HAL Id: hal-00895504 https://hal.archives-ouvertes.fr/hal-00895504 Submitted on 1 Jan 2003 HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. Isolation, characterisation and identification of lactobacilli focusing mainly on cheeses and other dairy products Valérie Coeuret, Ségolène Dubernet, Marion Bernardeau, Micheline Gueguen, Jean Vernoux To cite this version: Valérie Coeuret, Ségolène Dubernet, Marion Bernardeau, Micheline Gueguen, Jean Vernoux. Isolation, characterisation and identification of lactobacilli focusing mainly on cheeses and other dairy products. Le Lait, INRA Editions, 2003, 83 (4), pp.269-306. 10.1051/lait:2003019. hal-00895504
Transcript
HAL Id: hal-00895504https://hal.archives-ouvertes.fr/hal-00895504
Submitted on 1 Jan 2003
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
Isolation, characterisation and identification oflactobacilli focusing mainly on cheeses and other dairy
productsValérie Coeuret, Ségolène Dubernet, Marion Bernardeau, Micheline Gueguen,
Jean Vernoux
To cite this version:Valérie Coeuret, Ségolène Dubernet, Marion Bernardeau, Micheline Gueguen, Jean Vernoux. Isolation,characterisation and identification of lactobacilli focusing mainly on cheeses and other dairy products.Le Lait, INRA Editions, 2003, 83 (4), pp.269-306. �10.1051/lait:2003019�. �hal-00895504�
Isolation, characterisation and identification of lactobacilli focusing mainly on cheeses
and other dairy products
Valérie COEURET, Ségolène DUBERNET, Marion BERNARDEAU, Micheline GUEGUEN, Jean Paul VERNOUX*
Laboratoire de microbiologie alimentaire U.S.C INRA, Université de Caen Basse-Normandie, Esplanade de la Paix, 14032 CAEN Cedex, France
Received 11 April 2002 – Accepted 5 May 2003
Published online 5 July 2003
Abstract – During the last fifteen years, the Lactobacillus genus has evolved and contains to datemore than 80 species. They are present in raw milk and dairy products such as cheeses, yoghurtsand fermented milks. Quality assurance programmes associated with research, development,production and validation of the health or technological benefits of these bacteria require theirrelevant isolation, counting and identification. This review presents the different selective media toisolate lactobacilli, and the numerous different available tools to characterise lactobacilli at genus,species or strain level using either culture-dependent methods: phenotypical, molecular or globalmethods, or using new culture-independent advanced molecular methods. Enzymes used for PFGE,hybridisation probes and PCR-based method primers are listed in seven tables. In conclusion, themain advantages and disadvantages associated with these techniques are presented.
Lactobacillus / media / PFGE restriction enzyme / probe / primer / cheese / dairy product
Résumé – Isolement, caractérisation et identification des lactobacilles des produits laitiers.Durant les quinze dernières années, le genre Lactobacillus a subi de nombreux remaniements etcompte actuellement plus de 80 espèces. Les lactobacilles sont présents dans le lait cru, les produitslaitiers tels que les fromages, les yaourts et les laits fermentés. Pour s’inscrire dans une démarchequalité visant au développement de lactobacilles à effet santé ou d’intérêt technologique,l’isolement, le comptage et la caractérisation parfaite de ces bactéries sont nécessaires. Cetteanalyse bibliographique présente les différents milieux sélectifs pour isoler les lactobacilles, ainsique les outils disponibles à ce jour pour caractériser les lactobacilles au niveau du genre, de l’espèceou de la souche, aussi bien par des méthodes culture dépendantes : phénotypique, moléculaire,globale, que par les nouvelles méthodes se réalisant à partir d’échantillons bruts. Les enzymesutilisées en PFGE, les sondes d’hybridation et les oligonucléotides utilisés pour les différentestechniques de PCR sont répertoriés en sept tableaux. En conclusion, les avantages et inconvénientsde ces techniques sont présentés.
1. INTRODUCTION: BIODIVERSITY OF LACTOBA-CILLI IN DAIRY PRODUCTS
Lactic acid bacteria (LAB) comprise awide range of genera and include a consid-erable number of species. Their commontraits are: Gram-positive, usually catalase-negative, growth under microaerophilic tostrictly anaerobic conditions and lacticacid production. These bacteria are themajor component of the starters used infermentation, especially for dairy prod-ucts, and some of them are also naturalcomponents of the gastrointestinal micro-flora. Lactobacillus is one of the mostimportant genera of LAB. In raw milk anddairy products such as cheeses, yoghurtsand fermented milks, lactobacilli are natu-rally present or added intentionally, fortechnological reasons or to generate ahealth benefit for the consumer.
1.1. Cheeses and milks
In cheese, lactobacilli (Lb. helveticus andLb. delbrueckii ssp. and lactis) are present inindustrial starter cultures for processinghard cheese (e.g. Emmental, Comté, ItalianGrana and Argentinean hard cheeses) [21,43, 92]. Pasta filata cheeses (Mozarella)and Italian hard cheeses (CanestratoPugliese and Parmigiano Reggiano) arealso processed with Lb. delbrueckii ssp.bulgaricus [42, 45, 83]. Lactobacilli start-ers are normally present at levels of109 bacteria/g, contribute to the lactic fer-mentation, and are involved at the beginningof ripening [70]. For example, in Emmen-tal cheeses, lactobacilli ferment galactoseexcreted by Streptococcus thermophilus,achieve acidification processes, and con-tribute to primary proteolysis [31, 35, 59,74]. Their numbers decrease rapidly duringripening, at a rate depending to somedegree on the sensitivity of the starters tosalt [110], on the water activity, and on theautolysis power of the strains [185]. Somelactobacilli are also present in the natural
microflora of the dairy products (non-starter lactic acid bacteria: NSLAB) andoriginate from animals, farms and cheesedairies: Lb. casei ssp. casei/Lb. paracasei ssp.paracasei, Lb.rhamnosus, Lb. plantarum,Lb. fermentum, Lb. brevis, Lb. buchneri,Lb. curvatus, Lb. acidophilus and Lb. pen-tosus [43, 58, 83, 110, 120, 133]. NSLAB,which are initially present in small num-bers (102 to 103 NSLAB/g after pressing inCheddar cheese), increase to high numbersin cheese varieties that require long ripen-ing times (107 to 108 bacteria/g withinabout 3 months in Cheddar cheese) [10, 21,34, 58, 69, 73, 86, 110, 145].
Coppola et al. [44] studied the micro-biological characteristics of raw milk, nat-ural whey starter and cheese during thefirst months of ripening of ParmigianoReggiano: thermophilic lactobacilli – Lb.helveticus, Lb. delbrueckii ssp. lactis. Lb.delbrueckii ssp. bulgaricus – disappearedwithin 30 d. Rod-shaped mesophilic facul-tatively heterofermentative lactobacilli –Lb. casei, Lb. paracasei ssp. paracasei, Lb.paracasei ssp. tolerans and Lb. rhamnosus –progressively increased in number until thefifth month of ripening. The resultsshowed that the thermophilic lactobacilliand Lb. rhamnosus were derived from nat-ural whey starter, whereas the other com-ponents of non-starter lactobacilli werederived from raw milk. Similarly, Comtécheese contains mesophilic lactobacillistrains. They originate from the raw milk,and this source was probably more impor-tant than the factory environment [17, 58].Differences have been observed between rawmilk cheeses and pasteurised or microfilteredmilk cheeses [58, 70, 164]. Eliskases-Lechner et al. [70] found that Bergkäse(Austrian regional cheese) cheeses madefrom pasteurised milk contained less thanone-thousandth the number of facultativelyheterofermentative lactobacilli (FHL) presentin raw milk cheeses. Differences in citratemetabolism, which occur in raw milkcheeses, can be attributed to the presenceof FHL in these cheeses. Very little effort iscurrently devoted to controlling the numbers
Identification of dairy lactobacilli 271
and types of NSLAB present, althoughthese organisms are thought to have a sig-nificant influence on cheese flavour devel-opment and to participate directly in theproduction of some major aroma com-pounds such as acetic acid, formic acid andgas [31, 56, 59, 173]. However, NSLABmay also cause defects. For example, inEmmental cheeses, Lb. plantarum maydisturb the metabolism of propionic acidbacteria, resulting in lower quality cheesesdue to opening and changes in flavourdevelopment [34].
It has repeatedly been claimed that theuse of selected strains as adjunct culturesimproves and accelerates flavour develop-ment. The use of Lactobacillus adjunctcultures to Cheddar cheese or to cow’smilk cheeses ripened for short periods oftime (e.g. Azua-Ulloa cheese) has beenreported to result in higher levels of prote-olytic products and higher sensory qualityscores in many studies [120, 124, 134,175]. However, some adjunct cultures maycause high levels of acidity, bitterness, offflavours and open and crumbly textures,clearly demonstrating the importance ofculture selection [46, 69, 101]. A numberof studies have recently evaluated the suit-ability of probiotic cultures as adjunct cul-tures in various cheeses: Lb. acidophilusand Lb. casei in Argentinian Fresco cheese[191], Lb. paracasei in Cheddar cheeses[76, 172], and Lb. acidophilus in goat’smilk cheeses [85].
1.2. Yoghurts and fermented milks
Lactobacillus delbrueckii ssp. bulgari-cus is one of the two bacteria necessary forthe production of yoghurts, and Lb. kefir isessential for the production of Caucasiansour milk kefir [111]. A recent trend is toadd probiotic lactobacilli to fermentedmilks to generate health benefits. The bac-teria generally added are Lb. acidophilus,Lb. rhamnosus, Lb. reuteri, Lb. casei, Lb.plantarum, Lb. johnsonii, Lb. crispatus,
Lb. paracasei and Lb. gasseri [88, 94, 113,142, 159]. For these products, carefulstrain selection is necessary. Klein et al.[113] recently observed that the identity ofmost of the lactobacilli used as probioticsdiffers from that marked on the packaging.Shah [162] reported that it is important tomonitor the survival of probiotic lactoba-cilli because a number of products havebeen found to contain only a few viablebacteria by the time they reach the market[87, 163, 167].
Given the great potential economicvalue of lactobacilli, one of the main objec-tives of microbiologists is to develop aclear picture of the microflora present inthe various dairy products, and the way inwhich it changes during processing. Forexample, during cheese manufacture and rip-ening, complex interactions occur betweenindividual components of the cheese micro-flora, and identification of these bacteria isessential for understanding their individualcontribution to cheese manufacture. Thisallows the development of a more targetedapproach to starter/adjunct selection for theimprovement of cheese quality [14]. Qual-ity assurance programmes associated withresearch, development, production and val-idation of the health or technological bene-fits of these bacteria require the relevantisolation, counting and identification ofbacteria.
Depending on the taxonomic leveldesired, several phenotypical or molecularmethodologies (polyphasic analysis) canbe used for isolation, characterisation andidentification of lactobacilli. Recent meth-odologies, which are culture-independentsuch as single strand conformation poly-morphism (SSCP), temperature gradientgel electrophoresis (TGGE) and denatur-ing gradient gel electrophoresis (DGGE)have also been proposed for characterisa-tion of microbial diversity. This paperreviews how the literature proposes char-acterisation of lactobacilli from dairy prod-ucts at genus, species or strain level.
272 V. Coeuret et al.
2. STUDY OF LACTOBACILLI BY CULTURE-DEPENDENT METHODS
2.1. Isolation of lactobacilli: from crude sample to pure culture
Guidelines have been published for theextraction of lactobacilli from milk ordairy products [104]. It is often difficult tocount microorganisms from particulate orsolid samples. In such cases, the cells mustbe separated from the particles and theextraction efficiency must be assessed. Theseparation methods currently used toextract microorganisms from solid foodmatrices such as hard cheese and from drystarters and other types of cheese are basedon an initial crude homogenisation in ablender and/or a stomacher [165]. For hardcheeses, an ultra-turrax or a stomacher isrecommended.
2.1.1. Resuspension medium
Callichia et al. [28] increased the rate ofrecovery of dried microorganisms to alevel similar to that of wet cultures that hadnever been dried by resuspending the driedmicroorganisms in a medium containingphosphate, cysteine, antifoaming agentand agar before plating. McCann et al.[130] analysed a number of commercialdairy cultures, comparing methods of sam-ple preparation: duplicate samples of pro-biotic products were resuspended anddiluted in both peptone and CRM (Calic-chia resuspension medium). Followingresuspension in peptone and incubation for30 min at 23 °C before plating, the rate ofCFU recovery for each product wasapproximately 50% of that obtained if themicroorganisms were resuspended inCRM and incubated for 30 min at 37 °Cbefore plating. For cheeses, dilution inTrisodium citrate (2% w/v) is generallyrecommended, and peptone salt or phos-phate buffer salt is generally used for dairyproducts such as yoghurts and fermentedmilks.
2.1.2. Selective media
Several elective and selective mediahave been developed for the isolation andcounting of Lactobacillus species and forthe differential counting of mixed popula-tions of lactic acid bacteria (Tab. I). Oxy-gen tolerance, nutritional requirements,antibiotic susceptibility, colony morphol-ogy and colour are used to differentiatestrains in these methods. The differentialdetection and counting of lactobacilli cancurrently be achieved in a number of ways.However, only a few media are useful for thedifferential counting of lactobacilli becausenumerous other microorganisms includingLactococcus, Enterococcus, Leuconostoc,Weissella, Bifidobacterium and Pedio-cocci grow on media similar to those usedfor lactobacilli. No medium has yet beendescribed on which only lactobacilli areable to grow.
Lactobacilli are generally isolated on richmedia such as MRS [54], which is routinelyused for the isolation and counting of lacto-bacilli from most (fermented) food products.The addition of a reducing agent such ascysteine 0.05% to MRS improves the specif-icity of this medium for Lactobacillus isola-tion [90, 116, 162]. MRS and M17 are,respectively, the medium of choice for thedifferential counting of Lactobacillus del-brueckii ssp. bulgaricus and Streptococcusthermophilus in yoghurt [105]. Birollo et al.[18] proposed the use of a cheaper medium –skim milk agar – for counting and differenti-ating colonies of these two bacteria. Whenyogurt microflora was supplemented byother lactobacilli, specific methods wereneeded. Enumeration of Lb. delbrueckii ssp.bulgaricus in the presence of Lb. acidophiluswas possible, on acidified MRS at pH 5.2 oron reinforced clostridial agar (RCA) at pH5.3, at 45 °C for 72 h [162]. Selective count-ing of Lb. acidophilus has been developedusing several MRS media in which the dex-trose is replaced by maltose, raffinose, melibi-ose, trehalose, arabinose, galactose, sorbitol,ribose, gluconate or salicin or cellobiose-
Identification of dairy lactobacilli 273
Table I. Differential plating media for detection and counting of Lactobacillus species.
Product Population MediaIncubation conditions Isolation Notice Ref.
Yogurts
S.thermophilusLb. delb. ssp. bulgaricus
Acidified MRS37 °C, 3 d,
Ana
S.thermophilusLb. delb. ssp. bulgaricus
S. thermophilus = circular opalescent white colonies with well defined borders
Lb. delb. ssp. bulgaricus = irregular white colonies
[103]
[190]
Bile-MRS
OG-MRS (Oxgall, Gentamycin MRS)
G-MRS (Galactose MRS)
37 °C, 3 d, Aer
Lb. delb. ssp. bulgaricus
Lb. acidophilus
X-Glu 37 °C, 3 d, Ana
Lb. delb. ssp. bulgaricus
Lb. acidophilus
Lb. delb. ssp. bulgaricus = white colonies
Lb. acidophilus = blue colonies
[114]
S.thermophilusLb. acidophilusBifidobacterium
spp.
