Journal Methods Microbiological - ENVIS

Journal of Microbiological Methods 42 (2000) 175–184

Journal of Microbiological Methods www.elsevier.com / locate / jmicmeth

An evaluation of chelex-based DNA purification protocols for the typing of lactic acid bacteria Giorgio Giraffa*, Lia Rossetti, Erasmo Neviani Istituto Sperimentale Lattiero Caseario, Via A. Lombardo 11, 26900 Lodi, Italy Received 6 March 2000; received in revised form 3 May 2000; accepted 30 May 2000

Abstract An easy and rapid protocol to extract DNA to be used as template for polymerase chain reaction (PCR) fingerprinting experiments from cultivable lactic acid bacteria (LAB) is proposed. Different procedures for rapid extraction of DNA by chelex (iminodiacetid acid) ionic resin were compared. Factors affecting the quality and reproducibility of PCR fingerprinting profiles were also investigated. Two out of three chelex-based protocols allowed to obtain DNA samples which, after PCR amplification, provided electrophoretic patterns comparable with those obtained by classical lysozyme and phenol–chloroform DNA extraction. A good level of reproducibility and consistency of the InstaGene procedure was verified. The procedure is fast, practical, and the DNA is of quality similar to that obtained by phenol–chloroform extraction. Although applied to a little number of LAB strains, chelex-based protocols are potentially applicable to a vast array of organisms and / or biological materials.  2000 Elsevier Science B.V. All rights reserved. Keywords: Bacterial taxonomy; Diagnostic microbiology; DNA extraction; Polymerase chain reaction (PCR) fingerprinting; Lactic acid bacteria

1. Introduction Lactic acid bacteria (LAB) play an important role in the human and animal gastrointestinal tract as well as in the production of many foods, feeds, and beverages (Stiles and Holzapfel, 1997). In many fermented foods microbial starters are composed of artisan cultures, which are mixed microbial communities of LAB. All of these systems are complex and usually consist of mixed consortia of lactobacilli, the individual components of which are difficult to

*Corresponding author. Tel.: 139-371-45011; fax: 139-37135579. E-mail address: [email protected] (G. Giraffa).

trace within the system. Typing of individual strains, together with more sensitive tools for species identification, represent a way to overcome this problem (Dykes and von Holy, 1994). Development of molecular biology techniques has improved the knowledge on the taxonomy and ecology of LAB and other bacteria, while opening a new interdisciplinary field, the molecular microbial ecology, to study microbial communities. Nowadays, molecular techniques provide an outstanding tool for microbial detection, identification, and typing. Several PCR-based DNA fingerprinting methods, including rDNA-RFLP analysis (or ARDRA), rDNA spacer length polymorphism analysis, and arbitrarily-primed PCR (or RAPD) have been claimed to produce species-specific and strain-specific fingerprints in

0167-7012 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0167-7012( 00 )00172-X

176

G. Giraffa et al. / Journal of Microbiological Methods 42 (2000) 175 – 184

food-associated LAB (Klein et al., 1998; Tailliez et al., 1996, 1998; Moschetti et al., 1997, 1998; Vandamme et al., 1996; Van Reenen and Dicks, 1996). Because of the vast array of PCR-fingerprinting techniques which have been described during past decades, computer-assisted comparisons have replaced visual comparison of PCR-generated fingerprints (Vauterin and Vauterin, 1992). These advances, coupled with the progressive introduction of simple and rapid methods of extracting DNA suitable for PCR amplification from a variety of media, have undoubtedly facilitated the diagnostic microbiology thus allowing the rapid screening of several strains and the more effective ecological studies of population dynamics and typing of LAB in complex food systems to be undertaken (Cocconcelli et al., 1995; Drake et al., 1996; Vaneechoutte and van Eldere, 1997). In the present study we compared different protocols for rapid extraction of DNA from LAB culture lysates for use in PCR fingerprinting experiments by chelex (iminodiacetid acid) ionic resin. Chelex is an ionic resin that can bind compounds which inhibit PCR (Singer-Sam et al., 1989). Compatibility, repeatability, reproducibility, and consistency of the PCR-fingerprinting patterns obtained by one chelexbased DNA extraction protocol were evaluated and compared with the classical phenol–chloroform DNA extraction procedure.

