LactobaciJlus Streptococcus Enterococcus Streptococcus Enterococcus
Fitzsimons, et al. 1999; Simpson, et al. 2002 Mangin, et al. 1999; Nomura, et al. 2006 Matte-Tailliez, et al. 2002 Taffliez, et al. 1998
Cusick and O'Sullivan 2000; Jenkins, et al. 2002
Taffliez, et al. 1998; Mangin, et al. 1999; Cibik, et al. 2000; Bouton, et al. 2002; Nomura, et al. 2006
Poznanski, et al. 2004
Moschetti, et. al. 2000; Coppola, et al. 2006
Versalovic, et al. 1991; Callon, et al. 2004 Bouton, et al. 2002
Versalovic, et al. 1991; Callon, et al. 2004; Bouton, et al. 2002 Zamfir, et al. 2006
Paffetti, et al. 1995; Andrighetto, et al. 2000 Ward and Timmins 1999
Moschetti, et al. 1998
Moschetti, et al. 1998; Mora, et al. 2002a; Cosentino, et al. 2004; de Vin, et al. 2005
" Some primers were grouped (different authors named primers showing the same sequence with different names) some others were renamed (primers with different sequences were named with same name).
REA-PFGE, performed by rare cutting endonucleases, has also been widely applied to type LAB isolates. The choice to use the endonuclease is of crucial importance to obtain reliable differentiation of the isolates. Endonuclease Sma I was used to type streptococci (Moschetti, et al. 1997; O'Sullivan and Fitzgerald 1998), enterococci (Gelsomino, et al. 2001; Psoni, et al. 2006), L. lactis (Moschetti, et al. 2001; Mannu, et al. 2000b; Delgado and Mayo 2004) and Lb. helveticus (Coppola, et al. 2006), Apa I for typing Ln. mesenteroides (Villani, et al. 1997), Not I for typing Lb. delbrueckii (Moschetti, et al. 1997; Giraffa, et al. 2004) and Pediococcus spp. (Simpson, et al. 2002). In some cases more than one endonuclease was used. Patterns obtained by Sma I and Apa I were analyzed by Jenkins, et al. (2002) to differentiate Swiss cheese starter culture strains of Lb. helveticus, St. thermophilus and Prop. Freudenreichii, and by Delgado and Myo (2004) to evaluate genetic diversities of Lc. lactis and Enterococcus spp. isolated from Spanish starter-free cheeses. Sgr AI and Xho I were applied by Bouton, et al. (2002) to monitor Lb. helveticus and Lb. delbrueckii subsp lactis strains isolated during Comtè cheese ripening. Simpson, et al. (2002) evaluated the discrimination power of different endonucleases (Apa I, Sma I, Asc I, Not I, Sfi I) for differentiation of Pediococcus spp. strains. REA-PFGE, albeit a laborious and expensive method, is highly reproducible and is, therefore, considered to offer the highest resolution for strain differentiation of LAB. Generally, analysis of RAE-PFGE patterns obtained by one well-chosen enzyme can provide fine, reliable differentiation. However, it has been suggested that analysis of two or three restriction enzymes should be used to differentiate Lactobacillus strains (Vancanneyt, et al. 2006c).
Blaiotta, et al. (2001) used REA-PFGE to monitor the addition of LAB, used as starter, to Cacioricotta cheese. By analyzing isolates from different phases of the fermentation the technique made it possible to evaluate the growth kinetics of each starter strain during the process.
Bouton, et al. (2002) used fingerprinting PCR-based methods and PFGE for typing and monitoring homofermentative lactobacilli during Comté cheese ripening. Isolates, which exhibited unique patterns by RAPD or REP-PCR, were distinguishable by PFGE. By contrast, some strains which were distinguishable by RAPD or REP-PCR were related by PFGE. These discrepancies were explained by the different exploration of DNA polymorphism (the whole DNA chromosome for PFGE, and region amplified by primers for RAPD and the REP-PCR). The use of second restriction enzymes would certainly be useful in this case.
Jenkins, et al. (2002), in analyzing genetic diversity in Swiss cheese starter cultures, found that strains with > 87 percent similarity by REA-PFGE consistently had the same acidification rate.
