| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,1
* School of Molecular Sciences, Victoria University, PO Box 14428, Melbourne City Mail Centre, Victoria 8001, Australia
Australian Starter Culture Research Centre Limited, 180 Princes Highway, Werribee, Victoria 3030, Australia
1 Corresponding author: powell{at}ascrc.com.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: Lactobacillus casei • Lactobacillus paracasei • Lactobacillus rhamnosus • Lactobacillus zeae
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Strains from this group belong to the normal intestinal flora and potentially have beneficial probiotic effects on human and animal health (Ouwehand et al., 2002), but some strains have been observed to be involved in opportunistic infections (Adams, 1999). Some bacteria of this group are major components of starter cultures for cheeses and yogurts, and are also found naturally in raw milk and in high numbers in matured cheeses, where they contribute to positive and negative aspects of cheese maturation and flavor development (Khalid and Marth, 1990; Fitzsimons et al., 1999).
The classification scheme applied to these bacteria has changed considerably in recent decades. The changes are largely based on improved knowledge of gene sequences, and are intended to make the classification scheme accurately reflect natural bacterial groups and the evolutionary relationships between them. Such changes are entirely sensible to taxonomists and evolutionary biologists but can complicate interpretation of the literature and cause confusion for industrial users of these bacteria and for those marketing or consuming products containing them.
Several DNA homology studies (e.g., Collins et al., 1989) resulted in 3 species being recognized (L. casei, L. paracasei, and L. rhamnosus) instead of a single "L. casei" taxon. Within a few years, further studies resulted in the proposal of a new species, L. zeae (Dicks et al., 1996), and in the suggestion (Dicks et al., 1996; Dellaglio et al., 2002) that some strains currently classified as L. casei and L. zeae should be regarded as the same species (and given the name L. zeae) and that many strains currently known as L. casei and L. paracasei should be reclassified into a single species (and given the name L. casei, with the name L. paracasei abandoned). The Judicial Commission of the International Committee on Systematics of Prokaryotes has yet to publish an opinion on this proposal (De Vos et al., 2005). The recent literature must be interpreted carefully, because some authors have adopted this new classification (e.g., Dellaglio et al., 2002; Dobson et al., 2004) but others have not (e.g., Mori et al., 1997; Ward and Timms, 1999; Vásquez et al., 2005; this study). Commercial strain descriptions often ignore the recommendations of taxonomists, and the classification "L. casei" is sometimes loosely applied to strains of any of these species.
In industrial application of lactobacilli, it is important to have tools with which to identify particular strains, to understand the relationships between strains, to monitor the genetic stability of strains, and to classify them into recognizable species with due reference to the current taxonomy of the organisms. As part of broader work on the potential probiotic properties of lactobacilli (Desai et al., 2004), the aims of this study were to examine the effectiveness of carbohydrate fermentation tests, species-specific PCR, and pulsed-field gel electrophoresis (PFGE) of genomic restriction fragments for discriminating food industry strains of the L. casei group.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
) were cultured using de Man, Rogosa, and Sharpe (MRS) broth (static cultures) or MRS agar (Oxoid, Basingstoke, UK) at 37° C. All strains grew well without the use of anaerobic conditions. Type strains (i.e., recognized reference strains) were included in the study. Species epithets are used in this report in accordance with the 4-species scheme (L. casei, L. zeae, L. paracasei, L. rhamnosus), with cross-reference to the alternative 3-species scheme (L. casei, L. zeae, L. rhamnosus) where appropriate.
|
. Oligonucleotide primers used for PCR were synthesized by Life Technologies (Melbourne, Victoria, Australia). Polymerase chain reactions (50 µL, using reagents from Roche Diagnostics, Castle Hill, New South Wales, Australia) were performed in 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2, and 50 mM KCl in the presence of Taq DNA polymerase (1 U), deoxynucleotide triphosphates (200 µM each), 2 primers (0.5 µM each), and 1 µL of bacterial genomic DNA prepared (by the method of Pu et al., 2002) from a fresh colony grown on the surface of MRS agar. A PTC thermocycler (MJ Research, Waltham, MA) was used. Initial denaturation was carried out at 94° C for 3 min, followed by 30 cycles of 45 s at 50° C, 60 s at 72° C, and 45 s at 94° C, with a final product extension for 5 min at 72° C. The PCR products were analyzed by 1% (wt/vol) agarose gel electrophoresis in 0.5 x Tris-borate-EDTA buffer with ethidium bromide (0.5 µg/mL).
|
PFGE
Preparation of Genomic DNA in Agarose Blocks.
