.V. All rights reserved.

processing (Loessner et al., 1997). In most cases, such blown processed cheese loses its market value, leading to significant economic loss. C. tyrobutyricum spores are detected in the environment, for example, in soil, water, air, and unhygienic animal bedding. The main source of contamination of raw milk is thought to be poor quality silage (Stadhouders and Spoelstra, 1990; Julien et al., 2008). C. tyrobutyricum spores, which can pass through alimentary tract of the cow, are excreted within the feces to the farm environment. Transmission to milk occurs via fecal contamination of the cow's teats. (Stadhouders and Spoelstra, 1990). Among Clostridium beijerinckii, Clostridium sporogenes, Clostridium butyricum, and C. tyrobutyricum, which are capable of butyric acid fermentation, C. tyrobutyricum is most commonly isolated from late-blown cheese (Cocolin et al., 2004; Le Bourhis et al., 2005). However, no specific culture medium that can distinguish each Clostridium species is available and phenotypic discrimination is almost impossible. Thus, it is difficult to identify each strain (Ingham et al., 1998; Sperner et al., 1999). Recently, more reliable techniques using PCR-based methods (Klijn et al., 1995; Herman et al., 1995; Cremonesi et al., 2012) and 16S rRNA probe-based hybridization methods (Klijn et al., 1994; Knabel et al., 1997) have been developed to identify C. tyrobutyricum. Such techniques are required to identify contaminants of natural cheese when late blowing of cheese occurs. However, it is often difficult to obtain reliable information on the

62

M. Nishihara et al. / International Journal of Food Microbiology 190 (2014) 61–65

actual source of contamination using the results of species-level identification. Generally, the identification of the contaminating agent at the strain level requires DNA typing. Most of the previous reports on DNA typing of Clostridium are limited to pathogenic strains such as Clostridium difficile, Clostridium botulinum, and Clostridium perfringens (Alonso et al., 2005; Hielm et al., 1998; Chalmers et al., 2008). For C. tyrobutyricum, although late blowing of cheese leads to considerable economic losses, reports employing DNA typing for strain identification have been limited. Pulsed-field gel electrophoresis (PFGE) is considered the gold standard of DNA typing. Although PFGE is a complicated technique and is time consuming, it is widely used for a number of bacterial strains because of its high discriminability and reproducibility. However, for C. tyrobutyricum, several problems of PFGE, such as smearing of PFGE bands, have been reported (Garde et al., 2012), and improved protocol with no difficulty has not yet been established. Previous reports have examined C. tyrobutyricum strains isolated from milk and cheese using PFGE (Ingham et al., 1998; Christiansen et al., 2005; Garde et al., 2012). However, those reports did not include any discussion of the source of contamination or comparison of the discriminatory power of various DNA typing techniques. Multilocus variable-number of tandem repeat (VNTR) analysis (MLVA) is a DNA typing technique employed to detect the number of repeat units in repetitive loci that are interspersed along genomic DNA. Gel electrophoresis or sequence analysis of the PCR-amplified product of each VNTR locus is conducted to detect the number of repeats. MLVA is thus a DNA typing technique that is simple, rapid, and highly reproducible. Although earlier studies using MLVA have extensively examined pathogens such as Staphylococcus aureus (Sabat et al., 2003), Listeria monocytogenes (Murphy et al., 2007; Miya et al., 2008), and C. botulinum (Fillo et al., 2011), reports on food spoilage organisms other than Geobacillus species (Seale et al., 2012) and Bacillus licheniformis (Dhakal et al., 2013), which are isolated from milk powders, are scarce. To our knowledge, this study is the first to use MLVA of C. tyrobutyricum to trace the source of contamination during cheese production with high discriminability and reproducibility. In addition, we also validated the effectiveness of the MLVA with respect to the management of natural cheese during cheese production. 2 . Materials and methods 2.1 . Bacterial strains C. tyrobutyricum JCM11008T was used to generate the whole genome draft sequence. Table 1 lists the 25 strains used for the development of MLVA. Of these strains, two strains were purchased from the National Collection of Industrial Food and Marine Bacteria (NCIMB, Scotland, UK). Table 2 shows the 29 strains used to trace the source of contamination. 2.2 . DNA extraction C. tyrobutyricum strains were cultured in GAM semisolid without dextrose “Nissui” medium (Nissui Pharmaceutical, Tokyo, Japan) and filtered to remove the agarose, at 35 °C in anearobic condition. For JCM11008T, DNA extraction was conducted using phenol-chloroform extraction (Takahashi et al., 2004). In the other strains, DNA extraction was conducted using NucleoSpin Tissue Kit (MACHEREY-NAGEL GmbH & Co. KG, Duren, Germany) according to the manufacture's protocols.

