ARTICLE IN PRESS

FOODMICROBIOLOGY

0740-0020/$ - se

doi:10.1016/j.fm

�CorrespondE-mail addr

Marc.Vancanne

Food Microbiology 24 (2007) 120–127

www.elsevier.com/locate/fm

Biodiversity and identification of sourdough lactic acid bacteria

Luc De Vuysta,�, Marc Vancanneytb

aResearch Group of Industrial Microbiology and Food Biotechnology, Department of Biological Sciences and Engineering, Vrije Universiteit Brussel,

Pleinlaan 2, B-1050 Brussels, BelgiumbBCCM/LMG Bacteria Collection, Laboratory for Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium

Available online 11 September 2006

Abstract

This review deals with recent developments on the biodiversity of sourdough lactic acid bacteria (LAB) and the recent description of

new sourdough LAB species. One of the outcomes of biodiversity studies of particular sourdough ecosystems throughout Europe is the

description of new taxa of LAB. During the last 3 years, several new LAB species have been isolated from traditional sourdoughs:

Lactobacillus mindensis, Lactobacillus spicheri, Lactobacillus rossiae, Lactobacillus zymae, Lactobacillus acidifarinae, Lactobacillus

hammesii, and Lactobacillus nantensis. Some of these species have been described on one single isolate only. Isolation of novel taxa

mainly depends on the cultivation approach used, i.e. (selective) incubation media and conditions. The distribution of the taxa of LAB is

highly variable from one sourdough ecosystem to another. Therefore, it is difficult to define correlations between population composition

and both the type of sourdough or the geographic location.

Identification of isolated strains needs a polyphasic approach, including a combination of phenotypic and genotypic methods, the

latter often based on the polymerase chain reaction (PCR) and encompassing 16S rRNA sequencing and DNA–DNA hybridizations. A

main obstacle in current identification approaches of LAB strains is the lack of a robust and exchangeable identification system for all

LAB species. Recent studies based on complete genomes have provided the basis for establishing sets of genes useful for multi-locus

sequence analysis (MLSA).

Monitoring the population dynamics of sourdough ecosystems can be performed by both culture-dependent (plating and incubation)

and culture-independent (e.g. PCR-Denaturing Gradient Gel Electrophoresis) methods. Although highly valuable for community

fingerprinting, culture-independent methods do not always yield precise quantitative information.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Sourdough lactic acid bacteria; Polyphasic identification; MLSA; PCR-DGGE

1. Introduction

Sourdough is a mixture of flour (wheat, rye, rice, etc.)and water that is fermented with lactic acid bacteria (LAB)and yeasts (Vogel et al., 1999; De Vuyst and Ganzle, 2005).These micro-organisms usually originate from flour, doughingredients, or the environment. Sourdough is used for theproduction of sourdough bread, classical bread, snacks,pizza, and sweet baked products. Sourdough fermentationsenhance dough properties; improve volume, texture,flavour, and nutritional value of the bread; retard thestaling process of bread; and protect bread from mould and

e front matter r 2006 Elsevier Ltd. All rights reserved.

.2006.07.005

ing author. Tel.: +322 629 32 45; fax: +322 629 27 20.

esses: ldvuyst@vub.ac.be (L. De Vuyst),

yt@UGent.be (M. Vancanneyt).

bacterial spoilage. Hundreds of different types of tradi-tional sourdough breads exist in Europe, in particular inItaly. They differ in the type of flour, other ingredients, andthe applied technology and fermentation process. Becauseof their artisan and region-dependent handling, sour-doughs are an important source of diverse LAB speciesand strains that are metabolically active or can be re-activated upon addition of flour and water. Some of thesestrains play a crucial role during the sourdough fermenta-tion process and are or can be used as sourdough starters.Sourdough is a unique food ecosystem in that it (i)

selects for LAB strains that are adapted to their environ-ment, and (ii) harbours LAB species specific for sourdough(De Vuyst and Neysens, 2005; Gobbetti et al., 2005; DalBello et al., 2005). Adaptations of certain LAB to asourdough environment include (i) a unique central

ARTICLE IN PRESSL. De Vuyst, M. Vancanneyt / Food Microbiology 24 (2007) 120–127 121

metabolism and/or transport of sourdough-specific carbo-hydrates such as maltose and fructose, maltose being themost abundant fermentable carbohydrate and fructosebeing an important alternative electron acceptor; (ii) anactivated proteolytic activity and/or arginine deiminasepathway; (iii) particular stress responses; and (iv) produc-tion of antimicrobial compounds. For instance, doughacidification is a prerequisite for rye baking to inhibit theflour a-amylase. Acidification also promotes a solubiliza-tion of rye pentosans and thus enhances water-binding ofthe dough, since gluten is lacking in rye. In the abundanceof maltose and under stress conditions, maltose phosphor-ylase is constitutively expressed in Lactobacillus sanfran-

ciscensis, Lactobacillus reuteri, Lactobacillus fermentum,and Lactobacillus pontis, which enables maltose hydrolysiswithout the expenditure of ATP and extracellular accumu-lation of glucose, initiating glucose repression in competi-tors for maltose (Stolz et al., 1996). This unique maltosemetabolism, in particular of L. sanfranciscencis, favours amutualistic association with maltose-negative yeast speciessuch as Saccharomyces exiguus or Candida humilis, as is thecase in San Francisco sourdough and Panettone, respec-tively. The use of fructose as alternative electron acceptor,being converted to mannitol, favours acetic acid produc-tion, which in turn influences the sensorial properties of thebread (Stolz et al., 1995a, b). In addition, both constitutivemaltose consumption and arginine conversion enable extraATP production, and hence enhance the competitiveness ofsourdough LAB such as L. sanfranciscensis and L. reuteri

(Stolz et al., 1996; De Angelis et al., 2002; Rollan et al.,2003). Finally, several sourdough LAB contribute to breadaroma (for instance through the production of ornithine asaroma precursor), dough rheology (amino acid metabo-lism), texture (production of exopolysaccharides), anddelay of spoilage (production of acetic acid) (Gobbettiet al., 2005).

