Isolation, Characterisation and Identification of Lactobacilli Focusing

Home , Lactobacillus brevis, Lactobacillus buchneri, Lactobacillus gallinarum, Lactobacillus jensenii, Lactobacillus parakefiri, Lactobacillus pontis

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

HAL Id: hal-00895504 https://hal.archives-ouvertes.fr/hal-00895504 Submitted on 1 Jan 2003

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Lait 83 (2003) 269–306 © INRA, EDP Sciences, 2003 269 DOI: 10.1051/lait:2003019 Review article

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 date more than 80 species. They are present in raw milk and dairy products such as cheeses, yoghurts and fermented milks. Quality assurance programmes associated with research, development, production and validation of the health or technological benefits of these bacteria require their relevant isolation, counting and identification. This review presents the different selective media to isolate lactobacilli, and the numerous different available tools to characterise lactobacilli at genus, species or strain level using either culture-dependent methods: phenotypical, molecular or global methods, 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, the main 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 et compte actuellement plus de 80 espèces. Les lactobacilles sont présents dans le lait cru, les produits laitiers tels que les fromages, les yaourts et les laits fermentés. Pour s’inscrire dans une démarche qualité 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. Cette analyse bibliographique présente les différents milieux sélectifs pour isoler les lactobacilles, ainsi que les outils disponibles à ce jour pour caractériser les lactobacilles au niveau du genre, de l’espèce ou 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 enzymes utilisées en PFGE, les sondes d’hybridation et les oligonucléotides utilisés pour les différentes techniques de PCR sont répertoriés en sept tableaux. En conclusion, les avantages et inconvénients de ces techniques sont présentés.

Lactobacillus / milieu sélectif / enzyme PFGE / sonde moléculaire / amorce PCR / fromage / produit laitier

* Correspondence and reprints E-mail: [email protected] 270 V. Coeuret et al.

1. INTRODUCTION: microflora of the dairy products (non- BIODIVERSITY OF LACTOBA- starter lactic acid bacteria: NSLAB) and CILLI IN DAIRY PRODUCTS originate from animals, farms and cheese dairies: Lb. casei ssp. casei/Lb. paracasei ssp. paracasei, Lb.rhamnosus, Lb. plantarum, Lactic acid bacteria (LAB) comprise a Lb. fermentum, Lb. brevis, Lb. buchneri, wide range of genera and include a consid- Lb. curvatus, Lb. acidophilus and Lb. pen- erable number of species. Their common tosus [43, 58, 83, 110, 120, 133]. NSLAB, traits are: Gram-positive, usually catalase- which are initially present in small num- negative, growth under microaerophilic to bers (102 to 103 NSLAB/g after pressing in strictly anaerobic conditions and lactic Cheddar cheese), increase to high numbers acid production. These bacteria are the in cheese varieties that require long ripen- major component of the starters used in ing times (107 to 108 bacteria/g within fermentation, especially for dairy prod- about 3 months in Cheddar cheese) [10, 21, ucts, and some of them are also natural 34, 58, 69, 73, 86, 110, 145]. components of the gastrointestinal microflora. Lactobacillus is one of the most Coppola et al. [44] studied the micro- important genera of LAB. In raw milk and biological characteristics of raw milk, nat- dairy products such as cheeses, yoghurts ural whey starter and cheese during the and fermented milks, lactobacilli are natu- first months of ripening of Parmigiano rally present or added intentionally, for Reggiano: thermophilic lactobacilli – Lb. technological reasons or to generate a helveticus, Lb. delbrueckii ssp. lactis. Lb. health benefit for the consumer. delbrueckii ssp. bulgaricus – disappeared within 30 d. Rod-shaped mesophilic facultatively heterofermentative lactobacilli – 1.1. Cheeses and milks Lb. casei, Lb. paracasei ssp. paracasei, Lb. paracasei ssp. tolerans and Lb. rhamnosus – In cheese, lactobacilli (Lb. helveticus and progressively increased in number until the Lb. delbrueckii ssp. and lactis) are present in fifth month of ripening. The results industrial starter cultures for processing showed that the thermophilic lactobacilli hard cheese (e.g. Emmental, Comté, Italian and Lb. rhamnosus were derived from nat- Grana and Argentinean hard cheeses) [21, ural whey starter, whereas the other com- 43, 92]. Pasta filata cheeses (Mozarella) ponents of non-starter lactobacilli were and Italian hard cheeses (Canestrato derived from raw milk. Similarly, Comté Pugliese and Parmigiano Reggiano) are cheese contains mesophilic lactobacilli also processed with Lb. delbrueckii ssp. strains. They originate from the raw milk, bulgaricus [42, 45, 83]. Lactobacilli start- and this source was probably more impor- ers are normally present at levels of tant than the factory environment [17, 58]. 109 bacteria/g, contribute to the lactic fer- Differences have been observed between raw mentation, and are involved at the beginning milk cheeses and pasteurised or microfiltered of ripening [70]. For example, in Emmen- milk cheeses [58, 70, 164]. Eliskases- tal cheeses, lactobacilli ferment galactose Lechner et al. [70] found that Bergkäse excreted by Streptococcus thermophilus, (Austrian regional cheese) cheeses made achieve acidification processes, and con- from pasteurised milk contained less than tribute to primary proteolysis [31, 35, 59, one-thousandth the number of facultatively 74]. Their numbers decrease rapidly during heterofermentative lactobacilli (FHL) present ripening, at a rate depending to some in raw milk cheeses. Differences in citrate degree on the sensitivity of the starters to metabolism, which occur in raw milk salt [110], on the water activity, and on the cheeses, can be attributed to the presence autolysis power of the strains [185]. Some of FHL in these cheeses. Very little effort is lactobacilli are also present in the natural currently devoted to controlling the numbers Identification of dairy lactobacilli 271 and types of NSLAB present, although Lb. paracasei and Lb. gasseri [88, 94, 113, these organisms are thought to have a sig- 142, 159]. For these products, careful nificant influence on cheese flavour devel- strain selection is necessary. Klein et al. opment and to participate directly in the [113] recently observed that the identity of production of some major aroma com- most of the lactobacilli used as probiotics pounds such as acetic acid, formic acid and differs from that marked on the packaging. gas [31, 56, 59, 173]. However, NSLAB Shah [162] reported that it is important to may also cause defects. For example, in monitor the survival of probiotic lactoba- Emmental cheeses, Lb. plantarum may cilli because a number of products have disturb the metabolism of propionic acid been found to contain only a few viable bacteria, resulting in lower quality cheeses bacteria by the time they reach the market due to opening and changes in flavour [87, 163, 167]. development [34]. Given the great potential economic It has repeatedly been claimed that the value of lactobacilli, one of the main objec- use of selected strains as adjunct cultures tives of microbiologists is to develop a improves and accelerates flavour develop- clear picture of the microflora present in Lactobacillus ment. The use of adjunct the various dairy products, and the way in cultures to Cheddar cheese or to cow’s which it changes during processing. For milk cheeses ripened for short periods of example, during cheese manufacture and rip- time (e.g. Azua-Ulloa cheese) has been reported to result in higher levels of prote- ening, complex interactions occur between olytic products and higher sensory quality individual components of the cheese micro- scores in many studies [120, 124, 134, flora, and identification of these bacteria is 175]. However, some adjunct cultures may essential for understanding their individual cause high levels of acidity, bitterness, off contribution to cheese manufacture. This flavours and open and crumbly textures, allows the development of a more targeted clearly demonstrating the importance of approach to starter/adjunct selection for the culture selection [46, 69, 101]. A number improvement of cheese quality [14]. Qual- of studies have recently evaluated the suit- ity assurance programmes associated with ability of probiotic cultures as adjunct cul- research, development, production and val- tures in various cheeses: Lb. acidophilus idation of the health or technological bene- and Lb. casei in Argentinian Fresco cheese fits of these bacteria require the relevant [191], Lb. paracasei in Cheddar cheeses isolation, counting and identification of [76, 172], and Lb. acidophilus in goat’s bacteria. milk cheeses [85]. Depending on the taxonomic level desired, several phenotypical or molecular methodologies (polyphasic analysis) can 1.2. Yoghurts and fermented milks be used for isolation, characterisation and identification of lactobacilli. Recent meth- Lactobacillus delbrueckii ssp. bulgari- odologies, which are culture-independent cus is one of the two bacteria necessary for such as single strand conformation poly- the production of yoghurts, and Lb. kefir is morphism (SSCP), temperature gradient essential for the production of Caucasian gel electrophoresis (TGGE) and denatur- sour milk kefir [111]. A recent trend is to ing gradient gel electrophoresis (DGGE) add probiotic lactobacilli to fermented have also been proposed for characterisa- milks to generate health benefits. The bac- tion of microbial diversity. This paper teria generally added are Lb. acidophilus, reviews how the literature proposes char- Lb. rhamnosus, Lb. reuteri, Lb. casei, Lb. acterisation of lactobacilli from dairy prod- plantarum, Lb. johnsonii, Lb. crispatus, ucts at genus, species or strain level. 272 V. Coeuret et al.

2. STUDY OF LACTOBACILLI 2.1.2. Selective media BY CULTURE-DEPENDENT METHODS Several elective and selective media have been developed for the isolation and 2.1. Isolation of lactobacilli: counting of Lactobacillus species and for from crude sample to pure culture the differential counting of mixed populations of lactic acid bacteria (Tab. I). Oxy- Guidelines have been published for the gen tolerance, nutritional requirements, extraction of lactobacilli from milk or antibiotic susceptibility, colony morphol- dairy products [104]. It is often difficult to ogy and colour are used to differentiate count microorganisms from particulate or strains in these methods. The differential solid samples. In such cases, the cells must detection and counting of lactobacilli can be separated from the particles and the currently be achieved in a number of ways. extraction efficiency must be assessed. The However, only a few media are useful for the separation methods currently used to differential counting of lactobacilli because extract microorganisms from solid food numerous other microorganisms including matrices such as hard cheese and from dry Lactococcus, Enterococcus, Leuconostoc, starters and other types of cheese are based Weissella, Bifidobacterium and Pedio- on an initial crude homogenisation in a cocci grow on media similar to those used blender and/or a stomacher [165]. For hard for lactobacilli. No medium has yet been cheeses, an ultra-turrax or a stomacher is described on which only lactobacilli are recommended. able to grow. 2.1.1. Resuspension medium Lactobacilli are generally isolated on rich media such as MRS [54], which is routinely Callichia et al. [28] increased the rate of used for the isolation and counting of lacto- recovery of dried microorganisms to a bacilli from most (fermented) food products. level similar to that of wet cultures that had The addition of a reducing agent such as never been dried by resuspending the dried cysteine 0.05% to MRS improves the specif- microorganisms in a medium containing icity of this medium for Lactobacillus isola- phosphate, cysteine, antifoaming agent tion [90, 116, 162]. MRS and M17 are, and agar before plating. McCann et al. respectively, the medium of choice for the [130] analysed a number of commercial differential counting of Lactobacillus del- dairy cultures, comparing methods of sam- brueckii ssp. bulgaricus and Streptococcus ple preparation: duplicate samples of pro- thermophilus in yoghurt [105]. Birollo et al. biotic products were resuspended and [18] proposed the use of a cheaper medium – diluted in both peptone and CRM (Calic- skim milk agar – for counting and differenti- chia resuspension medium). Following ating colonies of these two bacteria. When resuspension in peptone and incubation for yogurt microflora was supplemented by 30 min at 23 °C before plating, the rate of other lactobacilli, specific methods were CFU recovery for each product was needed. Enumeration of Lb. delbrueckii ssp. approximately 50% of that obtained if the bulgaricus in the presence of Lb. acidophilus microorganisms were resuspended in was possible, on acidified MRS at pH 5.2 or CRM and incubated for 30 min at 37 °C on reinforced clostridial agar (RCA) at pH before plating. For cheeses, dilution in 5.3, at 45 °C for 72 h [162]. Selective count- Trisodium citrate (2% w/v) is generally ing of Lb. acidophilus has been developed recommended, and peptone salt or phos- using several MRS media in which the dex- phate buffer salt is generally used for dairy trose is replaced by maltose, raffinose, melibi- products such as yoghurts and fermented ose, trehalose, arabinose, galactose, sorbitol, milks. ribose, gluconate or salicin or cellobiose- Identification of dairy lactobacilli 273

Table I. Differential plating media for detection and counting of Lactobacillus species.

Incubation Product Population Media conditions Isolation Notice Ref. Yogurts S. thermophilus = circular opalescent white colonies S.thermophilus with well defined borders 37 °C, 3 d, Lb. delb. ssp. Lb. delb. ssp. bulgaricus = [104] Acidified MRS Ana bulgaricus rounded colonies, duller, flat with non defined borders

S.thermophilus Impossibility to differentiate S.thermophilus Skim milk agar 37 °C, 3 d, Lb. delb. ssp. between both types [18, 190] Lb. delb. ssp. Ana bulgaricus of colonies bulgaricus Acidified MRS S.thermophilus = Skim milk agar circular opalescent white S.thermophilus colonies with well defined 37 °C, 3 d, Lb. delb. ssp. borders Lb. delb. ssp. bulga- [104] Aer bulgaricus ricus = bigger irregular [18] translucent colonies with [190] non defined borders Acidified skim millk 47 °C, 2 d, Lb. delb. ssp. agar Ana bulgaricus [32] Fermented S.thermophilus CLBS milks and Lb. delb. ssp. (Cellobiose probiotic bulgaricus Lactobacillus Lb. acidophilus [142] products Lb. acidophilus selection agar) Lb. casei Lb. casei MRS Salicin37 °C, 2 d, [162] Ana MRS Sorbitol Lb. casei [151] LC (Lb. casei agar) S. thermophilus = blue colonies S.thermophilus HHD Lb. delb. ssp. bulgaricus = Lb. delb. ssp. (Homofermentative- blue colonies bulgaricus heterofermentative Lb. acidophilus = blue [29] Lb. acidophilus Differential medium) colonies with white sur- Bifidobacterium rounding spp. S.thermophilus Bifidobacterium spp. = 37 °C, 2 d, Lb. delb. ssp. white colonies LA agar Ana bulgaricus S. thermophilus = pin point Lb. acidophilus colonies Bifidobacterium Lb. delb. ssp. bulgaricus = sp. elevated and white colonies [32] Bifidus blood agar Lb. acidophilus = flat and grey colonies Bifidobacterium sp. = elevated and chocolate brown colonies S. thermophilus = circular or semi circular colonies, convex, opaque, white-violet, often with a dark centre TPPY (Tryptose Pro- S.thermophilus Lb. delb. ssp. bulgaricus = teose Peptone yeast 37 °C, 2 d, Lb. delb. ssp. flat, transparent, diffuse extract- Aer bulgaricus colonies, of undefined [23] eriochrome T agar) Lb. acidophilus shape, with an irregular edge Lb. acidophilus = small colonies, white to violet, slightly elevated and somewhat fuzzy 274 V. Coeuret et al.

Table I. Differential plating media for detection and counting of Lactobacillus species.

S. thermophilus = circular or semi circular colonies, convex, opaque, white-vio- S.thermophilus let, often with a dark centre TPPYPB (TPPY 37 °C, 2 d, Lb. delb. ssp. Lb. delb. ssp. bulgaricus = with Prussian blue) Aer bulgaricus small shiny white colonies [78] Lb. acidophilus surrounded by a wide royal blue zone Lb. acidophilus = large pale colonies surrounded by a wide royal blue zone Lb. acidophilus MRS maltose Bifidobacterium spp. Growth of Bifidobacterium Lb. acidophilus bifidum, B. infantis and B. MRS arabinose Bifidobacterium breve is inhibited, but not spp. for B. longum and B. pseudolongum 37 °C, 3 d, Ana All bifidobacterium species [117] Lb. acidophilus do not grow on this media Bile-MRS Bifidobacterium and some Lb. delb.bulg can spp. grow. Lb. delb.bulgaricus Only some strains of Lb. RCA pH 5.5 Bifdobacterium spp. acidophilus can grow on Lb. acidophilus this media T-MRS (Trehalose MRS) 37 °C, 3 d, Aer Lb. acidophilus Lb. acidophilus = round [103] Bile-MRS creamly colonies OG-MRS (Oxgall, [190] Gentamycin MRS) Lb. delb. ssp. bulgaricus = irregular white colonies G-MRS (Galactose 37 °C, 3 d, Lb. delb. ssp. MRS) Aer bulgaricus Lb. acidophilus Lb. delb. ssp. bulgaricus = Lb. delb. ssp. white colonies X-Glu 37 °C, 3 d, bulgaricus Lb. acidophilus = blue [114] Ana Lb. acidophilus colonies S.thermophilus Lb. acidophilus LBS (Lactobacillus 37 °C, 2 d, Bifidobacterium selection agar) Ana Lb. acidophilus [159] spp. 37 °C, 2 d, E. faecium E. faecium (24H) Brigg’s agar Aer Lb. acidophilus Lb. acidophilus (48H) E.faecium Modified Brigg’s [28] Lb. acidophilus agar (Brigg’s agar, 37 °C, 2 d, Lb. acidophilus All the strains are not Bifidobacterium plus Streptomycin Aer Streptomycin resistant spp. sulfate) 37 °C, 2 d, E. faecium E. faecium (24H) [130] 3 BLA Aer Lb. acidophilus Lb. acidophilus (48H) Lactobacilli, and Cheeses MRS 37 °C, 3 d, other lactic acid [105] Ana bacteria 42 °C, 3 d, Thermophilic [21] MRS Ana Lactobacilli [58] LBS medium plus 30 °C, 3 d, Cycloheximide Ana Lactobacilli [120] Identification of dairy lactobacilli 275

Table I. Differential plating media for detection and counting of Lactobacillus species.