Table 1 MLVA profile of the strains used for calculation of the DI value. Isolate no.

Origin

Isolation date TR1 TR2 TR3 TR4 TR5 MLVA (day/mo/year) type

IFH 2477 IFH 2479 IFH 2491 IFH 2476 IFH 2483 IFH 2493 IFH 2496 IFH 2487 IFH 2497 IFH 2484 IFH 2475 IFH 2492 IFH 2480 IFH 2481 IFH 2486 IFH 2488 IFH 2489 IFH 2494 IFH 2498 IFH 2500 IFH 2501 IFH 2502 IFH 2474 NCIMB 701790 NCIMB 9582

Environment Environment Environment Raw milk Environment Environment Environment Environment Environment Environment Raw milk Environment Environment Environment Environment Environment Environment Environment Environment Environment Environment Environment Raw milk silage

17-Dec-12 25-Dec-12 21-Jan-13 25-Dec-12 25-Dec-12 21-Jan-13 21-Jan-13 17-Dec-12 21-Jan-13 17-Dec-12 14-Dec-12 21-Jan-13 17-Dec-12 17-Dec-12 17-Dec-12 17-Dec-12 9-Jan-13 21-Jan-13 21-Jan-13 21-Jan-13 21-Jan-13 21-Jan-13 14-Dec-12 10-Oct-87

11 11 11 11 10 10 10 2 2 9 9 7 3 4 4 5 13 8 4 5 12 11 6 4

5 5 5 5 5 5 5 7 7 8 8 7 6 6 6 6 5 5 8 7 8 8 6 8

4 4 4 4 4 4 4 6 6 4 4 4 4 6 4 4 4 4 4 4 4 4 6 6

6 6 6 6 6 6 6 6 6 6 6 7 7 6 7 7 6 6 6 6 6 6 6 7

8 8 8 8 8 8 8 7 7 8 8 7 8 7 8 8 8 8 8 8 8 8 7 8

TYR 1 TYR 1 TYR 1 TYR 1 TYR 2 TYR 2 TYR 2 TYR 3 TYR 3 TYR 4 TYR 4 TYR 5 TYR 6 TYR 7 TYR 8 TYR 9 TYR 10 TYR 11 TYR 12 TYR 13 TYR 14 TYR 15 TYR 16 TYR 17

8-Jan-65

12

7

6

6

8

TYR 18

the manufactures' protocols. Construction of contigs was performed by using the GS De novo assembler (Roche, Basel, Switzerland). These sequences were annotated using Genome Traveler (In Silico Biology, Kanagawa, Japan).

2.4 . Design of the MLVA primers Using the Tandem Repeat Finder program (TRF) (http://tandem.bu. edu/trf/trf.html) (Benson, 1999), we searched for targeted VNTR loci in each contig of the whole-genome draft sequence of C. tyrobutyricum JCM11008T. Primers for amplifying each VNTR locus were designed using the Primer Express Software (Life Technologies, Foster City, CA). The primer was designed within 300 bp of the flanking regions of each VNTR locus. The composition of the PCR reaction was as follows: 50.0 gl of the final mixture consisted of 5.0 gl of 10 × PCR buffer, 4.0 gl of 2.5 mM dNTP mixture, 5.0 gl of each 10 gM primer (shown in Table 3), 0.25 gl (5 U/gl) of TaKaRa Taq DNA polymerase (Takara Bio Inc., Otsu, Japan), 1.0 gl of 25 ng/gl template DNA, and 29.75 gl of DW. PCR amplification was performed in a Veriti thermal cycler (Life Technologies, Foster City, CA) using the following conditions: 5 min of heating at 95 °C; 35 cycles each of 1 min at 95 °C, 30 s at the annealing temperature shown in Table 3, and 1 min at 72 °C; and 2 min at 72 °C. The PCR products were purified using the Fast Gene Gel/PCR Extraction Kit (NIPPON Genetics Co. Ltd., Tokyo, Japan). DNA sequencing was performed using the Big Dye Terminator v.3.1 Cycle Sequencing Kit (Life Technologies). After sequence analysis of the amplified VNTR loci using the 3500 Genetic Analyzer (Life Technologies), the number of repeats of the motif sequence was determined.