Besides these adaptations, dominance of sourdoughLAB may depend on the technology used for sourdoughproduction, for instance through the selective pressureexerted by the environmental conditions, as in the case ofL. sanfranciscensis in type I sourdoughs, in particular inSan Francisco bread and Panettone production (Gobbettiand Corsetti, 1997; Picozzi et al., 2006), and of Lactoba-

cillus amylovorus in type II rye sourdoughs (Muller et al.,2001). Also, the production of specific inhibitory sub-stances, such as reutericyclin in the case of L. reuteri, mayfavour the dominance of sourdough LAB, for instance inGerman, type II sourdoughs (Ganzle and Vogel, 2002;Corsetti et al., 2004; Dal Bello et al., 2005). This oftenresults in stable sourdough ecosystems, such as in the caseof the commercial sourdough starter Bocker–Reinzucht–Sauerteig (BRS), of which the composition remained stableover a period of at least two decades (Bocker et al., 1990;Ganzle et al., 1998).

As the recently published review papers of De Vuyst andNeysens (2005) and Ehrmann and Vogel (2005) havedescribed the taxonomy, biodiversity, and metabolic

interactions of the sourdough microbiota in detail, thismanuscript will deal with recent developments on thebiodiversity of sourdough LAB and the recent descriptionof new sourdough LAB species.

2. Novel validly named species isolated from sourdough since

2003

2.1. Taxa and sources

During the last 3 years, several new LAB species havebeen isolated from traditional sourdoughs that werecontinuously propagated by back-slopping (repeated cyclicre-inoculation) at ambient temperature. Lactobacillus

mindensis has been described based on four isolates, ofwhich one was isolated and purified from an industrial ryesourdough starter preparation (BRS) and the others werere-isolated after back-slopping of one particular sourdoughinitiated with that starter (Ehrmann et al., 2003). Lacto-

bacillus spicheri has been described on the basis of a singlestrain isolated from an industrially processed rice sour-dough (Meroth et al., 2004). Lactobacillus rossiae has beendescribed on the basis of six isolates, each of themoriginating from a different wheat sourdough from CentralItaly (Corsetti et al., 2005). Lactobacillus zymae has in itsoriginal description been isolated from both Belgian andGreek wheat sourdough, indicating a geographical spread,although the number of strains was restricted to a singleisolate in each case (De Vuyst et al., 2002; Vancanneytet al., 2005). Lactobacillus acidifarinae has been describedwith a single strain from Belgian wheat sourdough(Vancanneyt et al., 2005). Lactobacillus hammesii has beendescribed based on two phenotypically distinct isolates, forwhich it is not clear whether they originate from the sameor different French wheat sourdoughs (Valcheva et al.,2005). Lactobacillus nantensis has been described on thebasis of 14 isolates, originating from the same Frenchwheat sourdough (Valcheva et al., 2006). Some of thesespecies have a broader distribution than the original sourceof isolation, as has been demonstrated by the isolation ofL. rossiae, L. spicheri, L. hammesii, and L. nantensis fromBelgian sourdoughs (data not shown). In addition,L. spicheri has been isolated from both rice (Merothet al., 2004) and wheat (Valcheva et al., 2005) sourdoughs.Interestingly, all these bacteria seem to be endemic tosourdough fermentations, as they have not been isolatedfrom other environments up to now. However, it has to beunderlined that all species have been described in the pastfew years only.

2.2. Phylogenetic allocation of new species among other

sourdough LAB

Representatives of the LAB genera Lactobacillus,Leuconostoc, Pediococcus, and Weissella have been de-tected in sourdoughs (De Vuyst and Neysens, 2005;Ehrmann and Vogel, 2005). The variety within the genus

ARTICLE IN PRESSL. De Vuyst, M. Vancanneyt / Food Microbiology 24 (2007) 120–127122

Lactobacillus is most prevalent (Fig. 1). The latter genus isphylogenetically the most heterogeneous one and can besubdivided in at least nine species groups based on 16S

100

989694929088

.Lactobacillus johnsonii

.Lactobacillus delbrueck

.Lactobacillus acidophilu

.Lactobacillus gallinarum

.Lactobacillus suntoryeu

.Lactobacillus amylovoru

.Lactobacillus crispatus

.Lactobacillus rossiae

.Lactobacillus fermentum

.Lactobacillus reuteri

.Lactobacillus pontis

.Lactobacillus panis

.Lactobacillus frumenti

.Lactobacillus sakei subs

.Lactobacillus curvatus

.Lactobacillus plantarum

.Lactobacillus pentosus

.Lactobacillus alimentariu

.Lactobacillus paralimen

.Lactobacillus nantensis

.Lactobacillus farciminis

.Lactobacillus mindensis

.Pediococcus pentosace

.Pediococcus acidilactici

.Lactobacillus sanfrancis

.Lactobacillus fructivoran

.Lactobacillus parabuchn

.Lactobacillus buchneri

.Lactobacillus hammesii

.Lactobacillus acidifarina

.Lactobacillus zymae

.Lactobacillus brevis

.Lactobacillus spicheri

.Lactobacillus perolens

.Lactobacillus paracasei

.Leuconostoc mesentero

.Leuconostoc citreum

.Weissella paramesente

.Weissella confusa

.Weissella cibaria

.Enterococcus faecalis

.Staphylococcus aureus

Fig. 1. Phylogenetic tree based on 16S rRNA gene sequence comparisons of L

the genus Lactobacillus, Lactobacillus species with aberrant phylogenetic pos

genera have been included to illustrate the diversity within the genus La

Staphylococcus aureus subsp. aureus was included as outgroup.

rRNA gene sequence similarity. Representatives of mostspecies groups, i.e. the Lactobacillus alimentarius, Lacto-

bacillus buchneri, Lactobacillus casei, Lactobacillus

ii subsp. delbrueckii

s

s

s

p.sakei

subsp. plantarum

s

tarius

us

censis

s

eri

e

subsp.paracasei

ides subsp. mese.