FH agar (Facultativ Heterofermentativen [106] Laktobazillen agar) Facultative [58] Facultativ Heterofer- 38 °C, 3 d, heterofermentative mentativen Laktoba- Ana Lactobacilli zillen agar plus [21] nalidixic acid 30 °C, 2 d, Mesophilic Ana Lactobacilli MRS plus [45] Cycloheximide 44 °C, 2 d, Thermophilic Ana Lactobacilli LBS (Lactobacillus 30 °C, 3 d, selection agar) Ana Lactobacilli Facultativ Facultative Heterofermentativen 37 °C, 3 d, heterofermentative Laktobazillen agar Ana Lactobacilli [70] plus Vancomycin OH medium (Obli- Obligate gate Heterofermenta- 37 °C, 3 d, heterofermentative tive Lactobacilli Aer Lactobacilli medium) HHD (homofermentative Homofermentative LAB = and heterofermenta- 30 °C, 3 d, Lactic acid blue to green colonies [132] tive differential Aer bacteria Heterofermentative LAB = medium) white colonies 20 °C, 5 d, MRS agar pH 6.5 Ana Mesophilic FH agar 37 °C, 3 d, Lactobacilli [17] Ana Probiotic 37 °C, 3 d, Bifidobacterium cheeses with Bifidobacte- LP-MRS agar Aer spp., Lb. casei [190] rium spp., Lb. 37 °C, 3 d, acidophilus, Lb. Bile-MRS agar Lb. acidophilus casei Ana LBS (Lactobacillus 30 °C, 5 d, With Lb. selection agar) Ana Lactobacilli [76] paracasei Miscella- 37 °C, 3 d, nous LAMVAB Ana Lactobacilli [90] Aer: Aerobic conditions; Ana: Anaerobic conditions; d: day. esculine [52, 117, 162, 190]. An agar lactobacilli can be counted on modified medium based on X-Glu [114] has also been Brigg’s agar supplemented with streptomy- used. MRS with trehalose (T-MRS) is rec- cin sulphate [28], or on 3BLA medium [130]. ommended by the International Dairy Feder- However, the use of such specific media ation [104] for the counting of Lb. acido- is limited by the target species. For exam- philus if this organism occurs in mixed ple, MRS-salicin or MRS-sorbitol agar can populations with yoghurt bacteria and bifido- be used for the selective counting of Lb. bacteria. Dave and Shah [52] reported that acidophilus provided that Lb. casei is not Rogosa acetate or media containing bile present in the product (in which case a total salts, oxgall or NaCl strongly inhibited the count for both species is obtained). If Lb. growth of Lb. acidophilus. In probiotic prod- casei is present, a second medium, such as ucts containing Bifidobacterium bifidum, Lactobacillus casei agar (LC agar) [151], Enterococcus faecium and Lb. acidophilus, must be used. The selective counting of 276 V. Coeuret et al. other lactobacilli used in probiotic and specific medium to date described for dairy products, such as Lb. plantarum, Lb. lactobacilli. Unfortunately, some Lactoba- reuteri and Lb. rhamnosus has not been cillus species, such as Lactobacillus del- studied extensively. Moreover, all the brueckii spp. bulgaricus, and some strains members of the so-called Lb. acidophilus of Lb. acidophilus are vancomycin-sensi- group: Lb. crispatus, Lb. gasseri, Lb. john- tive [32, 90]. Hartemink et al. [90] pro- sonii, Lb. gallinarum and Lb. amylovorus, posed the use of this medium for the isolation are hardly discriminated by plating. of lactobacilli from probiotics containing For the lactobacilli found in cheeses, mixed populations of lactobacilli, bifido- HHD medium (homofermentative and het- bacteria and enterococci. Bifidobacteria erofermentative differential medium) can are susceptible to vancomycin and only a be used for the differential counting of few vancomycin-resistant enterococci have homofermentative and heterofermentative been isolated. lactobacilli [72]. Mesophilic and ther- The use of coloured indicators facili- mophilic lactobacilli are separated by the tates differentiation between microorgan- use of different incubation temperatures: isms. For example, in LAMVAB and HHD 30 °C or 42 °C to 45 °C. FH medium (fac- [132], bromocresol green (pH indicator) is ultative heterofermentative Lactobacillus used as an indicator of acid production. agar) contains vancomycin (50 mg·L–1), Ghoddusi and Robinson [78] added Prus- and has been used for the isolation of sian blue to TPPY agar, Tryptose Proteose NSLAB [13, 22, 127, 150]. LBS (Lactoba- Peptone Yeast extract agar (giving TPPYB cillus selection agar) medium, also known medium), to distinguish Lb. delbrueckii as Rogosa medium [156], incubated at ssp. bulgaricus and Lb. acidophilus from 30 °C has been used to isolate lactobacilli Bifidobacterium and Streptococcus ther- from hand-made cheeses such as Bergkäse mophilus. Kneifel and Pacher [114] devel- [70, 83, 120, 150] but MRS remains the oped an agar medium, X-Glu agar, for the most commonly used medium for the iso- selective counting of Lb. acidophilus in lation of lactobacilli. yoghurt-related milk products containing a Hartemink et al. [90] developed a new mixed flora of lactobacilli, streptococci selective medium, LAMVAB (Lactobacil- and bifidobacteria. In this medium, Lb. acidophilus is identified by testing for its lus anaerobic MRS with vancomycin and bromocresol green), for the isolation of -D-glucosidase activity by means of a Lactobacillus species. Firstly, they used it chromogenic reaction involving X-Glu, which is incorporated into Rogosa agar to isolate lactobacilli from faeces (in which they are present in small numbers), and medium. Based on a similar principle for - then successfully for various species of D-galactopyranosidase activity, Kneifel et al. [115] used X-Gal medium to differentiate lactobacilli from dairy products. The medium blue colonies of lactobacilli from white is highly selective, due to its low pH and colonies of pediococci and enterococci in the presence of vancomycin (20 mg·L–1). silage inoculants. The use of TTC (triphe- Unlike other Gram-positive bacteria, most nyl tetrazolium chloride) may also be help- lactobacilli are resistant to vancomycin ful for strain differentiation [140, 160]. and Gram-negative bacteria are generally sensitive. Vancomycin cannot be used to 2.2. Characterisation and identifica- select lactobacilli in products also contain- tion of lactobacilli from genus ing leuconostocs and pediococci because level to strain level these bacteria are also vancomycin-resistant, and careful morphological examination is For decades, differentiation between required in such cases for differentiation. genera has been based on phenotypic char- However, this medium remains the most acters. Under a light microscope, lactobacilli Identification of dairy lactobacilli 277 are generally regularly shaped, non-motile, tive) and III (obligate heterofermentative) – non-spore-forming, Gram-positive rods. which is still used today for phenotypical However, cell morphology varies widely, analysis. from long, straight or slighty crescent- The most recent version of Bergey’s shaped rods to coryneform coccobacilli. Manual of Systematics [111] included Numerous genera display such morpho- about 50 species of Lactobacillus. This logical features. However, we can separate manual reported taxonomic changes at by simple tests such as tests for the oxygen species level (e.g. Lb. bavaricus became tolerance, presence of catalase and growth Lb. sake) and at subspecies level (e.g. Lb. on acidified MRS Carnobacterium, Lacto- casei ssp. rhamnosus became Lb. rhamno- bacillus and Weissella (non-obligate aer- sus, Lb. bulgaricus became Lb. delbrueckii obe, catalase (-), growth on acidified MRS) ssp. bulgaricus, etc.). Moreover, some from Brochothrix, Caryophanon, Erysip- lactobacilli became members of the new elothrix, Kurthia, Listeria and Renibacte- genus Weissella (e.g. Lb. kandleri became rium [111]. Carnobacterium resembles W. kandleri), which also includes former lactobacilli but does not grow on acetate members of the genus Leuconostoc (e.g. media. The establishment of a new genus, Leuconoctoc paramesenteroides became Weissella [41], encompassing the Parame- W. paramesenteroides), whereas other senteroides group, which includes Leucon- lactobacilli were assigned to another new ostoc paramesenteroides and some hetero- genus, Carnobacterium (e.g. Lb. divergens fermentative Lactobacillus species, seems became Cb. divergens). Today, the genus to be justified on the basis of phylogenetic Lactobacillus contains 88 species and 15 analysis. A cell wall murein, based on subspecies according to a recent listing lysine with an interpeptide bridge contain- (Tab. II, www.bacterio.cict.fr, 6 January ing alanine or alanine plus serine or gly- 2003). Protein analysis such as protein fin- cine, can be used to distinguish Weissella gerprinting (analysis of total soluble cyto- from heterofermentative lactobacilli [111]. plasmic proteins), or multilocus enzyme Classical phenotypic tests for identifica- electrophoresis (analysis of electrophoretic tion of lactobacilli are based on physiolog- mobilities of certain enzymes) are advanced ical characteristics such as respiratory phenotypic methods in current use. Such type, motility, growth temperature and analysis can discriminate between bacteria growth in NaCl, and on biochemical char- to the species level and beyond. Lipid pro- acteristics such as homo/hetero-fermenta- filing has also been used. However, the tive, production of lactic acid isomers, identification of isolates to species level metabolism of carbohydrate substrates, can be difficult because of the considerable coagulation of milk and presence of partic- variations in biochemical attributes (fermen- ular enzymes (e.g. arginine dihydrolase, tation profiles) that seem to occur between antibiotic susceptibilities, and so on). strains currently considered to belong to Lactobacilli are typically chemoorgano- the same species, and some species are not trophic and ferment carbohydrates, pro- readily distinguishable in terms of pheno- ducing lactic acid as a major end product. typic characteristics. This is especially true In 1919, Orla-Jensen divided them into for the so-called Lactobacillus plantarum three subgenera – Thermobacterium, group (Lb. plantarum, Lb. paraplantarum Streptobacterium and Betabacterium – and Lb. pentosus), the Lactobacilllus casei according to optimal growth temperature and Lactobacilllus paracasei group (Lb. and hexose fermentation pathways. This casei, Lb. rhamnosus, Lb. zeae and Lb. classification was given up by Kandler and paracasei), Lb. brevis and Lb. buchneri. Weiss [111], who proposed a classification Recently Dellaglio et al. [57] proposed, on into three groups – I (obligate homofer- the basis of considerable published evi- mentative), II (facultative heterofermenta- 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 equi Lactobacillus paracasei ssp. Lactobacillus acidipiscis *Lactobacillus farciminis tolerans *Lactobacillus acidophilus Lactobacillus ferintoshensis Lactobacillus parakefiri Lactobacillus agilis Lactobacillus fermentum Lactobacillus paralimentarius Lactobacillus algidus Lactobacillus fornicalis Lactobacillus paraplantarum Lactobacillus alimentarius Lactobacillus fructivorans Lactobacillus pentosus Lactobacillus amylolyticus Lactobacillus frumenti Lactobacillus perolens Lactobacillus amylophilus Lactobacillus fuchuensis *Lactobacillus plantarum Lactobacillus amylovorus Lactobacillus gallinarum Lactobacillus pontis Lactobacillus animalis *Lactobacillus gasseri Lactobacillus psittaci Lactobacillus arizonensis Lactobacillus graminis *Lactobacillus reuteri Lactobacillus aviarius ssp. Lactobacillus hamsteri *Lactobacillus rhamnosus araffinosus Lactobacillus helveticus Lactobacillus rogosae Lactobacillus aviarius ssp. Lactobacillus heterohiochii Lactobacillus ruminis aviarius Lactobacillus hilgardii Lactobacillus sakei ssp. Lactobacillus bifermentans Lactobacillus homohiochii carnosus Lactobacillus brevis Lactobacillus iners Lactobacillus sakei ssp. Lactobacillus buchneri Lactobacillus intestinalis sakei *Lactobacillus casei Lactobacillus jensenii Lactobacillus salivarius ssp. Lactobacillus catenaformis *Lactobacillus johnsonii salicinius Lactobacillus cellobiosus Lactobacillus kefiranofaciens Lactobacillus salivarius ssp. Lactobacillus coleohominis Lactobacillus kefirgranum salivarius Lactobacillus collinoides Lactobacillus kefiri Lactobacillus sanfranciscensis Lactobacillus coryniformis ssp. Lactobacillus kimchii Lactobacillus sharpeae coryniformis Lactobacillus kunkeei Lactobacillus suebicus Lactobacillus coryniformis ssp. Lactobacillus leichmannii Lactobacillus trichodes torquens Lactobacillus lindneri Lactobacillus vaccinostercus *Lactobacillus crispatus Lactobacillus malefermentans Lactobacillus vaginalis Lactobacillus curvatus ssp. Lactobacillus mali Lactobacillus vitulinus curvatus Lactobacillus maltaromicus Lactobacillus zeae Lactobacillus curvatus ssp. Lactobacillus manihotivorans melibiosus Lactobacillus mucosae Lactobacillus cypricasei Lactobacillus murinus Lactobacillus delbrueckii ssp. Lactobacillus nagelii bulgaricus Lactobacillus oris *Lactobacillus delbrueckii ssp. Lactobacillus panis delbrueckii Lactobacillus pantheris Lactobacillus delbrueckii ssp. Lactobacillus parabuchneri lactis *Lactobacillus paracasei ssp. Lactobacillus diolivorans paracasei Lactobacillus durianis 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 judicial G+C content should vary by no more than commission and that the species Lactoba- a 10% range within a well-defined genus cillus casei is not correctly represented by [188]. The nucleotide sequences of Lacto- the strain ATCC 393. bacillus 16S ribosomal DNA (rDNA) pro- Studies based on 16S rDNA have led to vide an accurate basis for identification. the classification of Lactobacillus species The sequence obtained from an isolate can into three major groups: the Leuconostoc be compared with those of Lactobacillus group, the Delbrueckii group, and the Lb. species held in databases. Recently, Dubernet casei-Pediococcus group [40, 174, 188]. et al. [62] defined a genus-specific primer Recently Lb. fructosus (the only lactoba- by analysing similarities between the cilli member of the Leuconostoc group) nucleotide sequences of the spacer region was reclassified as Leuconostoc fructosum between the 16S and 23S ribosomal RNA [8]. Closely related species and strains genes of Lactobacillus. The specificity of with similar phenotypic features may now this genus-specific primer combined with a be reliably differentiated from each other universal primer was tested against 23 by DNA-based techniques. Molecular strains of lactobacilli of varied origin (cor- methods can be used for taxonomic analy- responding to 21 species) Escherichia coli, ses to firstly determine the species to two leuconoctocs species, Carnobacte- which a bacterium belongs, by DNA/DNA rium piscicola, Pediococcus pentosaceus, hybridisation, sequencing, polymorphism Bifidobacterium bifidum, Weissella confusa, chain reaction (PCR), ribotyping, poly- Enterococcus faecalis, Staphylococcus morphism chain reaction-restriction frag- aureus and Listeria monocytogenes. Posi- ment length polymorphism (PCR-RFLP), tive amplification was only obtained with and secondly permit strain differentiation the lactobacilli strains. by the use of techniques such as restriction enzyme analysis (REA), randomly ampli- 2.2.2. Analysis at species level fied polymorphic DNA (RAPD), repeated sequence extragenic palindrom PCR Phenotypical micro methods (REP-PCR), amplified fragment length polymorphism (AFLP), plasmid profiling Several combinations of tests and and pulsed field gel electrophoresis (PFGE). ready-to-inoculate identification kits such However, identification can be done with a as API 50 CH, LRA Zym and API Zym high degree of confidence only if a correct enzymatic tests can be used for the rapid validation of the method or probes or prim- and theoretically reproducible phenotypic ers have been checked using close genera, identification of pure cultures. They have species or strains. been used for the characterisation and Polyphasic taxonomy is increasingly identification of lactobacilli in milks [133], used [4, 75, 118, 167, 174, 188]. Lawson yoghurts and other fermented milks [6] and et al. [118] described, on the basis of phyl- in cheeses [6, 21, 53, 58, 92, 118, 120, 133, ogenetic and phenotypic evidence, a new 175]. However, the reliability of these tests species of Lactobacillus, Lb. cypricasei, has been questioned, especially for API which was isolated from Halloumi cheese, 50 CH, initially developed for the identifi- a semi-hard cheese from Cyprus. cation of medical Lactobacillus strains. In addition, the manufacturer’s database is 2.2.1. Analysis at genus level not updated and some Lactobacillus species are missing. Andrighetto et al. [6] used The genus Lactobacillus is heterogene- API 50 CH to analyse 25 strains of ther- ous, with the G+C content of the DNA of mophilic lactobacilli isolated from yoghurt its species varying from 33 to 55% [40, and from semi-hard and hard cheeses (Lb. 89]. However, it is generally thought that delbrueckii ssp. lactis and ssp. bulgaricus, 280 V. Coeuret et al.