2.5 . Discriminatory index (DI) 2.3 . Whole genome shot gun sequencing and de novo assembly The whole genome shot gun sequencing was performed using the Roche GS junior platform (Roche, Basel, Switzerland), the GS junior Rapid Library preparation kit, and GS junior emPCR kit (Lib-L), following

Discriminatory power, i.e., the ability to distinguish between unrelated strains, was calculated based on the Simpson's index of diversity (Hunter and Gaston, 1988). As the value approaches 1, the power of the method to discriminate unrelated strains increases.

M. Nishihara et al. / International Journal of Food Microbiology 190 (2014) 61–65

63

Table 2 MLVA profile of the strains related to the natural cheese factories and the processed cheese factories. Isolate No.

Origin

Factory code

Isolation date (day/m o/year)

TR1

TR2

TR3

TR4

TR5

MLVA type

IFH 960 IFH 967 IFH 1013 IFH 1016 IFH 944 IFH 2470 IFH 997 IFH 1000 IFH 2458 IFH 869 IFH 965 IFH 2419 IFH 961 IFH 946 IFH 2460 IFH 2466 IFH 2461 IFH 1031 IFH 942 IFH 959 IFH 958 IFH 996 IFH 2420 IFH 2459 IFH 2462 IFH 2464 IFH 2418 IFH 2421 IFH 2467

Raw milka Raw milka Natural cheesea Natural cheesea Processed cheese Processed cheese Natural cheese Natural cheese Natural cheese Processed cheese Natural cheese⁎ Natural cheese⁎ Raw milkb Natural cheeseb Natural cheesec Processed cheesec Natural cheese Processed cheese Natural cheese Raw milk Raw milk Natural cheese Natural cheese Natural cheese Natural cheese Natural cheese Processed cheese Processed cheese Processed cheese

N1 N1 N1 N1 P1 P1 N1 N1 N2 P1 N1 N1 N1 N1 N2 P1 N2 P1 N1 N1 N1 N1 N1 N2 N2 N2 P2 P2 P1

29-Nov-07 13-Feb-08 20-May-08 20-May-08 20-Nov-07 17-Dec-12 27-Mar-08 14-Apr-08 14-Dec-12 10-Jul-06 29-Nov-07 25-Apr-91 29-Nov-07 21-Nov-07 17-Dec-12 17-Dec-12 17-Dec-12 16-Jun-08 10-Aug-07 29-Nov-07 29-Nov-07 27-Mar-08 13-Jun-91 14-Dec-12 17-Dec-12 17-Dec-12 21-May-91 25-Jun-91 17-Dec-12

11 11 11 11 11 11 4 4 4 4 2 2 7 7 7 7 5 5 9 12 5 6 6 7 7 7 6 4 7

5 5 5 5 5 5 7 7 7 7 7 7 7 7 7 7 6 6 8 8 7 7 7 7 6 5 6 8 7

4 4 4 4 4 4 6 6 6 6 6 6 4 4 6 6 6 6 4 4 4 4 6 4 7 4 4 6 5

6 6 6 6 6 6 7 7 7 7 6 6 7 7 6 6 6 6 6 6 7 7 7 7 6 6 6 7 6

8 8 8 8 8 8 8 8 8 8 7 7 7 7 7 7 7 7 8 8 7 7 8 8 8 8 8 5 7

TYR 1 TYR 1 TYR 1 TYR 1 TYR 1 TYR 1 TYR 19 TYR 19 TYR 19 TYR 19 TYR 3 TYR 3 TYR 5 TYR 5 TYR 20 TYR 20 TYR 21 TYR 21 TYR 4 TYR 14 TYR 22 TYR 23 TYR 24 TYR 25 TYR 26 TYR 27 TYR 28 TYR 29 TYR 30

a,b

From natural cheese and raw milk that were direct raw materials of natural cheese, the strains that had identical MLVA types were detected. From processed cheese and natural cheese that were direct raw material of processed cheese, the strains that had identical MLVA type were detected. ⁎ Historical isolate. Sixteen years or more have passed since the strain was isolated.