Bacterial genomic DNA was prepared from 1 mL of MRS broth overnight culture, using a modification of the method of Tanskanen et al. (1990). Cells were harvested centrifugally, washed, and then suspended in 150 µL of NaCl-EDTA-Tris solution (1 M NaCl, 10 mM EDTA, 10 mM Tris-HCl, pH 8.0) at 45° C. The cell suspension was mixed with 200 µL of low-melting-temperature agarose (2.5% wt/vol; BioRad Laboratories, Regents Park, New South Wales, Australia) at 45° C, and then dispensed into molds (Amersham Pharmacia Bio-tech, Castle Hill, New South Wales, Australia) and allowed to solidify. Cells immobilized in agarose blocks were incubated overnight at 37° C in lysis buffer containing 100 mM EDTA, 1 M NaCl, 10 mM Tris-HCl (pH 8.0), 1% lauroylsarcosine, 10 mg/mL of lysozyme (Sigma, St. Louis, MO), and 30 U/mL of mutanolysin (Sigma). Treatment with proteinase K (5 mg/mL; ICN Biochemicals, Seven Hills, New South Wales, Australia) was performed for 4 d at 42° C in 100 mM EDTA (pH 8.0) and 1% lauroylsarcosine.
Restriction Enzyme Digestion and PFGE.
A 1-mm slice was cut from each agarose block using a sterile razor blade. Slices were equilibrated with 1 mM EDTA, 10 mM Tris-HCl (pH 8.0) for 1 h, then with 2 changes (45 min each) of NotI restriction enzyme buffer (Roche). Restriction enzyme digestion was carried out with 5 U of NotI at 37° C for 18 to 20 h.
Electrophoresis was carried out with an LKB Pulsaphor CHEF apparatus (Amersham Pharmacia Biotech) in 1.3% agarose with 0.5 x Tris-borate-EDTA buffer at 14° C and at 290 V. Field pulse times were increased from 1 to 10 s over 6 h, then from 10 to 20 s over 4 h. The agarose gel was stained with ethidium bromide (0.5 µg/mL) in 0.5 x Tris-borate-EDTA buffer for 1 h. Then, DNA bands were visualized using 302-nm UV transillumination and photographed with a MultiDoc-It digital imaging system (UVP, Cambridge, UK). Electrophoretic band patterns were compared by Jaccard-UPGMA (unweighted pair group method with arithmetic mean) analysis using GelCompar II software (Applied Maths, Sint-Martens-Latem, Belgium) with band-matching optimization and tolerance settings of 1%.
Carbohydrate Fermentation Tests
To classify strains biochemically, API 50CHL test plates and APILABplus software with identification database version 4.0 (BioMerieux, Marcy-l’Etoile, France) were used. A further Jaccard-UPGMA analysis was performed on tabulated data using FreeTree 0.9.1.50 (Pavlicek et al., 1999) and TreeView 1.6.6 (Page, 1996) software.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
) targeting the 16S rRNA gene and the spacer region between the 16S and 23S rRNA genes were used in this study. Some of these primers had previously been used in limited studies that did not include a full set of species type strains and in which the results of PCR were not corroborated by sequence analysis of the bacterial strains used. In the present study, a gene region that included the sequences targeted by primers W1, W2, W3, and D1 was PCR amplified from each of the type strains using the Y1-Y2 primer pair, and the DNA sequence from each strain was determined. Sequences were consistent with relevant GenBank sequence entries AY196978 (ATCC 393), D86517 (ATCC 334), D16552 (ATCC 7469), and D86516 (ATCC 15820).
Polymerase chain reaction was carried out using template DNA from each of the type strains (Table 1) with primer pairs W1-Y2, W2-Y2, W3-Y2, and D1-Y2. In these experiments, the common primer Y2 was used with different specific primers designed on the basis of sequence "signatures" from the V1 region of the 16S rRNA gene (Mori et al., 1997; Ward and Timmins, 1999; GenBank D86516). A PCR product of approximately 295 bp was amplified from each type strain, depending which specific primer was used (Table 3). It was therefore shown that these primer pairs could distinguish the 4 bacterial type strains.
|
). Most of these strains had been isolated and phenotypically classified by food industry laboratories between the years 1975 and 1998, and ASCC 277 was derived from a strain acquired by the National Collection of Industrial, Marine and Food Bacteria (Bucksburn, Aberdeen, UK in 1958. Ten of these strains had historically been classified as "L. casei," a classification that could at the time have included strains that now would be classified as L. casei (L. zeae under recent proposals), L. paracasei (L. casei under recent proposals), L. zeae, or L. rhamnosus.