Lb. helveticus and Lb. acidophilus). For inexpensive method that has already been most of the strains, clear assignment to a used for the identification and classifica- particular species or subspecies was not tion of lactobacilli. The entire procedure possible because ambiguous results were consists of several experimental steps, obtained for the sugar fermentation profile. from the growth of bacteria to the scanning Nigatu [141] also reported a lack of agree- of the electrophoregram. SDS-PAGE sep- ment between the API 50 CH grouping arates proteins exclusively according to pattern of isolates and RAPD clusters. molecular weight. Native (non-denaturing) Tynkkynen et al. [184] used API 50 CH for PAGE can be used as a complementary identifying strains of the Lb. casei group technique, separating cell proteins accord- (Lb. rhamnosus, Lb. zeae and Lb. casei). ing to their charge and size, providing high The exact identifications of these closely resolution and good band definition. In related species were not reliable; some highly standardised conditions, reproduci- were doubtful or unacceptable and some ble patterns can be obtained that are ame- strains were misidentified with a good nable to rapid, computer-based digital identification level. Furthermore, variabil- analysis. Protein profiles can be stored in ity may be observed within a single strain. database format and may be routinely used For example, the Lb. rhamnosus GG strain to confirm the identity of Lactobacillus has traditionally been detected, counted strains, to differentiate between unknown and identified on the basis of cultures in isolates and to evaluate classification selective anaerobic conditions on MRS or schemes, at species level or below [53, 75, Rogosa agar (37 °C for 78 h), colony mor- 92, 118, 122, 146, 147]. De Angelis et al. phology (large, white, creamy and [53] isolated NSLAB from 12 Italian ewe’s opaque), Gram staining and cell morphol- milk cheeses. Most of the species studied ogy (Gram-positive and uniform rods in gave specific protein profiles, except Lb. chains) and the carbohydrate fermentation plantarum and Lb. pentosus, which were profile in the API 50 CHL test. However, it grouped in the same cluster, confirming the has been pointed out that the colony mor- results previously obtained by Van Reenen phology and the carbohydrate fermenta- and Dicks [187]. Gancheva et al. [75] used tion pattern of strain GG are not always SDS-PAGE to analyse the cellular proteins typical, due to variation [32]. This varia- of a set of 98 strains belonging to nine species tion may result from the loss or gain of of the Lactobacillus acidophilus rRNA- plasmids, leading to inconsistency in the group of varied origin (Lb. acidophilus, metabolic traits of a strain, as most of the Lb. amylolyticus, Lb. crispatus, Lb. john- proteins involved in carbohydrate fermen- sonii, Lb. gasseri, Lb. gallinarum, Lb. hel- tation are plasmid-encoded [9]. veticus, Lb. iners and Lb. amylovorus). Most of these species can be differentiated Protein fingerprinting by SDS-PAGE, but poor discrimination was obtained between Lb. johnsonii and A bacterial strain always produces the Lb. gasseri strains, and between some strains same set of proteins if grown under stand- of Lb. amylovorus and Lb. gallinarum. ardised conditions. The electrophoregrams produced by zone electrophoresis of these proteins under well-defined conditions can Multilocus enzyme electrophoresis be considered as a sort of fingerprint of the bacterial strains from which they are About 50% of all enzymes investigated obtained. Sodium dodecylsulphate poly- to date exist in multiple molecular forms. acrylamide gel electrophoresis (SDS-PAGE) These isoenzymes usually differ in electro- of whole-cell proteins is one of the tech- phoretic mobility and catalytic parameters. niques used. It is a relatively simple and Enzyme multiplicity may depend on genetic Identification of dairy lactobacilli 281 factors directly (primary isoenzymes and strains [47, 66, 154]. Moreover, the relia- allozymes) or indirectly (secondary isoen- bility of lipid and polysaccharide profiling zymes, generated by post-translational for discriminating between Lactobacillus modifications). Isoenzymes may be dis- species has been questioned and fatty acid tributed between different cell compart- methyl ester (FAME) analysis does not ments and may be encoded by at least two seem to be reliable for LAB [113]. different genes. Multiple loci encoding enzymes of identical substrate specificities Hybridisation are usually thought to result from gene duplications. Point mutations then lead to The use of probes for hybridisation with divergence in the amino acid sequences of nucleic acid fragments is a technique with the proteins encoded by the duplicated great potential for the future. A nucleic genes, resulting in the production of differ- acid probe is a fragment (20-30 pb) of a ent enzyme forms, separable by electro- single-stranded nucleic acid fragment that phoresis. Differences in electrophoretic specifically hybridises to complementary mobility may result from differences in regions of a target nucleic acid. It can be charge and/or size. Multilocus enzyme used directly on a colony, or after DNA/ analysis has been shown to be of great RNA extraction. The target nucleic acid potential [161] in the differentiation of consists of single-stranded DNA or RNA LAB species [182]. The electrophoretic molecules. Molecular probes may be labelled mobility of LDH (lactate dehydrogenase) radioactively or non-radioactively. Radio- in starch gels [77] and in polyacrylamide active labelling involves the phosphoryla- gels [95] has been used to discriminate tion of the 5' terminus of the probe with between phenotypically very similar spe- [32P] ATP. Non-radioactive labelling may cies: Lb. acidophilus, Lb. crispatus, Lb. be direct, using alkaline phosphatase or gallinarum, Lb. gasseri and Lb. johnsonii. peroxidase, or indirect, by attachment of a Lortal et al. [122] studied peptidoglycan ligand-protein or a hapten-antibody. Fluo- hydrolases of industrial starters as a new rescent probes (FISH: fluorescent in situ tool for bacterial species identification. hybridisation) may also be used. The The peptidoglycan hydrolase patterns of extensive use of multiple oligonucleotide 94 strains of lactobacilli belonging to 10 probes has become possible following different species were determined (Lb. hel- major developments in the sequencing of veticus, Lb. acidophilus, Lb. delbrueckii, rRNA genes. Depending on the level of Lb. brevis, Lb. fermentum, Lb. jensenii, Lb. detection required (genus or species), dif- plantarum, Lb. sake, Lb. curvatus and Lb. ferent regions of the genome might be used reuteri). Each species gave its own specific as targets. The sequences of 16S and 23S pattern, with differences observed even rRNA molecules contain highly conserved between closely related species. It is also regions common to all eubacteria, and possible to type strains of Lactobacillus highly variable regions unique to a partic- acidophilus [146], or clinical isolates and ular species [32, 199]. Thus, nucleic acid biotechnologically-used strains of Lacto- probes, in particular probes targeting bacillus rhamnosus [112], or strains of rRNA sequences, can be used for the reliable Lactobacillus casei [55] isolated from identification of bacteria, for monitoring ensiled high-moisture corn grain. population changes during fermentation and detecting the presence of bacterial Lipid profiling contamination or spoilage bacteria [60]. Such probes have been extensively used in Lipid profiling by gas chromatography the analysis of dairy products [6, 98]. Oli- is more useful for the grouping of strains gonucleotide DNA probes, mostly target- than for the identification of individual ing variable regions of the 16S or 23S 282 V. Coeuret et al.

Table III. Oligonucleotide probes for the identification of Lactobacilli.

Probe 5’Sequence3’ Target Specificity Ref. Lba TCTTTCGATGCATCCACA 23S Lb. acidophilus [193] Lba AGCGAGCUGAACCAACAGAUUC 16S Lb. acidophilus [96] Lbam GTAAATCTGTTGGTTCCGC 16S Lb. amylovorus [68] Lbb TGTTGAAATCAAGTGCAAG 16S Lb. brevis [193] Lbc ATGATAATACCCGACTAA 23S Lb. curvatus [97] Lbco AGCACTTCATTTAACGGG 16S Lb. collinoides [68] Lbcp CAATCTCTTGGCTAGCAC 23S Lb. crispatus [67] Lbcr GCAGGCAATACACTGATG 23S Lb. casei /Lb. rhamnosus [98] Lbcrp CTGATGTGTACTGGGTTC 23S Lb. casei / Lb. paracasei / Lb. [98] rhamnosus Lbd AAGGATAGCATGTCTGCA 23S Lb. delbrueckii [98] Lbdb ATCCGAAAACCTTCT 16S Lb. delbrueckii ssp. bulgaricus [119] Lbdl ATCCGAAGACCTTCT 16S Lb delbrueckii ssp. lactis/delbrueckii [119] 33/2 CATCAACTGGCGCCTT 730pb EcoRI/PstI Lb. delbrueckii ssp. lactis [119] DNA fragment 34B CATCAACCGGGGCTTT 730pb EcoRI/PstI Lb. delbrueckii ssp. bulgaricus [119] DNA fragment Lbfe GCGACCAAAATCAATCAGG 16S Lb. fermentum [193] Lbfr CTCGCTGCTAACTTAAGTC 16S Lb. fructivorans /Lb. homohiochii [193] Lbg TCCTTTGATATGCATCCA 23S Lb. gasseri [146] Lbh ACTTACGTACATCCACAG 23S Lb. helveticus [98] Lbhi CTCAACTTCATTGACCAAG 16S Lb. hilgardii [68] Lbj ATAATATATGCATCCACAG 23S Lb. johnsonii [146] Lbk GTTTCATGTTAAATCATTCA 16S Lb. kefir [68] Lbkf TGCGGCTAGCCCTTCCGG 23S Lb. kefiranofaciens [68] Lbl TCGGTCAGATCTATCGTC 16S Lb. lindneri [68] Lbma CAAAAGCGACAGCTCGAAAG 16S Lb. manihotivorans [3] Lbp ATCTAGTGGTAACAGTTG 23S Lb pentosus / Lb plantarum [97] Lbpa CACTGACAAGCAATACAC 23S Lb paracasei [98] Lbpa TAACTCATTGACTGACTCG 23S Lb parabuchneri [68] Lbpe TCAAATGTAAATCATGATG 16S Lb pentosus / plantarum [68] Unamed GGTATTGGTGATGCAAG 16S Lb. perolens [11] Lbpp ATCTAGTCGTAACAGTTG 23S Lb plantarum / pentosus [97] Lbpo GGTAATCCATCGTCAAATC 16S Lb pontis [193] Lbre GATCCATCGTCAATCAGG 16S Lb reuteri [68] Lbru TTCGGTGAAAGAAAGCTTG 16S Lb ruminis [96] Lbs TTAATGATAATACTCGATT 23S Lb sake [97] Lbsa TAAGAATCAATTGGGCGAC 16S Lb sanfransiscensis [193] rRNA genes, have been widely used for due to the high level of similarity between species identification and strain detection their rRNA gene sequences. For example, (Tab. III). However, such rRNA probes such probes cannot distinguish Lb. plantarum cannot be used for closely related species from Lb. pentosus or Lb. paraplantarum Identification of dairy lactobacilli 283

[25]. These species are currently distin- comparative work. The spacer region guished by probing Southern blots with a method has the advantage of distinguishing pyrDFE gene fragment from Lb. plantarum between Lb. rhamnosus and Lb. casei or by DNA/DNA hybridisation. Particular strains [177]. It can be used to distinguish Lb. attention had to be done to their specificity plantarum, from Lb. paraplantarum, these since Roy et al. [158] demonstrated that two closely related species belonging to the probe defined by Hertel et al. [98] for the Lb. plantarum group [12]. Chen et al. Lb. helveticus also hybridise with Lb. gall- [33] analysed the 5S-23S rRNA intergenic inarum strains. spacer regions (ISRs) of the Lactobacillus In colony hybridisation, bacteria are group. This method was found to be an plated on membranes that are then placed effective way of discriminating Lb. rham- on nutrient agar, allowing the bacteria to nosus from Lb. casei/Lb. paracasei because form colonies. The colonies are lysed. spacer length polymorphism results in a Hybridisation with a labelled probe can be 76/80 bp insertion with respect to the 16S used for both qualitative and quantitative V2-V3 sequences. analyses, and has been used for LAB [37]. Polymerase chain reaction (PCR) Sequencing This method developed by Mullis [139] Comparison of rRNA gene sequences is uses primers, which are about 20 to 30 pb. currently considered to be the most power- Berthier and Ehrlich [15] studied the 16S/ ful and accurate method for determining 23S SRs DNA from six closely related species the degree to which microorganisms are of lactobacilli (Lb. curvatus, Lb. graminis, phylogenetically related [199]. Advances Lb. sake, Lb. plantarum, Lb. paraplantarum in molecular biological techniques have and Lb. pentosus). Only the larger frag- made it possible to sequence long stretches ment displayed differences in sequence of rRNA genes. Initially, reverse tran- between species, and primers derived from scriptase was used to generate DNA from this region were defined for the six species. rRNA, and this DNA was then sequenced. SR sequences could not be used to type It is now possible to sequence 16S or 23S strains within the two groups of Lactoba- rDNA molecules by direct PCR sequenc- cillus species. ing, and this method has generated large Numerous species-specific primers have sequence databases. Although the species- been derived from spacer regions [15, 170, specific sequences are located in the first 177, 179]. The species-specific primers half of the 16S rRNA gene (V1-V3 currently available are listed in Table IV. region), identification is more accurate if the whole gene is sequenced [171]. This Mannu et al. [126] used seven pairs of requires the sequencing of about 1.5 kb of specific primers designed by Berthier and DNA. Tannock et al. [177] showed that Ehrlich [15], Drake et al. [60] and Ward comparison of the16S-23S spacer region and Timmins [196] to analyse 457 isolates sequences of lactobacilli can be used in from Fiore Sardo cheese, a traditional hard practical situations for strain identification. cheese from Sardinia. Only seven isolates The spacer region sequences is sequencing were not successfully identified with this rapidly and accurately identifies Lactoba- method; 31 were obligate homofermenta- cillus isolates obtained from gastrointestinal, tive lactobacilli, 419 isolates were faculta- yoghurt and silage samples. The 16S-23S tive heterofermentative lactobacilli (Lb. spacer sequences of lactobacilli are small, plantarum, Lb. paracasei and Lb. curvatus). only about 200 bp in length. These short Many recent studies have been carried sequences are easy to sequence on both out by multiplex PCR (Tab. V), in which strands and provide reliable information for several primers are added to the same sample, 284 V. Coeuret et al.

Table IV. PCR primers used for lactobacilli identification.

Primers 5’–3’ Target Ref. LbLMA1-rev CTCAAAACTAAACAAAGTTTC 16S/23S spacer region Lactobacilli R16-1 CTTGTACACACCGCCCGTCA 16S rRNA gene [62] (/LbLMA1-rev) (/LbLMA1-rev) Y2 CCCACTGCTGCCTCCCGTAGGAGT 16S rRNA gene casei(/Y2) TGCACTGAGATTCGACTTAA (/ Y2) 16S Lb. casei [196] para(/Y2) CACCGAGATTCAACATGG (/Y2) 16S Lb. paracasei rham(/Y2) TGCATCTTGATTTAATTTTG (/Y2) 16S Lb. rhamnosus 16 GCTGGATCACCTCCTTTC 16S rRNA gene paracaseiITS/16 CGATGCGAATTTCTTTTTC /16 16S/23S spacer region Lb. paracasei rhamnosusITS/16 CGATGCGAATTTCTATTATT /16 16S/23S spacer region Lb. rhamnosus zeaeITS/16 CGATGCGAATTTCTAAATT /16 16S/23S spacer region Lb. zeae

16 reverse GAAAGGAGGTGATCCAGC 16S rRNA gene [17] paracasei CACCGAGATTCAACATGG/16 reverse 16S rRNA gene. Lb. paracasei 16S/16 reverse

rhamnosus TTGCATCTTGATTTAATTTTG/16 reverse 16S rRNA gene. Lb. rhamnosus 16S/16 reverse

zeae 16S/16 reverse GCATCGTGATTCAACTTAA/16 reverse 16S rRNA gene. Lb. zeae Ala CTGCTGGGACGATTTG Ala'(/Ala) CTGCTGGGACCATGTG (/Ala) Lb. curvatus (1840 pb) Alb CTGCTGGGACCATTATTG Alb'(/Alb) CTGCTGGGACACAATATG (/Alb) Not tested (1470 pb) Alc GGAGGGTGTTCAGGAC Alc'(/Alc) GGAGGGTGTTGATAGG (/Alc) Lb. curvatus (260 pb) Bla CTGCTGGGACCAATT [16] Bla'(/Bla) CTGCTGGGACGAAAAG (/Bla) Lb. sake B1 (750 pb) B2a CTGCTGGGACCTTAA B2a'(/B2a) CTGCTGGGACTGAAG (/B2a) Lb. sake B2 (1700 pb) 16 GCTGGATCACCTCCTTTC Lc(/16) TTGGTACTATTTAATTCTTAG (/16) Lb. curvatus (220 pb) Ls(/16) ATGAAACTATTAAATTGGTAC (/16) Lb. sake (220 pb) 16 GCTGGATCACCTCCTTTC 16S rRNA gene 23(/16) AGTGCCAAGGCATCCACC (/16) 23S rRNA gene(/16S) Lc(/16) TTGGTACTATTTAATTCTTAG (/16) 16S/23S spacer region Lb. curvatus Lg(/16) GTTGGTACATTTAATTCTTGA (/16) 16S/23S spacer region Lb. graminis Lpapl(/16) ATGAGGTATTCAACTTATT (/16) 16S/23S spacer region [15] Lb. paraplantarum/plantarum Lpe(/16) GTATTCAACTTATTAGAACG (/16) 16S/23S spacer region Lb. pentosus Lpl(/16) ATGAGGTATTCAACTTATG (/16) 16S/23S spacer region Lb. plantarum Ls(/16) ATGAAACTATTAAATTGGTAC (/16) 16S/23S spacer region Lb. sake LB1 AAAAATGAAGTTGTTTAAAGTAGGTA Lb. delbrueckii bulgaricus (1065 pb) [181] LLB1 (/LB1) AAGTCTGTCCTCTGGCTGG (/LB1) Lb. delbrueckii lactis (1600 pb) 20A AATTCCGTCAACTCCTCATC 23B (/20A) TGATCCGCTGCTTCATTTCA(/20A) Lb. delbrueckii ssp. (715 pb) 34/2 CGTCAACTCCTCATCAACCGGGGCT 37/1(/34/2) CGCCGCCCGGGTGAAGGTG(/34/2) Lb. delbrueckii ssp. bulgaricus (678 pb) [119] 33 CCTCATCAACTGGCGCC 37(/33) CGCCCGGGTAAAGGTA(/33) Lb. delbrueckii ssp. lactis (670 pb) Identification of dairy lactobacilli 285

Table IV. PCR primers used for lactobacilli identification.