roides

subsp.aureus

ATCC 33200T

ATCC 9649T

ATCC 4356T

ATCC 33199T

SAT

DSM 20531T

DSM 20584T

CS1T

ATCC 14931T

DSM 20016T

LTH 2587T

DSM 6035T

TMW 1.666T

DSM 20017T

DSM 20010T

JCM 1149T

ATCC 8041T

DSM 20249T

DSM 13238T

Lp33

ATCC 29644T

TMW 1.80T

DSM 20336T

DSM 20284T

ATCC 27651T

DSM 20203T

LMG 11457T

DSM 20057T

TMW 1.1236T

LMG 22200T

LMG 22198T

ATCC 14869T

LTH 5753T

L 532T

JCM 8130T

DSM 20343T

ATCC 49370T

NRIC 1542T

JCM 1093T

LMG 17699T

JCM 5803T

NCDO 949T

..... AJ002515

..... AY050172

..... M58802

..... AJ242968

...... AY445815

..... M58805

..... Y17362

..... AJ564009

..... M58819

..... X76328

..... X76329

..... X94230

..... AJ250074

..... M58829

..... AJ270951

..... D79210

..... D79211

..... M58804

..... AJ417500

..... AY690834

..... M58817

..... AJ313530

..... AJ305321

..... AJ305320

..... X76327

..... X76330

..... AY026751

..... M58811

..... AJ632219

...... AJ632158

...... AJ632157

..... M58810

...... AJ534844

..... Y19167

..... D79212

..... M23035

..... AF111948

..... AB023238

..... AB023241

..... AJ295989

..... AB012212

..... X70648

actobacillus type strains. Representatives of the nine species groups within

itions, novel sourdough species, and species representing closely related

ctobacillus and its intertwined relatedness towards other LAB genera.

ARTICLE IN PRESSL. De Vuyst, M. Vancanneyt / Food Microbiology 24 (2007) 120–127 123

delbrueckii, Lactobacillus fructivorans, Lactobacillus plan-

tarum, L. reuteri, and Lactobacillus sakei species groups,have been isolated from sourdoughs. Lacking, up to now,are isolates of the Lactobacillus salivarius species group.Four of the novel taxa, L. acidifarinae, L. hammesii,L. spicheri, and L. zymae, are members of the Lactobacillus

buchneri species group; L. mindensis and L. nantensis

enrich the species diversity of the L. alimentarius speciesgroup; and L. rossiae belongs to the L. reuteri speciesgroup.

2.3. Isolation conditions

Isolation of novel taxa mainly depends on thecultivation approach used, i.e. (selective) incubationmedia (such as de Man–Rogosa–Sharpe or MRSmedium for lactic acid bacteria, in particular lactobacilli)and conditions (temperature, pH, aerobic/anaerobic,etc.). For instance, it has been shown that the isolationof Lactobacillus pontis and Lactobacillus panis dependson the availability of specific amino acids and vitamins inthe growth medium. Supplemented media such as MRS5(Meroth et al., 2003) enable their isolation from Belgianwheat sourdoughs (data not shown), and hence they arenot unique for German rye sourdough batters with anextended fermentation period and higher temperatures(Vogel et al., 1994; Wiese et al., 1996). Moreover,L. pontis strains have been found as intestinal isolates(Leser et al., 2002).

L. mindensis has been isolated anaerobically at 30 1C onmodified MRS agar (pH 5.4), additionally containingfructose and maltose, as described for the isolation ofL. pontis (Vogel et al., 1994; Ehrmann et al., 2003).L. spicheri has been isolated at 30 1C under micro-aerophilic conditions on MRS5 agar, which merelycontains the components described above, but withlowered pH (Meroth et al., 2003, 2004). L. rossiae hasbeen isolated under aerobic conditions at 30 1C onmodified MRS agar, to which maltose and fresh yeastextract were added at 1% and 10%, respectively, and thefinal pH was adjusted to 5.6 (Corsetti et al., 2001, 2005).L. acidifarinae has been isolated at 37 1C on MRS agar (pH5.4) supplemented with 1% (w/v) of both maltose andfructose and 0.1% cycloheximide, and two isolates ofL. zymae have been isolated at 30 1C onMRS agar (pH 5.4)supplemented with 2% maltose, and as for L. acidifarinae,respectively, under aerobic conditions (Vancanneyt et al.,2005). L. hammesii and L. nantensis have been isolatedanaerobically at 301C on MRS agar (pH 6.2) supplementedwith 1% (w/v) maltose and 0.5% (w/v) fresh yeast extract(Valcheva et al., 2005, 2006). It appears that modified MRS(MRS supplemented with maltose and/or fructose as wellas amino acids and vitamins) with lowered pH (o6.0) andanaerobic conditions are successful for the isolation of(new) sourdough LAB.

2.4. Methods used for species delineation

The description of new LAB taxa is one of the outcomesof diversity studies of particular sourdough ecosystems.The general approach of these studies starts with theisolation and purification of different colony types and ascreening of the genotypic or phenotypic diversity, using afast fingerprinting methodology such as randomly ampli-fied polymorphic DNA (RAPD) or whole-cell proteinelectrophoresis, and subsequent comparison with an in-house reference database. Strains with aberrant profiles areselected for 16S rRNA gene sequence analysis to determinetheir phylogenetic position. Depending on the similarityand position of the strain, an alternative screening methodmay be used to confirm its separate position, beforedetermination of the DNA base content and performinglaborious DNA–DNA hybridization experiments andphenotypic studies. Such a polyphasic approach enablesthe final identification of the isolated strains.When considering the seven recent novel sourdough

taxa, different genotypic (or phenotypic) methodologieshave been used as initial approach for species identifica-tion. L. mindensis and L. spicheri isolates have beenselected for further study after an initial screening usingRAPD-polymerase chain reaction (PCR) (Ehrmann et al.,2003; Meroth et al., 2004). L. rossiae isolates could not beassigned to a known species using an API 50 CH gallery;RAPD-PCR using four primers has been used for straindifferentiation among isolates of this species (Corsettiet al., 2005). L. acidifarinae and L. zymae have beeninitially screened using poly-acrylamide gel electrophoresis(PAGE) of whole-cell proteins and more evidence for aseparate species status has been provided using amplifiedfragment length polymorphism (AFLP) fingerprinting(Vancanneyt et al., 2005). In the studies of L. hammesii

and L. nantensis, evidence for a separate species status hasbeen provided using RAPD-PCR and AFLP (Valchevaet al., 2005, 2006).