As it is a time-consuming technique, REA-PFGE was applied when fine strain typing was needed, when a small number of isolates have to be typed and when other strain typing techniques may be unreliable. Therefore, in many cases, it is applied as a supplementary technique to confirm or improve results obtained by other typing methods. Moschetti, et al. (1997) analyzed Not I-REA-PFGE patterns of Lb. delbrueckii subsp. bulgaricus isolated from commercial yogurt and showed that some strains isolated from products of different dairies displayed the same pattern, suggesting that different dairies used the same starter. Similar results were obtained by Vancanneyt, et al. (2006c) who applied REA-PFGE to confirm and/or improve results obtained by AFLP analyzing Lb. rhamnosus strains isolated from different commercial probiotic preparation. Coppola, et al. (2006) analyzed Lb. helveticus strains isolated during manufacture of fior di latte cheese by RAPD-PCR and SmaI-REA-PFGE. Of 55 strains only four RAPD-PCR profiles were found by using primer Primm 239 (reliably used to differentiate Lc. lactis strains). Therefore, for a more appropriate biotyping, SmaI-REA-PFGE was applied. Using this last technique, a total of 13 different patterns were found. Also in this case, as already shown in Lc. lactis strains, strains showing the same profile were found in milk, in curd at the beginning of ripening and in curd at the end of ripening.
Overall, the most reliable method for strain differentiation is still REA-PFGE analysis and, therefore, its application is going to be fundamental for the monitoring of microorganisms in dairy processing.
The AFLP technique was used only to differentiate and characterize some species of the Lb. plantarun group (Lb. plantarum, Lb. pentosus and Lb. para-plantarum) by Torriani, et al. (2001a). Fluorescent AFLP (FAFLP) was also applied to type probiotic Lb. rhamnosus strains by Vancanneyt, et al. (2006c). The AFLP technique normally displays good levels of reproducibility and reliability -apart from some reported problems related to the initial DNA concentration or to the endonuclease or ligase treatment efficiency - but it is quite laborious and time-consuming, given that it requires two enzymatic reactions and large polyacrylamide gels to reach a good level of band separation. Although the observed strain-to-strain variations in the FAFLP patterns within a given cluster may reflect strain-specific differences, such variations are, in most cases, introduced during data processing. Therefore, for strain typing, FAFLP should be complemented by other fingerprinting techniques such as PFGE (Vancanneyt, et al. 2006c). However, FAFLP performed by multiple primer combination has proved to be a valid and powerful tool to reveal intraspecies diversities (Vancanneyt, et al. 2002). Recently, a simplified AFLP technique, called Sau-PCR, was applied to LAB fingerprinting (Corich, et al. 2005). Results suggest that Sau-PCR may be considered for DNA fingerprinting based on analyses as a possible alternative to the RAPD technique in cases where reproducibility or polymorphism levels are not satisfactory, and as an alternative to the AFLP technique, but with lower costs in terms of time and equipment, when a restriction-plus-amplification approach is preferred. However, AFLP, FAFLP and Sau-PCR were never used for typing large numbers of isolates from the dairy environment.
SDS-PAGE of WCPs was also applied to characterize cultivable dairy microflora. Villani, et al. (1997) evaluated diversities of Ln. mesenteroides strains isolated from dairy and non-dairy environments; Moschetti, et al. (1997) of Lb. delbrueckii isolated from yogurt, raw and pasteurized milks; Rossi, et al. (1998) propionibacteria from different dairy sources; Silva, et al. (2003) isolated enterococci from an artisanal Portuguese cheese, and Delgado and Mayo (2004) isolated lactococci and enterococci from Spanish starter-free cheeses. Piraino, et al. (2005) and Zamfir, et al. (2006) applied SDS-PAGE of WCPs to identify and characterize LAB occurring in caciocavallo cheeses and Romanian dairy products, respectively. Finally, Piraino, et al. (2006) applied unsupervised and supervised artificial neural networks for the identification of LAB (Lactobacillus, Leuconostoc, Enterococcus, Lactococcus and Streptococcus) on the basis of their SDS-PAGE of the WCP pattern. SDS-PAGE of surface proteins was applied by Gatti, et al. (2004) to differentiate Lb. helveticus strains isolated from different natural whey starter cultures. However, there is some evidence of poor differentiation of some LAB species by this technique. De Angelis, et al. (2001), analyzing LAB of 12 Italian ewe's milk cheeses, showed that strains of Lb. plantarum and Lb. pentosus grouped in the same cluster. Gancheva, et al. (1999), analyzing a set of 98 strains belonging to nine species of the Lb. acidiphilus group had difficulty differentiating between L. johnsonnii and Lb. gasseri strains, and between those of Lb. gallinarum and Lb. amylovorus.
However, statistical analysis of SDS-PAGE of WCPs provides an effective tool for the classification and identification of LAB (Piraino, et al 2005 and 2006). By applying this technique Piraino, et al. (2005) demonstrated the possibility of discriminating PDO cheeses from non-PDO and showed that the microflora of PDO cheeses was less heterogeneous than that of non-PDO cheeses, and consisted mainly of non-starter LAB. Finally, in some cases the discrimination power of this technique was comparable to that of REA-PFGE (Delgado and Mayo 2004).