2. Materials and methods

2.1. Bacterial strains and growth conditions Strains of LAB used in this study are listed in Table 1. Bacterial strains were maintained as frozen stocks at 2 808C in the presence of 150 ml / l glycerol as cryoprotective agent. Working cultures were prepared through three transfers in MRS broth medium (Biokar, Beauvais, France). Strains were cultivated at 308C (for Lactococcus lactis subsp. lactis and Lactobacillus plantarum) and 378C (for Lactobacillus helveticus, Lactobacillus delbrueckii subsp. bulgaricus, and Lactobacillus delbrueckii subsp. lactis) for 16–18 h. Purity was checked by plating on MRS agar medium and microscopic examination. Confirmation of classification from purified cultures before DNA extraction was performed by species-specific PCR according to primers, methods, and amplification conditions previously described (Table 1).

2.2. Phenol–chloroform extraction of total DNA Approximately 100–200 ml of each strain grown overnight in MRS broth ( | 10 7 CFU) were pelleted by centrifugation at 12,500 3 g for 5 min. The pellets were washed twice with sterile water, or TE 0.1 buffer (10 mM Tris–HCl–0.1 mM EDTA, pH

Table 1 Strains investigated in the present study Organism

Lactococcus lactis subsp. lactis Lactococcus lactis subsp. lactis Lactobacillus helveticus Lactobacillus helveticus Lactobacillus delbrueckii subsp. bulgaricus Lactobacillus delbrueckii subsp. bulgaricus Lactobacillus delbrueckii subsp. lactis Lactobacillus delbrueckii subsp. lactis Lactobacillus plantarum a

Collection number a

ATCC 19435 T ISLCPT5 (cheese isolate) ATCC 15009 T LH30 (cheese starter) ATCC 11842 T LB2 (yoghurt) ATCC 12315 T LL5 (cheese whey) ATCC 14917 T

PCR identification Expected amplicons (size in bp)

References

933 933 About 200 About 200 1065 1065 1600 1600 250

Corroler et al. (1998) Corroler et al. (1998) Tilsala-Timisjarvi and Alatossava (1997) Tilsala-Timisjarvi and Alatossava (1997) Torriani et al. (1999) Torriani et al. (1999) Torriani et al. (1999) Torriani et al. (1999) Quere et al. (1997)

A superscript T denotes a type strain; the other strains come from the collection of Istituto Sperimentale Lattiero Caseario of Lodi, Italy (sources of isolation are indicated in parentheses).


8.0), in a clean 1.5-ml microcentrifuge tube and repelleted by centrifugation. Total DNA was extracted from washed cell pellets by a standard alkaline lysis method. Washed cell pellets were resuspended in 500 ml of TES buffer (50 mM Tris– HCl–1 mM EDTA–6.7% saccharose, pH 8.0). Following addition of 10 mg of lysozyme (SigmaAldrich, Milan, Italy), tubes were incubated at 378C for 30 min. In a second step, 125 ml of sodium dodecyl sulfate (20%) were added, followed by an incubation at 378C for 15 min. DNA was extracted three times with equal volumes of phenol and chloroform and precipitated with cold (2208C) isopropanol. After centrifugation at 12,500 3 g for 30 min at 48C, 1 ml of ETOH (70%) was added to pelleted DNA. After a brief spin at 12,500 3 g for 5 min at 48C, purified DNA was dried and resuspended overnight at 48C in 150 ml of TE buffer (10 mM Tris, 1 mM EDTA, pH 8.0). Ribonuclease A (10 mg / ml; Sigma-Aldrich) was added to resuspended DNA and, after incubation at 378C for 1 h, DNA samples were briefly stored at 2 208C until use.