LbP11 AATTGAGGCAGCTGGCCA LbP12 (/LbP11) GATTACGGGAGTCCAAGC (/LbP11) RAPD derived primer. Lb. plantarum [148] Lb1 AGAGTTTGATCATGGCTCAG Lb2 (/Lb1) CGGTATTAGCATCTGTTTCC (/Lb1) semi-universal primer. 16S rRNA Aci I TCTAAGGAAGCGAAGGAT Aci II (/Aci I) CTCTTCTCGGTCGCTCTA (/Aci I) 16S-23S SR. Lb. acidophilus Pr I CAGACTGAAAGTCTGACGG PrII 5/PrI) GTACTGACTTGCGTCAGCGG (/PrI) 16S-23S SR Lb. paracasei/rhamnosus Pcas I (/PrI) GCGATGCGAATTTCTTTTTC (/PrI) 16S-23S SR Lb. paracasei [179] Rha II(/PrI) GCGATGCGAATTTCTATTATT (/PrI) 16S-23S SR Lb. rhamnosus Del I ACGGATGGATGGAGAGCAG Del II (/Del I) GCAAGTTTGTTCTTTCGAACTC (/Del I) 16S-23S SR Lb. delbrueckii Hel I GAAGTGATGGAGAGTAGAGATA Hel II (/Hel I) CTCTTCTCGGTCGCCTTG (/Hel I) 16S-23S SR Lb. helveticus DB1 ACCTATCTCTAGGTGTAGCGCA SS1 (/DB1) GTGCTGCAGAGAGTTTGATCCTGGCT- 16S Lb. delbrueckii (1100 pb) CAG (/DB1) HE1 AGCAGATCGCATGATCAGCT [60] SS2 (/HE1) CACGGATCCTACGGGTACCTTGTTAC- 16S Lb. helveticus/Lb. acidophilus [6] GACTT(/HE1) (1400 pb)

CA1 (/SS1) TGATCTCTCAGGTGATCAAAA (/SS1) 16S Lb. casei, Lb. rhamnosus 16 SII ACTACCAGGGTATCTAATCC Aci 16SI (/16SII) AGCTAACCAACAGATTCAC (/16 SII) 16S Lb. acidophilus Cri 16SI(/16SII) GTAATGACGTTAGGAAAGCG (/16 SII) 16S Lb. crispatus GasI GAGTGCGAGAGCATAAAG GasII (/GasI) CTATTTCAAGTTGAGTTTCTCT (/GasI) 16S-23S SR Lb. gasseri Joh 16SI(/16SII) GAGCTTGCCTAGATGATTTTA (/16 SII) 16S-23S SR Lb. johnsonii Lpfr GCCGCCTAAGGTGGGACAGAT PlanII (/Lpfr) TTACCTAACGGTAAATGCGA (Lpfr) 16S-23S SR Lb. plantarum PrI CAGACTGAAAGTCTGACGG [194] CasII (/PrI) GCGATGCGAATTTCTTTTTC (/PrI) 16S-23S SR Lb. casei [184] ZeaeI TGTTTTGAGGGGACG ZeaeII (/ZeaeI) ATGCGATGCGAATTTCTAAATT (/ZeaeI) 16S-23S SR Lb. zeae RhaII(/PrI) GCGATGCGAATTTCTATTATT (/PrI) 16S-23S SR Lb. rhamnosus Reu(/lpfr) AACACTCAAGGATTGTCTGA (/lpfr) 16S-23S SR Lb. reuteri FermII(/lpfr) CTGATCGTAGATCAGTCAAG (/lpfr) 16S-23S SR Lb. fermentum ShaI GATAATCATGTAAGAAACCGC ShaII(/ShaII) ATATTGTTGGTCGCGATTCG (/ShaII) 16S-23S SR Lb. sharpae SAL1 ATTCACTCGTAAGAAGT 16S [30] LOWLAC(/SAL1) CGACGACCATGAACCACCTGT(/SAL1) 16S Lb. salivarius CbsA2F GTACCAAGCCAAAGCAAGAC CbsA2R(/CbsA2F) GTTTGAAGCCTTTACGTAAGTC CbsA (S-Layer protein gene) [99] (/CbsA2F) Lb. perolens 97K CTGCTGCCTCCCGTA 16S Universal [27] Lpacaf(/97K) CCGAGATTCAACATGG(/97K) 16S Lb. paracasei making it possible to detect several micro- samples [123] and in meat spoilage [200]. organisms or species at the same time. Song et al. [170] used multiplex PCR as a Multiplex PCR has been used to detect Lb. rapid, simple and reliable method for the pontis and Lb. panis in sourdough fermen- identification of common Lactobacillus tation [138], and Lactobacillus in faecal isolates from human stool samples, and 286 V. Coeuret et al.

Table V. Multiplex PCR primers used for lactobacilli identification.

Primers 5’–3’ Target Ref.

Lac-2 CCTCTTCGCTCGCCGCTACT Ldel-7 (/lac-2) ACAGATGGATGGAGAGCAGA (/lac-2) ISR/23S PCR-G Group I lactobacilli (450pb) LU-1'(/lac-2) ATTGTAGAGCGACCGAGAAG (/lac-2) ISR/23S PCR-G Group II lactobacilli (300 pb) LU-3'(/lac-2) AAACCGAGAACACCGCGTT (/lac-2) ISR/23S PCR-G Group IV lactobacilli (350 pb) LU-5(/lac-2) CTAGCGGGTGCGACTTTGTT (/lac-2) ISR/23S PCR-G Group III lactobacilli (400 pb)

23-10C CCTTTCCCTCACGGTACTG Laci-1 (23-10C) TGCAAAGTGGTAGCGTAAGC (/23-10C) ISR/23S PCR-II-1 Group II, Lb. acidophilus (210 pb) Ljen-3 (23-10C) AAGAAGGCACTGAGTACGGA (/23-10C) ISR/23S PCR-II-1 Group II, Lb. jensenii (700 pb) Lcri-1 AGGATATGGAGAGCAGGAAT Lcri-2(/Lcri-1) CAACTATCTCTTACACTGCC (Lcri-1) ISR/23S PCR-II-2 Group II, Lb. crispatus (522 pb) Lgas-1 AGCGACCGAGAAGAGAGAGA Lgas-2 (/Lgas-1) TGCTATCGCTTCAAGTGCTT (/Lgas-1) ISR/23S PCR-II-2 Group II, Lb. gasseri (360 pb) [170]

Lfer-3 ACTAACTTGACTGATCTACGA Lfer-4 (/Lfer-3) TTCACTGCTCAAGTAATCATC (/Lfer-3) ISR/23S PCR-IV Group IV, Lb. fermentum (192 pb) Lpla-3 ATTCATAGTCTAGTTGGAGGT Lpla-2 (Lpla-3) CCTGAACTGAGAGAATTTGA (/Lpla-3) ISR/23S PCR-IV Group IV, Lb. plantarum (248 pb) Lreu-1 CAGACAATCTTTGATTGTTTAG Lreu-4 (/Lreu-1) GCTTGTTGGTTTGGGCTCTTC (/Lreu-1) ISR/23S PCR-IV Group IV, Lb. reuteri (303 pb) Lsal-1 AATCGCTAAACTCATAACCT Lsal-2 (/lsa-2) CACTCTCTTTGGCTAATCTT (/lsa-2) ISR/23S PCR-IV Group IV, Lb. salivarius (411 pb)

Lpar-4 (/LU-5) GGCCAGCTATGTATTCACTGA (/LU-5) ISR/23S PCR-III Group III, Lb. paracasei (312 pb) RhaII (/LU-5) GCGATGCGAATTTCTATTATT (/LU-5) ISR/23S PCR-III Group III, Lb. rhamnosus (113 pb)

Y1 TGGCTCAGAACGAACGCTAGGCCCG 16S rRNA Y2 (/Y1) CCCACTGCTGCCTCCCGTAGGAGT (/Y1) 16S rRNA PCR A (350 pb)

16 GCTGGATCACCTCCTTTC 16S rRNA gene Ls (/16) ATGAAACTATTAAATTGGTAC (/16) 16S/23S SR Lb. sake PCR A (220 pb)

Lc (/16) TTGGTACTATTTAATTCTTAG (/16) 16S/23S SR Lb. curvatus PCR B (220 pb) [200] Lu1r CCACAGCGAAAGGTGCTTGCAC Leuconostoc 16S rRNA gene Lu2 (/lu1r) GATCCATCTCTAGGTGACGCCG (/lulr) Leuconostoc PCR B (175 pb)

Lw5 (/Y1) ACTAGAATCATTCCCTATTCTAGC (/Y1) Leuconostoc PCR C (470 pb)

Cb1 CCGTCAGGGGATGAGCAGTTAC Carnobacterium 16S rRNA gene Cb2r (/Cb1) ACATTCGGAAACGGATGCTAAT (/Cbl) Carnobacterium PCR D (340 pb)

616V AGAGTTTGATYMTGGCTCAG universal 609R (/616V) ACTACYNGGGTATCTAAKCC (/616V) universal (800 pb) LaponR (/616V) AGCCATCTTTGAAAT (/616V) Lb. pontis (236 pb) [138] LapanR (/616V) AACCATCTTTTATAC (/616V) Lb. panis(236 pb) LaspecR (/616V) AGCCTTCTTTTATAC (/616V) Lb. species(236 pb) established a two-step PCR method. In this Ribotyping method, lactobacilli are first classified into Southern blotting is carried out after the four groups, and then one or two multiplex restriction digestion of chromosomal DNA PCR assays are carried out for each group, and agarose electrophoresis. In this process, for species identification. Lb. delbrueckii DNA is transferred to a membrane for was identified in the first grouping multi- hybridisation with a labelled 23S, 16S or plex PCR and 10 species were identified in 5S rRNA gene probe. As bacteria have the second multiplex PCR for each group. multiple copies of rRNA operons in their Identification of dairy lactobacilli 287 chromosome, several fragments in the tional gel electrophoresis (CGE). RFLP restriction digest mixture hybridise with analysis of the 16S rRNA gene, chromo- the probe. In general, the fingerprint pat- somal DNA cleaved with EcoRI and terns obtained by this method are more sta- HindIII, gave identical patterns for most of ble and easier to interpret than those the strains of Lb. plantarum [109]. When obtained by restriction enzyme analysis the specific region corresponds to rDNA, (REA). Ribotyping has been used with then this method is called PCR-ARDRA some success to characterise strains of var- (amplified rDNA restriction analysis), and ious Lactobacillus species [155], strains of is derived from ribotyping. Andrighetto Lb. helveticus [80] and strains of Lb. del- et al. [6] analysed a 1500 bp polymorphism brueckii [135]. Miteva et al. [135] successfully in EcoRI-digested 16S rDNA fragments differentiated between the three subspecies from 25 strains of 4 species of lactobacilli of Lb. delbrueckii by EcoRI ribotyping. isolated from cheeses and yoghurts: Lb. Zhong et al. [202] used ribotyping for spe- delbrueckii ssp. lactis and Lb. delbrueckii cies discrimination (Lb. jensenii, Lb. casei, ssp. bulgaricus, Lb. helveticus and Lb. aci- Lb. rhamnosus, Lb. acidophilus, Lb. dophilus. Different patterns were observed, plantarum and Lb. fermentum). In general, making it possible to distinguish between the ribotyping has a greater discriminatory various Lactobacillus species and subspe- power at species level than at strain level. cies. Giraffa et al. [80] used PCR-ARDRA Tynkkynen et al. [184] analysed 24 lacto- to identify Lb. delbrueckii isolates to sub- bacilli strains biochemically identified as species level and to differentiate this species members of the Lactobacillus casei group from Lb. helveticus and Lb. acidophilus. (Lb. rhamnosus and Lb. casei): ribotyping As these species were present in the same by EcoRI digestion and southern blotting ecological niches, and displayed similar with a chemiluminescent probe correspond- phenotypic characteristics and a close ing to the rrnB rRNA operon from E. coli genetic relationship, PCR-ARDRA was resulted in the detection of a triplet, which efficient. Bouton et al. [22] confirmed by seems to be typical for most Lb. rhamnosus PCR-ARDRA strains isolated from Comté strains. cheese belonged to Lb. delbrueckii ssp. A fully automated ribotyping system, lactis. For six presumed strains of Lb. hel- the RiboPrinter microbial characterisation veticus, no cutting by EcoRI was obtained, system, has been developed for identifica- even if fermentation profiles suggested tion at the genus, species and strain levels that all strains belonged to Lb. helveticus [26]. This method is automated and based rather than Lb. acidophilus. Chromosomal on a standardised protocol, maximising inter- rearrangements [82] or cross-protection by laboratory reproducibility. It is easy to carry methylation could explain the loss of a out but the equipment is rather expensive. cleavage site. In the database supplied by the manufacturer (Qualicon), reference is made to sev- 2.2.3. Analysis at strain level eral bacterial genera: lactic acid bacteria (lactobacilli included), Salmonella, Liste- Restriction Enzyme Analysis (REA) ria, Escherichia, Pseudomonas and others. Restriction enzyme analysis (REA) PCR – Restriction Fragment Length involves the extraction and digestion of Polymorphism analysis (PCR-RFLP) chromosomal DNA with restriction endo- nucleases and separation of the fragments PCR-restriction fragment length poly- by conventional gel electrophoresis (CGE). morphism analysis involves the amplifica- The number of bands obtained, generally tion of a specific region, followed by ranging between 1000 and 20 000 bp in restriction enzyme digestion and conven- size, are dependent on the restriction 288 V. Coeuret et al. enzymes used. The complexity of the and Dicks [61] used RAPD to separate spe- banding pattern makes visual evaluation cies of the Lactobacillus acidophilus group difficult and necessitates the use of compu- (Lb. acidophilus, Lb. crispatus, Lb. amy- ter-assisted multivariate analysis [32]. lovorus, Lb. gallinarum, Lb. gasseri and Electrophoretic separation of the DNA Lb. johnsonii), which are difficult to distin- fragments obtained after restriction endo- guish on the basis of simple physiological nuclease digestion has been achieved for and biochemical tests. Johansson et al. [109] many bacterial species of the genus Lacto- evaluated the typing potential of RAPD for bacillus. REA has been successfully used Lb. plantarum strains from various sources: to differentiate between strains of Lb. aci- 50% of the strains could be individually dophilus [157]. Zhong et al. [202] used separated from all the other strains tested BclI and DraI to separate 64 strains of and REA could separate the rest. lactobacilli; this method allowed discrimi- Cocconcelli et al. [38] demonstrated nation between the strains, but the patterns that the community of thermophilic lacto- produced were very complex. bacilli that dominates in Parmesan cheese is composed of a limited number of bacte- RAPD/ AP-PCR rial strains belonging to the Lb. helveticus and Lb. delbrueckii ssp. lactis species. Polymerase chain reaction (PCR)-based Baruzzi et al. [12] reported strains of the DNA fingerprinting methods using arbi- following species: Lb. acetotolerans, Lb. trary primers (AP) have been developed alimentarius, Lb. brevis, Lb. gasseri, Lb. for studying genomic DNA polymorphism. kefiri, Lb. paracasei, Lb. plantarum and Arbitrarily primed PCR and randomly Lb. zeae in Ricotta forte cheese. RAPD amplified polymorphic DNA (RAPD) analysis was used to separate the strains, methods were first described in 1990 [197, which were previously grouped into one 198]. In these similar techniques, the protein profile cluster as Lb. plantarum primers are generally about 10 nucleotides and Lb. pentosus [187]. Quiberoni et al. long and are not directed at any known [149] used primers P1 and P2 to discrimi- sequence of the bacterial genome, as the nate between 25 isolates obtained from primers are selected arbitrarily. A single Sardo and Reggianito cheeses. Giraffa et al. arbitrary oligonucleotide primer directs the [81] characterised 23 strains of Lb. helveti- amplification of random segments of cus isolated from natural whey starter cul- genomic DNA. It generates a characteris- tures used for Italian hard cheese. Sohier tic spectrum of short DNA products of var- et al. [169] used RAPD primers (and REP- ious complexities. RAPD techniques have PCR) to separate isolates of Lb. brevis and been extensively used in the typing of lac- Lb. hilgardii. The two fingerprinting meth- tic acid bacteria [176]. Some of the prim- ods were equally suitable for revealing ers used are listed in Table VI. Randomly species-specific genetic profiles. RAPD amplified polymorphic DNA analysis has analysis may have the advantage of facili- been used to monitor population dynamics tating simultaneous strain typing, species in food fermentation and to estimate the affiliation determination and individual diversity of Lactobacillus strains in cheeses strain differentiation [16]. RAPD-derived [12, 17, 22] whey culture [38], sausage probes and primers have been described fermentation [137, 152] and maize dough for the identification of lactobacilli to spe- [91]. It can also be used to distinguish a cies level, and even to strain level [148]. particular strain from the natural flora, Tilsala-Timisjarvi and Alatossava [180] such as distinguishing Lactobacillus probi- have also developed strain-specific DNA- otic adjunct from the natural NSLAB pop- derived PCR primers for a probiotic strain ulation in Cheddar cheese [172]. Du Plessis of Lb. rhamnosus. Identification of dairy lactobacilli 289

Table VI. RAPD primers.

RAPD primers (5’–3’) Used for Ref. 1254 Lb. delb. bulgaricus, Lb. acidophilus, Lb. kéfiranofasciens, CCGCAGCCAA Lb. helveticus, Lb. delb. lactis, Lb. casei, Lb. rhamnosus, [181] M13 Lb. maltoromicus, Lb. buchneri, Lb. kéfir. GAGGGTGGCGGTTCT 9898 Lb. brevis, Lb. hilgardi [169] GCAGCCGG AGTCAGCCAC Lb. casei, Lb. rhamnosus, Lb. zeae [184] P1 GCGGCGTCGCTAATACATGC [38] Lactobacillus P4 [152] ATCTACGCATTTCACCGCTAC CTGCTGGGAC Lb. curvatus, Lb. graminis, Lb. sakei [16] GGAGGGTGTT OPL-01 GGCATGACCT Lb. acidophilus, Lb. crispatus, Lb. amylovorus, [61] OPL-04 Lb. gallinarum, Lb. gasseri, Lb. johnsonii GACTGCACAC UNAMED Lb. plantarum [109] ACGCGCCCT Lactobacillus [7] UNAMED Lb. acidophilus, Lb. helveticus, Lb. casei, Lb. reuteri, [36] AGCAGCGTGG Lb. plantarum. [12] OPL-01 GGCATGACCT OPL-04 GACTGCACAC Lb. pentosus, Lb. casei, Lb. sake, Lb. curvatus, [187] OPL-02 Lb. plantarum TGGGCGTCAA OPL-05 ACGCAGGCAC OPB-06 TGCTCTGCCC [80] Lb. helveticus OPB-10 [22] CTGCTGGGAC P1 TGCTCTGCCC Lb. helveticus [149] P2 CTGCTGGGAC RP Lb. paracasei, Lb. rhamnosus [196] CAGCACCCAC CRA 23 GCGATCCCCA Lactobacillus sp. [51] CRA25 AACGCGCAAC OPA-02 TGCCGAGCTG OPM-05 GGGAACGTGT Lb. acidophilus group [75] OPL-07 AGGCGGGAAC OPL-16 AGGTTGCAGG 290 V. Coeuret et al.

Table VII. REP/ERIC primers.