LBS (Lactobacillus selection agar)
37 °C, 2 d, Ana Lb. acidophilus [159]
E.faeciumLb. acidophilusBifidobacterium
spp.
Brigg’s agar37 °C, 2 d,
AerE. faecium
Lb. acidophilusE. faecium (24H)
Lb. acidophilus (48H)
[28]Modified Brigg’s agar (Brigg’s agar, plus Streptomycin
sulfate)
37 °C, 2 d, Aer
Lb. acidophilus All the strains are not Streptomycin resistant
3 BLA37 °C, 2 d,
AerE. faecium
Lb. acidophilusE. faecium (24H)
Lb. acidophilus (48H)[130]
Cheeses MRS 37 °C, 3 d, Ana
Lactobacilli, and other lactic acid
bacteria[105]
MRS42 °C, 3 d,
AnaThermophilic Lactobacilli
[21][58]
LBS medium plus Cycloheximide
30 °C, 3 d, Ana Lactobacilli [120]
Table I. Differential plating media for detection and counting of Lactobacillus species.
Identification of dairy lactobacilli 275
esculine [52, 117, 162, 190]. An agarmedium based on X-Glu [114] has also beenused. MRS with trehalose (T-MRS) is rec-ommended by the International Dairy Feder-ation [104] for the counting of Lb. acido-philus if this organism occurs in mixedpopulations with yoghurt bacteria and bifido-bacteria. Dave and Shah [52] reported thatRogosa acetate or media containing bilesalts, oxgall or NaCl strongly inhibited thegrowth of Lb. acidophilus. In probiotic prod-ucts containing Bifidobacterium bifidum,Enterococcus faecium and Lb. acidophilus,
lactobacilli can be counted on modifiedBrigg’s agar supplemented with streptomy-cin sulphate [28], or on 3BLA medium [130].
However, the use of such specific mediais limited by the target species. For exam-ple, MRS-salicin or MRS-sorbitol agar canbe used for the selective counting of Lb.acidophilus provided that Lb. casei is notpresent in the product (in which case a totalcount for both species is obtained). If Lb.casei is present, a second medium, such asLactobacillus casei agar (LC agar) [151],must be used. The selective counting of
FH agar (Facultativ Heterofermentativen Laktobazillen agar)
38 °C, 3 d, Ana
Facultative heterofermentative
Lactobacilli
[106][58]
Facultativ Heterofer-mentativen Laktoba-
zillen agar plus nalidixic acid
[21]
MRS plus Cycloheximide
30 °C, 2 d, Ana
Mesophilic Lactobacilli
[45]44 °C, 2 d, Ana
Thermophilic Lactobacilli
LBS (Lactobacillus selection agar)
30 °C, 3 d, Ana Lactobacilli
[70]
Facultativ Heterofermentativen Laktobazillen agar plus Vancomycin
Table I. Differential plating media for detection and counting of Lactobacillus species.
276 V. Coeuret et al.
other lactobacilli used in probiotic anddairy products, such as Lb. plantarum, Lb.reuteri and Lb. rhamnosus has not beenstudied extensively. Moreover, all themembers of the so-called Lb. acidophilusgroup: Lb. crispatus, Lb. gasseri, Lb. john-sonii, Lb. gallinarum and Lb. amylovorus,are hardly discriminated by plating.
For the lactobacilli found in cheeses,HHD medium (homofermentative and het-erofermentative differential medium) canbe used for the differential counting ofhomofermentative and heterofermentativelactobacilli [72]. Mesophilic and ther-mophilic lactobacilli are separated by theuse of different incubation temperatures:30 °C or 42 °C to 45 °C. FH medium (fac-ultative heterofermentative Lactobacillusagar) contains vancomycin (50 mg·L–1),and has been used for the isolation ofNSLAB [13, 22, 127, 150]. LBS (Lactoba-cillus selection agar) medium, also knownas Rogosa medium [156], incubated at30 °C has been used to isolate lactobacillifrom hand-made cheeses such as Bergkäse[70, 83, 120, 150] but MRS remains themost commonly used medium for the iso-lation of lactobacilli.
Hartemink et al. [90] developed a newselective medium, LAMVAB (Lactobacil-lus anaerobic MRS with vancomycin andbromocresol green), for the isolation ofLactobacillus species. Firstly, they used itto isolate lactobacilli from faeces (in whichthey are present in small numbers), andthen successfully for various species oflactobacilli from dairy products. The mediumis highly selective, due to its low pH andthe presence of vancomycin (20 mg·L–1).Unlike other Gram-positive bacteria, mostlactobacilli are resistant to vancomycinand Gram-negative bacteria are generallysensitive. Vancomycin cannot be used toselect lactobacilli in products also contain-ing leuconostocs and pediococci becausethese bacteria are also vancomycin-resistant,and careful morphological examination isrequired in such cases for differentiation.However, this medium remains the most
specific medium to date described forlactobacilli. Unfortunately, some Lactoba-cillus species, such as Lactobacillus del-brueckii spp. bulgaricus, and some strainsof Lb. acidophilus are vancomycin-sensi-tive [32, 90]. Hartemink et al. [90] pro-posed the use of this medium for the isolationof lactobacilli from probiotics containingmixed populations of lactobacilli, bifido-bacteria and enterococci. Bifidobacteriaare susceptible to vancomycin and only afew vancomycin-resistant enterococci havebeen isolated.
The use of coloured indicators facili-tates differentiation between microorgan-isms. For example, in LAMVAB and HHD[132], bromocresol green (pH indicator) isused as an indicator of acid production.Ghoddusi and Robinson [78] added Prus-sian blue to TPPY agar, Tryptose ProteosePeptone Yeast extract agar (giving TPPYBmedium), to distinguish Lb. delbrueckiissp. bulgaricus and Lb. acidophilus fromBifidobacterium and Streptococcus ther-mophilus. Kneifel and Pacher [114] devel-oped an agar medium, X-Glu agar, for theselective counting of Lb. acidophilus inyoghurt-related milk products containing amixed flora of lactobacilli, streptococciand bifidobacteria. In this medium, Lb.acidophilus is identified by testing for its�-D-glucosidase activity by means of achromogenic reaction involving X-Glu,which is incorporated into Rogosa agarmedium. Based on a similar principle for �-D-galactopyranosidase activity, Kneifel et al.[115] used X-Gal medium to differentiateblue colonies of lactobacilli from whitecolonies of pediococci and enterococci insilage inoculants. The use of TTC (triphe-nyl tetrazolium chloride) may also be help-ful for strain differentiation [140, 160].
2.2. Characterisation and identifica-tion of lactobacilli from genus level to strain level
For decades, differentiation betweengenera has been based on phenotypic char-acters. Under a light microscope, lactobacilli
Identification of dairy lactobacilli 277
are generally regularly shaped, non-motile,non-spore-forming, Gram-positive rods.However, cell morphology varies widely,from long, straight or slighty crescent-shaped rods to coryneform coccobacilli.Numerous genera display such morpho-logical features. However, we can separateby simple tests such as tests for the oxygentolerance, presence of catalase and growthon acidified MRS Carnobacterium, Lacto-bacillus and Weissella (non-obligate aer-obe, catalase (-), growth on acidified MRS)from Brochothrix, Caryophanon, Erysip-elothrix, Kurthia, Listeria and Renibacte-rium [111]. Carnobacterium resembleslactobacilli but does not grow on acetatemedia. The establishment of a new genus,Weissella [41], encompassing the Parame-senteroides group, which includes Leucon-ostoc paramesenteroides and some hetero-fermentative Lactobacillus species, seemsto be justified on the basis of phylogeneticanalysis. A cell wall murein, based onlysine with an interpeptide bridge contain-ing alanine or alanine plus serine or gly-cine, can be used to distinguish Weissellafrom heterofermentative lactobacilli [111].Classical phenotypic tests for identifica-tion of lactobacilli are based on physiolog-ical characteristics such as respiratorytype, motility, growth temperature andgrowth in NaCl, and on biochemical char-acteristics such as homo/hetero-fermenta-tive, production of lactic acid isomers,metabolism of carbohydrate substrates,coagulation of milk and presence of partic-ular enzymes (e.g. arginine dihydrolase,antibiotic susceptibilities, and so on).Lactobacilli are typically chemoorgano-trophic and ferment carbohydrates, pro-ducing lactic acid as a major end product.In 1919, Orla-Jensen divided them intothree subgenera – Thermobacterium,Streptobacterium and Betabacterium –according to optimal growth temperatureand hexose fermentation pathways. Thisclassification was given up by Kandler andWeiss [111], who proposed a classificationinto three groups – I (obligate homofer-mentative), II (facultative heterofermenta-
tive) and III (obligate heterofermentative) –which is still used today for phenotypicalanalysis.
The most recent version of Bergey’sManual of Systematics [111] includedabout 50 species of Lactobacillus. Thismanual reported taxonomic changes atspecies level (e.g. Lb. bavaricus becameLb. sake) and at subspecies level (e.g. Lb.casei ssp. rhamnosus became Lb. rhamno-sus, Lb. bulgaricus became Lb. delbrueckiissp. bulgaricus, etc.). Moreover, somelactobacilli became members of the newgenus Weissella (e.g. Lb. kandleri becameW. kandleri), which also includes formermembers of the genus Leuconostoc (e.g.Leuconoctoc paramesenteroides becameW. paramesenteroides), whereas otherlactobacilli were assigned to another newgenus, Carnobacterium (e.g. Lb. divergensbecame Cb. divergens). Today, the genusLactobacillus contains 88 species and 15subspecies according to a recent listing(Tab. II, www.bacterio.cict.fr, 6 January2003). Protein analysis such as protein fin-gerprinting (analysis of total soluble cyto-plasmic proteins), or multilocus enzymeelectrophoresis (analysis of electrophoreticmobilities of certain enzymes) are advancedphenotypic methods in current use. Suchanalysis can discriminate between bacteriato the species level and beyond. Lipid pro-filing has also been used. However, theidentification of isolates to species levelcan be difficult because of the considerablevariations in biochemical attributes (fermen-tation profiles) that seem to occur betweenstrains currently considered to belong tothe same species, and some species are notreadily distinguishable in terms of pheno-typic characteristics. This is especially truefor the so-called Lactobacillus plantarumgroup (Lb. plantarum, Lb. paraplantarumand Lb. pentosus), the Lactobacilllus caseiand Lactobacilllus paracasei group (Lb.casei, Lb. rhamnosus, Lb. zeae and Lb.paracasei), Lb. brevis and Lb. buchneri.Recently Dellaglio et al. [57] proposed, onthe basis of considerable published evi-dence, that the name of Lactobacillus
278 V. Coeuret et al.
Table II. Lactobacillus species (www.bacterio.cict.fr, 6 January 2003).
Lactobacillus acetotolerans
Lactobacillus acidipiscis
*Lactobacillus acidophilus
Lactobacillus agilis
Lactobacillus algidus
Lactobacillus alimentarius
Lactobacillus amylolyticus
Lactobacillus amylophilus
Lactobacillus amylovorus
Lactobacillus animalis
Lactobacillus arizonensis
Lactobacillus aviarius ssp.
araffinosus
Lactobacillus aviarius ssp.
aviarius
Lactobacillus bifermentans
Lactobacillus brevis
Lactobacillus buchneri
*Lactobacillus casei
Lactobacillus catenaformis
Lactobacillus cellobiosus
Lactobacillus coleohominis
Lactobacillus collinoides
Lactobacillus coryniformis ssp.
coryniformis
Lactobacillus coryniformis ssp.
torquens
*Lactobacillus crispatus
Lactobacillus curvatus ssp.
curvatus
Lactobacillus curvatus ssp.
melibiosus
Lactobacillus cypricasei
Lactobacillus delbrueckii ssp.
bulgaricus
*Lactobacillus delbrueckii ssp.
delbrueckii
Lactobacillus delbrueckii ssp.
lactis
Lactobacillus diolivorans
Lactobacillus durianis
Lactobacillus equi
*Lactobacillus farciminis
Lactobacillus ferintoshensis
Lactobacillus fermentum
Lactobacillus fornicalis
Lactobacillus fructivorans
Lactobacillus frumenti
Lactobacillus fuchuensis
Lactobacillus gallinarum
*Lactobacillus gasseri
Lactobacillus graminis
Lactobacillus hamsteri
Lactobacillus helveticus
Lactobacillus heterohiochii
Lactobacillus hilgardii
Lactobacillus homohiochii
Lactobacillus iners
Lactobacillus intestinalis
Lactobacillus jensenii
*Lactobacillus johnsonii
Lactobacillus kefiranofaciens
Lactobacillus kefirgranum
Lactobacillus kefiri
Lactobacillus kimchii
Lactobacillus kunkeei
Lactobacillus leichmannii
Lactobacillus lindneri
Lactobacillus malefermentans
Lactobacillus mali
Lactobacillus maltaromicus
Lactobacillus manihotivorans
Lactobacillus mucosae
Lactobacillus murinus
Lactobacillus nagelii
Lactobacillus oris
Lactobacillus panis
Lactobacillus pantheris
Lactobacillus parabuchneri
*Lactobacillus paracasei ssp.
paracasei
Lactobacillus paracasei ssp.
tolerans
Lactobacillus parakefiri
Lactobacillus paralimentarius
Lactobacillus paraplantarum
Lactobacillus pentosus
Lactobacillus perolens
*Lactobacillus plantarum
Lactobacillus pontis
Lactobacillus psittaci
*Lactobacillus reuteri
*Lactobacillus rhamnosus
Lactobacillus rogosae
Lactobacillus ruminis
Lactobacillus sakei ssp.
carnosus
Lactobacillus sakei ssp.
sakei
Lactobacillus salivarius ssp.
salicinius
Lactobacillus salivarius ssp.
salivarius
Lactobacillus sanfranciscensis
Lactobacillus sharpeae
Lactobacillus suebicus
Lactobacillus trichodes
Lactobacillus vaccinostercus
Lactobacillus vaginalis
Lactobacillus vitulinus
Lactobacillus zeae
In bold: Lactobacilli used in dairy products; with an *: Lactobacilli used in probiotic product.
Identification of dairy lactobacilli 279
paracasei had to be rejected by the judicialcommission and that the species Lactoba-cillus casei is not correctly represented bythe strain ATCC 393.
Studies based on 16S rDNA have led tothe classification of Lactobacillus speciesinto three major groups: the Leuconostocgroup, the Delbrueckii group, and the Lb.casei-Pediococcus group [40, 174, 188].Recently Lb. fructosus (the only lactoba-cilli member of the Leuconostoc group)was reclassified as Leuconostoc fructosum[8]. Closely related species and strainswith similar phenotypic features may nowbe reliably differentiated from each otherby DNA-based techniques. Molecularmethods can be used for taxonomic analy-ses to firstly determine the species towhich a bacterium belongs, by DNA/DNAhybridisation, sequencing, polymorphismchain reaction (PCR), ribotyping, poly-morphism chain reaction-restriction frag-ment length polymorphism (PCR-RFLP),and secondly permit strain differentiationby the use of techniques such as restrictionenzyme analysis (REA), randomly ampli-fied polymorphic DNA (RAPD), repeatedsequence extragenic palindrom PCR(REP-PCR), amplified fragment lengthpolymorphism (AFLP), plasmid profilingand pulsed field gel electrophoresis (PFGE).However, identification can be done with ahigh degree of confidence only if a correctvalidation of the method or probes or prim-ers have been checked using close genera,species or strains.