2.5. Bottlenecks in current identification approaches

A main obstacle in current identification approaches ofLAB strains is the lack of a robust and exchangeableidentification system for all LAB species, for instancethrough the use of shared databases. Nowadays a widevariety of genotypic (or phenotypic) ‘bar-code’ or finger-printing methods are used, which are optimized for subsetsof often highly related species, and which are notapplicable or provide a different taxonomic resolution formore distantly related taxa. ‘Optimized’ methods arecreated for particular subsets of strains and requireprevious knowledge of the phylogenetic affiliation for aproper identification. These genotypic identification ap-proaches are furthermore not reproducible betweenlaboratories. This means that data obtained by differentlaboratories, even when applying the same protocol, arenot comparable, implying that each laboratory should

ARTICLE IN PRESSL. De Vuyst, M. Vancanneyt / Food Microbiology 24 (2007) 120–127124

construct its own databases. Methods applicable for allLAB and linked to updated databases, comprising nearlyall currently recognized species, are only available inspecialized research laboratories. A disadvantage of these‘generally applicable’ fingerprinting methods is that theyare standardized to that extent that all instrumentalupgrades influence the patterns and require a reconstruc-tion of the database.

A further obstacle is the low number of isolates, oftenfrom a restricted geographic area, used for speciesdescriptions. This results in poor information on theintraspecific variation. This has consequences for reliableidentification of new field isolates. It particularly makes thephenotypic delineation very strict, and the discovery of newfield isolates may require an emendation of the descriptionof the species.

The poor reproducibility of the API 50 CH systems andincomplete commercial databases is another problem forroutine identification of LAB.

A last major bottleneck is the need for laboriousDNA–DNA hybridizations to unequivocally assign strainsto a particular species. Literature data demonstrate thatapplying different methods may largely influence the degreeof DNA binding. Furthermore, these methods often yield alarge standard deviation, which may result in differentinterpretations when obtaining intermediate hybridizationlevels. A first example among sourdough LAB is thefollowing. Lactobacillus suntoryeus, isolated from maltwhisky distilleries, has been recently described by Cachatand Priest (2005), based on DNA–DNA hybridizationvalues of less than 43% with their nearest neighboursLactobacillus helveticus and Lactobacillus gallinarum. In amore recent polyphasic approach, Naser et al. (2006) havedemonstrated that the type strains of L. suntoryeus andL. helveticus yield reassociation values above 70%, andthey hence have confirmed that these names are synonyms.A second example is the intermediate hybridization valueof 68% between Lactobacillus kimchii and Lactobacillus

paralimentarius (De Vuyst et al., 2002). Both species havebeen described more or less simultaneously and are sosimilar, although slightly below the level of speciesdelineation, that assignment of new borderline isolates toone of these species becomes questionable.

2.6. Modern genomic approaches for identification and its

future perspectives

Recent studies based on complete genomes haveprovided the basis for establishing sets of genes useful formulti-locus sequence analysis (MLSA) in large numbers ofbacterial lineages. Also, they have confirmed that se-quences from housekeeping genes can accurately predictgenome relatedness. Accordingly, such sequences can beused for species level identification (Gevers et al., 2004,2005; Konstantinidis and Tiedje, 2005; Santos and Och-man, 2004; Zeigler, 2003). The candidate genes used forspecies differentiation must be present in single-copy and

widely distributed among bacterial genomes, and thosegenes in which recombination might confer a selectiveadvantage or closely linked genes should be avoided(Zeigler, 2003). Furthermore, these genes should beinformative with adequate amount of resolution. Accord-ingly, this variability should be sufficient to differentiatedifferent species of a particular genus. Consequently, theMLSA approach differs extensively from all other non-sequence-based approaches by the generation of ‘absolute’,perfectly exchangeable sequences. Recently, the applicationof MLSA using protein-coding loci, such as pheS (geneencoding the phenylalanyl-tRNA synthase alpha-subunit)and rpoA (gene encoding the DNA-dependent RNApolymerase alpha-subunit) has been proven to be a robustsystem for identification of the known Enterococcus (Naseret al., 2005a,b) and Lactobacillus (unpublished data)species, and has provided high discrimination for differ-entiation of closely related species with almost identical 16SrRNA gene sequences. In addition, the use of these loci hasproven to be an efficient screening method for delineationof novel taxa (Naser et al., 2005c; Svec et al., 2005). Insome taxonomic studies, particular genes have beenselected to clarify the taxonomic position of particularsubsets of lactobacilli (Felis et al., 2001; Torriani et al.,2001). These approaches only remain valuable whencontinuously upgraded with new taxa and when consider-ing the gap between the intraspecies heterogeneity and theinterspecies distance for each taxon.