2.3. Chelex-based extraction of total DNA Total DNA from washed cell pellets of each strain was extracted by three chelex-based procedures according to: (A) the DNA extraction protocol described for Lb. helveticus by Drake et al. (1996) and using as the chelex reagent that sold from Sigma (Sigma-Aldrich); (B) the protocol described for the preparation of genomic DNA from bacteria using as the extraction reagent the InstaGene matrix (Bio-Rad Laboratories, Milan, Italy); and (C) the method for Gram positive (and acid-fast) bacteria described in the MicroSeqE protocol (Perkin Elmer, Monza, Italy) which uses the DNA extraction reagent included in the kit. In Fig. 1 are summarized the essential steps of the three chelex-based protocols and the phenol–chloroform extraction procedure.

2.4. Evaluation of quantity and purity of DNA Quantity and purity of DNA samples from alkaline extraction procedures were checked by optical reading at 230, 260, 280, and 320 nm, as described by Sambrook et al. (1989). This enabled us to add the

177

same quantity of DNA from different DNA extraction procedures to each PCR reaction tube. DNA integrity (e.g. no degraded DNA) before PCR and after quantification was visualised by agarose gel electrophoresis in the presence of DNA mass ladders.

2.5. PCR fingerprinting amplification Total DNA from different strains was used as a template for PCR fingerprinting using as a primer the M13 minisatellite core sequence (Huey and Hall, 1989) with sequence 59-GAGGGTGGCGGTTCT-39. Amplification conditions consisted of an initial denaturation step of 948C for 120 s, followed by 40 cycles of: 948C for 60 s, 458C for 20 s, and 728C for 120 s; a final elongation step of 728C for 10 min was performed. PCR was performed in 50-ml amplification mixtures in a Perkin Elmer, mod. 9700 thermoblock (Perkin Elmer) with 50 mM Tris–HCl (pH 8.3), 2.0 mmol / l of primer (Celbio, Milan, Italy), 2.5 ng / ml of total DNA (i.e. between 10 and 20 ml of chelex-based DNA preparations), 3.0 mM MgCl 2 , 1.25 U of Taq polymerase (Perkin Elmer), and 200 mM of each dNTP. PCR profiles were visualised after overnight electrophoresis (1.5 V/ cm) in Seakem GTG agarose gels (1.5% w / v; FMC Bio Products, SPA, Milan, Italy) and staining with ethidium bromide. A 1-kbp plus DNA Ladder (Life Technologies Italia, Milan, Italy) was used as a DNA molecular weight marker.

2.6. Data analysis The photographs of the gels were scanned (Scanjet 6100 C / T, Hewlett Packard, Milan, Italy), and the resulting densitometric traces of the band profiles were analyzed with the pattern analysis software package GelCompar Version 4.1 (Applied Maths, Kortrjik, Belgium). Calculation of similarity of the PCR fingerprinting profiles was based on the Pearson product-moment correlation coefficient. The Pearson correlation coefficient, which provides similarity based upon densitometric curves, was chosen because it is generally more appropriate to evaluate similarities between PCR fingerprinting profiles. A dendrogram was deduced from the matrix of simi-

178


Fig. 1. Schematic presentation of the four protocols, i.e. the three chelex-based methods and the phenol–chloroform method, used for DNA extraction.


larities by using the unweighted pair group method using arithmetic average (UPGMA) clustering algorithm (Vauterin and Vauterin, 1992). The quality of the cluster analysis (i.e. the consistence of the clusters) was verified by calculating the cophenetic correlation value (in percent) for each dendrogram using the GelCompar software. This value, which is calculated for the whole dendrogram, provides an estimation of the faithfulness of a cluster analysis by calculating the correlation between the dendrogramderived similarities and the matrix similarities. The minimum level of repeatability of the amplification conditions was calculated by running DNA samples from duplicated amplifications of each DNA extract. To limit problems of repeatability (and reproducibility) all the samples to be compared were processed at the same time. For each strain, concordance between genotypes obtained using the three chelex-based extraction methods and conventional phenol–chloroform extraction was determined. It was performed by comparison and cluster analysis of the chelex-obtained PCR fingerprinting profiles with those obtained by phenol–chloroform extraction, used as controls. Comparisons were possible because by using GelCompar software the four methods were made compatible, e.g. all data were obtained with one single experimental procedure, processed with a single, operator-independent, normalization procedure and had the same format. In addition, the normalization of the gels performed by the software before the cluster analysis allowed to minimize problems of repeatability and reproducibility caused by variability in the electrophoretic runs. Moreover, cluster analysis between profiles obtained by the chelex-based protocol B (Fig. 1) and the controls allowed to evaluate the following parameters. • Reproducibility. The level of reproducibility of the PCR fingerprinting patterns, which was defined as the level of precision of the method over time, was evaluated by repeated running of DNA samples from duplicated amplifications of each DNA extract at time 0 (T 0) and after 8 weeks of storage at 2208C (T 1). • Consistency. In a given electrophoretic method,