Polyphasic taxonomy is increasinglyused [4, 75, 118, 167, 174, 188]. Lawsonet al. [118] described, on the basis of phyl-ogenetic and phenotypic evidence, a newspecies of Lactobacillus, Lb. cypricasei,which was isolated from Halloumi cheese,a semi-hard cheese from Cyprus.
2.2.1. Analysis at genus level
The genus Lactobacillus is heterogene-ous, with the G+C content of the DNA ofits species varying from 33 to 55% [40,89]. However, it is generally thought that
G+C content should vary by no more thana 10% range within a well-defined genus[188]. The nucleotide sequences of Lacto-bacillus 16S ribosomal DNA (rDNA) pro-vide an accurate basis for identification.The sequence obtained from an isolate canbe compared with those of Lactobacillusspecies held in databases. Recently, Dubernetet al. [62] defined a genus-specific primerby analysing similarities between thenucleotide sequences of the spacer regionbetween the 16S and 23S ribosomal RNAgenes of Lactobacillus. The specificity ofthis genus-specific primer combined with auniversal primer was tested against 23strains of lactobacilli of varied origin (cor-responding to 21 species) Escherichia coli,two leuconoctocs species, Carnobacte-rium piscicola, Pediococcus pentosaceus,Bifidobacterium bifidum, Weissella confusa,Enterococcus faecalis, Staphylococcusaureus and Listeria monocytogenes. Posi-tive amplification was only obtained withthe lactobacilli strains.
2.2.2. Analysis at species level
Phenotypical micro methods
Several combinations of tests andready-to-inoculate identification kits suchas API 50 CH, LRA Zym and API Zymenzymatic tests can be used for the rapidand theoretically reproducible phenotypicidentification of pure cultures. They havebeen used for the characterisation andidentification of lactobacilli in milks [133],yoghurts and other fermented milks [6] andin cheeses [6, 21, 53, 58, 92, 118, 120, 133,175]. However, the reliability of these testshas been questioned, especially for API50 CH, initially developed for the identifi-cation of medical Lactobacillus strains. Inaddition, the manufacturer’s database isnot updated and some Lactobacillus spe-cies are missing. Andrighetto et al. [6] usedAPI 50 CH to analyse 25 strains of ther-mophilic lactobacilli isolated from yoghurtand from semi-hard and hard cheeses (Lb.delbrueckii ssp. lactis and ssp. bulgaricus,
280 V. Coeuret et al.
Lb. helveticus and Lb. acidophilus). Formost of the strains, clear assignment to aparticular species or subspecies was notpossible because ambiguous results wereobtained for the sugar fermentation profile.Nigatu [141] also reported a lack of agree-ment between the API 50 CH groupingpattern of isolates and RAPD clusters.Tynkkynen et al. [184] used API 50 CH foridentifying strains of the Lb. casei group(Lb. rhamnosus, Lb. zeae and Lb. casei).The exact identifications of these closelyrelated species were not reliable; somewere doubtful or unacceptable and somestrains were misidentified with a goodidentification level. Furthermore, variabil-ity may be observed within a single strain.For example, the Lb. rhamnosus GG strainhas traditionally been detected, countedand identified on the basis of cultures inselective anaerobic conditions on MRS orRogosa agar (37 °C for 78 h), colony mor-phology (large, white, creamy andopaque), Gram staining and cell morphol-ogy (Gram-positive and uniform rods inchains) and the carbohydrate fermentationprofile in the API 50 CHL test. However, ithas been pointed out that the colony mor-phology and the carbohydrate fermenta-tion pattern of strain GG are not alwaystypical, due to variation [32]. This varia-tion may result from the loss or gain ofplasmids, leading to inconsistency in themetabolic traits of a strain, as most of theproteins involved in carbohydrate fermen-tation are plasmid-encoded [9].
Protein fingerprinting
A bacterial strain always produces thesame set of proteins if grown under stand-ardised conditions. The electrophoregramsproduced by zone electrophoresis of theseproteins under well-defined conditions canbe considered as a sort of fingerprint of thebacterial strains from which they areobtained. Sodium dodecylsulphate poly-acrylamide gel electrophoresis (SDS-PAGE)of whole-cell proteins is one of the tech-niques used. It is a relatively simple and
inexpensive method that has already beenused for the identification and classifica-tion of lactobacilli. The entire procedureconsists of several experimental steps,from the growth of bacteria to the scanningof the electrophoregram. SDS-PAGE sep-arates proteins exclusively according tomolecular weight. Native (non-denaturing)PAGE can be used as a complementarytechnique, separating cell proteins accord-ing to their charge and size, providing highresolution and good band definition. Inhighly standardised conditions, reproduci-ble patterns can be obtained that are ame-nable to rapid, computer-based digitalanalysis. Protein profiles can be stored indatabase format and may be routinely usedto confirm the identity of Lactobacillusstrains, to differentiate between unknownisolates and to evaluate classificationschemes, at species level or below [53, 75,92, 118, 122, 146, 147]. De Angelis et al.[53] isolated NSLAB from 12 Italian ewe’smilk cheeses. Most of the species studiedgave specific protein profiles, except Lb.plantarum and Lb. pentosus, which weregrouped in the same cluster, confirming theresults previously obtained by Van Reenenand Dicks [187]. Gancheva et al. [75] usedSDS-PAGE to analyse the cellular proteinsof a set of 98 strains belonging to nine speciesof the Lactobacillus acidophilus rRNA-group of varied origin (Lb. acidophilus,Lb. amylolyticus, Lb. crispatus, Lb. john-sonii, Lb. gasseri, Lb. gallinarum, Lb. hel-veticus, Lb. iners and Lb. amylovorus).Most of these species can be differentiatedby SDS-PAGE, but poor discriminationwas obtained between Lb. johnsonii andLb. gasseri strains, and between some strainsof Lb. amylovorus and Lb. gallinarum.
Multilocus enzyme electrophoresis
About 50% of all enzymes investigatedto date exist in multiple molecular forms.These isoenzymes usually differ in electro-phoretic mobility and catalytic parameters.Enzyme multiplicity may depend on genetic
Identification of dairy lactobacilli 281
factors directly (primary isoenzymes andallozymes) or indirectly (secondary isoen-zymes, generated by post-translationalmodifications). Isoenzymes may be dis-tributed between different cell compart-ments and may be encoded by at least twodifferent genes. Multiple loci encodingenzymes of identical substrate specificitiesare usually thought to result from geneduplications. Point mutations then lead todivergence in the amino acid sequences ofthe proteins encoded by the duplicatedgenes, resulting in the production of differ-ent enzyme forms, separable by electro-phoresis. Differences in electrophoreticmobility may result from differences incharge and/or size. Multilocus enzymeanalysis has been shown to be of greatpotential [161] in the differentiation ofLAB species [182]. The electrophoreticmobility of LDH (lactate dehydrogenase)in starch gels [77] and in polyacrylamidegels [95] has been used to discriminatebetween phenotypically very similar spe-cies: Lb. acidophilus, Lb. crispatus, Lb.gallinarum, Lb. gasseri and Lb. johnsonii.Lortal et al. [122] studied peptidoglycanhydrolases of industrial starters as a newtool for bacterial species identification.The peptidoglycan hydrolase patterns of94 strains of lactobacilli belonging to 10different species were determined (Lb. hel-veticus, Lb. acidophilus, Lb. delbrueckii,Lb. brevis, Lb. fermentum, Lb. jensenii, Lb.plantarum, Lb. sake, Lb. curvatus and Lb.reuteri). Each species gave its own specificpattern, with differences observed evenbetween closely related species. It is alsopossible to type strains of Lactobacillusacidophilus [146], or clinical isolates andbiotechnologically-used strains of Lacto-bacillus rhamnosus [112], or strains ofLactobacillus casei [55] isolated fromensiled high-moisture corn grain.
Lipid profiling
Lipid profiling by gas chromatographyis more useful for the grouping of strainsthan for the identification of individual
strains [47, 66, 154]. Moreover, the relia-bility of lipid and polysaccharide profilingfor discriminating between Lactobacillusspecies has been questioned and fatty acidmethyl ester (FAME) analysis does notseem to be reliable for LAB [113].
Hybridisation
The use of probes for hybridisation withnucleic acid fragments is a technique withgreat potential for the future. A nucleicacid probe is a fragment (20-30 pb) of asingle-stranded nucleic acid fragment thatspecifically hybridises to complementaryregions of a target nucleic acid. It can beused directly on a colony, or after DNA/RNA extraction. The target nucleic acidconsists of single-stranded DNA or RNAmolecules. Molecular probes may be labelledradioactively or non-radioactively. Radio-active labelling involves the phosphoryla-tion of the 5' terminus of the probe with[32P] ATP. Non-radioactive labelling maybe direct, using alkaline phosphatase orperoxidase, or indirect, by attachment of aligand-protein or a hapten-antibody. Fluo-rescent probes (FISH: fluorescent in situhybridisation) may also be used. Theextensive use of multiple oligonucleotideprobes has become possible followingmajor developments in the sequencing ofrRNA genes. Depending on the level ofdetection required (genus or species), dif-ferent regions of the genome might be usedas targets. The sequences of 16S and 23SrRNA molecules contain highly conservedregions common to all eubacteria, andhighly variable regions unique to a partic-ular species [32, 199]. Thus, nucleic acidprobes, in particular probes targetingrRNA sequences, can be used for the reliableidentification of bacteria, for monitoringpopulation changes during fermentationand detecting the presence of bacterialcontamination or spoilage bacteria [60].Such probes have been extensively used inthe analysis of dairy products [6, 98]. Oli-gonucleotide DNA probes, mostly target-ing variable regions of the 16S or 23S
282 V. Coeuret et al.
rRNA genes, have been widely used forspecies identification and strain detection(Tab. III). However, such rRNA probescannot be used for closely related species
due to the high level of similarity betweentheir rRNA gene sequences. For example,such probes cannot distinguish Lb. plantarumfrom Lb. pentosus or Lb. paraplantarum
Table III. Oligonucleotide probes for the identification of Lactobacilli.
[25]. These species are currently distin-guished by probing Southern blots with apyrDFE gene fragment from Lb. plantarumor by DNA/DNA hybridisation. Particularattention had to be done to their specificitysince Roy et al. [158] demonstrated thatthe probe defined by Hertel et al. [98] forLb. helveticus also hybridise with Lb. gall-inarum strains.
In colony hybridisation, bacteria areplated on membranes that are then placedon nutrient agar, allowing the bacteria toform colonies. The colonies are lysed.Hybridisation with a labelled probe can beused for both qualitative and quantitativeanalyses, and has been used for LAB [37].
Sequencing
Comparison of rRNA gene sequences iscurrently considered to be the most power-ful and accurate method for determiningthe degree to which microorganisms arephylogenetically related [199]. Advancesin molecular biological techniques havemade it possible to sequence long stretchesof rRNA genes. Initially, reverse tran-scriptase was used to generate DNA fromrRNA, and this DNA was then sequenced.It is now possible to sequence 16S or 23SrDNA molecules by direct PCR sequenc-ing, and this method has generated largesequence databases. Although the species-specific sequences are located in the firsthalf of the 16S rRNA gene (V1-V3region), identification is more accurate ifthe whole gene is sequenced [171]. Thisrequires the sequencing of about 1.5 kb ofDNA. Tannock et al. [177] showed thatcomparison of the16S-23S spacer regionsequences of lactobacilli can be used inpractical situations for strain identification.The spacer region sequences is sequencingrapidly and accurately identifies Lactoba-cillus isolates obtained from gastrointestinal,yoghurt and silage samples. The 16S-23Sspacer sequences of lactobacilli are small,only about 200 bp in length. These shortsequences are easy to sequence on bothstrands and provide reliable information for
comparative work. The spacer regionmethod has the advantage of distinguishingbetween Lb. rhamnosus and Lb. caseistrains [177]. It can be used to distinguish Lb.plantarum, from Lb. paraplantarum, thesetwo closely related species belonging tothe Lb. plantarum group [12]. Chen et al.[33] analysed the 5S-23S rRNA intergenicspacer regions (ISRs) of the Lactobacillusgroup. This method was found to be aneffective way of discriminating Lb. rham-nosus from Lb. casei/Lb. paracasei becausespacer length polymorphism results in a76/80 bp insertion with respect to the 16SV2-V3 sequences.
Polymerase chain reaction (PCR)
This method developed by Mullis [139]uses primers, which are about 20 to 30 pb.Berthier and Ehrlich [15] studied the 16S/23S SRs DNA from six closely related speciesof lactobacilli (Lb. curvatus, Lb. graminis,Lb. sake, Lb. plantarum, Lb. paraplantarumand Lb. pentosus). Only the larger frag-ment displayed differences in sequencebetween species, and primers derived fromthis region were defined for the six species.SR sequences could not be used to typestrains within the two groups of Lactoba-cillus species.
Numerous species-specific primers havebeen derived from spacer regions [15, 170,177, 179]. The species-specific primerscurrently available are listed in Table IV.
Mannu et al. [126] used seven pairs ofspecific primers designed by Berthier andEhrlich [15], Drake et al. [60] and Wardand Timmins [196] to analyse 457 isolatesfrom Fiore Sardo cheese, a traditional hardcheese from Sardinia. Only seven isolateswere not successfully identified with thismethod; 31 were obligate homofermenta-tive lactobacilli, 419 isolates were faculta-tive heterofermentative lactobacilli (Lb.plantarum, Lb. paracasei and Lb. curvatus).
Many recent studies have been carriedout by multiplex PCR (Tab. V), in whichseveral primers are added to the same sample,
284 V. Coeuret et al.
Table IV. PCR primers used for lactobacilli identification.
16S rRNA gene23S rRNA gene(/16S)16S/23S spacer region Lb. curvatus16S/23S spacer region Lb. graminis16S/23S spacer region Lb. paraplantarum/plantarum16S/23S spacer region Lb. pentosus16S/23S spacer region Lb. plantarum16S/23S spacer region Lb. sake
making it possible to detect several micro-organisms or species at the same time.Multiplex PCR has been used to detect Lb.pontis and Lb. panis in sourdough fermen-tation [138], and Lactobacillus in faecal
samples [123] and in meat spoilage [200].Song et al. [170] used multiplex PCR as arapid, simple and reliable method for theidentification of common Lactobacillusisolates from human stool samples, and
Table IV. PCR primers used for lactobacilli identification.
286 V. Coeuret et al.
established a two-step PCR method. In thismethod, lactobacilli are first classified intofour groups, and then one or two multiplexPCR assays are carried out for each group,for species identification. Lb. delbrueckiiwas identified in the first grouping multi-plex PCR and 10 species were identified inthe second multiplex PCR for each group.
RibotypingSouthern blotting is carried out after the
restriction digestion of chromosomal DNAand agarose electrophoresis. In this process,DNA is transferred to a membrane forhybridisation with a labelled 23S, 16S or5S rRNA gene probe. As bacteria havemultiple copies of rRNA operons in their
Table V. Multiplex PCR primers used for lactobacilli identification.