3. New insights in the biodiversity of sourdough LAB

3.1. Correlation between population composition and the

type of sourdough

Although LAB initially isolated from sourdough are notnecessarily unique for sourdough ecosystems, some corre-lations can be seen between specific LAB species and thetype of sourdough, and sometimes the origin of thesourdough. In practice, sourdoughs are either continuouslypropagated by back-slopping using the mother spongetaken from the preceding fermentation process, or pro-duced by using once a week a commercial starter followedby back-slopping for several days. Therefore, largedifferences can often be seen in species composition withinand among sourdough types.Sourdough LAB usually belong to the genus Lactoba-

cillus, but occasionally Leuconostoc spp., Weissella spp.,Pediococcus spp., and Enterococcus spp. have been found(De Vuyst and Neysens, 2005). In general, heterofermen-tative Lactobacillus species dominate the sourdoughmicrobiota, in particular in type I or traditional sour-doughs that are manufactured by repeated inoculation atambient temperature (20–30 1C) and consist of naturalmixed cultures of LAB (responsible for dough acidification,aroma formation, and leavening) and/or yeast (doughleavening and aroma formation). In type II or accelerated(industrial) sourdoughs, one-step processes incubated at

ARTICLE IN PRESSL. De Vuyst, M. Vancanneyt / Food Microbiology 24 (2007) 120–127 125

higher temperatures (usually above 30 1C) with longerfermentation times (2–5 days) and higher water contents,leavening is sustained by the addition of baker’s yeast;LAB not only contribute to acidification but are alsoresponsible for the organoleptic properties of the fermentedend product.

Lactobacillus sanfranciscensis is considered one of themost important species in type I sourdoughs (Gobbetti andCorsetti, 1997; Picozzi et al., 2006). This obligatelyheterofermentative LAB species is dominant in SanFrancisco sourdough and Panettone production, typicaltype I sourdoughs, as well as in many traditional Italian-baked products such as Pandoro, Colomba, Cornettobrioche, and snacks. Traditional type I sourdoughs withmultiple stages of fermentation frequently select forL. brevis, L. fermentum, L. paralimentarius, L. plantarum,and L. pontis. Type II sourdoughs are characterized by theoccurrence of L. brevis, L. fermentum, L. frumenti,L. pontis, L. panis, and L. reuteri.

3.2. Correlation between population composition and

geographic location

The microbial ecology of a sourdough is not a result ofgeography itself, but it is the result of traditional practicesin specific geographical regions, which certainly affect theprocesses themselves. Therefore, distribution of the taxa ofLAB is highly variable from one sourdough ecosystem toanother; many sourdoughs contain associations of hetero-and homofermentative LAB (De Vuyst and Neysens,2005). The eventually established LAB consortia com-monly reflect the media resources (carbohydrates, aminoacids, vitamins), environmental conditions (temperature,pH, redox potential) and technology applied [processparameters, use of (a) starter(s), use of baker’s yeast].For instance, Italian sourdoughs harbour complex associa-tions of LAB species. The dominant LAB in Italian wheatsourdoughs have been reported to be the obligatelyheterofermentative L. sanfranciscensis, L. brevis, L. fer-

mentum, and L. fructivorans, the facultatively heterofer-mentative L. plantarum and Lactobacillus alimentarius, andthe obligately homofermentative L. acidophilus, L. del-

brueckii subsp. delbrueckii, and Lactobacillus farciminis

(Gobbetti et al., 1994; Corsetti et al., 2001, 2003). Insourdoughs from Umbria (Central Italy), L. sanfranciscen-

sis and L. plantarum are the dominant species (Corsettiet al., 2001). Sourdoughs from Foggia and Lecce (SouthernItaly) seem to typically contain associations betweenL. sanfranciscensis and Leuconostoc citreum, and betweenL. sanfranciscensis and L. alimentarius, respectively (Cor-setti et al., 2001). For the production of Altamura bread, atraditional durum wheat bread produced in Altamura(Southern Italy), L. plantarum has recently been found todominate the sourdough microbiota (Ricciardi et al., 2005).Finally, L. rossiae seems to have a wide distribution insourdoughs from Central and Southern Italy (Settanniet al., 2005, 2006). Most Sardinian sourdoughs are

dominated by the facultative heterofermentative lactoba-cilli L. plantarum and L. pentosus (Catzeddu et al., 2006).As mentioned above, it remains unclear whether or not themajor and minor differences between species compositionof traditional sourdoughs actually reflect typical differencesbetween the regions.

3.3. Monitoring of sourdough-associated LAB species

Monitoring the population dynamics of ecosystems canbe performed by both culture-dependent (plating andincubation) and culture-independent methods. Culture-independent methods such as denaturing gradient gelelectrophoresis (DGGE) of PCR-amplified 16S rRNAfragments have led to an improved understanding of thenatural microbial populations present in a variety ofsourdough ecosystems. PCR-DGGE enables the monitor-ing of the microbial population dynamics as well as thedetection of non-cultivable micro-organisms. Five suchstudies have been published for sourdough. Meroth et al.(2003) have investigated the bacterial population of ryesourdoughs produced under different ecological conditionsby using commercially available starter cultures. Specificcommunity profiles correlated well with certain environ-mental conditions and the PCR-DGGE results wereconsistent with those obtained with culturing. In particu-lar, during continuous propagation of type II ryesourdoughs at higher temperature (40 1C instead of30 1C), L. frumenti and L. panis were dominating insteadof L. pontis (Meroth et al., 2003). L. fermentum could notbe detected by PCR-DGGE, although it was found bycultivation in the fermentation at 30 1C. This bias may becaused by differences in the efficiency of DNA extractionor by preferred amplification of certain templates in PCR.Similarly, Settanni et al. (2006) reported on the importanceof the sensitivity of the DNA extraction method for thedetection of L. rossiae and L. paralimentarius, whichoccurred in lower levels than L. plantarum andL. sanfranciscensis, in artisan wheat sourdoughs collectedin Abruzzo (Italy). Fluctuations within the LAB commu-nity during a rice sourdough fermentation revealed thatL. pontis decreased in numbers over time and Lactobacillus

curvatus became dominant after 3 days of fermentation asrevealed by PCR-DGGE (Meroth et al., 2004). Gatto andTorriani (2004) have monitored changes in the LABpopulations during a traditional Italian wheat sourdoughfermentation process. It turned out that L. sanfranciscensis,L. brevis, Lactobacillus arizonensis/L. plantarum group, andL. kimchii/L. paralimentarius prevailed throughout thefermentation. Conventional culture-dependent methodspartially reflected the PCR-DGGE results, as a lowerdiversity in LAB species was detected. Randazzo et al.(2005) have reported on the dominance of L. sanfrancis-

censis and L. fermentum during traditional Sicilian wheatsourdough production as revealed by PCR-DGGE and ofL. sanfranciscensis, L. pentosus, and L. plantarum through