Using the PCR assay, 13 of the 17 strains were classified as L. rhamnosus, including 9 of the 10 strains historically described as L. casei. Confirmation of the classification of these strains was achieved by testing all strains in PCR with primer pair Rha(F)–Rha(R) (Table 2). This primer pair and the W3–Y2 pair gave the same classification of L. rhamnosus strains (results not shown).
Four strains gave positive PCR with the W2–Y2 primer pair, resulting in classification as L. paracasei (L. casei under recent proposals). None of the industry isolates gave positive PCR with primer pairs W1–Y2 or D1–Y2. This is consistent with molecular taxonomic studies that have defined few strains of L. casei or L. zeae.
PFGE
Pulsed-field gel electrophoresis of genomic NotI restriction fragments was carried out to discriminate strains at the genetic level. Tynkkynen et al. (1999) used this method and concluded that it was a powerful technique for discriminating L. rhamnosus strains. In their study involving human isolates of L. rhamnosus, they found several isolates to be indistinguishable from probiotic strain L. rhamnosus GG (Saxelin, 1997), perhaps indicating widespread consumption of this commercially produced strain.
In the present study, inspection revealed 18 different NotI restriction fragment patterns from the 21 independent strains studied (Figure 1), showing that most of the strains were genuinely different from each other and distinguishable by this method. This complements the PCR methods used, which distinguished L. paracasei (L. casei under recent proposals) and L. rhamnosus strains at the species level.
|
Comparison of small-fragment restriction enzyme patterns was previously shown to give a useful species distinction of L. paracasei from L. rhamnosus and of the L. zeae type strain from other species, but the L. casei type strain could not be distinguished from L. rhamnosus (Vásquez et al., 2005). In the present study using large-fragment PFGE (with an enzyme chosen because it shows differences between strains), the species could not be separated (data not shown). However, after using PCR to make a species assignment, pattern similarities (close strain relationships) within L. rhamnosus were readily identified (Figure 1B) and the distinction of L. paracasei strains and the type strains of L. casei and L. zeae could be achieved (Figure 1A). The similarity index (Figure 1) indicates that L. casei and L. zeae are perhaps no more distant from each other than are some strains within L. paracasei or L. rhamnosus.
Biochemical Classification
Biochemical identification results from API 50CHL testing are shown in Table 1. Lactobacillus paracasei (L. casei under recent proposals) and L. rhamnosus classifications were consistent with classifications from the PCR assays with one clear exception (ASCC 279). The internal quality assessment feature of the API software judged 2 classifications (ASCC 526 and CSCC 2603) as "unacceptable," implying that the pattern analysis software was unable to make a clear classification. Neither L. casei nor L. zeae is classifiable using the API proprietary database and software, presumably because few strains have been available for study. The L. casei and L. zeae type strain fermentation patterns differed from typical L. paracasei fermentation patterns by 6 to 9 sugars fermented or not fermented, and by 5 sugars from each other. Even so, Jaccard-UPGMA analysis of the fermentation test results was also unable to separate these strains from the other species or from each other (not shown). Such observations illustrate the limitations of biochemical profiling when applied to closely related bacteria with intraspecies heterogeneity.
Taxonomic Considerations
The phylogenetic relationships between the various bacteria of this group are generally not in dispute, but there is debate over where species boundaries should be drawn and what names should be attached to those species. The results of this study are consistent with the prevailing view that 1) L. rhamnosus is a common and readily distinguishable species, 2) a second common and readily distinguishable species exists, herein referred to as L. paracasei, but proposed by some (Dellaglio et al., 2002) to be known as L. casei, and 3) relatively uncommon types also exist. These L. casei–L. zeae types are undoubtedly closely related to each other (Mori et al., 1997; Dobson et al., 2004; Vásquez et al., 2005), plausibly justifying reclassification of these bacteria into a single species (Dellaglio et al., 2002).