ISR/23S PCR-G Group I lactobacilli (450pb)ISR/23S PCR-G Group II lactobacilli (300 pb)ISR/23S PCR-G Group IV lactobacilli (350 pb)ISR/23S PCR-G Group III lactobacilli (400 pb)
ISR/23S PCR-II-1 Group II, Lb. acidophilus (210 pb)ISR/23S PCR-II-1 Group II, Lb. jensenii (700 pb)
ISR/23S PCR-II-2 Group II, Lb. crispatus (522 pb)
ISR/23S PCR-II-2 Group II, Lb. gasseri (360 pb)
ISR/23S PCR-IV Group IV, Lb. fermentum (192 pb)
ISR/23S PCR-IV Group IV, Lb. plantarum (248 pb)
ISR/23S PCR-IV Group IV, Lb. reuteri (303 pb)
ISR/23S PCR-IV Group IV, Lb. salivarius (411 pb)
ISR/23S PCR-III Group III, Lb. paracasei (312 pb) ISR/23S PCR-III Group III, Lb. rhamnosus (113 pb)
chromosome, several fragments in therestriction digest mixture hybridise withthe probe. In general, the fingerprint pat-terns obtained by this method are more sta-ble and easier to interpret than thoseobtained by restriction enzyme analysis(REA). Ribotyping has been used withsome success to characterise strains of var-ious Lactobacillus species [155], strains ofLb. helveticus [80] and strains of Lb. del-brueckii [135]. Miteva et al. [135] successfullydifferentiated between the three subspeciesof Lb. delbrueckii by EcoRI ribotyping.Zhong et al. [202] used ribotyping for spe-cies discrimination (Lb. jensenii, Lb. casei,Lb. rhamnosus, Lb. acidophilus, Lb.plantarum and Lb. fermentum). In general,ribotyping has a greater discriminatorypower at species level than at strain level.Tynkkynen et al. [184] analysed 24 lacto-bacilli strains biochemically identified asmembers of the Lactobacillus casei group(Lb. rhamnosus and Lb. casei): ribotypingby EcoRI digestion and southern blottingwith a chemiluminescent probe correspond-ing to the rrnB rRNA operon from E. coliresulted in the detection of a triplet, whichseems to be typical for most Lb. rhamnosusstrains.
A fully automated ribotyping system,the RiboPrinter microbial characterisationsystem, has been developed for identifica-tion at the genus, species and strain levels[26]. This method is automated and basedon a standardised protocol, maximising inter-laboratory reproducibility. It is easy to carryout but the equipment is rather expensive.In the database supplied by the manufac-turer (Qualicon), reference is made to sev-eral bacterial genera: lactic acid bacteria(lactobacilli included), Salmonella, Liste-ria, Escherichia, Pseudomonas and others.
PCR – Restriction Fragment Length Polymorphism analysis (PCR-RFLP)
PCR-restriction fragment length poly-morphism analysis involves the amplifica-tion of a specific region, followed byrestriction enzyme digestion and conven-
tional gel electrophoresis (CGE). RFLPanalysis of the 16S rRNA gene, chromo-somal DNA cleaved with EcoRI andHindIII, gave identical patterns for most ofthe strains of Lb. plantarum [109]. Whenthe specific region corresponds to rDNA,then this method is called PCR-ARDRA(amplified rDNA restriction analysis), andis derived from ribotyping. Andrighettoet al. [6] analysed a 1500 bp polymorphismin EcoRI-digested 16S rDNA fragmentsfrom 25 strains of 4 species of lactobacilliisolated from cheeses and yoghurts: Lb.delbrueckii ssp. lactis and Lb. delbrueckiissp. bulgaricus, Lb. helveticus and Lb. aci-dophilus. Different patterns were observed,making it possible to distinguish between thevarious Lactobacillus species and subspe-cies. Giraffa et al. [80] used PCR-ARDRAto identify Lb. delbrueckii isolates to sub-species level and to differentiate this speciesfrom Lb. helveticus and Lb. acidophilus.As these species were present in the sameecological niches, and displayed similarphenotypic characteristics and a closegenetic relationship, PCR-ARDRA wasefficient. Bouton et al. [22] confirmed byPCR-ARDRA strains isolated from Comtécheese belonged to Lb. delbrueckii ssp.lactis. For six presumed strains of Lb. hel-veticus, no cutting by EcoRI was obtained,even if fermentation profiles suggestedthat all strains belonged to Lb. helveticusrather than Lb. acidophilus. Chromosomalrearrangements [82] or cross-protection bymethylation could explain the loss of acleavage site.
2.2.3. Analysis at strain level
Restriction Enzyme Analysis (REA)
Restriction enzyme analysis (REA)involves the extraction and digestion ofchromosomal DNA with restriction endo-nucleases and separation of the fragmentsby conventional gel electrophoresis (CGE).The number of bands obtained, generallyranging between 1000 and 20 000 bp insize, are dependent on the restriction
288 V. Coeuret et al.
enzymes used. The complexity of thebanding pattern makes visual evaluationdifficult and necessitates the use of compu-ter-assisted multivariate analysis [32].Electrophoretic separation of the DNAfragments obtained after restriction endo-nuclease digestion has been achieved formany bacterial species of the genus Lacto-bacillus. REA has been successfully usedto differentiate between strains of Lb. aci-dophilus [157]. Zhong et al. [202] usedBclI and DraI to separate 64 strains oflactobacilli; this method allowed discrimi-nation between the strains, but the patternsproduced were very complex.
RAPD/ AP-PCR
Polymerase chain reaction (PCR)-basedDNA fingerprinting methods using arbi-trary primers (AP) have been developedfor studying genomic DNA polymorphism.Arbitrarily primed PCR and randomlyamplified polymorphic DNA (RAPD)methods were first described in 1990 [197,198]. In these similar techniques, theprimers are generally about 10 nucleotideslong and are not directed at any knownsequence of the bacterial genome, as theprimers are selected arbitrarily. A singlearbitrary oligonucleotide primer directs theamplification of random segments ofgenomic DNA. It generates a characteris-tic spectrum of short DNA products of var-ious complexities. RAPD techniques havebeen extensively used in the typing of lac-tic acid bacteria [176]. Some of the prim-ers used are listed in Table VI. Randomlyamplified polymorphic DNA analysis hasbeen used to monitor population dynamicsin food fermentation and to estimate thediversity of Lactobacillus strains in cheeses[12, 17, 22] whey culture [38], sausagefermentation [137, 152] and maize dough[91]. It can also be used to distinguish aparticular strain from the natural flora,such as distinguishing Lactobacillus probi-otic adjunct from the natural NSLAB pop-ulation in Cheddar cheese [172]. Du Plessis
and Dicks [61] used RAPD to separate spe-cies of the Lactobacillus acidophilus group(Lb. acidophilus, Lb. crispatus, Lb. amy-lovorus, Lb. gallinarum, Lb. gasseri andLb. johnsonii), which are difficult to distin-guish on the basis of simple physiologicaland biochemical tests. Johansson et al. [109]evaluated the typing potential of RAPD forLb. plantarum strains from various sources:50% of the strains could be individuallyseparated from all the other strains testedand REA could separate the rest.
Cocconcelli et al. [38] demonstratedthat the community of thermophilic lacto-bacilli that dominates in Parmesan cheeseis composed of a limited number of bacte-rial strains belonging to the Lb. helveticusand Lb. delbrueckii ssp. lactis species.Baruzzi et al. [12] reported strains of thefollowing species: Lb. acetotolerans, Lb.alimentarius, Lb. brevis, Lb. gasseri, Lb.kefiri, Lb. paracasei, Lb. plantarum andLb. zeae in Ricotta forte cheese. RAPDanalysis was used to separate the strains,which were previously grouped into oneprotein profile cluster as Lb. plantarumand Lb. pentosus [187]. Quiberoni et al.[149] used primers P1 and P2 to discrimi-nate between 25 isolates obtained fromSardo and Reggianito cheeses. Giraffa et al.[81] characterised 23 strains of Lb. helveti-cus isolated from natural whey starter cul-tures used for Italian hard cheese. Sohieret al. [169] used RAPD primers (and REP-PCR) to separate isolates of Lb. brevis andLb. hilgardii. The two fingerprinting meth-ods were equally suitable for revealingspecies-specific genetic profiles. RAPDanalysis may have the advantage of facili-tating simultaneous strain typing, speciesaffiliation determination and individualstrain differentiation [16]. RAPD-derivedprobes and primers have been describedfor the identification of lactobacilli to spe-cies level, and even to strain level [148].Tilsala-Timisjarvi and Alatossava [180]have also developed strain-specific DNA-derived PCR primers for a probiotic strainof Lb. rhamnosus.
It is also possible to use a combinationof two or more 10-mer oligonucleotides(multiplex RAPD) in a single PCR to gen-erate RAPD profiles, making it possible todiscriminate between Lactobacillus strains[51].
RAPD analysis has been shown to beless effective than other molecular meth-ods, although in some cases, it allowed theseparation of strains indistinguishable byother techniques. The screening of newRAPD primers might increase the specifi-city of this technique for strain typing.RAPD analysis is a rapid and cheap method,but careful optimisation is required toobtain reproducible results.
REP-PCR/ERIC-PCR
Repeated sequences are present in thegenomes of all organisms. The firstdescribed and most extensively studiedrepeated sequence is the extragenic palin-drome (REP) or palindromic unit (PU), ini-tially identified in Salmonella typhimu-rium and Escherichia coli [79]. Thissequence has a copy number of 500–1000and consists of a 35-40 bp inverted repeat.It is found in clusters, with successive cop-ies arranged in opposite orientations. Asecond family of repetitive elements,
called IRU (intergenic repeat units) orERIC (enterobacterial repetitive intergenicconsensus) has been described; IRU are124-127 bp long and have a copy numberof about 30-50 in E. coli and 150 in S. typh-imurium [100, 166]. The members of boththe PU and IRU families are located innon-coding but probably transcribedregions of the chromosome and both havea potential stem-loop structure. Thesesequences have been searched for andstudied in lactobacilli [17, 22, 102, 169],and repetitive primers were designed forPCR. Some of the primers used are listedin Table VII.
Berthier et al. [17] analysed 488 isolatesof mesophilic lactobacilli isolated fromComté cheese. These isolates were identifiedto species and strain level with a combina-tion of two PCR-based methods: amplifica-tion with pairs of repetitive primers (ERICand REP), and amplification with specificprimers. REP-PCR fingerprints were used toassign strains to species. Combined REPand ERIC fingerprints have the advantageover RAPD analysis that the sequencesconsidered are longer and are therefore lesssensitive to minor changes in reaction condi-tions [17]. Hyytiä-Trees et al. [102] concludedthat REP-PCR has a discriminatory powersimilar to that of RAPD analysis, but
weaker than that of pulsed-field gel elec-trophoresis (PFGE). However, if the resultsof REP-PCR and RAPD analysis werecombined, the discriminatory power insome cases equalled that of PFGE.
Amplified fragment lengthpolymorphism (AFLP)
Another method, although not widelyused except in a systematic approach, iscalled amplified fragment length polymor-phism (AFLP). This method is sophisti-cated and combines PCR and restrictionenzyme techniques. AFLP templates arefirst prepared by cutting with two enzymes(for example, the six-base cutter HindIIIand the four-base cutter MseI, for Lactoba-cillus), resulting in DNA fragments withtwo different types of sticky ends. Appro-priate adaptors (short oligonucleotides) areligated to these ends to form templates forPCR, using two different primers contain-ing the adaptor sequence extended toinclude one or more selective bases next tothe restriction site of the primer. Only frag-ments that completely match the primersequence are amplified, resulting in selec-tive amplification according to the initialDNA structure and cutting. The amplifica-tion process results in an array of 30 to40 DNA fragments that are group- and/orspecies- and/or strain-specific [108]. How-ever, AFLP analysis, which involves alarge number of experimental steps, has tobe carefully monitored by a specialist toensure a high level of reproducibility, evenif an automated AFLP-fingerprinting sys-tem is used [75]. RAPD-PCR and AFLPanalyses have been used for the rapid typ-ing of strains of Lb. acidophilus and relatedspecies, using at least 3 different primers,and digital analysis of the combined pat-tern for all the primers. These techniqueswere found to discriminate between thestrains at a much finer taxonomic levelthan SDS-PAGE fingerprinting, althoughtheir reproducibility was found to be a mat-ter of debate [75].
Pulsed-field gel electrophoresis (PFGE)
Like REA, this method involves restric-tion enzyme digestion, but the enzymesused for PFGE must have a low cutting fre-quency, as is the case for SmaI and SgrAI.However, in this case, the restriction frag-ments are resolved by pulsed-field gelelectrophoresis. This technique, involvingthe application of an alternating electric fieldin two defined directions, is used to sepa-rate very large fragments, from 5 � 104 pbto 2 � 106 pb. This method is highly dis-criminatory and reproducible, and gener-ates a banding pattern that is easy to inter-pret.
PFGE analysis alone, with two or threeappropriate enzymes, can be used for reli-able strain typing. In several Lactobacillusstudies, PFGE has been shown to be apowerful method for strain typing(Tab. VIII). However, one drawback ofthis method is that only a limited numberof samples can be analysed. Tynkkynenet al. [184] identified, by PCR with spe-cific primers, 24 strains of Lactobacillusbiochemically related to the Lactobacilluscasei group. These strains were typed byRAPD, ribotyping and PFGE, to make pos-sible comparison of the discriminatorypower of the methods. Twelve RAPD gen-otypes were detected among the 24 Lacto-bacillus strains; ribotyping with EcoRIproduced 15 different fingerprint patternsand PFGE revealed 17 different genotypes.
Blaiotta et al. [20] used PFGE to moni-tor the addition of Lactobacillus, used asstarter, to Cacioricotta cheese. This tech-nique made it possible to analyse thegrowth kinetics of each starter strain dur-ing the process. Similarly, Jacobsen et al.[107] monitored the survival of probioticstrains of Lb. rhamnosus, Lb. reuteri andLb. delbrueckii ssp. lactis in faeces.
Bouton et al. [22] used a PCR-basedmethod and PFGE for typing and monitor-ing homofermentative lactobacilli duringComté cheese ripening. Isolates, whichexhibited unique patterns by RAPD or
292 V. Coeuret et al.
REP-PCR, were distinguishable by PFGE,but some strains which were distinguisha-ble by RAPD or REP-PCR were related byPFGE; these discrepancies were explainedby the different exploration of DNA poly-morphism (the whole DNA chromosomefor PFGE, and region amplified by primersfor RAPD and the REP-PCR). The use ofsecond restriction enzymes would cer-tainly be useful in this case.
Plasmid profiling
Plasmid-based typing methods are use-ful, particularly if large numbers of strainsare to be examined. Plasmids are notevenly distributed among the various iso-lates belonging to different species ofLactobacillus. A rapid mechanical lysismethod for the routine analysis of plasmids
from Lactobacillus isolated from sour-dough has been reported [153].
The plasmid profile of a particular LABor Lactobacillus strain may be used as anidentification marker. It has been per-formed on Lactobacillus strains [63, 64,136]. However, a plasmid DNA replicatesindependently from the chromosome, andplasmidic genes are usually more unstablethan chromosomal gene function [65].Although the plasmid profile of a strainremains stable under laboratory condi-tions, plasmids may be lost during fermen-tation in unfavourable growth conditions,or may undergo rearrangements by conju-gative transfer. Therefore, plasmid profilingtechniques have several potential disad-vantages, such as the ability of a strain togain, lose or modify its plasmids [192], and
Table VIII. PFGE Restriction enzymes and migration conditions used for lactobacilli.