In addition to the above typing options, some authors have developed assays targeting genes encoding for key proteins or enzymes in food-borne bacteria. Gatti, et al. (2005) evaluated diversities of surface layer (S-layer) protein genes in Lb. helveticus strains and demonstrated that heterogeneity exists in genes of this species. However, cluster analysis of the sequences separated strains into only two main clusters. Ercolini, et al. (2005) evaluated sequence diversities of lacZS operon of dairy St. thermophilus strains. Due to sequence polymorphism it was possible to design PCR-DGGE and PCR-SSCP systems allowing four and two groups, respectively, to be detected among strains analyzed. Moreover, a specific PCR system allowing detection of only one group of strains was designed. De Vin, et al. (2005), analyzing galR-galK (regulator and galactokinase genes, respectively) intergenic region of 49 St. thermophilus strains, found eight different genotypes. Of the latter, only four were related to the Gal-positive phenotype.
MLST (multi-locus sequence typing), which exploits the genetic variation present in six housekeeping loci, was recently applied to determine the genetic relationship among Lb. plantarum isolates (De Las Rivas, et al. 2006). Of the 16 strains analyzed, there were 14 different allelic combinations, with 12 of them represented by only one strain. MLHT (multi-locus hybridization typing) performed by five housekeeping gene probes was used by Mora, et al. (2002b) to subgroup P. acidilactici strains. MLRT (multi-locus restriction typing) analyzing restriction patterns from eight loci of housekeeping genes was applied by Borgo, et al. (2007) to characterize Lb. helveticus strains isolated from whey starter cultures and cheeses. High heterogeneity among strains was shown and an excellent association was observed between restriction profiles and origin of most of the isolates analyzed.
These last typing or sub-grouping approaches (Gatti, et al. 2005; Ercolini, et al. 2005; De Las Rivas, et al. 2006; Mora, et al. 2002b) have not yet been applied to characterize or monitor wild strains isolated from dairy ecosystems. However, these techniques have reached a great level of automation and will surely have an important role in the rapid typing of bacteria of dairy origin.
Ribotyping was applied to evaluate genetic diversity of Leuconostoc spp. (Villani, et al. 1997), to differentiate Lb. delbrueckii subspecies (Moschetti, et al. 1997), to differentiate and characterize strains of Lb. casei group species (Svec, et al. 2005) and to subgroup Lb. plantarum strains (De Las Rivas, et al. 2006). Originally, ribotyping was intended for taxonomic use (Grimont and Grimont 1986), but it was also later applied for typing strains. However, due to its weak discriminatory power as a typing method, other techniques have replaced it. With a commercially available system, all stages of manual ribotyping can be performed and the basic protocol takes at least five days. However, the development of an automated ribotyping system, the RiboPrinter®, (Qualicon Inc., Wilmington, Del., U.S.) made it possible to shorten the procedure to eight hours. The process is highly standardized and data are stored electronically. In addition, data can be exchanged between different laboratories. Using more than one enzyme, the RiboPrinter® proved to be a valuable primary typing method for pathogens (Grif, et al. 2003). Research performed by Brunner, et al. (2000) provides evidence that PFGE and automated ribotyping are two reliable methods that can be useful for epidemiologic investigations on group A streptococci. Most strains belonging to the Lb. casei group and the Lb. acidophilus group were discriminated at the species level by automated ribotyping (Chun, et al. 2001). Massi, et al. (2004) compared automated ribopatterns of seven probiotic Lactobacillus strains (Lb. acidophilus, Lb. delbrueckii subsp. bulgaricus, Lb. casei, Lb. plantarum, Lb. brevis, Lb. salivarius subsp. salicinius, Lb. gasseri) with those reported in the RiboPrinter® database. All probiotic Lactobacillus strains gave specific new fingerprinting patterns, as none of them was included in the pre-existing ribogroups of the RiboPrinter® database. Due to the ribotyping specificity, the authors concluded that the method represents a powerful tool for strain-specific detection of these lactobacilli. However, Kitahara, et al. (2005), analyzing automated ribopatterns of 22 Lb. sanfranciscensis strains, obtained only four clusters at less than 80 percent similarity, while Basaran, et al. (2001) obtained only 10 different ribopatterns analyzing 20 lactococci. However, cluster analysis of data allowed differentiation of Lc. lactis subsp. cremoris strains from those of Lc. lactis subsp. lactis. Beaslay and Saris (2004) applied RiboPrinter® technology to differentiate nisin producing Lc. lactis strains isolated from human and cow's milk.
Was this article helpful?