At the heart of the nomenclature debate is the proposal to reclassify the type strain of L. casei (ATCC 393 and equivalents, in this study ASCC 290 and CSCC 5203) as L. zeae. Species are conventionally defined largely in terms of their type strains, and so this reclassification would be a highly unusual step. The phylogenetic relationships could equally be accommodated by merging L. casei and L. zeae under the name L. casei (with ATCC 393 as the type strain) and reclassifying some strains currently held in collections as L. casei to L. paracasei. This would not involve the controversial act of removing a type strain from its species.
Practical Species and Strain Discrimination
Mori et al. (1997) proposed that the sequences of 16S rRNA genes could distinguish species within the L. casei group, and they established "signature" sequence features that could be used as PCR targets for species-specific product amplification (e.g., Ward and Timmins, 1999). This approach has been extended in the present work to permit 4 groups to be discriminated. Polymerase chain reaction conditions (particularly the annealing temperature; data not shown) are crucial to achieving clear discrimination, and might require adjustment for use with different PCR reaction mixes or thermal cycling devices. The use of 4 separate PCR reactions (plus a negative control to which no template DNA is added) provides a set of positive-negative test results that are easily interpreted and that provide immediate confirmation of PCR specificity or of any cross-reaction. The methods used in this study (PCR followed by electrophoresis) are now available to many food laboratories, and the ease of interpretation is a considerable advantage over conventional tests such as fermentation pattern analysis. The species-specific primers could also form the basis of a "probe oligonucleotide" set for use in a single-tube quantitative PCR assay using fluorescence-detection real-time PCR. The positive-negative aspect of the test would also be a useful confirmation of specificity in a real-time assay system.
The combination of 4-reaction PCR and PFGE is a workable approach for strain classification and identification, but the results must be interpreted carefully. The results obtained in this study are consistent with the conclusions of Mori et al. (1997), Ward and Timmins (1999), and Vásquez et al. (2001) that L. paracasei and L. rhamnosus could readily be identified and that L. casei could be distinguished from L. zeae on the basis of aspects of their 16S rRNA gene sequences. Both are readily distinguished from L. paracasei and L. rhamnosus. However, Vásquez et al. (2005) reported a more detailed sequence analysis showing that the multiple copies of the 16S rRNA gene within a single strain are not always identical. Notably, intrastrain sequence heterogeneity was observed within the type strains of L. casei, L. paracasei, and L. zeae. Interstrain sequence polymorphism was observed within these species and within L. rhamnosus, further complicating interpretation of available sequence data and identification of unambiguous signature sequences. Some of this heterogeneity is within the region targeted by the primers W1, W2, W3, and D1 (at positions 96, 97, 99, 105, and 110 as described by Vásquez et al., 2005). The results of the present study suggest this heterogeneity does not interfere with the PCR-based classification of strains of L. paracasei and L. rhamnosus, nor does it interfere with distinguishing between the type strains of L. casei and L. zeae. The full extent of inter- and intrastrain heterogeneity within the latter species (whether they are regarded as 2 species or merged into a single species) is not known because so few strains have been studied. It is possible that the primers W1 and D1 target only 2 of many possible sequence variants. Evaluation with other strains would be required to ensure that no problems existed, but the multiple positive—negative aspect of the test scheme would alert the experimenter to any cross-reactions that might occur.
Reclassification of Archive Isolates
Strains described in the historical literature or held in archival collections require reexamination and possible reclassification before they can properly be described and compared with recent isolates and the recent literature. In this study, 9 strains historically described as L. casei were reclassified as L. rhamnosus.
Ward and Timmins (1999) examined "L. casei group" isolates from cheeses and reported L. rhamnosus and a preponderance of L. paracasei, but no L. casei. The isolates tested in the present study came from cheeses but also from a range of noncheese sources, with L. rhamnosus as the most common species. Considered together, these studies indicate that L. casei (as currently defined) is not common in dairy environments or in dairy manufacturing. Lactobacillus paracasei and L. rhamnosus are the common cheese isolates from the L. casei group and, in this study, L. rhamnosus was identified from cultures and other dairy sources.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication August 22, 2005. Accepted for publication March 27, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Coudeyras, H. Marchandin, C. Fajon, and C. Forestier Taxonomic and Strain-Specific Identification of the Probiotic Strain Lactobacillus rhamnosus 35 within the Lactobacillus casei Group Appl. Envir. Microbiol., May 1, 2008; 74(9): 2679 - 2689. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