Primers (5’–3’) Used for Ref. REP1R-I IIIICGICGICATCIGGC Lb. hilgardii, Lb. brevis [169] REP2-I ICGICTTATCIGGCCTAC REP-1R-Dt IIINCGNCGNCATCNGGC [17] Lactobacillus sp. REP2-D [22] NCGNCTTATCNGGCCTAC REP-1R-Dt IIINCGNCGNCATCNGGC REP2-Dt NCGNCTTATCNGGCCTAC Lb. sakei [102] BOXA1R CTACGGCAAGGCGACGCTGACG RW3A TCGCTCAAAACAACGACACC

It is also possible to use a combination called IRU (intergenic repeat units) or of two or more 10-mer oligonucleotides ERIC (enterobacterial repetitive intergenic (multiplex RAPD) in a single PCR to gen- consensus) has been described; IRU are erate RAPD profiles, making it possible to 124-127 bp long and have a copy number discriminate between Lactobacillus strains of about 30-50 in E. coli and 150 in S. typh- [51]. imurium [100, 166]. The members of both RAPD analysis has been shown to be the PU and IRU families are located in less effective than other molecular meth- non-coding but probably transcribed ods, although in some cases, it allowed the regions of the chromosome and both have separation of strains indistinguishable by a potential stem-loop structure. These other techniques. The screening of new sequences have been searched for and RAPD primers might increase the specifi- studied in lactobacilli [17, 22, 102, 169], city of this technique for strain typing. and repetitive primers were designed for RAPD analysis is a rapid and cheap method, PCR. Some of the primers used are listed but careful optimisation is required to in Table VII. obtain reproducible results. Berthier et al. [17] analysed 488 isolates of mesophilic lactobacilli isolated from REP-PCR/ERIC-PCR Comté cheese. These isolates were identified to species and strain level with a combina- Repeated sequences are present in the tion of two PCR-based methods: amplifica- genomes of all organisms. The first tion with pairs of repetitive primers (ERIC described and most extensively studied and REP), and amplification with specific repeated sequence is the extragenic palin- primers. REP-PCR fingerprints were used to drome (REP) or palindromic unit (PU), ini- assign strains to species. Combined REP tially identified in Salmonella typhimu- and ERIC fingerprints have the advantage rium and Escherichia coli [79]. This over RAPD analysis that the sequences sequence has a copy number of 500–1000 considered are longer and are therefore less and consists of a 35-40 bp inverted repeat. sensitive to minor changes in reaction condi- It is found in clusters, with successive cop- tions [17]. Hyytiä-Trees et al. [102] concluded ies arranged in opposite orientations. A that REP-PCR has a discriminatory power second family of repetitive elements, similar to that of RAPD analysis, but Identification of dairy lactobacilli 291 weaker than that of pulsed-field gel elec- Pulsed-field gel electrophoresis trophoresis (PFGE). However, if the results (PFGE) of REP-PCR and RAPD analysis were combined, the discriminatory power in Like REA, this method involves restric- some cases equalled that of PFGE. tion enzyme digestion, but the enzymes used for PFGE must have a low cutting fre- quency, as is the case for SmaI and SgrAI. Amplified fragment length However, in this case, the restriction frag- polymorphism (AFLP) ments are resolved by pulsed-field gel Another method, although not widely electrophoresis. This technique, involving used except in a systematic approach, is the application of an alternating electric field called amplified fragment length polymor- in two defined directions, is used to separate very large fragments, from 5 P 104 pb phism (AFLP). This method is sophisti- to 2 P 106 pb. This method is highly dis- cated and combines PCR and restriction criminatory and reproducible, and gener- enzyme techniques. AFLP templates are ates a banding pattern that is easy to inter- first prepared by cutting with two enzymes pret. (for example, the six-base cutter HindIII and the four-base cutter MseI, for Lactoba- PFGE analysis alone, with two or three cillus), resulting in DNA fragments with appropriate enzymes, can be used for reliable strain typing. In several Lactobacillus two different types of sticky ends. Appro- studies, PFGE has been shown to be a priate adaptors (short oligonucleotides) are powerful method for strain typing ligated to these ends to form templates for (Tab. VIII). However, one drawback of PCR, using two different primers contain- this method is that only a limited number ing the adaptor sequence extended to of samples can be analysed. Tynkkynen include one or more selective bases next to et al. [184] identified, by PCR with spe- the restriction site of the primer. Only frag- cific primers, 24 strains of Lactobacillus ments that completely match the primer biochemically related to the Lactobacillus sequence are amplified, resulting in selec- casei group. These strains were typed by tive amplification according to the initial RAPD, ribotyping and PFGE, to make pos- DNA structure and cutting. The amplifica- sible comparison of the discriminatory tion process results in an array of 30 to power of the methods. Twelve RAPD gen- 40 DNA fragments that are group- and/or otypes were detected among the 24 Lacto- species- and/or strain-specific [108]. How- bacillus strains; ribotyping with EcoRI ever, AFLP analysis, which involves a produced 15 different fingerprint patterns large number of experimental steps, has to and PFGE revealed 17 different genotypes. be carefully monitored by a specialist to Blaiotta et al. [20] used PFGE to moni- ensure a high level of reproducibility, even tor the addition of Lactobacillus, used as if an automated AFLP-fingerprinting sys- starter, to Cacioricotta cheese. This tech- tem is used [75]. RAPD-PCR and AFLP nique made it possible to analyse the analyses have been used for the rapid typ- growth kinetics of each starter strain during of strains of Lb. acidophilus and related ing the process. Similarly, Jacobsen et al. species, using at least 3 different primers, [107] monitored the survival of probiotic and digital analysis of the combined pat- strains of Lb. rhamnosus, Lb. reuteri and tern for all the primers. These techniques Lb. delbrueckii ssp. lactis in faeces. were found to discriminate between the Bouton et al. [22] used a PCR-based strains at a much finer taxonomic level method and PFGE for typing and monitor- than SDS-PAGE fingerprinting, although ing homofermentative lactobacilli during their reproducibility was found to be a mat- Comté cheese ripening. Isolates, which ter of debate [75]. exhibited unique patterns by RAPD or 292 V. Coeuret et al.

Table VIII. PFGE Restriction enzymes and migration conditions used for lactobacilli.

Lactobacilli species Restriction enzyme Running conditions Ref. –1 Lb. acidophilus SmaI, ApaI (0.5–15) s, 24 h, 6 V·cm [157] (5–25) s, 24 h, 6 V·cm–1 (0.3s: 1h; 0.5s: 1h; 0.7s: Lb. sake SmaI, AvrII 1h; 2s: 5h; 4s: 6h) [19] 200 V (1–5) s, 20 h, 6 V·cm–1 Lb. casei/rhamnosus SmaI, BglI, SfiI (1–20) s, 24 h, 4.5 V·cm–1 [71] Lb. paracasei (40–80) s, 20 h, 6V·cm–1 Lactobacilli ApaI (1–12) s, 17 h, 5 V·cm–1 [131] –1 Lb. sanfranciscensis ApaI, SacII, SgrAI (0.5–8) s, 17 h, 6 V·cm [201] (3–40) s, 24 h, 6 V·cm–1 Lb. acidophilus (2 s, 6 h, 350 mA / 5 s, 6 h, 370 mA / 10 s, Lb. gallinarum 4h, 390mA / 15s, 4h, 410mA / 30s, 4h Lb. gasseri ApaI, SmaI, SgrAI 430 mA / 60 s, 3 h, 450 mA) [158] Lb. helveticus Lb. helveticus SmaI, SgrAI (1–13) s, 20 h, 200 V [121] ApaI, SmaI, NotI, SfiI, 5 s, 16 h, 140 mA Lb. plantarum SwaI 10 s, 16 h, 140 mA [50] Lb. casei Lb. rhamnosus NotI, SfiI (1–15) s, 22 h, 5 V·cm–1 [184] Lb. zeae Lb. delbrueckii ssp. bulgaricus NotI [20] Lb.rhamnosus ApaI Lb. reuteri (1–15) s, 20 h, 200 V [107] Lb. delbrueckii ssp. lactis Lb. helveticus Lb. delbrueckii SgrA1, XhoI (2–13) s, 22 h, 200 V [22]

REP-PCR, were distinguishable by PFGE, from Lactobacillus isolated from sour- but some strains which were distinguisha- dough has been reported [153]. ble by RAPD or REP-PCR were related by The plasmid profile of a particular LAB PFGE; these discrepancies were explained or Lactobacillus strain may be used as an by the different exploration of DNA poly- identification marker. It has been per- morphism (the whole DNA chromosome formed on Lactobacillus strains [63, 64, for PFGE, and region amplified by primers 136]. However, a plasmid DNA replicates for RAPD and the REP-PCR). The use of independently from the chromosome, and second restriction enzymes would cer- plasmidic genes are usually more unstable tainly be useful in this case. than chromosomal gene function [65]. Although the plasmid profile of a strain Plasmid profiling remains stable under laboratory condi- Plasmid-based typing methods are use- tions, plasmids may be lost during fermen- ful, particularly if large numbers of strains tation in unfavourable growth conditions, are to be examined. Plasmids are not or may undergo rearrangements by conju- evenly distributed among the various iso- gative transfer. Therefore, plasmid profiling lates belonging to different species of techniques have several potential disad- Lactobacillus. A rapid mechanical lysis vantages, such as the ability of a strain to method for the routine analysis of plasmids gain, lose or modify its plasmids [192], and Identification of dairy lactobacilli 293 there is often no correlation between plas- used in the cheese industry. Wild strains of mid content and species identification lactobacilli isolated from raw milk cheeses [202]. from Normandy were well identified. The spectral database makes it possible to iden- Phage-related DNA polymorphism tify new strains, with a high percentage of good results: 100% at the genus and spe- Bacterial lysogeny, that is, the presence cies level for collection strains; and 100% of prophages as genetic elements in the at the genus level and 69% at the species bacterial chromosome, is common in level for wild isolates previously identified lactobacilli, especially in the Lb. casei by RAPD and the phenotypic method. The group [72]. Brandt et al. [24] investigated results obtained were as reliable as those whether further intraspecies DNA poly- obtained by genomic methods such as morphism could be revealed by screening RAPD analysis [2]. for the presence and distribution of phage- Another global technique, pyrolysis- related DNA sequences among the strains mass spectrometry [125], is also very of a single bacterial species. He studied 11 promising but has not yet been applied to Lb. rhamnosus strains used as starter and lactobacilli. probiotic cultures in the Finnish dairy industry. Six different PCR product patterns were obtained by amplification with primer pairs derived from the nucleotide 3. CHARACTERISATION sequence of a HindIII fragment of the AND IDENTIFICATION phage Lc-Nu genome. Phage Lc-Nu DNA- OF LACTOBACILLI derived PCR was found to be an effective BY CULTURE-INDEPENDENT tool for detecting polymorphism in the Lb. METHODS rhamnosus strains. Culture-independent methods involve 2.2.4. Global techniques extraction of nucleic acids (DNA or RNA) from raw samples and the use of probes for The recent development of whole bacte- hybridisation and primers for denaturing rium analysis by FTIR (Fourier transform gradient gel electrophoresis (DGGE), temper- infrared spectroscopy) is of great potential ature gradient gel electrophoresis (TGGE) for rapid identification of lactobacilli [1, 2, and single strand conformation polymor- 48, 49, 128]. Bacterial spectra are usually phism (SSCP). recorded in the mid-infrared. They are spe- These techniques give a picture of the cific to a bacterial strain and show the populations present in a complex matrix, vibrational characteristics of the whole cel- bypassing problems related to injured, and lular components: fatty acids, intracellular viable but non-cultivable bacteria. and membrane proteins, polysaccharides and nucleic acids. The statistical treatment of spectral data makes it possible to dis- 3.1. Hybridisation criminate between different genera, species and strains. The reproducibility problems Ehrmann et al. [68] developed a tech- initially encountered have been resolved, nique facilitating the direct identification resulting in standardised conditions for cell of LAB, without prior culture, in cheese, growth and sample preparation. The princi- yoghurt, sausages, sauerkraut and sour- pal advantage of this technique, as pointed dough and based on a reverse dot-blot out by almost all authors, is its simplicity assay. Oligonucleotide probes specific to the with respect to genome analysis. Amiel various Lactobacillus species are extended et al. [1] established libraries of species by adding a polythymidine phosphate tail. 294 V. Coeuret et al.

Table IX. DGGE/ TGGE primers.

Primers 5’–3’ Specificity Ref. [178]

HDA1/ HDA2 (GCclamp)ACTCCTACGGGAGGCAGCAGT/GTATTACCGCGGCTGCTGGCA V2-V3/16S [194] {[144] V3F/ V3R (GCclamp)CCTACGGGAGGCAGCA/ ATTACCGCGGCTGCTGG V3/16S rDNA [44] (GCclamp)ACGGGGGGACTCCTACGGGAGGCAGCAG/ gc338f/ 518r TCTGTGATGCCCTTAGATGTTCTGGG V3/16S rDNA [5] P1/P2 (GCclamp)TACGGGAGGCAGCAG/ ATTACCGCGGCTGCTGG V3/16S rDNA [39] Lac1/Lac2GC AGCAGTAGGGAATCTTCCA/(GCclamp)ATTYCACCGCTACACATG 16S rDNA [195] Unnamed CGCCGGGGGCGCGCCCCGGGCGGG/ GCGGGGGCACGGGGGG 16S rDNA [150]