Lactobacilli species Restriction enzyme Running conditions Ref.
there is often no correlation between plas-mid content and species identification[202].
Phage-related DNA polymorphism
Bacterial lysogeny, that is, the presenceof prophages as genetic elements in thebacterial chromosome, is common inlactobacilli, especially in the Lb. caseigroup [72]. Brandt et al. [24] investigatedwhether further intraspecies DNA poly-morphism could be revealed by screeningfor the presence and distribution of phage-related DNA sequences among the strainsof a single bacterial species. He studied 11Lb. rhamnosus strains used as starter andprobiotic cultures in the Finnish dairyindustry. Six different PCR product pat-terns were obtained by amplification withprimer pairs derived from the nucleotidesequence of a HindIII fragment of thephage Lc-Nu genome. Phage Lc-Nu DNA-derived PCR was found to be an effectivetool for detecting polymorphism in the Lb.rhamnosus strains.
2.2.4. Global techniques
The recent development of whole bacte-rium analysis by FTIR (Fourier transforminfrared spectroscopy) is of great potentialfor rapid identification of lactobacilli [1, 2,48, 49, 128]. Bacterial spectra are usuallyrecorded in the mid-infrared. They are spe-cific to a bacterial strain and show thevibrational characteristics of the whole cel-lular components: fatty acids, intracellularand membrane proteins, polysaccharidesand nucleic acids. The statistical treatmentof spectral data makes it possible to dis-criminate between different genera, speciesand strains. The reproducibility problemsinitially encountered have been resolved,resulting in standardised conditions for cellgrowth and sample preparation. The princi-pal advantage of this technique, as pointedout by almost all authors, is its simplicitywith respect to genome analysis. Amielet al. [1] established libraries of species
used in the cheese industry. Wild strains oflactobacilli isolated from raw milk cheesesfrom Normandy were well identified. Thespectral database makes it possible to iden-tify new strains, with a high percentage ofgood results: 100% at the genus and spe-cies level for collection strains; and 100%at the genus level and 69% at the specieslevel for wild isolates previously identifiedby RAPD and the phenotypic method. Theresults obtained were as reliable as thoseobtained by genomic methods such asRAPD analysis [2].
Another global technique, pyrolysis-mass spectrometry [125], is also verypromising but has not yet been applied tolactobacilli.
3. CHARACTERISATION AND IDENTIFICATION OF LACTOBACILLI BY CULTURE-INDEPENDENT METHODS
Culture-independent methods involveextraction of nucleic acids (DNA or RNA)from raw samples and the use of probes forhybridisation and primers for denaturinggradient gel electrophoresis (DGGE), temper-ature gradient gel electrophoresis (TGGE)and single strand conformation polymor-phism (SSCP).
These techniques give a picture of thepopulations present in a complex matrix,bypassing problems related to injured, andviable but non-cultivable bacteria.
3.1. Hybridisation
Ehrmann et al. [68] developed a tech-nique facilitating the direct identificationof LAB, without prior culture, in cheese,yoghurt, sausages, sauerkraut and sour-dough and based on a reverse dot-blotassay. Oligonucleotide probes specific to thevarious Lactobacillus species are extendedby adding a polythymidine phosphate tail.
294 V. Coeuret et al.
These extended oligonucleotides bind nat-urally to filter membranes and serve asspecies-specific capture probes for thelabelled, PCR-amplified rRNA gene frag-ment [68].
A simple, rapid method for whole-cellhybridisation with fluorescent-labeled rRNA-targeted peptide nucleic acid (PNA) probeswas recently developed for the detectionand identification of thermophilic Lacto-bacillus cells growing in milk or present inindustrial starter culture [129]. The proto-col involves filtration of the samples, andepifluorescence microscopy is used fordetection. Its detection limit is 104 to 106
cells per mL, and specific oligonucleotidesare available for Lb. delbrueckii, Lb. helve-ticus, Lb. pentosus and Lb. plantarum. TherRNA-targeted oligonucleotide probescurrently available for the identification ofLAB are listed in Table III. Various typesof assay can be used. In dot-blot assays thetarget nucleic acid must be extracted andimmobilised on a membrane. Such assayshave been used for the simultaneous iden-tification of various lactobacilli, such asLb. curvatus, Lb. sake, Lb. pentosus, Lb.plantarum, Lb. delbrueckii and Lb. helveti-cus [97, 143, 192].
3.2. PCR-DGGE, PCR-TGGE
Methods such as denaturing gradient gelelectrophoresis (DGGE) and temperaturegradient gel electrophoresis (TGGE) have
been developed for the analysis of micro-bial communities without culture, by thesequence-specific separation of amplified16S rDNA fragments. Separation is basedon the lower electrophoretic mobility of apartially melted double-stranded DNAmolecule in polyacrylamide gels contain-ing a linear gradient of DNA denaturants, amixture of urea and formamide (DGGE),or subjected to a linear temperature gradi-ent (TTGE). The melting of fragments pro-ceeds in discrete melting domain stretchesof base pairs with an identical melting tem-perature. Once the domain with the lowestmelting temperature reaches its meltingtemperature (Tm), in the denaturing ortemperature gradient gel, the moleculeundergoes a transition from a helical to apartially melted structure, and its migrationstops. Optimal resolution occurs whenamplicons are not completely denaturedand when the region to be screened is in thelowest melting domain. This is achievedby adding a 30-40 bp GC-rich clamp to oneof the PCR primers (Tab. IX). This resultsin sequence variants of particular frag-ments ceasing to migrate at different posi-tions in the denaturing gradient, facilitat-ing their effective separation by TGGE orDGGE.
The members of the bacterial commu-nity are often amplified using primers cor-responding to the 16 S rDNA sequence [5,39, 93, 144, 168, 186, 194, 195]. Speciescan then be distinguished by comparing the
migration distance of the PCR ampliconsin gels with those of reference strains [3].
These techniques have recently beenused in the evaluation of microbial diver-sity, particularly the diversity of lactoba-cilli in cheeses [44, 144, 150], sausages[39], starch fermentation [5], malt whiskyfermentation [186], beer [183], faeces[168, 178, 195], and the gastrointestinaltract [194], and for the identification ofLactobacillus species [189]. Coppola et al.[44] used PCR 16S-23S rDNA spacer pol-ymorphism and PCR-DGGE analysis ofthe V3 region of 16S rDNA to study themicrobial diversity of unripened PastaFilata cheeses. The number of bands in thePCR profiles obtained made it possible todistinguish between industrial, cottageindustry and traditional dairy products.
Ogier et al. [144] identified by PCR-TGGE of the 16S rDNA V3 regions bacte-rial species (Lactobacillus, Lactococcus,Leuconostoc, Enterococcus, Pediococcus,Streptococcus and Staphylococcus) presentin home-made or commercial dairy prod-ucts. V3-TGGE sequences differentiatebetween bacteria belonging to the differentgenera. However, V3-TGGE did not distin-guish between members of the Lactobacilluscasei group (Lb. casei, Lb. paracasei andLb. rhamnosus), or members of the Lacto-bacillus acidophilus group (Lb. galli-narum, Lb. crispatus, Lb. amylovorus, Lb.acidophilus and closely Lb. helveticus).Lb. pentosus and Lb. plantarum or Lb.johnsonii and Lb. gasseri have similar V3sequences and co-migrate. Only Lb. reuteri,Lb. brevis, Lb. fermentum, Lb. delbrueckiissp. bulgaricus, and Lb. delbrueckii ssp.lactis can be easily distinguished. Cheeseswith similar production procedures (e.g.Brie and Camembert cheeses, or Emmen-tal and Comté…) produced commonTGGE bands. TGGE provides a descrip-tion of the dominant bacterial species in acomplex ecosystem, but bacterial minority(less than 1%) cannot be detected. So,TGGE seems to be an excellent moleculartool to analyse diversity within complex
bacterial communities, but screening innew primers in a more discriminating areathan V3 16S rDNA is necessary for lacto-bacilli species.
Randazzo et al. [150] obtained DGGEprofiles derived from PCR and RT-PCR(reverse transcriptase-PCR) of DNA andRNA, and compared them in order todetermine the expression level of the 16SrRNA genes of the most predominant bacte-ria during the manufacturing of Ragusanocheese. The evolution of the Lactobacilluscommunity during the manufacturing andripening process was reflected in the unsta-ble DGGE profiles generated by usingLactobacillus-specific primers, which tar-get all members of the Lactobacillusgroup, but include Leuconostoc and Pedio-coccus spp. too. Concurrently microbialenumeration on different media was done,and after cultivation, isolates were identi-fied by both classical phenotypic methodsand 16S rDNA sequence analysis. Lb. del-brueckii, that was dominant as indicated byDGGE, was never isolated on the selectivemedia.
Tannock et al. [178] assessed the impactof probiotic consumption on the intestinalmicroflora by monitoring the faecal com-munity by FISH (fluorescence in situhybridisation) and DGGE. Heilig et al.[93] used DGGE to study the stability ofthe bacterial community of the gastrointes-tinal tract in various age groups, over vari-ous time periods, and successive changesin the Lactobacillus community wereobserved. They also assessed the specifi-city of the PCR and DGGE approach forstudying the retention in faecal samples ofa Lactobacillus strain administered duringa clinical trial.
Van Beek and Priest [186] adopted apolyphasic approach, using light and elec-tron microscopy and denaturing gradientgel electrophoresis (DGGE) of PCR-ampli-fied fragments of 16S ribosomal DNA tomonitor the development of the lactic acidbacterial community during malt whiskyfermentation. Their results revealed that
296 V. Coeuret et al.
culture-dependent methods underestimatedbacterial diversity and demonstrated thepresence of novel lactobacilli and othertaxa in malt whisky during fermentation.
These new techniques have the advan-tage of facilitating the direct study andanalysis of population dynamics withincomplex microbial systems. These meth-ods are only just beginning to be applied toecological studies of cheese microflora.
3.3. PCR-SSCP
SSCP (single strand conformation poly-morphism) analysis for molecular identifi-cation in microbial ecosystems is based on16S rDNA sequences. No culture isrequired. SSCP detects sequence varia-tions between DNA fragments, usuallyamplified by PCR from variable regions ofthe 16S rRNA gene and uses neutral, non-denaturing polyacrylamide gels. Shortdouble-stranded DNA fragments are firstgenerated using standard PCR protocols.Single-stranded DNA is created by com-bining a small aliquot of PCR product withan equal volume of formamide (80–95%),and then it is heat denatured. The two com-plementary strands of DNA will migratedifferently and will therefore separate duringgel electrophoresis. This fingerprintingmethod characterises microbial diversity,by comparing closed microflora [84]. SSCPanalysis has been adapted for the rapididentification of both Gram-positive andGram-negative bacteria to genus and spe-cies level. Analysis of the microflora ofAOC Salers cheese resulted in the co-elution of Lc. lactis, Lb. plantarum, Lb.pentosus and St. thermophilus, all of whichwere associated with the main peak [84].
4. CONCLUSION
It is widely recognised that the identifi-cation of lactobacilli to species or strainlevel on the basis of physiological and bio-chemical criteria is very ambiguous andcomplicated. Numerous taxonomic changeshave been observed in the Lactobacillusgenus as qualification of old species in new
genera or description of new species. Thisleads to a problematic genus characterisationby phenotypic tests and to an increasinguse of classical culture-based molecularmethods. New molecular techniques formicrobial community analysis that do notrequire isolation of the microorganisms arevery promising. They provide a complemen-tary picture of the population obtained usingculture-based techniques when applied tothe analysis of milks and dairy products.However, these molecular approaches haveseveral limitations, including the design ofadequate primers, and the possibility thatDNA isolation, amplification and cloningmight be biased by certain strains andsequences. There is also a dependence onthe detection threshold and on the numberof lactobacilli, unfortunately low in high-quality raw milks. Nevertheless, thesemethods provide an overview of the diver-sity of microorganisms present in a partic-ular sample. They provide qualitative andpossibly semi-quantitative information,which should be complemented by quanti-tative PCR to obtain results as close as pos-sible to reality.
We intended to produce a guide for thereader, covering pertinent techniques usedfor a polyphasic analysis of lactobacilli.This guide was not produced, since pro-posing a technique for a specified use is notcompletely reliable. In fact, the choice ofthe technique to use is variable dependingon:– the level of discrimination required;– the type of product and matrix, the nature
of undesirable organisms and the diver-sity of the lactobacilli present;
– the amount of time available;– the available staff and equipment;– the number of strains and isolates to be
studied.Furthermore, numerous techniques, cul-
ture-dependent or culture-independent, arebased on the use of probes and primers. Forthese techniques the discrimination leveldepends on the existence or not of the
Identification of dairy lactobacilli 297
probes and primers at the taxonomic leveldesired. To date we are very far from hav-ing specific primers and probes for the 88lactobacilli species, and regarding thosewhich have been designed, their specificityand validity should be checked one by onewith closed genera, species or strains.Another problem results from the given listof 88 lactobacilli species since it is not anofficial list (it does not exist) and thus tobypass possible misidentification allprobes and primers should be validatedagainst the same reference strain at thebeginning to ensure their common specifi-city. Moreover, all the techniques men-tioned in this review have not been appliedto the lactobacilli using the same objec-tives. The genus primer designed byDubernet et al. [62], has been used for PCRand PCR-TGGE, but not for hybridisation,but it is clear that it could be used. The dif-ficulty of choosing a technique that hasgood dicrimination power depends notonly on the techniques but also on the spe-cies or strains. Results also depend on thequality and exhaustivity of a database.Very good results at genus, species andstrain level could be obtained by usingFTIR, but if the database is not complete(not enough strains of different species, orof different origin), results will not be asgood as they could be. Finally, only a fewlimited techniques can be applied with ahigh degree of confidence although theyare dependent on database robustness:sequencing to identify at genus and specieslevel, and sequencing or pulsed field gelelectrophoresis to discriminate strains.
In conclusion, analysis of lactobacilli incheeses and other dairy products is very com-plicated and the use of different techniques,especially molecular-based phenotypic orgenomic techniques, is recommended.
REFERENCES
[1] Amiel C., Mariey L., Curk-Daubié M.C.,Pichon P., Travert J., Potentiality of FourierTransform Infrared Spectroscopy (FTIR) fordiscrimination and identification of dairylactic acid bacteria, Lait 80 (2000) 445–459.
[2] Amiel C., Mariey L., Denis C., Pichon P.,Travert J., FTIR spectroscopy and taxono-mic purpose: Contribution to the classifica-tion of lactic acid bacteria, Lait 81 (2001)249–255.
[3] Ampe F., Design and evaluation of a Lacto-bacillus manihotivorans species-specificrRNA-targeted hybridization probe and itsapplication to the study of sour cassava fer-mentation, Appl. Environ. Microbiol. 66(2000) 2224–2226.
[4] Ampe F., ben Omar N., Moizan C., WacherC., Guyot J.P., Polyphasic study of the spa-tial distribution of microorganisms in Mexi-can pozol, a fermented maize dough,demonstrates the need for cultivation-inde-pendent methods to investigate traditionalfermentations, Appl. Environ. Microbiol.65 (1999) 5464–5473.