ARTICLE IN PRESSL. De Vuyst, M. Vancanneyt / Food Microbiology 24 (2007) 120–127126

a cultivation-dependent approach. However, L. fermentum

could not be recovered from culture plates.In general, PCR-DGGE using a part of the 16S rRNA

gene is often seen as an alternative or better a complemen-tary culture-independent approach for studying the biodi-versity of microbial ecosystems. Disadvantages are: (i) thepresence of multiple and slightly variable 16S rRNAoperons provides complex and different patterns forparticular strains of a single species; (ii) the detection limitis restricted and strain-dependent; and (iii) the intensity ofbands does not necessarily correspond with total countspresent in the ecosystem.

3.4. Alternative perspectives for studying population

composition

Although highly valuable for community fingerprinting,culture-independent methods do not always yield quanti-tative information. Yet, PCR-DGGE may provide inter-esting semi-quantitative data. Culture-independentmethods often need to be combined with culture-dependenttechniques to obtain total counts and to have a completepicture on the biodiversity of the ecosystem.

Real-time PCR (or quantitative PCR), a culture-independent method that provides quantitative informa-tion, enables the simultaneous detection and quantificationof micro-organisms by measuring the relative amount ofamplicon generated throughout the PCR reaction, using acombination of specific primers and intercalating dyes orspecific fluorescently labelled probes (Furet et al., 2004).Real-time PCR approaches may be targeting various genes.A drawback of this method is that for enumeration of allspecies a separate probe is required or a primer set for eachspecies, which creates an enormous increase in cost andworkload.

Fluorescent in situ hybridization (FISH) enables thedirect enumeration of whole bacterial cells in samples usingeither fluorescence microscopy or flow cytometry (Brunseret al., 2006). Flow cytometry is a rapid and sensitivetechnique that can determine cell numbers and measurevarious physiological characteristics of each individual cell,using appropriate fluorescent dyes.

Finally, in particular with respect to taxonomy and‘omics’ studies, miniaturization of DNA hybridizationtechniques has led to the development of DNA chips orDNA micro-arrays (Bae et al., 2005). With these tools, anarray of oligonucleotides, cDNA, or genomic DNA can beimmobilized on a small glass slide in such a way that onesample can be tested simultaneously against a large numberof probes.

4. Conclusions

Identification of sourdough LAB by traditional meth-ods, such as sugar fermentation patterns, is considered notreliable. A number of molecular methods have been used toidentify LAB strains isolated from sourdoughs, in parti-

cular methods based on the polymerase chain reaction,such as RAPD-PCR. In most cases a polyphasic approachis required for a valid identification of new sourdough LABstrains, including 16S rRNA sequencing and DNA–DNAhybridizations. New tools such as MLSA and micro-arraysare becoming available.

References

Bae, J.-W., Rhee, S.-K., Park, J.R., Chung, W.-H., Nam, Y.-D., Lee, I.,

Kim, H., Park, Y.-H., 2005. Development and evaluation of genome-

probing microarrays for monitoring lactic acid bacteria. Appl.

Environ. Microbiol. 71, 8825–8835.

Bocker, G., Vogel, R.F., Hammes, W.P., 1990. Lactobacillus sanfrancisco

als stabiles Element in einem Reinzucht–Sauerteig–Praparat. Getreide

Mehl Brot 44, 269–274.

Brunser, O., Gotteland, M., Cruchet, S., Figueroa, G., Garrido, D.,

Steenhout, P., 2006. Effect of a milk formula with prebiotics on the

intestinal microbiota of infants after an antibiotic treatment. Pediatr.

Res. 59, 451–456.

Cachat, E., Priest, F.G., 2005. Lactobacillus suntoryeus sp. nov., isolated

from malt whisky distilleries. Int. J. System. Evol. Microbiol. 55,

31–34.

Catzeddu, P., Mura, E., Parente, E., Sanna, M., Farris, G.A., 2006.

Molecular characterization of lactic acid bacteria from sourdough

breads produced in Sardinia (Italy) and multivariate statistical

analyses of results. System. Appl. Microbiol. 29, 138–144.

Corsetti, A., Lavermicocca, P., Morea, M., Baruzzi, F., Tosti, N.,

Gobbetti, M., 2001. Phenotypic and molecular identification and

clustering of lactic acid bacteria and yeasts from wheat (species

Triticum durum and Triticum aestivum) sourdoughs of Southern Italy.

Int. J. Food Microbiol. 64, 95–104.

Corsetti, A., De Angelis, M., Dellaglio, F., Paparella, A., Fox, P.F.,

Settanni, L., Gobbetti, M., 2003. Characterization of sourdough lactic

acid bacteria based on genotypic and cell-wall protein analyses. J.

Appl. Microbiol. 94, 641–654.

Corsetti, A., Settanni, L., van Sinderen, D., 2004. Characterization of

bacteriocin-like inhibitory substances (BLIS) from sourdough lactic

acid bacteria and evaluation of their in vitro and in situ activity. J.

Appl. Microbiol. 96, 521–534.

Corsetti, A., Settanni, L., van Sinderen, D., Felis, G.E., Dellaglio, F.,

Gobbetti, M., 2005. Lactobacillus rossii sp. nov., isolated from wheat

sourdough. Int. J. System. Evol. Microbiol. 55, 35–40.

Dal Bello, F., Walter, J., Roos, S., Jonsson, H., Hertel, C., 2005. Inducible

gene expression in Lactobacillus reuteri LTH5531 during type II

sourdough fermentation. Appl. Environ. Microbiol. 71, 5873–5878.

De Angelis, M., Mariotti, L., Rossi, J., Servili, M., Fox, P.F., Rollan, G.,

Gobbetti, M., 2002. Arginine catabolism by sourdough lactic acid

bacteria: purification and characterization of the arginine deiminase

pathway enzymes from Lactobacillus sanfranciscensis CB1. Appl.