c

2.6 . Nucleotide sequence accession numbers

3.2 . Development of MLVA

The JCM 11008T draft genome data have been deposited into the DDBJ/EMBL/GenBank, with bioproject ID PRJDB1432.

Each contig was analyzed using the TRF, identifying a total of 54 targeted VNTR loci. Five VNTR loci with the sequence length and the number of repeats suitable for the MLVA were PCR amplified for sequence analysis. Differences in the number of the repeats were observed among the strains. Table 4 shows the characteristics of VNTR loci used for MLVA. Five VNTR loci were constructed using a 7-base sequence motif (Table 3). In terms of the number of alleles at each VNTR locus among the tested 25 strains, TR1 was the highest (12), followed by TR2 (4), TR3 (2), TR4 (2), and TR5 (2). The number of the repeats at each VNTR locus was 2–13, 5–8, 4–6, 6–7, and 7–8 at TR1, TR2, TR3, TR4, and TR5, respectively. For TR1, TR2, TR4, and TR5, each repeat

3 . Results 3.1 . Whole-genome draft sequence of C. tyrobutyricum JCM11008 JCM 11008T genome data consists of 697 contigs, with a total nucleotide

sequence 2,528,478 bp, an N50-contig of 4687 bp length, a peak-depth of 4.0, and a 30.93% GC-content.

Table 3 Primers used for the PCR amplification of VNTR loci. Locus

Repeat motif a

Contig no. b

Accession no.

TR1

CACAATT

contig247

BASR01000247

TR2

AATTCAC

contig14

BASR01000014

TR3

AATAACA

contig190

BASR01000190

TR4

CAATTCA

contig281

BASR01000281

TR5

TCAGAGT

contig14

BASR01000014

a b c

Primer sets

PCR products

Annealing Temp(°C)

Primer name

Sequence(5′-3′) c

Location on contig

Length

contig 247-F contig 247-R contig 14-01-F contig 14-01-R contig 190-F contig 190-R contig 281-F contig 281-R contig 14-02-F contig 14-02-R

AGATGCCATCATACCTCT GATTATCTATATAAACACTGTCTCA GTGCAAGATTTGGAGTCAC GTATTCAACTTATTGTCAGGGA GGTAATCTGTGAGGAATTT AAGTGAAAGAAGTGACTGTAG TGACAACGTCCTTGTAACT AGTCCATCCTTATGAAGAAC ACAGAACATATGGTGCAGC AGCAAAGTTCAGAGAACAAAG

2369–2629

261

53.0

5789–6007

219

55.0

2017–2231

215

49.5

322–532

211

52.0

9571–9820

250

54.0

Repeat motifs were based on a draft genome sequence of JCM11008T. Contigs were constructed using the GS De novo assembler (Roche). Primers were 100% matched with the genomic sequence data of JCM11008T.

64

M. Nishihara et al. / International Journal of Food Microbiology 190 (2014) 61–65

Table 4 Characteristics of VNTR loci used for MLVA in the 25 strains. Locus

TR1 TR2 TR3 TR4 TR5

No. of repeats Smallest

Largest

2 5 4 6 7

13 8 6 7 8

No. of alleles

DI a

No. of repeats in JCM11008T b

12 4 2 2 2

0.923 0.763 0.380 0.333 0.333

5 8 6 7 8

a DI (Simpson's index of discrimination) was calculated using the 25 strains shown in Table 1. b The number of repeats was based on the results of a draft genome sequence of JCM11008T.