[5] Ampe F., Sirvent A., Zakhia N., Dynamicsof the microbial community responsible fortraditional sour cassava starch fermentationstudied by denaturing gradient gel electro-phoresis and quantitative rRNA hybridiza-tion, Int. J. Food Microbiol. 65 (2001)45–54.
[6] Andrighetto C., De Dea P., Lombardi A.,Neviani E., Rossetti L., Giraffa G., Molecu-lar identification and cluster analysis ofhomofermentative thermophilic lactobacilliisolated from dairy products, Res. Micro-biol. 149 (1998) 631–643.
[7] Antonsson M., Ardo Y., Molin G., A com-parison between the microflora of Herrgardcheese from three different dairies, Int.Dairy J. 11 (2001) 285–291.
[8] Antunes A., Rainey F.A., Nobre M.F.,Schumann P., Ferreira A.M., Ramos A.,Santos H., da Costa M.S., Leuconostocficulneum sp. nov., a novel lactic acid bac-terium isolated from a ripe fig, and reclassi-fication of Lactobacillus fructosus as Leu-conostoc fructosum comb. nov., Int. J. Syst.Evol. Microbiol. 52 (2002) 647–655.
[10] Arizcun C., Barcina Y., Torr P., Identifica-tion of lactic acid bacteria isolated fromRoncal and Idiazabal cheeses, Lait 77(1997) 729–736.
[11] Back W., BohaK I., Ehrmann M., LudwigW., Pot B., Kersters K., Schleifer K.H.,Lactobacillus perolens sp. nov., a soft drinkspoilage bacterium, Syst. Appl. Microbiol.22 (1999) 354–359.
[12] Baruzzi F., Morea M., Matarante A., Coc-concelli P.S., Changes in the Lactobacilluscommunity during Ricotta forte cheese
298 V. Coeuret et al.
natural fermentation, J. Appl. Microbiol. 89(2000) 807–814.
[13] Beresford T., Pelaez C., Jimeno J., Natureand growth of non starter flora, in: Euro-pean Commission, Quality and microbio-logy of traditional and raw milk cheeses,Dijon, France, 1998, pp. 225–238.
[15] Berthier F., Ehrlich S.D., Rapid speciesidentification within two groups of closelyrelated lactobacilli using PCR primers thattarget the 16S/23S rRNA spacer region,FEMS Microbiol. Lett. 161 (1998) 97–106.
[16] Berthier F., Ehrlich S.D., Genetic diversitywithin Lactobacillus sakei and Lactobacil-lus curvatus and design of PCR primers forits detection using randomly amplifiedpolymorphic DNA, Int. J. Syst. Bacteriol.49 (1999) 997–1007.
[17] Berthier F., Beuvier E., Dasen A., GrappinR., Origin and diversity of mesophilic lac-tobacilli in Comté cheese, as revealed byPCR with repetitive and species-specificprimers, Int. Dairy J. 11 (2001) 293–305.
[18] Birollo G.A., Reinheimer J.A., VinderolaC.G., Viability of lactic acid microflora indifferent types of yoghurt, Food Res. Int. 33(2000) 799–805.
[19] Bjorkroth J., Ridell J., Korkeala H., Charac-terization of Lactobacillus sake strainsassociating with production of ropy slimeby randomly amplified polymorphic DNA(RAPD) and pulsed-field gel electrophore-sis (PFGE) patterns, Int. J. Food Microbiol.31 (1996) 59–68.
[20] Blaiotta B.G., Simeoli G.M., Andolfi R.,Villani F., Coppola S., Monitoring lacticacid bacteria strains during “Cacioricotta”cheese production by restriction endonu-clease analysis and pulsed-field gel electro-phoresis, J. Dairy Res. 68 (2001) 139–144.
[21] Bouton Y., Guyot P., Grappin R., Preliminarycharacterization of microflora of Comtécheese, J. Appl. Microbiol. 85 (1998) 123–131.
[22] Bouton Y., Guyot P., Beuvier E., TailliezP., Grappin R., Use of PCR-based methodsand PFGE for typing and monitoring homo-fermentative lactobacilli during Comtécheese ripening, Int. J. Food. Microbiol. 76(2002) 27–38.
[23] Bracquart P., An agar medium for the diffe-rential of Streptococcus thermophilus andLactobacillus bulgaricus in yoghurt, J.Appl. Bacteriol. 51 (1981) 303–305.
[24] Brandt K., Tilsala-Timisjarvi A., Alatos-sava T., Phage-related DNA polymorphismin dairy and probiotic Lactobacillus,Micron 32 (2001) 59–65.
[25] Bringel F., Curk M.C., Hubert J.C., Charac-terization of lactobacilli by southern-typehybridization with a Lactobacillus planta-rum pyrDFE probe, Int. J. Syst. Bacteriol.46 (1996) 588–594.
[26] Bruce J., Automated system rapidly identi-fies and characterises microorganisms infood, Food Technol. 50 (1996) 77–81.
[27] Bunte C., Hertel C., Hammes W.P., Moni-toring and survival of Lactobacillus para-casei LTH 2579 in food and the humanintestinal tract, Syst. Appl. Microbiol. 23(2000) 260–266.
[28] Calicchia M.L., Wang C.I.E., Nomura T.,Yotsuzuka F., Osato D.W., Selective enu-meration of Bifidobacterium bifidum, Ente-rococcus faecium, and Streptomycin-resis-tant Lactobacillus acidophilus from amixed probiotic product, J. Food Prot. 56(1993) 954–957.
[29] Camaschella P., Pirovano F., Mognot O.,Sozzi T., Method for the detection and enu-meration of different thermophilic lacticcultures and bifidobacteria utilising onlyone culture medium, in: Amgar A. (Ed.),Food Safety’94, Laval France, 1994, p. 335.
[30] Chagnaud P., Machinis K., Coutte L.A.,Marecat A., Mercenier A., Rapid PCR-based procedure to identify lactic acid bac-teria: application to six common Lactoba-cillus species, J. Microbiol. Meth. 44(2001) 139–148.
[31] Chamba J.F., Emmental cheese: a complexmicrobial ecosystem. Consequences onselection and use of starters, Sci. Aliments20 (2000) 37–54.
[32] Charteris W.P., Kelly P.M., Morelli L.,Collins J.K., Selective detection, enumera-tion and identification of potentially probio-tic Lactobacillus and Bifidobacterium spe-cies in mixed bacterial populations, Int. J.Food Microbiol. 35 (1997) 1–27.
[33] Chen H., Lim C.K., Lee Y.K., Chan Y.N.,Comparative analysis of the genes encoding23S-5S rRNA intergenic spacer regions ofLactobacillus casei-related strains, Int. J.Evol. Microbiol. 50 (2000) 471–478.
[34] Choisy C., Desmazeaud M., Guéguen M.,Lenoir J., Schmidt J.L., Tourneur C., Lesphénomènes microbiens, in: Eck A., GillisJ.C. (Eds.), Le fromage, Lavoisier Tec &Doc, Paris, France, 1997, pp. 377–446.
[35] Chopard M.A., Schmitt M., Perreard E.,Chamba J.F., Qualitative aspect of proteolyticactivity of thermophilic lactobacilli using inSwiss cheeses, Lait 81 (2001) 183–194.
Identification of dairy lactobacilli 299
[36] Cocconcelli P.S., Porro D., Galandini S.,Senini L., Development of RAPD protocolfor typing of strains of lactic acid bacteriaand enterococci, Lett. Appl. Microbiol. 21(1995) 376–379.
[37] Cocconcelli P.S., Senini L., Bottazzi V.,Study of the lactic microflora of Tomacheese: use of rRNA targeted oligonucleo-tides for the identification of isolatedstrains, Ann. Microbiol. Enzymol. 46(1996) 21–28.
[38] Cocconcelli P.S., Parisi M.G., Senini L.,Bottazzi V., Use of RAPD and 16S rDNAsequencing for the study of Lactobacilluspopulation dynamics in natural whey cul-ture, Lett. Appl. Microbiol. 25 (1997) 8–12.
[39] Cocolin L., Manzano M., Cantoni C., ComiG., Development of a rapid method for theidentification of Lactobacillus spp. isolatedfrom naturally fermented Italian sausagesusing a polymerase chain reaction-tempera-ture gradient gel electrophoresis, Lett.Appl. Microbiol. 30 (2000) 126–129.
[40] Collins M.D., Rodrigues U., Ash C.,Aguirre M., Farrow J.A.E., Martinez-Murcia A., Phillips B.A., Williams A.M.,Wallbanks S., Phylogenetic analysis of thegenus Lactobacillus and related lactic acidbacteria as determined by reverse transcrip-tase sequencing of 16S rRNA, FEMSMicrobiol. Lett. 77 (1991) 5–12.
[41] Collins M.D., Samelis J., Metaxopoulos J.,Wallbanks S., Taxonomic studies on someleuconostoc-like organisms from fermentedsausages: description of a new genus Weis-sella for the Leuconostoc paramesenteroi-des group of species, J. Appl. Bacteriol. 75(1993) 595–603.
[42] Coppola R., Nanni M., Iorizzo M., Sorren-tino A., Sorrentino E., Grazia L., Survey oflactic acid bacteria during the advanced sta-ges of the ripening of Parmigiano Reggianocheese, J. Dairy Res. 64 (1997) 305–310.
[43] Coppola R., Nanni M., Iorizzo M., SorrentinoA., Sorrentino E., Chiavari C., Grazia L.,Microbiological characteristics of Parmi-giano Reggiano cheese during the cheese-making and the first months of the ripening,Lait 80 (2000) 479–490.
[44] Coppola R., Nanni M., Succi M., SorrentinoA., Iorizzo M., Chiavari C., Grazia L., Enu-meration of thermophilic lactic acid bacte-ria in ripened cheeses manufactured fromraw milk, Milchwissenschaft 56 (2001)140–142.
[45] Corbo M.R., Albenzio M., De Angelis M.,Sevi A., Gobbetti M., Microbiological andbiochemical properties of Canestrato Pugliesehard cheese supplemented with bifidobac-teria, J. Dairy Sci. 84 (2001) 551–561.
[46] Crow V., Curry B., Hayes M., The ecologyof non-starter lactic acid bacteria (NSLAB)and their use as adjuncts in New ZealandCheddar, Int. Dairy J. 11 (2001) 275–283.
[47] Curk M.C., Caractérisation des espècesbrassicoles du genre Lactobacillus, Thèsede Doctorat, Université Louis Pasteur,Strasbourg, France, 1994.
[48] Curk M.C., Peladan F., Hubert J.C., Fouriertransform infrared spectroscopy for identi-fying Lactobacillus species, FEMS Micro-biol. Lett. 123 (1994) 241–248.
[49] Curk M.C., Amiel C., Denis C., Reyrolle J.,Pichon P., Travert J., Discrimination desouches de bactéries lactiques d’intérêtlaitier par spectroscopie infrarouge à trans-formée de Fourier (IRTF), in: ADRIA-Nor-mandie, les bactéries lactiques, Lactic 97,Caen, 10–12 Septembre, France, 1997,420–421.
[50] Daniel P., Sizing of the Lactobacillus plan-tarum genome and other lactic acid bacteriaspecies by transverse alternating field elec-trophoresis, Curr. Microbiol. 30 (1995)243–246.
[51] Daud Khaled A.K., Neilan B.A., Henriks-son A., Conway P.L., Identification andphylogenetic analysis of Lactobacillususing multiplex RAPD-PCR, FEMS Micro-biol. Lett. 153 (1997) 191–197.
[52] Dave R.I., Shah N.P., Evaluation of mediafor selective enumeration of Streptococcusthermophilus, Lactobacillus delbrueckiissp. bulgaricus, Lactobacillus acidophilus,and bifidobacteria, J. Dairy Sci. 79 (1996)1529–1536.
[53] De Angelis M., Corsetti A., Tosti N., RossiJ., Corbo M.R., Gobbetti M., Characteriza-tion of non-starter lactic acid bacteria fromItalian ewe cheeses based on phenotypic,genotypic, and cell wall protein analyses, Appl.Environ. Microbiol. 67 (2001) 2011–2020.
[54] De Man J.C., Rogosa M., Sharpe M.E., Amedium for cultivation of lactobacilli, J.Appl. Bacteriol. 23 (1960) 130–135.
[56] Dellaglio F., Lombardi A., New develop-ments in the identification of non startermicroorganisms in cheese, in: EuropeanCommission, Quality and microbiology oftraditional and raw milk cheeses, Dijon,France, 1998, pp. 53–63.
[57] Dellaglio F., Felis G.E., Torriani S., Thestatus of the species Lactobacillus casei(Orla-Jensen 1916) Hansen and Lessel
300 V. Coeuret et al.
1971 and Lactobacillus paracasei Collinset al. 1989. Request for an Opinion, Int. J.Syst. Evol. Microbiol. 52 (2002) 285–287.
[58] Demarigny Y., Beuvier E., Dasen A.,Duboz G., Influence of raw milk microfloraon the characteristics of Swiss-type chee-ses. I. Evolution of microflora during ripe-ning and characterization of facultativelyheterofermentative lactobacilli, Lait 76(1996) 371–387.
[59] Demarigny Y., Beuvier E., Buchin S.,Pochet S., Grappin R., Influence of rawmilk microflora on the characteristics ofSwiss-type cheeses. II. Biochemical andsensory characteristics, Lait 77 (1997)151–167.
[60] Drake M., Small C.L., Spence K.D., SwansonB.G., Rapid detection and identification ofLactobacillus spp. in dairy Products byusing the polymerase chain reaction, J.Food Prot. 59 (1996) 1031–1036.
[61] Du Plessis E.M., Dicks L.M., Evaluation ofrandom amplified polymorphic DNA(RAPD)-PCR as a method to differentiateLactobacillus acidophilus, Lactobacilluscrispatus, Lactobacillus amylovorus, Lac-tobacillus gallinarum, Lactobacillus gas-seri, and Lactobacillus johnsonii, Curr.Microbiol. 31 (1995) 114–118.
[62] Dubernet S., Desmasures N., Guéguen M.,A PCR-based method for identification oflactobacilli at genus level, FEMS Micro-biol. Lett. 214 (2002) 271.
[63] Duffner F., O’Connell M., A comparativeevaluation of plasmid profiling and riboty-ping in the analysis of Lactobacillus planta-rum strain heterogeneity in silage, J. Appl.Bacteriol. 78 (1995) 20–27.
[64] Dykes G.A., Burgess A.Y., von Holy A.,Plasmid profiles of lactic acid bacteria asso-ciated with vacuum-packaged Vienna sau-sage manufactured and spoilage, Lett.Appl. Microbiol. 17 (1993) 182–184.
[65] Dykes G.A., von Holy A., Strain typing inthe genus Lactobacillus, Lett. Appl. Micro-biol. 19 (1994) 63–66.
[66] Dykes G.A., Cloete T.E., von Holy A.,Taxonomy of lactic acid bacteria associatedwith vacuum-packaged processed meatspoilage by multivariate analysis of cellularfatty acids, Int. J. Food Microbiol. 28(1995) 89–100.
[67] Ehrmann M., Ludwig W., Schleifer K.H.,Species specific oligonucleotide probe forthe identification of Streptococcus thermo-philus., Syst. Appl. Microbiol. 15 (1992)453–455.