179

the final strain identification is only reliable if the variations caused by differences in growth conditions and DNA extraction parameters are found to be within the ranges of repeatability and reproducibility of the amplification conditions and DNA stability evaluated previously (Pot et al., 1994). Consistency was estimated by running DNA samples from amplifications of duplicated DNA extracts of each strain to determine overlap. Duplicated DNA extracts (DDE) came from two different cultures of the particular strain, grown as described above.

3. Results and discussion PCR fingerprinting profiles using the M13 minisatellite sequence consisted of DNA bands sized between |300 and 2000 bp (Fig. 2, PCR profiles of four type strains shown as example). The cluster analysis applied to all the DNA extracts using Pearson as correlation coefficient allowed us to calculate a minimum repeatability of about 80% for our PCR fingerprinting experiments, which corresponds well to the minimum level of repeatability for a similar technique (i.e. the RAPD technique) verified in a previous investigation (Tailliez et al., 1996). Hence, it was deduced that only clusters with values of the correlation coefficient (expressed as a percentage value) above 80% were considered identical. The numerical analysis of the normalised PCR fingerprinting patterns of each strain (shown as dendrograms in Figs. 3 and 4) allowed to separate profiles into distinct clusters. In Fig. 3 are shown experiments to determine the best overlapping between chelex-based and conventional phenol–chloroform DNA extraction procedures. Seven out of nine PCR fingerprinting profiles from DNAs extracted by chelex-based protocols B and / or C clustered together with controls at percent correlation coefficient values higher than 80. Indeed, controls of only two strains (L. helveticus LH30 and Lc. lactis subsp. lactis ISLCPT5) remained unclustered (Fig. 3). Conversely, the most part (i.e. six out of nine) of the PCR profiles obtained from DNA extracted with protocol A either remained unclustered or did not cluster with the controls (Fig. 3). The calculated value of the

180


Fig. 2. Agarose gel electrophoresis, shown as example, of amplified DNA of some of the strains used in the present study. DNA was extracted by protocols A (lanes 2, 7, 12, and 17), B (lanes 3, 8, 13, and 18), C (lanes 4, 9, 14, and 19) and phenol–chloroform (lanes 5, 10, 15, and 20) as described in Fig. 1. Lanes: 2–5, Lactobacillus delbrueckii subsp. bulgaricus ATCC 11842 T ; 7–10, L. delbrueckii subsp. lactis ATCC 12315 T ; 12–15, L. helveticus ATCC 15009 T ; 17–20, Lactococcus lactis subsp. lactis ATCC 19435 T . Lanes 1, 6, 11, 16, and 21, molecular size DNA marker 1-kbp plus DNA ladder (Life Technologies).

cophenetic correlation coefficient for this dendrogram was 87.5%, indicating good reliability. In most cases, therefore, both B and C protocols allowed to obtain DNAs of a PCR quality similar to that of conventional DNA extraction, as indicated by identical profiles obtained after numerical analysis; the protocol B was chosen for further tests because it was simpler, faster, and more practical than protocol C (Fig. 1). The reproducibility of protocol B was tested by duplicated amplifications of DNAs at time 0 (T 0) and after 8 weeks of storage at 2208C (T 1). A pattern of reproducibility over time for almost all the strains was observed. Except for Lc. lactis subsp. lactis ATCC 19435 T , the duplicate T 0 and T 1 profiles of each strain obtained by procedure B (T 0 and T 1 profiles) and by phenol–chloroform extraction (T 0 and T 1 control profiles) were usually grouped at percent correlation coefficient above the 80% threshold (Fig. 4). A good consistency of the procedure B was also verified, as shown by comparing PCR fingerprinting profiles of DNA samples obtained using freshly