These extended oligonucleotides bind nat- been developed for the analysis of micro- urally to filter membranes and serve as bial communities without culture, by the species-specific capture probes for the sequence-specific separation of amplified labelled, PCR-amplified rRNA gene frag- 16S rDNA fragments. Separation is based ment [68]. on the lower electrophoretic mobility of a A simple, rapid method for whole-cell partially melted double-stranded DNA hybridisation with fluorescent-labeled rRNA- molecule in polyacrylamide gels contain- targeted peptide nucleic acid (PNA) probes ing a linear gradient of DNA denaturants, a was recently developed for the detection mixture of urea and formamide (DGGE), and identification of thermophilic Lacto- or subjected to a linear temperature gradi- bacillus cells growing in milk or present in ent (TTGE). The melting of fragments pro- industrial starter culture [129]. The proto- ceeds in discrete melting domain stretches col involves filtration of the samples, and of base pairs with an identical melting tem- epifluorescence microscopy is used for perature. Once the domain with the lowest detection. Its detection limit is 104 to 106 melting temperature reaches its melting cells per mL, and specific oligonucleotides temperature (Tm), in the denaturing or are available for Lb. delbrueckii, Lb. helve- temperature gradient gel, the molecule ticus, Lb. pentosus and Lb. plantarum. The undergoes a transition from a helical to a rRNA-targeted oligonucleotide probes partially melted structure, and its migration currently available for the identification of stops. Optimal resolution occurs when LAB are listed in Table III. Various types amplicons are not completely denatured of assay can be used. In dot-blot assays the and when the region to be screened is in the target nucleic acid must be extracted and lowest melting domain. This is achieved immobilised on a membrane. Such assays by adding a 30-40 bp GC-rich clamp to one have been used for the simultaneous iden- of the PCR primers (Tab. IX). This results tification of various lactobacilli, such as in sequence variants of particular frag- Lb. curvatus, Lb. sake, Lb. pentosus, Lb. ments ceasing to migrate at different posi- plantarum, Lb. delbrueckii and Lb. helveti- tions in the denaturing gradient, facilitat- cus [97, 143, 192]. ing their effective separation by TGGE or DGGE. 3.2. PCR-DGGE, PCR-TGGE The members of the bacterial community are often amplified using primers cor- Methods such as denaturing gradient gel responding to the 16 S rDNA sequence [5, electrophoresis (DGGE) and temperature 39, 93, 144, 168, 186, 194, 195]. Species gradient gel electrophoresis (TGGE) have can then be distinguished by comparing the Identification of dairy lactobacilli 295 migration distance of the PCR amplicons bacterial communities, but screening in in gels with those of reference strains [3]. new primers in a more discriminating area These techniques have recently been than V3 16S rDNA is necessary for lacto- used in the evaluation of microbial diver- bacilli species. sity, particularly the diversity of lactoba- Randazzo et al. [150] obtained DGGE cilli in cheeses [44, 144, 150], sausages profiles derived from PCR and RT-PCR [39], starch fermentation [5], malt whisky (reverse transcriptase-PCR) of DNA and fermentation [186], beer [183], faeces RNA, and compared them in order to [168, 178, 195], and the gastrointestinal determine the expression level of the 16S tract [194], and for the identification of rRNA genes of the most predominant bacte- Lactobacillus species [189]. Coppola et al. ria during the manufacturing of Ragusano [44] used PCR 16S-23S rDNA spacer pol- cheese. The evolution of the Lactobacillus ymorphism and PCR-DGGE analysis of community during the manufacturing and the V3 region of 16S rDNA to study the ripening process was reflected in the unsta- microbial diversity of unripened Pasta ble DGGE profiles generated by using Filata cheeses. The number of bands in the Lactobacillus-specific primers, which tar- PCR profiles obtained made it possible to get all members of the Lactobacillus distinguish between industrial, cottage group, but include Leuconostoc and Pedio- industry and traditional dairy products. coccus spp. too. Concurrently microbial Ogier et al. [144] identified by PCR- enumeration on different media was done, TGGE of the 16S rDNA V3 regions bacte- and after cultivation, isolates were identi- rial species (Lactobacillus, Lactococcus, fied by both classical phenotypic methods Leuconostoc, Enterococcus, Pediococcus, and 16S rDNA sequence analysis. Lb. del- Streptococcus and Staphylococcus) present brueckii, that was dominant as indicated by in home-made or commercial dairy prod- DGGE, was never isolated on the selective ucts. V3-TGGE sequences differentiate media. between bacteria belonging to the different Tannock et al. [178] assessed the impact genera. However, V3-TGGE did not distin- of probiotic consumption on the intestinal guish between members of the Lactobacillus microflora by monitoring the faecal com- casei group (Lb. casei, Lb. paracasei and munity by FISH (fluorescence in situ Lb. rhamnosus), or members of the Lacto- hybridisation) and DGGE. Heilig et al. bacillus acidophilus group (Lb. galli- [93] used DGGE to study the stability of narum, Lb. crispatus, Lb. amylovorus, Lb. the bacterial community of the gastrointes- acidophilus and closely Lb. helveticus). tinal tract in various age groups, over vari- Lb. pentosus and Lb. plantarum or Lb. ous time periods, and successive changes johnsonii and Lb. gasseri have similar V3 in the Lactobacillus community were sequences and co-migrate. Only Lb. reuteri, observed. They also assessed the specifi- Lb. brevis, Lb. fermentum, Lb. delbrueckii city of the PCR and DGGE approach for ssp. bulgaricus, and Lb. delbrueckii ssp. studying the retention in faecal samples of lactis can be easily distinguished. Cheeses a Lactobacillus strain administered during with similar production procedures (e.g. a clinical trial. Brie and Camembert cheeses, or Emmen- Van Beek and Priest [186] adopted a tal and Comté…) produced common polyphasic approach, using light and elec- TGGE bands. TGGE provides a descrip- tron microscopy and denaturing gradient tion of the dominant bacterial species in a gel electrophoresis (DGGE) of PCR-ampli- complex ecosystem, but bacterial minority fied fragments of 16S ribosomal DNA to (less than 1%) cannot be detected. So, monitor the development of the lactic acid TGGE seems to be an excellent molecular bacterial community during malt whisky tool to analyse diversity within complex fermentation. Their results revealed that 296 V. Coeuret et al. culture-dependent methods underestimated genera or description of new species. This bacterial diversity and demonstrated the leads to a problematic genus characterisation presence of novel lactobacilli and other by phenotypic tests and to an increasing taxa in malt whisky during fermentation. use of classical culture-based molecular These new techniques have the advan- methods. New molecular techniques for tage of facilitating the direct study and microbial community analysis that do not analysis of population dynamics within require isolation of the microorganisms are complex microbial systems. These meth- very promising. They provide a complemen- ods are only just beginning to be applied to tary picture of the population obtained using ecological studies of cheese microflora. culture-based techniques when applied to the analysis of milks and dairy products. 3.3. PCR-SSCP However, these molecular approaches have several limitations, including the design of SSCP (single strand conformation poly- adequate primers, and the possibility that morphism) analysis for molecular identifi- DNA isolation, amplification and cloning cation in microbial ecosystems is based on might be biased by certain strains and 16S rDNA sequences. No culture is sequences. There is also a dependence on required. SSCP detects sequence varia- the detection threshold and on the number tions between DNA fragments, usually of lactobacilli, unfortunately low in high- amplified by PCR from variable regions of the 16S rRNA gene and uses neutral, non- quality raw milks. Nevertheless, these denaturing polyacrylamide gels. Short methods provide an overview of the diver- double-stranded DNA fragments are first sity of microorganisms present in a partic- generated using standard PCR protocols. ular sample. They provide qualitative and Single-stranded DNA is created by com- possibly semi-quantitative information, bining a small aliquot of PCR product with which should be complemented by quanti- an equal volume of formamide (80–95%), tative PCR to obtain results as close as pos- and then it is heat denatured. The two com- sible to reality. plementary strands of DNA will migrate We intended to produce a guide for the differently and will therefore separate during reader, covering pertinent techniques used gel electrophoresis. This fingerprinting for a polyphasic analysis of lactobacilli. method characterises microbial diversity, This guide was not produced, since pro- by comparing closed microflora [84]. SSCP posing a technique for a specified use is not analysis has been adapted for the rapid completely reliable. In fact, the choice of identification of both Gram-positive and the technique to use is variable depending Gram-negative bacteria to genus and spe- on: cies level. Analysis of the microflora of – the level of discrimination required; AOC Salers cheese resulted in the co- – the type of product and matrix, the nature elution of Lc. lactis, Lb. plantarum, Lb. pentosus and St. thermophilus, all of which of undesirable organisms and the diver- were associated with the main peak [84]. sity of the lactobacilli present; – the amount of time available; 4. CONCLUSION – the available staff and equipment; It is widely recognised that the identifi- – the number of strains and isolates to be cation of lactobacilli to species or strain studied. level on the basis of physiological and bio- Furthermore, numerous techniques, cul- chemical criteria is very ambiguous and ture-dependent or culture-independent, are complicated. Numerous taxonomic changes based on the use of probes and primers. For have been observed in the Lactobacillus these techniques the discrimination level genus as qualification of old species in new depends on the existence or not of the Identification of dairy lactobacilli 297 probes and primers at the taxonomic level [2] Amiel C., Mariey L., Denis C., Pichon P., desired. To date we are very far from hav- Travert J., FTIR spectroscopy and taxonomic purpose: Contribution to the classifica- ing specific primers and probes for the 88 tion of lactic acid bacteria, Lait 81 (2001) lactobacilli species, and regarding those 249–255. which have been designed, their specificity [3] Ampe F., Design and evaluation of a Lacto- and validity should be checked one by one bacillus manihotivorans species-specific with closed genera, species or strains. rRNA-targeted hybridization probe and its application to the study of sour cassava fer- Another problem results from the given list mentation, Appl. Environ. Microbiol. 66 of 88 lactobacilli species since it is not an (2000) 2224–2226. official list (it does not exist) and thus to [4] Ampe F., ben Omar N., Moizan C., Wacher bypass possible misidentification all C., Guyot J.P., Polyphasic study of the spa- probes and primers should be validated tial distribution of microorganisms in Mexi- can pozol, a fermented maize dough, against the same reference strain at the demonstrates the need for cultivation-inde- beginning to ensure their common specifi- pendent methods to investigate traditional city. Moreover, all the techniques men- fermentations, Appl. Environ. Microbiol. tioned in this review have not been applied 65 (1999) 5464–5473. [5] Ampe F., Sirvent A., Zakhia N., Dynamics to the lactobacilli using the same objec- of the microbial community responsible for tives. The genus primer designed by traditional sour cassava starch fermentation Dubernet et al. [62], has been used for PCR studied by denaturing gradient gel electro- and PCR-TGGE, but not for hybridisation, phoresis and quantitative rRNA hybridization, Int. J. Food Microbiol. 65 (2001) but it is clear that it could be used. The dif- 45–54. ficulty of choosing a technique that has [6] Andrighetto C., De Dea P., Lombardi A., good dicrimination power depends not Neviani E., Rossetti L., Giraffa G., Molecu- only on the techniques but also on the spe- lar identification and cluster analysis of cies or strains. Results also depend on the homofermentative thermophilic lactobacilli isolated from dairy products, Res. Micro- quality and exhaustivity of a database. biol. 149 (1998) 631–643. Very good results at genus, species and [7] Antonsson M., Ardo Y., Molin G., A com- strain level could be obtained by using parison between the microflora of Herrgard FTIR, but if the database is not complete cheese from three different dairies, Int. (not enough strains of different species, or Dairy J. 11 (2001) 285–291. of different origin), results will not be as [8] Antunes A., Rainey F.A., Nobre M.F., Schumann P., Ferreira A.M., Ramos A., good as they could be. Finally, only a few Santos H., da Costa M.S., Leuconostoc limited techniques can be applied with a ficulneum sp. nov., a novel lactic acid bac- high degree of confidence although they terium isolated from a ripe fig, and reclassi- are dependent on database robustness: fication of Lactobacillus fructosus as Leu- conostoc fructosum comb. nov., Int. J. Syst. sequencing to identify at genus and species Evol. Microbiol. 52 (2002) 647–655. level, and sequencing or pulsed field gel [9] Arhné S., Molin G., Stahl S., Plasmids in electrophoresis to discriminate strains. Lactobacillus strains isolated from meat In conclusion, analysis of lactobacilli in and meat products, Syst. Appl. Microbiol. 11 (1989) 320–325. cheeses and other dairy products is very com- [10] Arizcun C., Barcina Y., Torr P., Identifica- plicated and the use of different techniques, tion of lactic acid bacteria isolated from especially molecular-based phenotypic or Roncal and Idiazabal cheeses, Lait 77 genomic techniques, is recommended. (1997) 729–736. [11] Back W., BohaK I., Ehrmann M., Ludwig REFERENCES W., Pot B., Kersters K., Schleifer K.H., Lactobacillus perolens sp. nov., a soft drink [1] Amiel C., Mariey L., Curk-Daubié M.C., spoilage bacterium, Syst. Appl. Microbiol. Pichon P., Travert J., Potentiality of Fourier 22 (1999) 354–359. Transform Infrared Spectroscopy (FTIR) for [12] Baruzzi F., Morea M., Matarante A., Coc- discrimination and identification of dairy concelli P.S., Changes in the Lactobacillus lactic acid bacteria, Lait 80 (2000) 445–459. community during Ricotta forte cheese 298 V. Coeuret et al.

natural fermentation, J. Appl. Microbiol. 89 [24] Brandt K., Tilsala-Timisjarvi A., Alatos- (2000) 807–814. sava T., Phage-related DNA polymorphism [13] Beresford T., Pelaez C., Jimeno J., Nature in dairy and probiotic Lactobacillus, and growth of non starter flora, in: Euro- Micron 32 (2001) 59–65. pean Commission, Quality and microbio- [25] Bringel F., Curk M.C., Hubert J.C., Charac- logy of traditional and raw milk cheeses, terization of lactobacilli by southern-type Dijon, France, 1998, pp. 225–238. hybridization with a Lactobacillus plantarum pyrDFE probe, Int. J. Syst. Bacteriol. [14] Beresford T.P., Fitzsimons N.A., Brennan 46 (1996) 588–594. N.L., Cogan T.M., Recent advances in cheese microbiology, Int. Dairy J. 11 [26] Bruce J., Automated system rapidly identi- (2001) 259–274. fies and characterises microorganisms in food, Food Technol. 50 (1996) 77–81. [15] Berthier F., Ehrlich S.D., Rapid species [27] Bunte C., Hertel C., Hammes W.P., Moni- identification within two groups of closely toring and survival of Lactobacillus para- related lactobacilli using PCR primers that casei LTH 2579 in food and the human target the 16S/23S rRNA spacer region, intestinal tract, Syst. Appl. Microbiol. 23 FEMS Microbiol. Lett. 161 (1998) 97–106. (2000) 260–266. [16] Berthier F., Ehrlich S.D., Genetic diversity [28] Calicchia M.L., Wang C.I.E., Nomura T., within Lactobacillus sakei and Lactobacil- Yotsuzuka F., Osato D.W., Selective enu- lus curvatus and design of PCR primers for meration of Bifidobacterium bifidum, Ente- its detection using randomly amplified rococcus faecium, and Streptomycin-resis- polymorphic DNA, Int. J. Syst. Bacteriol. tant Lactobacillus acidophilus from a 49 (1999) 997–1007. mixed probiotic product, J. Food Prot. 56 [17] Berthier F., Beuvier E., Dasen A., Grappin (1993) 954–957. R., Origin and diversity of mesophilic lac- [29] Camaschella P., Pirovano F., Mognot O., tobacilli in Comté cheese, as revealed by Sozzi T., Method for the detection and enu- PCR with repetitive and species-specific meration of different thermophilic lactic primers, Int. Dairy J. 11 (2001) 293–305. cultures and bifidobacteria utilising only one culture medium, in: Amgar A. (Ed.), [18] Birollo G.A., Reinheimer J.A., Vinderola Food Safety’94, Laval France, 1994, p. 335. C.G., Viability of lactic acid microflora in [30] Chagnaud P., Machinis K., Coutte L.A., different types of yoghurt, Food Res. Int. 33 Marecat A., Mercenier A., Rapid PCR- (2000) 799–805. based procedure to identify lactic acid bac- [19] Bjorkroth J., Ridell J., Korkeala H., Charac- teria: application to six common Lactoba- terization of Lactobacillus sake strains cillus species, J. Microbiol. Meth. 44 associating with production of ropy slime (2001) 139–148. by randomly amplified polymorphic DNA [31] Chamba J.F., Emmental cheese: a complex (RAPD) and pulsed-field gel electrophore- microbial ecosystem. Consequences on sis (PFGE) patterns, Int. J. Food Microbiol. selection and use of starters, Sci. Aliments 31 (1996) 59–68. 20 (2000) 37–54. [20] Blaiotta B.G., Simeoli G.M., Andolfi R., [32] Charteris W.P., Kelly P.M., Morelli L., Villani F., Coppola S., Monitoring lactic Collins J.K., Selective detection, enumera- acid bacteria strains during “Cacioricotta” tion and identification of potentially probio- cheese production by restriction endonu- tic Lactobacillus and Bifidobacterium spe- clease analysis and pulsed-field gel electro- cies in mixed bacterial populations, Int. J. phoresis, J. Dairy Res. 68 (2001) 139–144. Food Microbiol. 35 (1997) 1–27. [21] Bouton Y., Guyot P., Grappin R., Preliminary [33] Chen H., Lim C.K., Lee Y.K., Chan Y.N., characterization of microflora of Comté Comparative analysis of the genes encoding cheese, J. Appl. Microbiol. 85 (1998) 123–131. 23S-5S rRNA intergenic spacer regions of Lactobacillus casei-related strains, Int. J. [22] Bouton Y., Guyot P., Beuvier E., Tailliez Evol. Microbiol. 50 (2000) 471–478. P., Grappin R., Use of PCR-based methods [34] Choisy C., Desmazeaud M., Guéguen M., and PFGE for typing and monitoring homo- Lenoir J., Schmidt J.L., Tourneur C., Les fermentative lactobacilli during Comté phénomènes microbiens, in: Eck A., Gillis cheese ripening, Int. J. Food. Microbiol. 76 J.C. (Eds.), Le fromage, Lavoisier Tec & (2002) 27–38. Doc, Paris, France, 1997, pp. 377–446. [23] Bracquart P., An agar medium for the diffe- [35] Chopard M.A., Schmitt M., Perreard E., rential of Streptococcus thermophilus and Chamba J.F., Qualitative aspect of proteolytic Lactobacillus bulgaricus in yoghurt, J. activity of thermophilic lactobacilli using in Appl. Bacteriol. 51 (1981) 303–305. Swiss cheeses, Lait 81 (2001) 183–194. Identification of dairy lactobacilli 299

[36] Cocconcelli P.S., Porro D., Galandini S., [46] Crow V., Curry B., Hayes M., The ecology Senini L., Development of RAPD protocol of non-starter lactic acid bacteria (NSLAB) for typing of strains of lactic acid bacteria and their use as adjuncts in New Zealand and enterococci, Lett. Appl. Microbiol. 21 Cheddar, Int. Dairy J. 11 (2001) 275–283. (1995) 376–379. [47] Curk M.C., Caractérisation des espèces [37] Cocconcelli P.S., Senini L., Bottazzi V., brassicoles du genre Lactobacillus, Thèse Study of the lactic microflora of Toma de Doctorat, Université Louis Pasteur, cheese: use of rRNA targeted oligonucleo- Strasbourg, France, 1994. tides for the identification of isolated [48] Curk M.C., Peladan F., Hubert J.C., Fourier strains, Ann. Microbiol. Enzymol. 46 transform infrared spectroscopy for identi- (1996) 21–28. fying Lactobacillus species, FEMS Micro- [38] Cocconcelli P.S., Parisi M.G., Senini L., biol. Lett. 123 (1994) 241–248. Bottazzi V., Use of RAPD and 16S rDNA [49] Curk M.C., Amiel C., Denis C., Reyrolle J., sequencing for the study of Lactobacillus Pichon P., Travert J., Discrimination de population dynamics in natural whey cul- souches de bactéries lactiques d’intérêt ture, Lett. Appl. Microbiol. 25 (1997) 8–12. laitier par spectroscopie infrarouge à trans- [39] Cocolin L., Manzano M., Cantoni C., Comi formée de Fourier (IRTF), in: ADRIA-Nor- G., Development of a rapid method for the mandie, les bactéries lactiques, Lactic 97, identification of Lactobacillus spp. isolated Caen, 10–12 Septembre, France, 1997, from naturally fermented Italian sausages 420–421. using a polymerase chain reaction-tempera- [50] Daniel P., Sizing of the Lactobacillus plan- ture gradient gel electrophoresis, Lett. tarum genome and other lactic acid bacteria Appl. Microbiol. 30 (2000) 126–129. species by transverse alternating field elec- [40] Collins M.D., Rodrigues U., Ash C., trophoresis, Curr. Microbiol. 30 (1995) Aguirre M., Farrow J.A.E., Martinez- 243–246. Murcia A., Phillips B.A., Williams A.M., [51] Daud Khaled A.K., Neilan B.A., Henriks- Wallbanks S., Phylogenetic analysis of the son A., Conway P.L., Identification and genus Lactobacillus and related lactic acid phylogenetic analysis of Lactobacillus bacteria as determined by reverse transcrip- using multiplex RAPD-PCR, FEMS Micro- tase sequencing of 16S rRNA, FEMS biol. Lett. 153 (1997) 191–197. Microbiol. Lett. 77 (1991) 5–12. [41] Collins M.D., Samelis J., Metaxopoulos J., [52] Dave R.I., Shah N.P., Evaluation of media Wallbanks S., Taxonomic studies on some for selective enumeration of Streptococcus thermophilus Lactobacillus delbrueckii leuconostoc-like organisms from fermented , bulgaricus Lactobacillus acidophilus sausages: description of a new genus Weis- ssp. , , and bifidobacteria, J. Dairy Sci. 79 (1996) sella for the Leuconostoc paramesenteroi- 1529–1536. des group of species, J. Appl. Bacteriol. 75 (1993) 595–603. [53] De Angelis M., Corsetti A., Tosti N., Rossi [42] Coppola R., Nanni M., Iorizzo M., Sorren- J., Corbo M.R., Gobbetti M., Characteriza- tino A., Sorrentino E., Grazia L., Survey of tion of non-starter lactic acid bacteria from lactic acid bacteria during the advanced sta- Italian ewe cheeses based on phenotypic, ges of the ripening of Parmigiano Reggiano genotypic, and cell wall protein analyses, Appl. cheese, J. Dairy Res. 64 (1997) 305–310. Environ. Microbiol. 67 (2001) 2011–2020. [43] Coppola R., Nanni M., Iorizzo M., Sorrentino [54] De Man J.C., Rogosa M., Sharpe M.E., A A., Sorrentino E., Chiavari C., Grazia L., medium for cultivation of lactobacilli, J. Microbiological characteristics of Parmi- Appl. Bacteriol. 23 (1960) 130–135. giano Reggiano cheese during the cheese- [55] Dellaglio F., Dicks L.M.T., Du Toit M., making and the first months of the ripening, Torriani S., Lactic acid bacteria in ensiled Lait 80 (2000) 479–490. high-moisture corn grain: physiological and [44] Coppola R., Nanni M., Succi M., Sorrentino genetic characterization, Syst. Appl. Micro- A., Iorizzo M., Chiavari C., Grazia L., Enu- biol. 5 (1991) 534–544. meration of thermophilic lactic acid bacte- [56] Dellaglio F., Lombardi A., New develop- ria in ripened cheeses manufactured from ments in the identification of non starter raw milk, Milchwissenschaft 56 (2001) microorganisms in cheese, in: European 140–142. Commission, Quality and microbiology of [45] Corbo M.R., Albenzio M., De Angelis M., traditional and raw milk cheeses, Dijon, Sevi A., Gobbetti M., Microbiological and France, 1998, pp. 53–63. biochemical properties of Canestrato Pugliese [57] Dellaglio F., Felis G.E., Torriani S., The hard cheese supplemented with bifidobac- status of the species Lactobacillus casei teria, J. Dairy Sci. 84 (2001) 551–561. (Orla-Jensen 1916) Hansen and Lessel 300 V. Coeuret et al.