unit was stable. However, for TR3, deficiency of bases was detected in repeat unit. The DI value of each TR was 0.923, 0.763, 0.380, 0.333, and 0.333 at TR1, TR2, TR3, TR4, and TR5, respectively. The alleles of TR1 to TR5 were connected, and the 25 strains were classified into 18 types (Table 1). The calculated DI value was 0.963. 3.3 . Tracing the source of contamination with MLVA Twenty-nine strains related to the natural cheese factories (N1, N2) and the processed cheese factories (P1, P2) were classified into 17 MLVA types (Table 2). The strains isolated from raw milk, natural cheese, and blown processed cheese were classified into 4 MLVA types, 12 MLVA types, and 7 MLVA types, respectively. Of these, the strains identified as TYR1 of the MLVA type were detected in raw milk, natural cheese, and processed cheese; the strains identified as TYR5 were detected in raw milk and natural cheese; those identified as TYR19, 20, and 21 were detected in natural cheese and processed cheese. However, only TYR20 marked in c (Table 2) showed an identical MLVA type between the strains isolated from blown processed cheese and those isolated from natural cheese, which was the direct raw material for blown processed cheese. For TYR1 marked in a and TYR5 marked in b (Table 2), the strains isolated from natural cheese and the strains isolated from raw milk, which were the direct raw materials of natural cheese, showed identical MLVA types. Two strains with TYR3 marked in * (Table 2) were detected in natural cheese, and at least 16 years have passed since each strain has been isolated. Other strains showed different MLVA types. 4 . Discussion Butyric blowing due to outgrowth of C. tyrobutyricum spores is the most common defect observed in processed cheese. During production, cheese is heated at 80 °C or higher, inactivating vegetative microbial cells present in the material. However, spores survive such heating temperatures. In addition, germination of spores may be even stimulated by heat treatment, which increases the risk of butyric blowing in processed cheese (Loessner et al., 1997). Of Clostridium spp. isolated from late-blown cheese, C. tyrobutyricum shows high frequency of isolation and produces high amount of butyric acid (Le Bourhis et al., 2005). Therefore, it is considered important to monitor the contamination of C. tyrobutyricum spores from natural cheese to processed cheese. In this study, we developed an MLVA typing method to trace the source of contamination with C. tyrobutyricum. The MLVA profile of 25 strains isolated from the environment, raw milk, and silage showed that MLVA had high discriminability (DI = 0.963). In addition to the genome data generated in this study, the genome sequence of the C. tyrobutyricum type strain is now available in GenBank (Jiang et al., 2013). We searched VNTR loci in those genomic data with the TRF. In those genomic data, we found the same five VNTR loci which were used by our MLVA. The MLVA scheme was the same as far as these genome data were used. However, it is known that the discriminability of MLVA