prepared cultures between the procedure B (DDE profiles) and phenol–chloroform extraction (DDE control profiles). Except for L. helveticus LH30, whose DDE control profiles grouped separately from their corresponding DDE profiles at percent correlation coefficient value of 78, the DDE and DDE control profiles of the other strains were included in either single clusters or merged into separate subclusters at percent correlation coefficients higher than 80 (Fig. 4). The cophenetic correlation coefficient (about 90.0%) indicated a good reliability of this cluster analysis. It can also be observed that the PCR fingerprinting analysis performed was confirmed to possess a good capacity to resolve strains into expected taxa. Distinct clusters (and / or sub-clusters), corresponding well with species and subspecies established by species-specific PCR, were in fact obtained at percent similarity values below 80 (Fig. 4). Previous studies have demonstrated the importance of colony age and other growth conditions as major factors affecting the quality and reproducibility of bacterial fingerprints obtained by PCR


181

Fig. 3. PCR fingerprinting profiles of strains and reference strains of various lactic acid bacteria obtained by DNA extracted with different chelex-based protocols and with phenol–chloroform and corresponding dendrogram derived from the unweighted pair group average linkage of Pearson correlation coefficients (expressed as a percentage value). The similarity value of 80% is indicated by a vertical dotted line. On the right hand side of the figure are the strain numbers, the indication of the protocol used to obtain the corresponding profile, and the cluster delineation by species. Abbreviations: L., Lactobacillus; Lc., Lactoccocus; subsp., subspecies; prot. A, prot. B, and prot. C, profiles obtained with DNA extracted by protocols A, B, and C as described in Fig. 1; control, profiles obtained with DNA extracted by phenol–chloroform as described in Section 2.

(Coutinho et al., 1993; Wilson, 1997). Therefore, provided that bacterial growth conditions were optimised and bacterial purity was checked by speciesspecific PCR, the procedure B: (i) provides stable DNA, giving over time reproducible PCR fingerprinting profiles; (ii) provides DNA giving PCR fingerprinting profiles similar to those obtained by DNA extracted with a phenol–chloroform technique; (iii) has a high consistency, similar to that observed using a phenol–chloroform technique, thus providing a reliable identification because the variations caused by differences in growth conditions were found to be within the limits of repeatability and reproducibility of the procedure. The use of new stocks of freshly prepared M13 primer gave comparable results and no loss of

repeatability and reproducibility of the PCR fingerprinting profiles were observed (data not shown). The use of the PCR with arbitrary nucleotide primers to produce DNA fingerprints of cultivable bacterial strains has revolutionised population studies of microorganisms. Concerning LAB, there is now a large literature on procedures for analysing a variety of species (Andrighetto et al., 1998; Cocconcelli et al., 1995; Corroler et al., 1998; Drake et al., 1996; Klein et al., 1998; Moschetti et al., 1998; Tailliez et al., 1996, 1998; Torriani et al., 1999; Van Reenen and Dicks, 1996). However, although most reports involve the use of purified DNA as templates, recent studies describe the use of chelating agents, such as chelex or other synthetic resins, for rapid extraction of a DNA suitable for reproducible PCR amplifica-

182


Fig. 4. PCR fingerprinting profiles of strains and reference strains of various lactic acid bacteria obtained by DNA extracted with chelex-based protocol B (see Fig. 1) and phenol–chloroform extraction and corresponding dendrogram derived from the unweighted pair group average linkage of Pearson correlation coefficients (expressed as a percentage value). The similarity value of 80% is indicated by a vertical dotted line. On the right hand side of the figure are the strain numbers, the indication of the DNA sample used to obtain the corresponding profile, and the cluster delineation by species. Abbreviations: L., Lactobacillus; Lc., Lactoccocus; subsp., subspecies; T 0, DNA extracted at time 0; T 1, DNA extracted after 8 weeks of storage at 2208C (see Section 2); DDE, duplicated DNA extract; control, profiles obtained with DNA extracted by phenol–chloroform as described in Section 2.