Environ. Microbiol. 68, 6193–6201.

De Vuyst, L., Ganzle, M.G., 2005. Trends in Food Science & Technology

Special Issue—Second International Symposium on Sourdough: from

Fundamentals to Applications, vol. 16. Elsevier, Amsterdam.

De Vuyst, L., Neysens, P., 2005. The sourdough microflora: biodiversity

and metabolic interactions. Trends Food Sci. Technol. 16, 43–56.

De Vuyst, L., Schrijvers, V., Paramithiotis, S., Hoste, B., Vancanneyt, M.,

Swings, J., Kalantzopoulos, G., Tsakalidou, E., Messens, W., 2002.

The biodiversity of lactic acid bacteria in Greek traditional wheat

sourdoughs is reflected in both composition and metabolite formation.

Appl. Environ. Microbiol. 68, 6059–6069.

Ehrmann, M.A., Vogel, R.F., 2005. Molecular taxonomy and genetics of

sourdough lactic acid bacteria. Trends Food Sci. Technol. 16, 31–42.

Ehrmann, M.A., Muller, M.R.A., Vogel, R.F., 2003. Molecular analysis

of sourdough reveals Lactobacillus mindensis sp. nov. Int. J. System.

Evol. Microbiol. 53, 7–13.

ARTICLE IN PRESSL. De Vuyst, M. Vancanneyt / Food Microbiology 24 (2007) 120–127 127

Felis, G.E., Dellaglio, F., Mizzi, L., Torriani, S., 2001. Comparative

sequence analysis of a recA gene fragment brings new evidence for a

change in the taxonomy of the Lactobacillus casei group. Int. J.

System. Evol. Microbiol. 51, 2113–2117.

Furet, J.-P., Quenee, P., Tailliez, P., 2004. Molecular quantification of

lactic acid bacteria in fermented milk products using real-time

quantitative PCR. Int. J. Food Microbiol. 97, 197–207.

Ganzle, M.G., Vogel, R.F., 2002. Contribution of reutericyclin production

to the stable persistence of Lactobacillus reuteri in an industrial

sourdough fermentation. Int. J. Food Microbiol. 80, 31–45.

Ganzle, M.G., Ehrmann, M., Hammes, W.P., 1998. Modeling of growth

of Lactobacillus sanfranciscensis and Candida milleri in response to

process parameters of sourdough fermentation. Appl. Environ.

Microbiol. 64, 2616–2623.

Gatto, V., Torriani, S., 2004. Microbial population changes during

sourdough fermentation monitored by DGGE analysis of 16S and 26S

rRNA gene fragments. Ann. Microbiol. 54, 31–42.

Gevers, D., Vandepoele, K., Simillion, C., Van de Peer, Y., 2004. Gene

duplication and biased functional retention of paralogs in bacterial

genomes. Trends Microbiol. 12, 148–154.

Gevers, D., Cohan, F.M., Lawrence, J.G., Spratt, B.G., Coenye, T., Feil,

E.J., Stackebrandt, E., Van de Peer, Y., Vandamme, P., Thompson,

F.L., Swings, J., 2005. Re-evaluating prokaryotic species. Nat. Rev.

Microbiol. 3, 733–739.

Gobbetti, M., Corsetti, A., 1997. Lactobacillus sanfrancisco a key sourdough

lactic acid bacterium: a review. Food Microbiol. 14, 175–187.

Gobbetti, M., Corsetti, A., Rossi, J., LaRosa, F., DeVincenzi, S., 1994.

Identification and clustering of lactic acid bacteria and yeasts from

wheat sourdoughs of central Italy. Ital. J. Food Sci. 1, 85–94.

Gobbetti, M., De Angelis, M., Corsetti, A., Di Cagno, R., 2005.

Biochemistry and physiology of sourdough lactic acid bacteria. Trends

Food Sci. Technol. 16, 57–69.

Konstantinidis, K.T., Tiedje, J.M., 2005. Towards a genome-based

taxonomy for prokaryotes. J. Bacteriol. 187, 6258–6264.

Leser, T.D., Amenuvor, J.Z., Jensen, T.K., Lindecrona, R.H., Boye, M.,

Møller, K., 2002. Culture-independent analysis of gut bacteria: the pig

gastrointestinal tract microbiota revisited. Appl. Environ. Microbiol.

68, 673–690.

Meroth, C.B., Walter, J., Hertel, C., Brandt, M.J., Hammes, W.P., 2003.

Monitoring the bacterial population dynamics in sourdough fermenta-

tion processes by using PCR-denaturing gradient gel electrophoresis.

Appl. Environ. Microbiol. 69, 475–482.

Meroth, C.B., Hammes, W.P., Hertel, C., 2004. Characterisation of the

microbiota of rice sourdoughs and description of Lactobacillus spicheri

sp. nov. System. Appl. Microbiol. 27, 151–159.

Muller, M.R.A., Wolfrum, G., Stolz, P., Ehrmann, M.A., Vogel, R.F.,

2001. Monitoring the growth of Lactobacillus species during a rye flour

fermentation. Food Microbiol. 18, 217–227.

Naser, S., Thompson, F.L., Hoste, B., Gevers, D., Vandemeulebroecke,

K., Cleenwerck, I., Thompson, C.C., Vancanneyt, M., Swings, J.,

2005a. Phylogeny and identification of enterococci using atpA gene

sequence analysis. J. Clin. Microbiol. 43, 2224–2230.

Naser, S.M., Thompson, F.L., Hoste, B., Gevers, D., Dawyndt, P.,

Vancanneyt, M., Swings, J., 2005b. Application of multilocus sequence

analysis (MLSA) for rapid identification of Enterococcus species based

on rpoA and pheS genes. Microbiology 151, 2141–2150.