improves with specific VNTR loci (Fillo et al., 2011). There is a possibility that VNTR loci with potentially large mutations may exist, other than the five VNTR loci that were used for MLVA in this study. Consequently, the discriminability may be further improved if such VNTR loci are used. The DI value of PFGE was calculated based on previous typing analyses of C. tyrobutyricum by PFGE: according to Christiansen et al. (2005), 22 strains could be classified into 18 types using ApaI (DI = 0.952), whereas according to Garde et al. (2012), 20 strains could be classified into 11 types using XhoI (DI = 0.926). However, to evaluate the efficiency of the technique, it is necessary to evaluate the same strain group using multiple typing techniques and compare each DI value. For S. aureus and L. monocytogenes, several studies have compared the discriminatory power of DNA typing techniques, including MLVA, PFGE, and multilocus sequencing typing (MLST). According to these studies, the discriminatory power of MLVA was equivalent or higher than that provided by PFGE and MLST (Malachowa et al., 2005; Miya et al., 2008). Compared to PFGE, MLVA shows it is not only relatively more discriminatory but also easy, simple, and rapid, leading to speedy source tracking of the causative strains of butyric blowing in processed cheese. Therefore, MLVA can be considered a suitable technique in resolving contamination problems associated with processed cheese production. The MLVA type of TYR20 was found in both blown processed cheese and natural cheese which was its direct raw material, indicating that this strain shifted from natural cheese to processed cheese (Table 2). On the other hand, TYR1, TYR19 and TYR21 could not be traced because these MLVA types were not identical between blown processed cheese and natural cheese which were its direct raw materials. For TYR28, TYR29, and TYR30 in the strains isolated from blown processed cheese, a common MLVA type was not detected in natural cheese, preventing effective tracing of the route of contamination. However, multiple MLVA types were detected from the strains isolated from blown processed cheese. Consequently, it was found that butyric blowing in processed cheese was not caused by contamination of strains showing a certain MLVA type. In this study, using our stock isolates, we have shown that this method can be used effectively as a tracing tool of C. tyrobutyricum. We believe our method would help in tracing the source of strains caused butyric blowing in processed cheese. In addition, C. tyrobutyricum IFH965 and IFH2419 of the TYR3 of MLVA type were isolated in 2007 and 1991, respectively. Accordingly, it was suggested that contamination of natural cheese with C. tyrobutyricum in the factory N1 might have been reoccurring for an extended period of time. This MLVA technique showed high reproducibility of data because the number of repeats was determined by the analysis of sequences at the VNTR loci. Comparison of the current data with those of previous data is therefore feasible and suitable for monitoring. The compiling of the information about the types of MLVA associated with butyric blowing in processed cheese is important for the future problems during cheese production. To prevent butyric blowing in processed cheese caused by C. tyrobutyricum, quality control strategies should be implemented. The risk of late blowing varies with salt, pH, water activity, and temperature (McSweeney and Fox, 2004), as well as pharmaceutical technologies such as bacterial centrifugation and the addition of nitrates (Stadhouders, 1990; Waes and Van Heddeghem, 1990). Therefore, we believe that stringent selection of high-quality natural cheese as raw material is the most effective approach for preventing butyric blowing in processed cheese. MLVA results may be also used as the selection criterion for purchasing raw materials because cheese processing entails mixing of various domestic and foreign natural cheeses. In conclusion, MLVA which we developed with the draft genome by next-generation sequencing is highly discriminable and reproducible. It allowed us to trace the source of contamination. In addition, this approach has some potential to become an effective tool for monitoring C. tyrobutyricum contamination of raw materials used for cheese production.