[68] Ehrmann M., Ludwig W., Schleifer K.H.,Reverse dot blot hybridization: a useful
method for the direct identification of lacticacid bacteria in fermented food, FEMSMicrobiol. Lett. 117 (1994) 143–149.
[69] El Soda M., Madkor S.A., Tong P.S.,Adjunct cultures: recent developments andpotential significance to the cheese indus-try, J. Dairy Sci. 83 (2000) 609–619.
[70] Eliskases-Lechner F., Ginzinger W., RohmH., Tschager E., Raw milk flora affectscomposition and quality of Bergkase. 1.Microbiology and fermentation com-pounds, Lait 79 (1999) 385–396.
[71] Ferrero M., Cesena C., Morelli L., ScolariG., Vescovo M., Molecular characteriza-tion of Lactobacillus casei strains, FEMSMicrobiol. Lett. 140 (1996) 215–219.
[72] Forsman P., Tanskanen J., Alatossava T.,Structural similarity and genetic homologybetween Lactobacillus casei bacteriopha-ges isolated in Japan and in Finland, Biosci.Biotechnol. Biochem. 57 (1993) 2042–2048.
[73] Fox P.F., Mc Sweeney P.L.H., Lynch C.M.,Significance on nonstarter lactic acid bacte-ria in Cheddar cheese, Aust. J. Dairy Tech-nol. 53 (1998) 83–89.
[74] Gagnaire V., Thierry A., Léonil J., Propioni-bacteria and facultatively heterofermentativelactobacilli weakly contribute to secondaryproteolysis of Emmental cheese, Lait 81(2001) 339–353.
[75] Gancheva A., Pot B., Vanhonacker K., HosteB., Kersters K., A polyphasic approachtowards the identification of strains belon-ging to Lactobacillus acidophilus and rela-ted species, Syst. Appl. Microbiol. 22(1999) 573–585.
[76] Gardiner G., Ross R.P., Collins J.K.,Fitzgerald G., Stanton C., Development of aprobiotic Cheddar cheese containinghuman-derived Lactobacillus paracaseistrains, Appl. Environ. Microbiol. 64 (1998)2192–2199.
[77] Gasser F., Electrophoretic characterizationof lactic dehydrogenases in the genus Lac-tobacillus, J. Gen. Microbiol. 62 (1970)223–239.
[78] Ghoddusi H.B., Robinson R.K., Enumera-tion of starter cultures in fermented milks, J.Dairy Res. 63 (1996) 151–158.
[79] Gilson E., Clement J.M., Brutlag D.,Hofnung M., A family of dispersed repeti-tive extragenic palindromic DNA sequencesin E. coli, EMBO J. 3 (1984) 1417–1421.
[80] Giraffa G., De Vecchi P., Rossetti L., Note:identification of Lactobacillus delbrueckiisubspecies bulgaricus and subspecies lactisdairy isolates by amplified rDNA restrictionanalysis, J. Appl. Microbiol. 85 (1998)918–924.
Identification of dairy lactobacilli 301
[81] Giraffa G., De Vecchi P., Rossi P., NicastroG., Fortina M.G., Genotypic heterogeneityamong Lactobacillus helveticus strainsisolated from natural cheese starters, J.Appl. Microbiol. 85 (1998) 411–416.
[82] Giraffa G., Gatti M., Rossetti L., Senini L.,Neviani E., Molecular diversity within L.helveticus as revealed by genotypic charac-terization, Appl. Environ. Microbiol. 65(2000) 1259–1265.
[83] Gobbetti M., Morea M., Baruzzi F., CorboM.R., Matarante A., Considine T., DiCagno R., Guinee T., Fox P.F., Microbiolo-gical, compositional, biochemical and tex-tural characterisation of CaciocavalloPuglise cheese during ripening, Int. Dairy J.12 (2002) 511–523.
[84] Godon J.J., Duthoit F., Delbès C., Millet L.,Montel M.C., Use of molecular fringerprintfor the study of complex microbial ecosys-tem. Application to AOC Salers cheese,Lait 81 (2001) 257–262.
[85] Gomes A.M., Malcata F.X., Developmentof probiotic cheese manufactured from goatmilk: response surface analysis via techno-logical manipulation, J. Dairy Sci. 81 (1998)1492–1507.
[86] Grappin R., Beuvier E., Bouton Y., PochetS., Advances in the biochemistry andmicrobiology of Swiss-type cheeses, Lait79 (1999) 1–20.
[87] Hamilton-Miller J.M.T., Shah S., Deficien-cies in microbiological quality and labellingof probiotic supplements, Int. J. FoodMicrobiol. 72 (2002) 175–176.
[88] Hamilton-Miller J.M.T., Shah S., WinklerJ.T., Public health issues arising frommicrobiological and labelling quality offoods and supplements containing probioticmicroorganims, Public Health Nutr. 2(1999) 223–229.
[89] Hammes W.P., Vogel R.F., The genus Lac-tobacillus, in: Wood B.J.B., HolzapfelW.H. (Eds.), The lactic acid bacteria. Thegenera of lactic acid bacteria, Blackie Aca-demic, London, UK, 1995, pp. 19–54.
[90] Hartemink R., Domenech V.R., RomboutsF.M., LAMVAB – A new selective mediumfor the isolation of lactobacilli from faeces,J. Microbiol. Meth. 29 (1997) 77–84.
[91] Hayford A.E., Petersen A., Vogensen F.K.,Jakobsen M., Use of conserved randomlyamplified polymorphic DNA (RAPD) frag-ments and RAPD pattern for characteriza-tion of Lactobacillus fermentum in Gha-naian fermented maize dough, Appl.Environ. Microbiol. 65 (1999) 3213–3221.
[92] Hébert E.M., Raya R.R., Tailliez P., deGiori G.S., Characterization of natural iso-
lates of Lactobacillus strains to be used asstarter cultures in dairy fermentation, Int. J.Food Microbiol. 59 (2000) 19–27.
[93] Heilig H.G.H.J., Zoetendal E.G., VaughanE.E., Marteau P., Akkermans A.D.L., deVos W.M., Molecular diversity of Lactoba-cillus spp. and other lactic acid bacteria inthe human intestine as determined by speci-fic amplification of 16S ribosomal DNA,Appl. Environ. Microbiol. 68 (2002)114–123.
[94] Heller K.J., Probiotic bacteria in fermentedfoods: product characteristics and starterorganisms, Am. J. Clin. Nutr. 73 (2001)374S–379S.
[95] Hensel R., Mayr U., Stetter K.O., KandlerO., Comparative studies of lactic acid dehy-drogenases in lactic acid bacteria. I. Purifi-cation and kinetics of the allosteric L-lacticacid dehydrogenase from Lactobacilluscasei ssp. casei and Lactobacillus curvatus,Arch. Microbiol. 112 (1977) 81–93.
[96] Hensiek R., Krupp G., Stackebrandt E.,Development of diagnostic oligonucleotideprobes for four Lactobacillus species occu-ring in the intestinal tract, Syst. Appl.Microbiol. 15 (1992) 123–128.
[97] Hertel C., Ludwig W., Obst M.,Vogel R.F.,Hammes W.P., Schleifer K.H., 23S rRNA-targeted oligonucleotide probes for therapid identification of meat lactobacilli,Syst. Appl. Microbiol. 14 (1991) 173–177.
[98] Hertel C., Ludwig W., Pot B., Kersters K.,Schleifer K.H., Differentiation of lactoba-cilli occuring in fermented milk products byusing oligonucleotides probes and electro-phoretic protein profiles, Syst. Appl.Microbiol. 16 (1993) 463–467.
[99] Horie M., Kajikawa H.S., Toba T., Identifi-cation of Lactobacillus crispatus by poly-merase chain reaction targeting S-layerprotein gene, Lett. Appl. Microbiol. 35(2002) 57–61.
[100] Hulton C.S., Higgins C.F., Sharp P.M.,ERIC sequences: a novel family of repeti-tive elements in the genomes of Escheri-chia coli, Salmonella typhimurium andother enterobacteria, Mol. Microbiol. 5(1991) 825–834.
[101] Hynes E., Ogier J.C., Delacroix-Buchet A.,Proteolysis during ripening of miniaturewashed-curd cheeses manufactured withdifferent strains of starter bacteria and aLactobacillus plantarum adjunct culture,Int. Dairy J. 11 (2001) 587–597.
[102] Hyytiä-Trees E., Lyhs U., Korkeala H.,Björkroth J., Characterisation of ropyslime-producing Lactobacillus sakei usingrepetitive element sequence-based PCR,Int. J. Food Microbiol. 50 (1999) 215–219.
302 V. Coeuret et al.
[103] IDF, Detection and enumeration of Lacto-bacillus acidophilus, Standard 306, Int.Dairy Fed. Brussels, Belgium, 1995.
[104] IDF, Yoghurt-Enumeration of characteris-tic micro-organisms-colony count techni-que at 37 °C Standard 117 B, Int. Dairy Fed.Brussels, Belgium, 1997.
[105] IDF, Dairy Stater cultures of Lactic AcidBacteria (LAB), Standard 149A, Int. DairyFed. Brussels, Belgium, 1997.
[106] Isolini D., Grand M., Glättli H., Selektiv-medien zum nachweis von obligat undfakultativ heterofermentativen laktoba-zillen, Schweiz Milchw. Forschung. 19(1990) 57–59.
[107] Jacobsen C.N., Rosenfeldt Nielsen V., Hay-ford A.E., Moller P.L., Michaelsen K.F.,Paerregaard A., Sandstrom B., Tvede M.,Jakobsen M., Screening of probiotic activi-ties of forty-seven strains of Lactobacillusspp. by in vitro techniques and evaluation ofthe colonization ability of five selectedstrains in humans, Appl. Environ. Micro-biol. 65 (1999) 4949–4956.
[108] Janssen P., Coopman R., Huys G., SwingsJ., Bleeker M., Vos P., Zabeau M., Kersters K.,Evaluation of the DNA fingerprinting methodAFLP as a new tool in bacterial taxonomy,Microbiology 142 (1996) 1881–1893.
[109] Johansson M.L., Molin G., Pettersson B.,Uhlén M., Ahrné S., Characterization andspecies recognition of Lactobacillus plan-tarum strains by restriction fragment lengthpolymorphism (RFLP) of the 16S rRNAgene, J. Appl. Bacteriol. 79 (1995) 536–541.
[110] Jordan K.N., Cogan T.M., Identificationand growth of non-starter lactic acid bacte-ria in Irish Cheddar cheese, Irish J. Agric.Food Res. 32 (1993) 47–55.
[111] Kandler O., Weiss N., Genus Lactobacillus,in: Sneath P.H.A., Mair N.S., Sharpe M.E.,Holt J.G. (Eds.), Bergey’s Manual of Syste-matics Bacteriology, Williams and Wilkins,Baltimore M.D., USA, 1984, pp. 1209–1234.
[112] Klein G., Hack B., Hanstein S., ZimmermannK., Reuter G., Intra-species characterizationof clinical isolates and biotechnologicallyused strains of Lactobacillus rhamnosus byanalysis of the total soluble cytoplasmaticproteins with silver staining, Int. J. FoodMicrobiol. 25 (1995) 263–275.
[113] Klein G., Pack A., Bonaparte C., Reuter G.,Taxonomy and physiology of probiotic lac-tic acid bacteria, Int. J. Food Microbiol. 41(1998) 103–125.
[114] Kneifel W., Pacher B., An X-Glu basedagar medium for the selective enumerationof Lactobacillus acidophilus in yogurt-rela-
ted milk products, Int. Dairy J. 3 (1993)277–291.
[115] Kneifel W., Toros A., Viernstein H., Diffe-rential enumeration of silage inoculantsbased on utilization of enzymatic activityand antibiotic sensitivity of bacteria, J.Appl. Bacteriol. 77 (1994) 42–48.
[116] Lankaputhra W.E.V., Shah N.P., Survivalof Lactobacillus and Bifidobacterium spp.in the presence of acid and bile salts, Cult.Dairy Prod. J. 30 (1995) 2–7.
[117] Lankaputhra W.E.V., Shah N.P., BritzM.L., Evaluation of media for selectiveenumeration of Lactobacillus acidophilusand Bifidobacterium species, Food Austr.48 (1996) 113–118.
[118] Lawson P.A., Papademas P., Wacher C.,Falsen E., Robinson R., Collins M.D., Lac-tobacillus cypricasei sp. nov., isolated fromHalloumi cheese, Int. J. Syst. Evol. Micro-biol. 51 (2001) 45–49.
[119] Lick S., Brockmann E., Heller K.J., Identi-fication of Lactobacillus delbrueckii andsubspecies by hybridization probes andPCR, Syst. Appl. Microbiol. 23 (2000)251–259.
[120] Lopez S., Mayo B., Identification and cha-racterization of homofermentative meso-philic Lactobacillus strains isolated fromartisan starter-free cheeses, Lett. Appl.Microbiol. 25 (1997) 233–238.
[122] Lortal S., Valence F., Bizet C., Maubois J.L.,Electrophoretic pattern of peptidoglycanhydrolases, a new tool for bacterial speciesidentification: application to 10 Lactobacil-lus species, Res. Microbiol. 148 (1997)461–474.
[123] Lucchini F., Kmet V., Cesena C., Coppi L.,Bottazzi V., Morelli L., Specific detectionof a probiotic Lactobacillus strain in faecalsamples by using multiplex PCR, FEMSMicrobiol. Lett. 158 (1998) 273–278.
[124] Madkor S.A., Tong P.S., El Soda M., Ripe-ning of Cheddar cheese with added attenua-ted adjunct cultures of lactobacilli, J. DairySci. 83 (2000) 1684–1691.
[125] Manchester L.N., Toole A., Goodacre R.,Characterization of Carnobacterium spe-cies by pyrolysis mass spectrometry, J.Appl. Bacteriol. 78 (1995) 88–96.
[126] Mannu L., Comunian R., Scintu M.F.,Mesophilic lactobacilli in Fiore Sardocheese: PCR-identification and evolutionduring cheese ripening, Int. Dairy J. 10(2000) 383–389.
Identification of dairy lactobacilli 303
[127] Mannu L., Riu G., Comunian R., FozziM.C., Scintu M.F., A preliminary study oflactic acid bacteria in whey starter cultureand industrial Pecorino Sardo ewe’s milkcheese: PCR-identification and evolutionduring ripening, Int. Dairy J. 12 (2002) 17–26.
[129] Matte-Tailliez O., Quénée P., Cibik R., VanOpstal J., Dessevre F., Firmesse O., TailliezP., Detection and identification of lacticacid bacteria in milk and industrial starterculture with fluorescently labeled rRNA-targeted peptide nucleic acid probes, Lait81 (2001) 237–248.
[130] Mc Cann T., Egan T., Weber G.H., Assayprocedures for commercial probiotic cultu-res, J. Food Prot. 59 (1995) 41–45.
[131] Mc Cartney A.L., Wenzhi W., TannockG.W., Molecular analysis of the composi-tion of the bifidobacterial and Lactobacillusmicroflora of humans, Appl. Environ.Microbiol. 62 (1996) 4608–4613.
[132] Mc Donald L.C., Mc Feeters R.F., DaeschelM.A., Fleming H.P., A differential mediumfor the enumeration of homofermentativeand heterofermentative lactic acid bacteria,Appl. Environ. Microbiol. 53 (1987)1382–1384.