tions from LAB (Cocconcelli et al., 1995; Drake et al., 1996). The continuously accumulating set of PCR fingerprinting data and the construction of a reliable database require a high degree of standardisation in experimental methodology. If different protocols for DNA extraction are used, concordance studies between methods should be performed, whereas using a common protocol there probably exists as many electrophoretic variants in experimental procedures as there are research groups. Even within a single laboratory, reproducibility of the DNA extraction procedure, PCR assay, and running conditions have to be checked and corrected if necessary. Concerning DNA extraction by chelex, a growing number of chelex-based protocols and kits are now commercially available with different degrees of complexity and reproducibility, but concordance studies between techniques are still scarce. Therefore, comparative evaluation of these DNA extraction and purification techniques may be of interest. In the present study, factors affecting the quality and reproducibility of PCR fingerprinting patterns obtained from different DNA samples extracted from MRS-broth-cultured LAB were investigated. We compared first different chelex-based protocols for PCR fingerprinting of LAB strains (either typestrains or food isolates) belonging to food-associated species with classical phenol–chloroform DNA extraction technique. Generally, chelex-based protocols enabled to extract PCR-grade DNAs which, after amplification, allowed to obtain electrophoretic patterns comparable with those obtained by classical lysozyme and phenol–chloroform extractions. This confirmed previous findings on PCR fingerprinting typing of Lb. helveticus strains using a chelex-based protocol to extract DNA and the same PCR primer (Drake et al., 1996). Experiments to determine pattern reproducibility and consistency after amplification of DNA extracted by the InstaGene, chelex-based procedure indicated its reliability as a valid alternative to conventional DNA extraction. As compared with other commercially or literature available chelex-based techniques, the greatest advantage of the InstaGene procedure is that it is faster and more practical. Although applied to a little number of LAB strains in the present study, chelex-based protocols are applicable to a

183

wider range of organisms (Singer-Sam et al., 1989; Walsh et al., 1991; Drake et al., 1996). Preliminary experiments in our laboratory with different strains and PCR primers seem to confirm the suitability of this procedure to type a large number of LAB isolates (unpublished results). Therefore, we propose this as an easy protocol to extract DNA from cultivable LAB to be used as template for PCR fingerprinting. The protocol is also rapid, e.g. 40 min are needed to process 15 samples by the proposed InstaGene, chelex-based method and 240 min by the phenol–chloroform purification method. We suggest also a methodology to follow for demonstrating reproducibility and consistency of PCR fingerprinting profiles obtained with differently extracted DNA samples.

Acknowledgements We are very grateful to Catia Ferri (University of Parma, Italy) for her technical assistance.

References Andrighetto, C., De Dea, P., Lombardi, A., Neviani, E., Rossetti, L., Giraffa, G., 1998. Molecular identification and cluster analysis of homofermentative thermophilic lactobacilli isolated from dairy products. Res. Microbiol. 149, 631–643. Cocconcelli, P.S., Porro, D., Galandini, S., Senini, L., 1995. Development of RAPD protocol for typing strains of lactic acid bacteria and enterococci. Lett. Appl. Microbiol. 21, 376–379. Corroler, D., Mangin, I., Desmasures, N., Gueguen, M., 1998. An ecological study of lactococci isolated from raw milk in the Camembert cheese registered designation of origin area. Appl. Environ. Microbiol. 64, 4729–4735. Coutinho, H.L.C., Handley, B.A., Kay, H.E., Stevenson, L., Beringer, J.E., 1993. The effect of colony age on PCR fingerprinting. Lett. Appl. Microbiol. 17, 282–284. Drake, M., Small, C.L., Spence, K.D., Swanson, B.G., 1996. Differentiation of Lactobacillus helveticus strains using molecular typing methods. Food Res. Int. 29, 451–455. Dykes, G.A., von Holy, A., 1994. Strain typing in the genus Lactobacillus. Lett. Appl. Microbiol. 19, 63–66. Huey, B., Hall, J., 1989. Hypervariable DNA fingerprinting in E. coli minisatellite probe from bacteriophage M13. J. Bacteriol. 171, 2528–2532. Klein, G., Pack, A., Bonaparte, C., Reuter, G., 1998. Taxonomy and physiology of probiotic lactic acid bacteria. Int. J. Food Microbiol. 41, 103–125. Moschetti, G., Blaiotta, G., Aponte, M., Mauriello, G., Villani, F.,