1971 and Lactobacillus paracasei Collins method for the direct identification of lactic et al. 1989. Request for an Opinion, Int. J. acid bacteria in fermented food, FEMS Syst. Evol. Microbiol. 52 (2002) 285–287. Microbiol. Lett. 117 (1994) 143–149. [58] Demarigny Y., Beuvier E., Dasen A., [69] El Soda M., Madkor S.A., Tong P.S., Duboz G., Influence of raw milk microflora Adjunct cultures: recent developments and on the characteristics of Swiss-type chee- potential significance to the cheese indus- ses. I. Evolution of microflora during ripe- try, J. Dairy Sci. 83 (2000) 609–619. ning and characterization of facultatively [70] Eliskases-Lechner F., Ginzinger W., Rohm heterofermentative lactobacilli, Lait 76 H., Tschager E., Raw milk flora affects (1996) 371–387. composition and quality of Bergkase. 1. [59] Demarigny Y., Beuvier E., Buchin S., Microbiology and fermentation com- Pochet S., Grappin R., Influence of raw pounds, Lait 79 (1999) 385–396. milk microflora on the characteristics of [71] Ferrero M., Cesena C., Morelli L., Scolari Swiss-type cheeses. II. Biochemical and G., Vescovo M., Molecular characteriza- sensory characteristics, Lait 77 (1997) tion of Lactobacillus casei strains, FEMS 151–167. Microbiol. Lett. 140 (1996) 215–219. [60] Drake M., Small C.L., Spence K.D., Swanson [72] Forsman P., Tanskanen J., Alatossava T., B.G., Rapid detection and identification of Structural similarity and genetic homology Lactobacillus spp. in dairy Products by between Lactobacillus casei bacteriopha- using the polymerase chain reaction, J. ges isolated in Japan and in Finland, Biosci. Food Prot. 59 (1996) 1031–1036. Biotechnol. Biochem. 57 (1993) 2042–2048. [61] Du Plessis E.M., Dicks L.M., Evaluation of [73] Fox P.F., Mc Sweeney P.L.H., Lynch C.M., random amplified polymorphic DNA Significance on nonstarter lactic acid bacte- (RAPD)-PCR as a method to differentiate ria in Cheddar cheese, Aust. J. Dairy Tech- Lactobacillus acidophilus, Lactobacillus nol. 53 (1998) 83–89. crispatus, Lactobacillus amylovorus, Lac- tobacillus gallinarum, Lactobacillus gas- [74] Gagnaire V., Thierry A., Léonil J., Propioni- seri, and Lactobacillus johnsonii, Curr. bacteria and facultatively heterofermentative Microbiol. 31 (1995) 114–118. lactobacilli weakly contribute to secondary proteolysis of Emmental cheese, Lait 81 [62] Dubernet S., Desmasures N., Guéguen M., (2001) 339–353. A PCR-based method for identification of lactobacilli at genus level, FEMS Micro- [75] Gancheva A., Pot B., Vanhonacker K., Hoste biol. Lett. 214 (2002) 271. B., Kersters K., A polyphasic approach towards the identification of strains belon- [63] Duffner F., O’Connell M., A comparative ging to Lactobacillus acidophilus and rela- evaluation of plasmid profiling and riboty- ted species, Syst. Appl. Microbiol. 22 ping in the analysis of Lactobacillus planta- (1999) 573–585. rum strain heterogeneity in silage, J. Appl. Bacteriol. 78 (1995) 20–27. [76] Gardiner G., Ross R.P., Collins J.K., Fitzgerald G., Stanton C., Development of a [64] Dykes G.A., Burgess A.Y., von Holy A., probiotic Cheddar cheese containing Plasmid profiles of lactic acid bacteria asso- human-derived Lactobacillus paracasei ciated with vacuum-packaged Vienna sau- strains, Appl. Environ. Microbiol. 64 (1998) sage manufactured and spoilage, Lett. 2192–2199. Appl. Microbiol. 17 (1993) 182–184. [77] Gasser F., Electrophoretic characterization [65] Dykes G.A., von Holy A., Strain typing in of lactic dehydrogenases in the genus Lac- the genus Lactobacillus, Lett. Appl. Micro- tobacillus, J. Gen. Microbiol. 62 (1970) biol. 19 (1994) 63–66. 223–239. [66] Dykes G.A., Cloete T.E., von Holy A., [78] Ghoddusi H.B., Robinson R.K., Enumera- Taxonomy of lactic acid bacteria associated tion of starter cultures in fermented milks, J. with vacuum-packaged processed meat Dairy Res. 63 (1996) 151–158. spoilage by multivariate analysis of cellular [79] Gilson E., Clement J.M., Brutlag D., fatty acids, Int. J. Food Microbiol. 28 Hofnung M., A family of dispersed repeti- (1995) 89–100. tive extragenic palindromic DNA sequences [67] Ehrmann M., Ludwig W., Schleifer K.H., in E. coli, EMBO J. 3 (1984) 1417–1421. Species specific oligonucleotide probe for [80] Giraffa G., De Vecchi P., Rossetti L., Note: the identification of Streptococcus thermo- identification of Lactobacillus delbrueckii philus., Syst. Appl. Microbiol. 15 (1992) subspecies bulgaricus and subspecies lactis 453–455. dairy isolates by amplified rDNA restriction [68] Ehrmann M., Ludwig W., Schleifer K.H., analysis, J. Appl. Microbiol. 85 (1998) Reverse dot blot hybridization: a useful 918–924. Identification of dairy lactobacilli 301

[81] Giraffa G., De Vecchi P., Rossi P., Nicastro lates of Lactobacillus strains to be used as G., Fortina M.G., Genotypic heterogeneity starter cultures in dairy fermentation, Int. J. among Lactobacillus helveticus strains Food Microbiol. 59 (2000) 19–27. isolated from natural cheese starters, J. [93] Heilig H.G.H.J., Zoetendal E.G., Vaughan Appl. Microbiol. 85 (1998) 411–416. E.E., Marteau P., Akkermans A.D.L., de [82] Giraffa G., Gatti M., Rossetti L., Senini L., Vos W.M., Molecular diversity of Lactoba- Neviani E., Molecular diversity within L. cillus spp. and other lactic acid bacteria in helveticus as revealed by genotypic charac- the human intestine as determined by speci- terization, Appl. Environ. Microbiol. 65 fic amplification of 16S ribosomal DNA, (2000) 1259–1265. Appl. Environ. Microbiol. 68 (2002) [83] Gobbetti M., Morea M., Baruzzi F., Corbo 114–123. M.R., Matarante A., Considine T., Di [94] Heller K.J., Probiotic bacteria in fermented Cagno R., Guinee T., Fox P.F., Microbiolo- foods: product characteristics and starter gical, compositional, biochemical and tex- organisms, Am. J. Clin. Nutr. 73 (2001) tural characterisation of Caciocavallo 374S–379S. Puglise cheese during ripening, Int. Dairy J. [95] Hensel R., Mayr U., Stetter K.O., Kandler 12 (2002) 511–523. O., Comparative studies of lactic acid dehy- [84] Godon J.J., Duthoit F., Delbès C., Millet L., drogenases in lactic acid bacteria. I. Purifi- Montel M.C., Use of molecular fringerprint cation and kinetics of the allosteric L-lactic for the study of complex microbial ecosys- acid dehydrogenase from Lactobacillus tem. Application to AOC Salers cheese, casei ssp. casei and Lactobacillus curvatus, Lait 81 (2001) 257–262. Arch. Microbiol. 112 (1977) 81–93. [85] Gomes A.M., Malcata F.X., Development [96] Hensiek R., Krupp G., Stackebrandt E., of probiotic cheese manufactured from goat Development of diagnostic oligonucleotide milk: response surface analysis via techno- probes for four Lactobacillus species occu- logical manipulation, J. Dairy Sci. 81 (1998) ring in the intestinal tract, Syst. Appl. 1492–1507. Microbiol. 15 (1992) 123–128. [86] Grappin R., Beuvier E., Bouton Y., Pochet [97] Hertel C., Ludwig W., Obst M.,Vogel R.F., S., Advances in the biochemistry and Hammes W.P., Schleifer K.H., 23S rRNA- microbiology of Swiss-type cheeses, Lait targeted oligonucleotide probes for the 79 (1999) 1–20. rapid identification of meat lactobacilli, Syst. Appl. Microbiol. 14 (1991) 173–177. [87] Hamilton-Miller J.M.T., Shah S., Deficien- cies in microbiological quality and labelling [98] Hertel C., Ludwig W., Pot B., Kersters K., of probiotic supplements, Int. J. Food Schleifer K.H., Differentiation of lactoba- Microbiol. 72 (2002) 175–176. cilli occuring in fermented milk products by using oligonucleotides probes and electro- [88] Hamilton-Miller J.M.T., Shah S., Winkler phoretic protein profiles, Syst. Appl. J.T., Public health issues arising from Microbiol. 16 (1993) 463–467. microbiological and labelling quality of foods and supplements containing probiotic [99] Horie M., Kajikawa H.S., Toba T., Identifi- microorganims, Public Health Nutr. 2 cation of Lactobacillus crispatus by poly- (1999) 223–229. merase chain reaction targeting S-layer protein gene, Lett. Appl. Microbiol. 35 [89] Hammes W.P., Vogel R.F., The genus Lac- (2002) 57–61. tobacillus, in: Wood B.J.B., Holzapfel W.H. (Eds.), The lactic acid bacteria. The [100] Hulton C.S., Higgins C.F., Sharp P.M., genera of lactic acid bacteria, Blackie Aca- ERIC sequences: a novel family of repeti- demic, London, UK, 1995, pp. 19–54. tive elements in the genomes of Escheri- chia coli, Salmonella typhimurium and [90] Hartemink R., Domenech V.R., Rombouts other enterobacteria, Mol. Microbiol. 5 F.M., LAMVAB – A new selective medium (1991) 825–834. for the isolation of lactobacilli from faeces, [101] Hynes E., Ogier J.C., Delacroix-Buchet A., J. Microbiol. Meth. 29 (1997) 77–84. Proteolysis during ripening of miniature [91] Hayford A.E., Petersen A., Vogensen F.K., washed-curd cheeses manufactured with Jakobsen M., Use of conserved randomly different strains of starter bacteria and a amplified polymorphic DNA (RAPD) frag- Lactobacillus plantarum adjunct culture, ments and RAPD pattern for characteriza- Int. Dairy J. 11 (2001) 587–597. tion of Lactobacillus fermentum in Gha- [102] Hyytiä-Trees E., Lyhs U., Korkeala H., naian fermented maize dough, Appl. Björkroth J., Characterisation of ropy Environ. Microbiol. 65 (1999) 3213–3221. slime-producing Lactobacillus sakei using [92] Hébert E.M., Raya R.R., Tailliez P., de repetitive element sequence-based PCR, Giori G.S., Characterization of natural iso- Int. J. Food Microbiol. 50 (1999) 215–219. 302 V. Coeuret et al.

[103] IDF, Detection and enumeration of Lacto- ted milk products, Int. Dairy J. 3 (1993) bacillus acidophilus, Standard 306, Int. 277–291. Dairy Fed. Brussels, Belgium, 1995. [115] Kneifel W., Toros A., Viernstein H., Diffe- [104] IDF, Yoghurt-Enumeration of characteris- rential enumeration of silage inoculants tic micro-organisms-colony count techni- based on utilization of enzymatic activity que at 37 °C Standard 117 B, Int. Dairy Fed. and antibiotic sensitivity of bacteria, J. Brussels, Belgium, 1997. Appl. Bacteriol. 77 (1994) 42–48. [105] IDF, Dairy Stater cultures of Lactic Acid [116] Lankaputhra W.E.V., Shah N.P., Survival Bacteria (LAB), Standard 149A, Int. Dairy of Lactobacillus and Bifidobacterium spp. Fed. Brussels, Belgium, 1997. in the presence of acid and bile salts, Cult. Dairy Prod. J. 30 (1995) 2–7. [106] Isolini D., Grand M., Glättli H., Selektiv- medien zum nachweis von obligat und [117] Lankaputhra W.E.V., Shah N.P., Britz fakultativ heterofermentativen laktoba- M.L., Evaluation of media for selective zillen, Schweiz Milchw. Forschung. 19 enumeration of Lactobacillus acidophilus Bifidobacterium (1990) 57–59. and species, Food Austr. 48 (1996) 113–118. [107] Jacobsen C.N., Rosenfeldt Nielsen V., Hay- [118] Lawson P.A., Papademas P., Wacher C., ford A.E., Moller P.L., Michaelsen K.F., Falsen E., Robinson R., Collins M.D., Lac- Paerregaard A., Sandstrom B., Tvede M., tobacillus cypricasei sp. nov., isolated from Jakobsen M., Screening of probiotic activi- Halloumi cheese, Int. J. Syst. Evol. Micro- ties of forty-seven strains of Lactobacillus biol. 51 (2001) 45–49. spp. by in vitro techniques and evaluation of the colonization ability of five selected [119] Lick S., Brockmann E., Heller K.J., Identi- strains in humans, Appl. Environ. Micro- fication of Lactobacillus delbrueckii and biol. 65 (1999) 4949–4956. subspecies by hybridization probes and PCR, Syst. Appl. Microbiol. 23 (2000) [108] Janssen P., Coopman R., Huys G., Swings 251–259. J., Bleeker M., Vos P., Zabeau M., Kersters K., [120] Lopez S., Mayo B., Identification and cha- Evaluation of the DNA fingerprinting method racterization of homofermentative meso- AFLP as a new tool in bacterial taxonomy, philic Lactobacillus strains isolated from Microbiology 142 (1996) 1881–1893. artisan starter-free cheeses, Lett. Appl. [109] Johansson M.L., Molin G., Pettersson B., Microbiol. 25 (1997) 233–238. Uhlén M., Ahrné S., Characterization and [121] Lortal S., Rouault A., Guézénec S., Gautier species recognition of Lactobacillus plan- M., Lactobacillus helveticus: strain typing tarum strains by restriction fragment length and genome size estimation by pulsed field polymorphism (RFLP) of the 16S rRNA gel electrophoresis, Curr. Microbiol. 34 gene, J. Appl. Bacteriol. 79 (1995) 536–541. (1997) 180–185. [110] Jordan K.N., Cogan T.M., Identification [122] Lortal S., Valence F., Bizet C., Maubois J.L., and growth of non-starter lactic acid bacte- Electrophoretic pattern of peptidoglycan ria in Irish Cheddar cheese, Irish J. Agric. hydrolases, a new tool for bacterial species Food Res. 32 (1993) 47–55. identification: application to 10 Lactobacil- lus species, Res. Microbiol. 148 (1997) [111] Kandler O., Weiss N., Genus Lactobacillus, 461–474. in: Sneath P.H.A., Mair N.S., Sharpe M.E., Holt J.G. (Eds.), Bergey’s Manual of Syste- [123] Lucchini F., Kmet V., Cesena C., Coppi L., matics Bacteriology, Williams and Wilkins, Bottazzi V., Morelli L., Specific detection Baltimore M.D., USA, 1984, pp. 1209–1234. of a probiotic Lactobacillus strain in faecal samples by using multiplex PCR, FEMS [112] Klein G., Hack B., Hanstein S., Zimmermann Microbiol. Lett. 158 (1998) 273–278. K., Reuter G., Intra-species characterization [124] Madkor S.A., Tong P.S., El Soda M., Ripe- of clinical isolates and biotechnologically ning of Cheddar cheese with added attenua- used strains of Lactobacillus rhamnosus by ted adjunct cultures of lactobacilli, J. Dairy analysis of the total soluble cytoplasmatic Sci. 83 (2000) 1684–1691. proteins with silver staining, Int. J. Food Microbiol. 25 (1995) 263–275. [125] Manchester L.N., Toole A., Goodacre R., Characterization of Carnobacterium spe- [113] Klein G., Pack A., Bonaparte C., Reuter G., cies by pyrolysis mass spectrometry, J. Taxonomy and physiology of probiotic lac- Appl. Bacteriol. 78 (1995) 88–96. tic acid bacteria, Int. J. Food Microbiol. 41 [126] Mannu L., Comunian R., Scintu M.F., (1998) 103–125. Mesophilic lactobacilli in Fiore Sardo [114] Kneifel W., Pacher B., An X-Glu based cheese: PCR-identification and evolution agar medium for the selective enumeration during cheese ripening, Int. Dairy J. 10 of Lactobacillus acidophilus in yogurt-rela- (2000) 383–389. Identification of dairy lactobacilli 303