[133] Medina R., Katz M., Gonzalez S., OliverG., Characterization of the lactic acid bacte-ria in ewe’s milk and cheese from northwestArgentina, J. Food Prot. 64 (2001) 559–563.
[134] Menendez S., Centeno J.A., Godinez R.,Rodriguez-Otero J.L., Effects of Lactoba-cillus strains on the ripening and organolep-tic characteristics of Arzua-Ulloa cheese,Int. J. Food Microbiol. 59 (2000) 37–46.
[135] Miteva V., Boudakov I., Ivanova-Stoyan-cheva G., Marinova B., Mitev V., Mengaud J.,Differentiation of Lactobacillus delbrueckiisubspecies by ribotyping and amplified ribo-somal DNA restriction analysis (ARDRA),J. Appl. Microbiol. 90 (2001) 909–918.
[136] Molin G., Arhné S., Johansson M.L., Com-parative evaluation of plasmid profiling andribotyping in the analysis of Lactobacillusplantarum strain heterogeneity in silage, J.Appl. Bacteriol. 80 (1996) 114–116.
[137] Morea M., Baruzzi F., Cappa F., Coccon-celli P.S., Molecular characterization of theLactobacillus community in traditional pro-cessing of Mozzarella cheese, Int. J. FoodMicrobiol. 43 (1998) 53–60.
[138] Muller M.R., Ehrmann M.A., Vogel R.F.,Multiplex PCR for the detection of Lacto-bacillus pontis and two related species in asourdough fermentation, Appl. Environ.Microbiol. 66 (2000) 2113–2116.
[139] Mullis K.B., The unusual origin of the poly-merase chain reaction, Sci. Am. 262 (1990)56-61, 64–65.
[140] Nagasaki H., Ito K., Matsuzaki S., TanakaS., An improved tetrazolium agar mediumfor testing sugar fermentation in lactoba-cilli, Nippon Saikingaku Zasshi 47 (1992)511–514.
[141] Nigatu A., Evaluation of numerical analy-ses of RAPD and API 50 CH patterns to dif-ferentiate Lactobacillus plantarum, Lact.fermentum, Lact. rhamnosus, Lact. sake,Lact. parabuchneri, Lact. gallinarum, Lact.casei, Weissella minor and related taxa iso-lated from kocho and tef, J. Appl. Micro-biol. 89 (2000) 969–978.
[142] Nighswonger B.D., Brashears M.M., Gilli-land S.E., Viability of Lactobacillus acido-philus and Lactobacillus casei in fermentedmilk products during refrigerated storage, J.Dairy Sci. 79 (1996) 212–219.
[143] Nissen H., Dainty R., Comparison of theuse of rRNA probes and conventionalmethods in identifying strains of Lactoba-cillus sake and L. curvatus isolated frommeat, Int. J. Food Microbiol. 25 (1995)311–315.
[144] Ogier J.C., Son O., Gruss A., Tailliez P.,Delacroix-Buchet A., Identification of bac-terial microflora in dairy products by tem-poral temperature gradient gel electropho-resis, Appl. Environ. Microbiol. 68 (2002)3691–3701.
[145] Peterson S.D., Marshall R.T., Nonstarterlactobacilli in Cheddar cheese: a review, J.Dairy Sci. 73 (1990) 1395–1410.
[146] Pot B., Hertel C., Ludwig W., DescheemaekerP., Kersters K., Schleifer K.H., Identifica-tion and classification of Lactobacillus aci-dophilus, L. gasseri and L. johnsonii strainsby SDS-PAGE and rRNA-targeted oligo-nucleotide probe hybridization, J. Gen.Microbiol. 139 (1993) 513–517.
[147] Pot B., Vandamme P., Kersters K., Analy-sis of electrophoretic whole-organisms pro-tein fingerprints, in: Goodfellow M.,O’Donnell A.G. (Eds.), Modern microbialmethods. Chemical methods in prokaryoticsystematics, John Wiley & Sons Ltd, Chi-chester, UK, 1994, pp. 493–521.
[148] Quere F., Deschamps A., Urdaci M.C.,DNA probe and PCR-specific reaction forLactobacillus plantarum, J. Appl. Micro-biol. 82 (1997) 783–790.
304 V. Coeuret et al.
[149] Quiberoni A., Tailliez P., Quénée P., Sua-rez V., Reinheimer J., Genetic (RAPD-PCR) and technological diversities amongwild Lactobacillus helveticus strains, J.Appl. Microbiol. 85 (1998) 591–596.
[150] Randazzo C.L., Torriani S., AkkermansA.D.L., de Vos W.M., Vaughan E.E.,Diversity, dynamics, and activity of bacte-rial communities during production of anartisanal Sicilian cheese as evaluated by16S rRNA analysis, Appl. Environ. Micro-biol. 68 (2002) 1882–1892.
[151] Ravula R.R., Shah N.P., Selective enumera-tion of Lactobacillus casei from yogurt andfermented milk drinks, Biotechnol. Tech.12 (1998) 819–822.
[152] Rebecchi A., Crivori S., Sarra P.G., Coc-concelli P.S., Physiological and moleculartechniques for the study of bacterial com-munity development in sausage fermentation,J. Appl. Microbiol. 84 (1998) 1043–1049.
[153] Reinkemeier M., Rocken W., LeitzmannC., A rapid mechanical lysing procedure forroutine analysis of plasmids from lactoba-cilli, isolated from sourdoughs, Int. J. FoodMicrobiol. 29 (1996) 93–104.
[154] Rizzo F.A., Korkeala H., Mononen I., Gaschromatography analysis of cellular fattyacids and neutral monosaccharides in theidentification of lactobacilli. Appl. Environ.Microbiol. 53 (1987) 2883–2888.
[162] Shah N.P., Probiotic bacteria: selective enu-meration and survival in dairy foods, J. DairySci. 83 (2000) 894–907.
[163] Shah N.P., Lankaputhra W.E.V., BritzM.L., Kyle W.S.A., Survival of Lactobacil-lus acidophilus and Bifidobacterium bifi-dum in commercial yoghurt during refrigera-ted storage, Int. Dairy J. 5 (1995) 515–521.
[164] Shakeel-Ur-Rehman S., Mc Sweeney P.L.H.,Fox P.F., A study of the role of the indige-nous microflora of raw milk on the ripeningof Cheddar cheese, Milchwissenschaft 57(1999) 388–392.
[165] Sharpe A.N., Jackson A.K., Stomaching: anew concept in bacteriologica sample pre-paration, Appl. Microbiol. 24 (1972) 175–178.
[166] Sharples G.J., Lloyd R.G., A novel repeatedDNA sequence located in the intergenicregions of bacterial chromosomes, Nucl.Acid. Res. 18 (1990) 6503–6508.
[167] Shin H.S., Lee J.H., Pestka J.J., Ustunol Z.,Viability of bifidobacteria in commercialdairy products during refrigerated storage,J. Food Prot. 63 (2000) 327–331.
[168] Simpson J.M., McCracken V.J., GaskinsH.R., Mackie R.I., Denaturing gradient gelelectrophoresis analysis of 16S ribosomalDNA amplicons to monitor changes in fecalbacterial populations of weaning pigs afterintroduction of Lactobacillus reuteri strainMM53, Appl. Environ. Microbiol. 66(2000) 4705–4714.
[169] Sohier D., Coulon J., Lonvaud-Funel A.,Molecular identification of Lactobacillushilgardii and genetic relatedness with Lac-tobacillus brevis, Int. J. Syst. Bacteriol. 49(1999) 1075–1081.
[170] Song Y., Kato N., Liu C., Matsumiya Y.,Kato H., Watanabe K., Rapid identificationof 11 human intestinal Lactobacillus spe-cies by multiplex PCR assays using group-and species-specific primers derived fromthe 16S-23S rRNA intergenic spacer regionand its flanking 23S rRNA, FEMS Micro-biol. Lett. 187 (2000) 167–173.
[171] Stackebrandt E., Goebel B.M., Taxonomicnote: a place for DNA-DNA reassociationand 16S rRNA sequence analysis in the pre-sent species definition in bacteriology, Int.J. Syst. Bacteriol. 44 (1994) 846–849.
[172] Stanton C., Lynch P.B., Collins J.K.,Fitzgerald G., Ross R.P., Probiotic cheese,Int. Dairy J. 8 (1998) 491–496.
[173] Steele J.L., Johnson M.E., Broadbent J.R.,Weimer B.C., Starter culture attributeswhich affect cheese flavor development, in:Adria-Normandie, Les bactéries lactiques,
Identification of dairy lactobacilli 305
Lactic 97, Caen, France, 10–12 Septembre,1997, pp. 157–170.
[174] Stiles M.E., Holzapfel W.H., Lactic acidbacteria of foods and their current taxo-nomy, Int. J. Food Microbiol. 36 (1997) 1–29.
[176] Tailliez P., Quénée P., Chopin A., Estima-tion de la diversité parmi les souches de lacollection CNRZ : application de la RAPDà un groupe de lactobacilles, Lait 76 (1996)147–158.
[177] Tannock G.W., Tilsala-Timisjarvi A., RodtongS., Ng J., Munro K., Alatossava T., Identifi-cation of Lactobacillus isolates from thegastrointestinal tract, silage, and yoghurt by16S-23S rRNA gene intergenic spacerregion sequence comparisons, Appl. Envi-ron. Microbiol. 65 (1999) 4264–4267.
[178] Tannock G.W., Munro K., Harmsen H.J.,Welling G.W., Smart J., Gopal P.K., Analy-sis of the fecal microflora of human sub-jects consuming a probiotic product contai-ning Lactobacillus rhamnosus DR20, Appl.Environ. Microbiol. 66 (2000) 2578–2588.
[179] Tilsala-Timisjarvi A., Alatossava T., Deve-lopment of oligonucleotide primers fromthe 16S-23S rRNA intergenic sequences foridentifying different dairy and probioticlactic acid bacteria by PCR, Int. J. FoodMicrobiol. 35 (1997) 49–56.
[180] Tilsala-Timisjarvi A., Alatossava T., Strain-specific identification of probiotic Lactobacil-lus rhamnosus with randomly amplified poly-morphic DNA-derived PCR primers, Appl.Environ. Microbiol. 64 (1998) 4816–4819.
[181] Torriani S., Zapparoli G., Dellaglio F., Useof PCR-based methods for rapid differentia-tion of Lactobacillus delbrueckii subsp. bul-garicus and L. delbrueckii subsp. lactis, Appl.Environ. Microbiol. 65 (1999) 4351–4356.
[182] Tsakalidou E., Manolopoulou E., KabarakiE., Zoidou E., Pot B., Kersters K.,Kalantzopoulos G., The combined use ofwhole-cell protein extracts for the identifi-cation (SDS-PAGE) and enzyme activityscreening of lactic acid bacteria isolatedfrom traditional Greek dairy products, Syst.Appl. Microbiol. 17 (1994) 444–458.
[183] Tsuchiya Y., Kano Y., Koshino S., Identifi-cation of lactic acid bacteria using tempera-ture gradient gel electrophoresis for DNAfragments amplified by polymerase chainreaction, J. Amer. Soc. Brew Chem. 52(1994) 95–99.
[184] Tynkkynen S., Satokari R., Saarela M.,Mattila-Sandholm T., Saxelin M., Compari-son of ribotyping, randomly amplified poly-morphic DNA analysis, and pulsed-fieldgel electrophoresis in typing of Lactobacil-lus rhamnosus and L. casei strains, Appl.Environ. Microbiol. 65 (1999) 3908–3914.
[185] Valence F., Richoux R., Thierry A., PalvaA., Lortal S., Autolysis of Lactobacillushelveticus and Propionibacterium freuden-reichii in Swiss cheeses: first evidence byusing species-specific lysis markers, J.Dairy Res. 65 (1998) 609–620.
[186] Van Beek S., Priest F.G., Evolution of thelactic acid bacterial community during maltwhisky fermentation: a polyphasic study,Appl. Environ. Microbiol. 68 (2002)297–305.
[187] Van Reenen C.A., Dicks L.M., Evaluationof numerical analysis of random amplifiedpolymorphic DNA (RAPD)-PCR as amethod to differentiate Lactobacillus plan-tarum and Lactobacillus pentosus, Curr.Microbiol. 32 (1996) 183–187.
[188] Vandamme P., Pot B., Gillis M., de Vos P.,Kersters K., Swings J., Polyphasic taxo-nomy, a consensus approach to bacterialsystematics, Microbiol. Rev. 60 (1996)407–438.
[189] Vasquez A., Ahrne S., Pettersson B., MolinG., Temporal temperature gradient gel elec-trophoresis (TTGE) as a tool for identifica-tion of Lactobacillus casei, Lactobacillusparacasei, Lactobacillus zeae and Lactoba-cillus rhamnosus, Lett. Appl. Microbiol. 32(2001) 215–219.
[190] Vinderola C.G., Reinheimer J.A., Culturemedia for the enumeration of Bifidobacte-rium bifidum and Lactobacillus acidophilusin the presence of yoghurt bacteria, Int.Dairy J. 9 (1999) 497–505.
[191] Vinderola C.G., Prosello W., GhibertoT.D., Reinheimer J.A., Viability of probio-tic (Bifidobacterium, Lactobacillus acido-philus and Lactobacillus casei) and nonpro-biotic microflora in Argentinian Frescocheese, J. Dairy Sci. 83 (2000) 1905–1911.
[192] Vogel R.F., Lohmann M., Nguyen M., Wel-ler A.N., Hammes W.P., Molecular charac-terization of Lactobacillus curvatus andLact. sake isolated from sauerkraut andtheir application in sausage fermentations,J. Appl. Bacteriol. 74 (1993) 295–300.
[193] Vogel R., Böcker G., Stolz P., Ehrmann M.,Fanta D., Ludwig W., Pot B., Kersters K.,Schleifer K.H., Hammes W., Identificationof lactobacilli from sourdough and descrip-tion of Lactobacillus pontis sp. nov., Int. J.Syst. Bacteriol. 44 (1994) 223–239.
306 V. Coeuret et al.
[194] Walter J., Tannock G.W., Tilsala-TimisjarviA., Rodtong S., Loach D.M., Munro K.,Alatossava T., Detection and identificationof gastrointestinal Lactobacillus species byusing denaturing gradient gel electrophore-sis and species-specific PCR primers, Appl.Environ. Microbiol. 66 (2000) 297–303.
[195] Walter J., Hertel C., Tannock G.W., LisC.M., Munro K., Hammes W.P., Detectionof Lactobacillus, Pediococcus, Leuconos-toc, and Weissella species in human fecesby using group-specific PCR primers anddenaturing gradient gel electrophoresis,Appl. Environ. Microbiol. 67 (2001)2578–2585.
[200] Yost C.K., Nattress F.M., The use of multi-plex PCR reactions to characterize popula-tions of lactic acid bacteria associated withmeat spoilage, Lett. Appl. Microbiol. 31(2000) 129–133.
[201] Zapparoli G., Torriani S., Dellaglio F., Dif-ferentiation of Lactobacillus sanfranciscen-sis strains by randomly amplified polymor-phic DNA and pulsed-field gel electrophoresis,FEMS Microbiol. Lett. 166 (1998) 325–332.
[202] Zhong W., Millsap K., Bialkowska-Hobrzanska H., Reid G., Differentiation ofLactobacillus species by molecular typing,Appl. Environ. Microbiol. 64 (1998)2418–2423.