M. Nishihara et al. / International Journal of Food Microbiology 190 (2014) 61–65

References Alonso, R., Martin, A., Pelaez, T., Marin, M., Rodriguez-Creixems, M., Bouza, E., 2005. An improved protocol for pulsed-field gel electrophoresis typing of Clostridium difficile. J. Med. Microbiol. 54, 155–157. Benson, G., 1999. Tandem repeats finder: a program to analyze DNA sequences. Nucleic Acids Res. 27, 573–580. Bergeres, J.L., Sivela, S., 1990. Detection and enumeration of clostridial spores related to cheese quality-classical and new methods. In: International Dairy Federation (Ed.), Methods of detection and prevention of anaerobic spore formers in relation to the quality of cheeseBulletin of the International Dairy Federation, No. 251. International Dairy Federation, Brussels, pp. 18–23. Chalmers, G., Martin, S.W., Hunter, D.B., Prescott, J.F., Weber, L.J., Boerlin, P., 2008. Genetic diversity of Clostridiumperfringens isolated from healthy broiler chickens at a commercial farm. Vet. Microbiol. 127, 116–127. Christiansen, P., Petersen, M.H., Kask, S., Mnller, P.L., Petersen, M., Nielsen, E.W., Vogensena, F.K., Ardo, Y., 2005. Anticlostridial activity of Lactobacillus isolated from semi-hard cheeses. Int. Dairy J. 15, 901–909. Cocolin, L., Innocente, N., Biasutti, M., Comi, G., 2004. The late blowing in cheese: a new molecular approach based on PCR and DGGE to study the microbia1 ecology of the alteration process. Int. J. Food Microbiol. 90, 83–91. Cremonesi, P., Vanoni, L., Silvetti, T., Morandi, S., Brasca, M., 2012. Identification of Clostridium beijerinckii, Cl. butyricum, Cl. sporogenes, Cl. tyrobutyricum isolated from silage, raw milk and hard cheese by a multiplex PCR assay. J. Dairy Res. 79, 318–323. Dhakal, R., Chauhan, K., Seale, R.B., Deeth, H.C., Pillidge, C.J., Powell, I.B., Craven, H., Turner, M.S., 2013. Genotyping of dairy Bacillus licheniformis isolates by high resolution melt analysis of multiple variable number tandem repeat loci. Food Microbiol. 34, 344–351. Fillo, S., Giordani, F., Anniballi, F., Gorge, O., Ramisse, V., Vergnaud, G., Riehm, J.M., Scholz, H.C., Splettstoesser, W.D., Kieboom, J., Olsen, J.S., Fenicia, L., Lista, F., 2011. Clostridium botulinum group I strain genotyping by 15-locus multilocus variable-number tandem-repeat analysis. J. Clin. Microbiol. 49, 4252–4263. Garde, S., Gaya, P., Arias, R., Nunez, M., 2012. Enhanced PFGE protocol to study the genomic diversity of Clostridium spp. isolated from Manchego cheeses with late blowing defect. Food Control 28, 392–399. Herman, L.M., De Block, J.H., Waes, G.M., 1995. A direct PCR detection method for Clostridium tyrobutyricum spores in up to 100 milliliters of raw milk. Appl. Environ. Microbiol. 61, 4141–4146. Hielm, S., Bjorkroth, J., Hyytia, E., Korkeala, H., 1998. Genomic analysis of Clostridium botulinum group ii by pulsed-field gel electrophoresis. Appl. Environ. Microbiol. 64, 703–708. Hunter, P.R., Gaston, M.A., 1988. Numerical index of the discriminatory ability of typing systems: an application of Simpson's index of diversity. J. Clin. Microbiol. 26, 2465–2466. Ingham, S.C., Hassler, J.R., Tsai, Y.W., Ingham, B.H., 1998. Differentiation of lactatefermenting, gas-producing Clostridium spp. isolated from milk. Int. J. Food Microbiol. 43, 173–183. Jiang, L., Zhu, L., Xu, X., Li, Y., Li, S., Huang, H., 2013. Genome sequence of Clostridium tyrobutyricum ATCC 25755, a butyric acid-overproducing strain. Genome Announc. 1, e00308–e00313. Julien, M.C., Dion, P., Lafreniere, C., Antoun, H., Drouin, P., 2008. Sources of clostridia in raw milk on farms. Appl. Environ. Microbiol. 74, 6348–6357. Klijn, N., Bovie, C., Dommes, J., Hoolwerf, J.D., Van Der Waals, C.B., Weerkamp, A.H., Nieuwenhof, F.F.J., 1994. Identification of Clostridium tyrobutyricum and related species using sugar fermentation, organic acid formation and DNA probes based on specific 16S rRNA sequences. Syst. Appl. Microbiol. 17, 249–256. Klijn, N., Nieuwenhof, F.F., Hoolwerf, J.D., van der Waals, C.B., Weerkamp, A.H., 1995. Identification of Clostridium tyrobutyricum as the causative agent of late blowing in cheese by species-specific PCR amplification. Appl. Environ. Microbiol. 61, 2919–2924.