Naser, S.M., Vancanneyt, M., De Graef, E., Devriese, L.A., Snauwaert, C.,

Lefebvre, K., Hoste, B., Svec, P., Decostere, A., Haesebrouck, F., Swings,

J., 2005c. Enterococcus canintestini sp. nov., from faecal samples of

healthy dogs. Int. J. System. Evol. Microbiol. 55, 2177–2182.

Naser, S.M., Hagen, K.E., Vancanneyt, M., Cleenwerck, I., Swings, J.,

Tompkins, T.A., 2006. Lactobacillus suntoryeus Cachat and Priest 2005

is a later synonym of Lactobacillus helveticus (Orla-Jensen 1919)

Bergey et al. 1925 (Approved Lists 1980). Int. J. System. Evol.

Microbiol. 56, 355–360.

Picozzi, C., D’Anchise, F., Foschino, R., 2006. PCR detection of

Lactobacillus sanfranciscensis in sourdough and Panettone baked

product. Eur. Food Res. Technol. 222, 330–335.

Randazzo, C.L., Heilig, H., Restuccia, C., Giudici, P., Caggia, C., 2005.

Bacterial population in traditional sourdough evaluated by molecular

methods. J. Appl. Microbiol. 99, 251–258.

Ricciardi, A., Parente, E., Piraino, P., Paraggio, M., Romano, P., 2005.

Phenotypic characterization of lactic acid bacteria from sourdoughs

for Altamura bread produced in Apulia (Southern Italy). Int. J. Food

Microbiol. 98, 63–72.

Rollan, G., Lorca, G.L., Font de Valdez, G., 2003. Arginine catabolism

and acid tolerance response in Lactobacillus reuteri isolated from

sourdough. Food Microbiol. 20, 313–319.

Santos, S.R., Ochman, H., 2004. Identification and phylogenetic sorting of

bacterial lineages with universally conserved genes and proteins.

Environ. Microbiol. 6, 754–759.

Settanni, L., van Sinderen, D., Rossi, J., Corsetti, A., 2005. Rapid

differentiation and in situ detection of 16 sourdough Lactobacillus

species by multiplex PCR. Appl. Environ. Microbiol. 71,

3049–3059.

Settanni, L., Valmorri, S., van Sinderen, D., Suzzi, G., Paparella, A.,

Corsetti, A., 2006. Combination of multiplex PCR and PCR-

denaturing gradient gel electrophoresis for monitoring common

sourdough-associated Lactobacillus species. Appl. Environ. Microbiol.

72, 3793–3796.

Stolz, P., Bocker, G., Hammes, W.P., Vogel, R.F., 1995a. Utilization

of electron acceptors by lactobacilli isolated from sourdough.

I. Lactobacillus sanfrancisco. Z. Lebensm.-Unters. Forsch. 201,

91–96.

Stolz, P., Vogel, R.F., Hammes, W.P., 1995b. Utilization of electron

acceptors by lactobacilli isolated from sourdough. II. Lactobacillus

pontis, L. reuteri, L. amylovorus, and L. fermentum. Z. Lebensm.-

Unters. Forsch. 201, 402–410.

Stolz, P., Hammes, W.P., Vogel, R.F., 1996. Maltose-phosphorylase and

hexokinase activity in lactobacilli from traditionally prepared sour-

doughs. Adv. Food Sci. 18, 1–6.

Svec, P., Vancanneyt, M., Koort, J., Naser, S.M., Hoste, B., Vihavainen,

E., Vandamme, P., Swings, J., Bjorkroth, J., 2005. Enterococcus

devriesei sp. nov., associated with animal sources. Int. J. System. Evol.

Microbiol. 55, 2479–2484.

Torriani, S., Felis, G.E., Dellaglio, F., 2001. Differentiation of Lactoba-

cillus plantarum, L. pentosus, and L. paraplantarum by recA gene

sequence analysis and multiplex PCR assay with recA gene-derived

primers. Appl. Environ. Microbiol. 67, 3450–3455.

Valcheva, R., Korakli, M., Onno, B., Prevost, H., Ivanova, I., Ehrmann,

M.A., Dousset, X., Ganzle, M.G., Vogel, R.F., 2005. Lactobacillus

hammesii sp. nov., isolated from French sourdough. Int. J. System.

Evol. Microbiol. 55, 763–767.

Valcheva, R., Ferchichi, M., Korakli, M., Ivanova, I., Ganzle, M.G.,

Vogel, R.F., Prevost, H., Onno, B., Dousset, X., 2006. Lactobacillus

nantensis sp. nov., isolated from French wheat sourdough. Int. J.

System. Evol. Microbiol. 56, 587–591.

Vancanneyt, M., Neysens, P., Dewachter, M., Engelbeen, K., Snauwaert,

C., Cleenwerck, I., Van der Meulen, R., Hoste, B., Tsakalidou, E., De

Vuyst, L., Swings, J., 2005. Lactobacillus acidifarinae sp. nov. and

Lactobacillus zymae sp. nov., from wheat sourdoughs. Int. J. System.

Evol. Microbiol. 55, 615–620.

Vogel, R.F., Bocker, G., Stolz, P., Ehrmann, M., Fanta, D., Ludwig, W.,

Pot, B., Kersters, K., Schleifer, K.H., Hammes, W.P., 1994.

Identification of lactobacilli from sourdough and description

of Lactobacillus pontis sp. nov. Int. J. System. Bacteriol. 44,

223–229.

Vogel, R.F., Knorr, R., Muller, M.R.A., Steudel, U., Ganzle, M.G.,

Ehrmann, M., 1999. Non-dairy lactic fermentations: the cereal world.

Antonie van Leeuwenhoek 76, 403–411.

Wiese, B.G., Strohmar, W., Rainey, F.A., Diekmann, H., 1996.

Lactobacillus panis sp. nov., from sourdough with a long fermentation

period. Int. J. System. Bacteriol. 46, 449–453.

Zeigler, D.R., 2003. Gene sequences useful for predicting relatedness of

whole genomes in bacteria. Int. J. System. Evol. Microbiol. 53,

1893–1900.