184


Coppola, S., 1997. Genotyping of Lactobacillus delbrueckii subsp. bulgaricus and determination of the number and forms of rrn operons in L. delbrueckii and its subspecies. Res. Microbiol. 148, 501–510. Moschetti, G., Blaiotta, G., Aponte, M., Catzeddu, P., Villani, F., Deiana, P. et al., 1998. Random amplified polymorphic DNA and amplified ribosomal DNA spacer polymorphism: powerful methods to differentiate Streptococcus thermophilus strains. J. Appl. Microbiol. 85, 25–36. Pot, B., Vandamme, P., Kersters, K., 1994. Analysis of electrophoretic whole-organism protein fingerprints. In: Goodfellow, M., O’Donnell, A.G. (Eds.), Chemical Methods in Prokaryotic Systematics. John Wiley, New York, pp. 493–521. Quere, F., Deschamps, A., Urdaci, M.C., 1997. DNA probe and PCR-specific reaction for Lactobacillus plantarum. J. Appl. Microbiol. 82, 783–790. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. In: 2nd Edition. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, New York. Singer-Sam, J., Tanguay, R.L., Riggs, A.D., 1989. Use of chelex to improve the PCR signal from a small number of cells. Amplifications 3, 11. Stiles, M.E., Holzapfel, W.E., 1997. Lactic acid bacteria of foods and their current taxonomy. Int. J. Food Microbiol. 36, 1–29. ´ ´ P., Chopin, A., 1996. Estimation de la Tailliez, P., Quenee, diversite` parmi les souches de la collection CNRZ: application de la RAPD a` un groupe de lactobacilles. Lait 76, 147–158. Tailliez, P., Tremblay, J., Ehrlich, D., Chopin, A., 1998. Molecular diversity and relationship within Lactococcus lactis as revealed

by randomly amplified polymorphic DNA (RAPD). System. Appl. Microbiol. 21, 530–538. Tilsala-Timisjarvi, A., Alatossava, T., 1997. Development of oligonucleotide primers from the 16S-23S rRNA intergenic sequences for identifying different dairy and probiotic lactic acid bacteria by PCR. Int. J. Food Microbiol. 35, 49–56. Torriani, S., Zapparoli, G., Dellaglio, F., 1999. Use of PCR-based methods for rapid differentiation of Lactobacillus delbrueckii subsp. bulgaricus and L. delbrueckii subsp. lactis. Appl. Environ. Microbiol. 65, 4351–4356. Vandamme, P., Pot, B., Gillis, M., De Vos, P., Kersters, K., Swings, J., 1996. Polyphasic taxonomy, a consensus approach to bacterial systematics. Microbiol. Rev. 60, 407–438. Vaneechoutte, M., van Eldere, J., 1997. The possibilities and limitations of nucleic acid amplification technology in diagnostic microbiology. J. Med. Microbiol. 46, 188–194. Van Reenen, C.A., Dicks, L.M.T., 1996. Evaluation of numerical analysis of random amplified polymorphic DNA (RAPD)-PCR as a method to differentiate Lactobacillus plantarum and Lactobacillus pentosus. Curr. Microbiol. 32, 183–187. Vauterin, L., Vauterin, P., 1992. Computer aided objective comparison of electrophoresis patterns for grouping and identification of microorganisms. Eur. Microbiol. 1, 37–41. Walsh, P.S., Metzger, D.A., Higuchi, R., 1991. Chelex  100 as a medium for simple extraction of DNA for PCR-based typing from forensic material. Biotechniques 10, 506–513. Wilson, I.G., 1997. Inhibition and facilitation of nucleic acid amplification. Appl. Environ. Microbiol. 63, 3741–3751.