65

Knabel, S., Tatzel, R., Ludwig, W., Wallnofer, P.R., 1997. Identification of Clostridium butyricum, Clostridium sporogenes and Clostridium tyrobutyricum by hybridization with 16S rRNA-targeted oligonucleotide probes. Syst. Appl. Microbiol. 20, 85–88. Le Bourhis, A.G., Saunier, K., Dore, J., Carlier, J.P., Chamba, J.F., Popoff, M.R., Tholozan, J.L., 2005. Development and validation of PCR primers to assess the diversity of Clostridium spp. in cheese by temporal temperature gradient gel electrophoresis. Appl. Environ. Microbiol. 71, 29–38. Le Bourhis, A.G., Dore, J., Carlier, J.P., Chamba, J.F., Popoff, M.R., Tholozan, J.L., 2007. Contribution of C. beijerinckii and C. sporogenes in association with C. tyrobutyricum to the butyric fermentation in Emmental type cheese. Int. J. Food Microbiol. 113, 154–163. Loessner, M.J., Maier, S.K., Schiwek, P., Scherer, S., 1997. Long-chain polyphosphates inhibit growth of Clostridium tyrobutyricum in processed cheese spreads. J. Food Prot. 60, 493–498. Malachowa, N., Sabat, A., Gniadkowski, M., Krzyszton-Russjan, J., Empel, J., Miedzobrodzki, J., Kosowska-Shick, K., Appelbaum, P.C., Hryniewicz, W., 2005. Comparison of multiplelocus variable-number tandem-repeat analysis with pulsed-field gel electrophoresis, spa-typing, and multilocus sequence typing for clonal characterization of Staphylococcus aureus isolates. J. Clin. Microbiol. 43, 3095–3100. McSweeney, P.L.H., Fox, P.F., 2004. Metabolism of residual lactose and of lactate and citrate, In: Fox, P.F., McSweeney, P.L.H., Cogan, T.M., Guinee, T.P. (Eds.), third ed. Cheese: Chemistry, Physics and Microbiology. vol. 1. Elsevier, London, pp. 365–366. Miya, S., Kimura, B., Sato, M., Takahashi, H., Ishikawa, T., Suda, T., Takakura, C., Fujii, T., Wiedmann, M., 2008. Development of a multilocus variable-number of tandem repeat typing method for Listeria monocytogenes serotype 4b strains. Int. J. Food Microbiol. 124, 239–249. Murphy, M., Corcoran, D., Buckley, J.F., O'Mahony, M., Whyte, P., Fanning, S., 2007. Development and application of multiple-locus variable number of tandem repeat analysis (MLVA) to subtype a collection of Listeria monocytogenes. Int. J. Food Microbiol. 115, 187–194. Sabat, A., Krzyszton-Russjan, J., Strzalka, W., Filipek, R., Kosowska, K., Hryniewicz, W., Travis, J., Potempa, J., 2003. New method for typing Staphylococcus aureus strains: multiple-locus variable-number tandem repeat analysis of polymorphism and genetic relationships of clinical isolates. J. Clin. Microbiol. 41, 1801–1804. Seale, R.B., Dhakal, R., Chauhan, K., Craven, H.M., Deeth, H.C., Pillidge, C.J., Powell, I.B., Turner, M.S., 2012. Genotyping of present-day and historical Geobacillus species isolates from milk powders by high-resolution melt analysis of multiple variable-number tandem-repeat loci. Appl. Environ. Microbiol. 78, 7090–7097. Sorensen, H.H., 2005. The world market for cheese 1995–2004, Bulletin of the International Dairy Federation, No. 402, 6th ed. International Dairy Federation, Brussels, pp. 33–73. Sperner, B., Eisgruber, H., Stolle, A., 1999. Use of the RAPID ID 32 A system for rapid identification of Clostridium species important in food hygiene. Int. J. Food Microbiol. 52, 169–180. Stadhouders, J., 1990. Prevention of butyric acid fermentation by the use of nitrate. In: International Dairy Federation (Ed.), Methods of detection and prevention of anaerobic spore formers in relation to the quality of cheeseBulletin of the International Dairy Federation No. 251. International Dairy Federation, Brussels, pp. 40–46. Stadhouders, J., Spoelstra, S.F., 1990. Prevention of the contamination of raw milk by making a good silage. In: International Dairy Federation (Ed.), Methods of detection and prevention of anaerobic spore formers in relation to the quality of cheeseBulletin of the International Dairy Federation No. 251. International Dairy Federation, Brussels, pp. 24–31. Takahashi, H., Kimura, B., Yoshikawa, M., Gotou, S., Watanabe, I., Fujii, T., 2004. Direct detection and identification of lactic acid bacteria in a food processing plant and in meat products using denaturing gradient gel electrophoresis. J. Food Prot. 67, 2515–2520. Waes, G., Van Heddeghem, A., 1990. Prevention of butyric acid fermentation by bacterial centrifugation of the cheese milk. In: International Dairy Federation (Ed.), Methods of detection and prevention of anaerobic spore formers in relation to the quality of cheeseBulletin of the International Dairy Federation No. 251. International Dairy Federation, Brussels, pp. 47–50.