[127] Mannu L., Riu G., Comunian R., Fozzi [138] Muller M.R., Ehrmann M.A., Vogel R.F., M.C., Scintu M.F., A preliminary study of Multiplex PCR for the detection of Lacto- lactic acid bacteria in whey starter culture bacillus pontis and two related species in a and industrial Pecorino Sardo ewe’s milk sourdough fermentation, Appl. Environ. cheese: PCR-identification and evolution Microbiol. 66 (2000) 2113–2116. during ripening, Int. Dairy J. 12 (2002) 17–26. [139] Mullis K.B., The unusual origin of the poly- [128] Mariey L., Signolle J.P., Amiel C., Travert J., merase chain reaction, Sci. Am. 262 (1990) Discrimination, classification, identification 56-61, 64–65. of microorganisms using FTIR spectros- [140] Nagasaki H., Ito K., Matsuzaki S., Tanaka copy and chemometrics, Vib. Spectrosc. 26 S., An improved tetrazolium agar medium (2001) 151–159. for testing sugar fermentation in lactoba- [129] Matte-Tailliez O., Quénée P., Cibik R., Van cilli, Nippon Saikingaku Zasshi 47 (1992) Opstal J., Dessevre F., Firmesse O., Tailliez 511–514. P., Detection and identification of lactic [141] Nigatu A., Evaluation of numerical analy- acid bacteria in milk and industrial starter ses of RAPD and API 50 CH patterns to dif- culture with fluorescently labeled rRNA- ferentiate Lactobacillus plantarum, Lact. targeted peptide nucleic acid probes, Lait fermentum, Lact. rhamnosus, Lact. sake, 81 (2001) 237–248. Lact. parabuchneri, Lact. gallinarum, Lact. casei, Weissella minor and related taxa iso- [130] Mc Cann T., Egan T., Weber G.H., Assay lated from kocho and tef, J. Appl. Micro- procedures for commercial probiotic cultu- biol. 89 (2000) 969–978. res, J. Food Prot. 59 (1995) 41–45. [142] Nighswonger B.D., Brashears M.M., Gilli- [131] Mc Cartney A.L., Wenzhi W., Tannock land S.E., Viability of Lactobacillus acido- G.W., Molecular analysis of the composi- philus and Lactobacillus casei in fermented tion of the bifidobacterial and Lactobacillus milk products during refrigerated storage, J. microflora of humans, Appl. Environ. Dairy Sci. 79 (1996) 212–219. Microbiol. 62 (1996) 4608–4613. [143] Nissen H., Dainty R., Comparison of the [132] Mc Donald L.C., Mc Feeters R.F., Daeschel use of rRNA probes and conventional M.A., Fleming H.P., A differential medium methods in identifying strains of Lactoba- for the enumeration of homofermentative cillus sake and L. curvatus isolated from and heterofermentative lactic acid bacteria, meat, Int. J. Food Microbiol. 25 (1995) Appl. Environ. Microbiol. 53 (1987) 311–315. 1382–1384. [144] Ogier J.C., Son O., Gruss A., Tailliez P., [133] Medina R., Katz M., Gonzalez S., Oliver Delacroix-Buchet A., Identification of bac- G., Characterization of the lactic acid bacte- terial microflora in dairy products by tem- ria in ewe’s milk and cheese from northwest poral temperature gradient gel electropho- Argentina, J. Food Prot. 64 (2001) 559–563. resis, Appl. Environ. Microbiol. 68 (2002) 3691–3701. [134] Menendez S., Centeno J.A., Godinez R., Rodriguez-Otero J.L., Effects of Lactoba- [145] Peterson S.D., Marshall R.T., Nonstarter cillus strains on the ripening and organolep- lactobacilli in Cheddar cheese: a review, J. tic characteristics of Arzua-Ulloa cheese, Dairy Sci. 73 (1990) 1395–1410. Int. J. Food Microbiol. 59 (2000) 37–46. [146] Pot B., Hertel C., Ludwig W., Descheemaeker P., Kersters K., Schleifer K.H., Identifica- [135] Miteva V., Boudakov I., Ivanova-Stoyan- tion and classification of Lactobacillus aci- cheva G., Marinova B., Mitev V., Mengaud J., dophilus, L. gasseri and L. johnsonii strains Differentiation of Lactobacillus delbrueckii by SDS-PAGE and rRNA-targeted oligo- subspecies by ribotyping and amplified ribo- nucleotide probe hybridization, J. Gen. somal DNA restriction analysis (ARDRA), Microbiol. 139 (1993) 513–517. J. Appl. Microbiol. 90 (2001) 909–918. [147] Pot B., Vandamme P., Kersters K., Analy- [136] Molin G., Arhné S., Johansson M.L., Com- sis of electrophoretic whole-organisms pro- parative evaluation of plasmid profiling and tein fingerprints, in: Goodfellow M., ribotyping in the analysis of Lactobacillus O’Donnell A.G. (Eds.), Modern microbial plantarum strain heterogeneity in silage, J. methods. Chemical methods in prokaryotic Appl. Bacteriol. 80 (1996) 114–116. systematics, John Wiley & Sons Ltd, Chi- [137] Morea M., Baruzzi F., Cappa F., Coccon- chester, UK, 1994, pp. 493–521. celli P.S., Molecular characterization of the [148] Quere F., Deschamps A., Urdaci M.C., Lactobacillus community in traditional pro- DNA probe and PCR-specific reaction for cessing of Mozzarella cheese, Int. J. Food Lactobacillus plantarum, J. Appl. Micro- Microbiol. 43 (1998) 53–60. biol. 82 (1997) 783–790. 304 V. Coeuret et al.

[149] Quiberoni A., Tailliez P., Quénée P., Sua- systematics, Appl. Environ. Microbiol. 51 rez V., Reinheimer J., Genetic (RAPD- (1986) 873–884. PCR) and technological diversities among [162] Shah N.P., Probiotic bacteria: selective enu- wild Lactobacillus helveticus strains, J. meration and survival in dairy foods, J. Dairy Appl. Microbiol. 85 (1998) 591–596. Sci. 83 (2000) 894–907. [150] Randazzo C.L., Torriani S., Akkermans [163] Shah N.P., Lankaputhra W.E.V., Britz A.D.L., de Vos W.M., Vaughan E.E., M.L., Kyle W.S.A., Survival of Lactobacil- Diversity, dynamics, and activity of bacte- lus acidophilus and Bifidobacterium bifi- rial communities during production of an dum in commercial yoghurt during refrigera- artisanal Sicilian cheese as evaluated by ted storage, Int. Dairy J. 5 (1995) 515–521. 16S rRNA analysis, Appl. Environ. Micro- biol. 68 (2002) 1882–1892. [164] Shakeel-Ur-Rehman S., Mc Sweeney P.L.H., Fox P.F., A study of the role of the indige- [151] Ravula R.R., Shah N.P., Selective enumera- nous microflora of raw milk on the ripening tion of Lactobacillus casei from yogurt and of Cheddar cheese, Milchwissenschaft 57 fermented milk drinks, Biotechnol. Tech. (1999) 388–392. 12 (1998) 819–822. [165] Sharpe A.N., Jackson A.K., Stomaching: a [152] Rebecchi A., Crivori S., Sarra P.G., Coc- new concept in bacteriologica sample pre- concelli P.S., Physiological and molecular paration, Appl. Microbiol. 24 (1972) 175–178. techniques for the study of bacterial com- [166] Sharples G.J., Lloyd R.G., A novel repeated munity development in sausage fermentation, DNA sequence located in the intergenic J. Appl. Microbiol. 84 (1998) 1043–1049. regions of bacterial chromosomes, Nucl. [153] Reinkemeier M., Rocken W., Leitzmann Acid. Res. 18 (1990) 6503–6508. C., A rapid mechanical lysing procedure for [167] Shin H.S., Lee J.H., Pestka J.J., Ustunol Z., routine analysis of plasmids from lactoba- Viability of bifidobacteria in commercial cilli, isolated from sourdoughs, Int. J. Food dairy products during refrigerated storage, Microbiol. 29 (1996) 93–104. J. Food Prot. 63 (2000) 327–331. [154] Rizzo F.A., Korkeala H., Mononen I., Gas [168] Simpson J.M., McCracken V.J., Gaskins chromatography analysis of cellular fatty H.R., Mackie R.I., Denaturing gradient gel acids and neutral monosaccharides in the electrophoresis analysis of 16S ribosomal identification of lactobacilli. Appl. Environ. DNA amplicons to monitor changes in fecal Microbiol. 53 (1987) 2883–2888. bacterial populations of weaning pigs after [155] Rodtong S., Tannock G.W., Differentiation introduction of Lactobacillus reuteri strain of Lactobacillus strains by ribotyping, Appl. MM53, Appl. Environ. Microbiol. 66 Environ. Microbiol. 59 (1993) 3480–3484. (2000) 4705–4714. [156] Rogosa M., Mitchell J.A., Wiseman R.F., A [169] Sohier D., Coulon J., Lonvaud-Funel A., selective medium and enumeration of oral Molecular identification of Lactobacillus and fecal lactobacilli, J. Bacteriol. 62 hilgardii and genetic relatedness with Lac- (1951) 132–133. tobacillus brevis, Int. J. Syst. Bacteriol. 49 (1999) 1075–1081. [157] Roussel Y., Colmin C., Simonet J.M., Decaris B., Strain characterization, genome [170] Song Y., Kato N., Liu C., Matsumiya Y., size and plasmid content in the Lactobacil- Kato H., Watanabe K., Rapid identification lus acidophilus group (Hansen and Mocquot), of 11 human intestinal Lactobacillus spe- J. Appl. Bacteriol. 74 (1993) 549–556. cies by multiplex PCR assays using group- and species-specific primers derived from [158] Roy D., Ward P., Vincent D., Mondou F., the 16S-23S rRNA intergenic spacer region Molecular identification of potentially pro- and its flanking 23S rRNA, FEMS Micro- biotic lactobacilli, Curr. Microbiol. 40 biol. Lett. 187 (2000) 167–173. (2000) 40–46. [171] Stackebrandt E., Goebel B.M., Taxonomic [159] Sanders M.E., Walker D.C., Walker K.M., note: a place for DNA-DNA reassociation Aoyama K., Klaenhammer T.R., Perfor- and 16S rRNA sequence analysis in the pre- mance of commercial cultures in fluid milk sent species definition in bacteriology, Int. applications, J. Dairy Sci. 79 (1996) 943–955. J. Syst. Bacteriol. 44 (1994) 846–849. [160] Schaeffer W.I., Friend K., Efficient detec- [172] Stanton C., Lynch P.B., Collins J.K., tion of soft agar grown colonies using a tetra- Fitzgerald G., Ross R.P., Probiotic cheese, zolium salt, Cancer Lett. 1 (1976) 259–262. Int. Dairy J. 8 (1998) 491–496. [161] Selander R.K., Caugent D.A., Ochman H., [173] Steele J.L., Johnson M.E., Broadbent J.R., Musser J.M., Gilmour M.N., Whittam T.S., Weimer B.C., Starter culture attributes Methods of multilocus enzyme electropho- which affect cheese flavor development, in: resis for bacterial population genetics and Adria-Normandie, Les bactéries lactiques, Identification of dairy lactobacilli 305

Lactic 97, Caen, France, 10–12 Septembre, [184] Tynkkynen S., Satokari R., Saarela M., 1997, pp. 157–170. Mattila-Sandholm T., Saxelin M., Compari- [174] Stiles M.E., Holzapfel W.H., Lactic acid son of ribotyping, randomly amplified poly- bacteria of foods and their current taxo- morphic DNA analysis, and pulsed-field nomy, Int. J. Food Microbiol. 36 (1997) 1–29. gel electrophoresis in typing of Lactobacil- lus rhamnosus and L. casei strains, Appl. [175] Swearingen P.A., O’Sullivan D.J., Warthesen Environ. Microbiol. 65 (1999) 3908–3914. J.J., Isolation, characterization, and influence [185] Valence F., Richoux R., Thierry A., Palva of native, nonstarter lactic acid bacteria on A., Lortal S., Autolysis of Lactobacillus Cheddar cheese quality, J. Dairy Sci. 84 helveticus and Propionibacterium freuden- (2001) 50–59. reichii in Swiss cheeses: first evidence by [176] Tailliez P., Quénée P., Chopin A., Estima- using species-specific lysis markers, J. tion de la diversité parmi les souches de la Dairy Res. 65 (1998) 609–620. collection CNRZ : application de la RAPD [186] Van Beek S., Priest F.G., Evolution of the à un groupe de lactobacilles, Lait 76 (1996) lactic acid bacterial community during malt 147–158. whisky fermentation: a polyphasic study, [177] Tannock G.W., Tilsala-Timisjarvi A., Rodtong Appl. Environ. Microbiol. 68 (2002) S., Ng J., Munro K., Alatossava T., Identifi- 297–305. cation of Lactobacillus isolates from the [187] Van Reenen C.A., Dicks L.M., Evaluation gastrointestinal tract, silage, and yoghurt by of numerical analysis of random amplified 16S-23S rRNA gene intergenic spacer polymorphic DNA (RAPD)-PCR as a region sequence comparisons, Appl. Envi- method to differentiate Lactobacillus plan- ron. Microbiol. 65 (1999) 4264–4267. tarum and Lactobacillus pentosus, Curr. [178] Tannock G.W., Munro K., Harmsen H.J., Microbiol. 32 (1996) 183–187. Welling G.W., Smart J., Gopal P.K., Analy- [188] Vandamme P., Pot B., Gillis M., de Vos P., sis of the fecal microflora of human sub- Kersters K., Swings J., Polyphasic taxo- jects consuming a probiotic product contai- nomy, a consensus approach to bacterial ning Lactobacillus rhamnosus DR20, Appl. systematics, Microbiol. Rev. 60 (1996) Environ. Microbiol. 66 (2000) 2578–2588. 407–438. [179] Tilsala-Timisjarvi A., Alatossava T., Deve- [189] Vasquez A., Ahrne S., Pettersson B., Molin lopment of oligonucleotide primers from G., Temporal temperature gradient gel elec- the 16S-23S rRNA intergenic sequences for trophoresis (TTGE) as a tool for identifica- identifying different dairy and probiotic tion of Lactobacillus casei, Lactobacillus lactic acid bacteria by PCR, Int. J. Food paracasei, Lactobacillus zeae and Lactoba- Microbiol. 35 (1997) 49–56. cillus rhamnosus, Lett. Appl. Microbiol. 32 [180] Tilsala-Timisjarvi A., Alatossava T., Strain- (2001) 215–219. specific identification of probiotic Lactobacil- [190] Vinderola C.G., Reinheimer J.A., Culture lus rhamnosus with randomly amplified poly- media for the enumeration of Bifidobacte- morphic DNA-derived PCR primers, Appl. rium bifidum and Lactobacillus acidophilus Environ. Microbiol. 64 (1998) 4816–4819. in the presence of yoghurt bacteria, Int. [181] Torriani S., Zapparoli G., Dellaglio F., Use Dairy J. 9 (1999) 497–505. of PCR-based methods for rapid differentia- [191] Vinderola C.G., Prosello W., Ghiberto tion of Lactobacillus delbrueckii subsp. bul- T.D., Reinheimer J.A., Viability of probio- garicus and L. delbrueckii subsp. lactis, Appl. tic (Bifidobacterium, Lactobacillus acido- Environ. Microbiol. 65 (1999) 4351–4356. philus and Lactobacillus casei) and nonpro- [182] Tsakalidou E., Manolopoulou E., Kabaraki biotic microflora in Argentinian Fresco E., Zoidou E., Pot B., Kersters K., cheese, J. Dairy Sci. 83 (2000) 1905–1911. Kalantzopoulos G., The combined use of [192] Vogel R.F., Lohmann M., Nguyen M., Wel- whole-cell protein extracts for the identifi- ler A.N., Hammes W.P., Molecular charac- cation (SDS-PAGE) and enzyme activity terization of Lactobacillus curvatus and screening of lactic acid bacteria isolated Lact. sake isolated from sauerkraut and from traditional Greek dairy products, Syst. their application in sausage fermentations, Appl. Microbiol. 17 (1994) 444–458. J. Appl. Bacteriol. 74 (1993) 295–300. [183] Tsuchiya Y., Kano Y., Koshino S., Identifi- [193] Vogel R., Böcker G., Stolz P., Ehrmann M., cation of lactic acid bacteria using tempera- Fanta D., Ludwig W., Pot B., Kersters K., ture gradient gel electrophoresis for DNA Schleifer K.H., Hammes W., Identification fragments amplified by polymerase chain of lactobacilli from sourdough and descrip- reaction, J. Amer. Soc. Brew Chem. 52 tion of Lactobacillus pontis sp. nov., Int. J. (1994) 95–99. Syst. Bacteriol. 44 (1994) 223–239. 306 V. Coeuret et al.

[194] Walter J., Tannock G.W., Tilsala-Timisjarvi [198] Williams J.G., Kubelik A.R., Livak K.J., A., Rodtong S., Loach D.M., Munro K., Rafalski J.A., Tingey S.V., DNA polymor- Alatossava T., Detection and identification phisms amplified by arbitrary primers are of gastrointestinal Lactobacillus species by useful as genetic markers, Nucl. Acid. Res. using denaturing gradient gel electrophore- 18 (1990) 6531–6535. sis and species-specific PCR primers, Appl. [199] Woese C.R., Bacterial evolution, Micro- Environ. Microbiol. 66 (2000) 297–303. biol. Rev. 51 (1987) 221–271. [195] Walter J., Hertel C., Tannock G.W., Lis C.M., Munro K., Hammes W.P., Detection [200] Yost C.K., Nattress F.M., The use of multi- of Lactobacillus, Pediococcus, Leuconos- plex PCR reactions to characterize popula- toc, and Weissella species in human feces tions of lactic acid bacteria associated with by using group-specific PCR primers and meat spoilage, Lett. Appl. Microbiol. 31 denaturing gradient gel electrophoresis, (2000) 129–133. Appl. Environ. Microbiol. 67 (2001) [201] Zapparoli G., Torriani S., Dellaglio F., Dif- 2578–2585. ferentiation of Lactobacillus sanfranciscen- [196] Ward L.J., Timmins M.J., Differentiation of sis strains by randomly amplified polymor- Lactobacillus casei, Lactobacillus paracasei phic DNA and pulsed-field gel electrophoresis, and Lactobacillus rhamnosus by polyme- FEMS Microbiol. Lett. 166 (1998) 325–332. rase chain reaction, Lett. Appl. Microbiol. [202] Zhong W., Millsap K., Bialkowska- 29 (1999) 90–92. Hobrzanska H., Reid G., Differentiation of [197] Welsh J., McClelland M., Fingerprinting Lactobacillus species by molecular typing, genomes using PCR with arbitrary primers, Appl. Environ. Microbiol. 64 (1998) Nucl. Acids Res. 18 (1990) 7213–7218. 2418–2423.

To access this journal online: www.edpsciences.org