Lactobacillus Sanfrancisco a Key Sourdough Lactic Acid Bacterium: a Review

Food Microbiology, 1997, 14, 175–187

ORIGINAL ARTICLE

Lactobacillus sanfrancisco a key sourdough lactic acid bacterium: a review

M. Gobbetti* and A. Corsetti

Sourdough may be considered a traditional product with a great future. Studies on the sourdough microflora have recently been improved. In this paper, research on Lactobacillus sanfrancisco, a key sourdough lactic acid bacterium, is reviewed. The physiology, trophic relationships with sourdough related organisms and the genetics are considered in order to explain and improve the biotechnological performances of Lb. sanfrancisco.  1997 Academic Press Limited

Introduction that Lb. sanfrancisco is synonymous with Lactobacillus¨ brevis subsp. lindneri (Spicher Lactobacillus sanfrancisco, named after the and Schroder 1978). city from where the sourdough micro- Lb. sanfrancisco growth is very fastidious: organism was first isolated, is an obligatory it requires fresh yeast extractives, unsatu- heterofermentative Lactobacillus with phylo- rated fatty acids (mainly oleic acid) genetic relationship to the Lactobacillis casei- (Sriranganathan et al. 1973) and it preferen- –Pediococcus group (Hammes and Vogel tially ferments maltose rather than glucose. 1995). It was first isolated by Kline and Occasionally, strains that ferment sucrose, Sugihara (1971) and subsequently revised by ribose, gluconate, galactose, raffinose or fruc- Weiss and Schillinger (1984) for inclusion in tose have been identified (Gobbetti et al. Approved List of Bacterial Names. 1994a, Onno and Roussel 1994, Vogel et al. Characteristics of Lb. sanfrancisco are a 1994, Hammes and Vogel 1995). 36–38 mol% G+C of the DNA (Kandler and Lb. sanfrancisco has been widely isolated Weiss 1986), a Lys–Ala type murein without from rye and wheat sourdoughs of several teichoic acid in the cell wall (Weiss and bread-producing areas and from sourdoughs Schillinger 1984) and one to three large used to make Panettone (Table 1). Tra- mesosomes at the site of the membrane–wall ditional productions by various stages of con- interface (Nelson et al. 1971). Different tinuous propagation and sourdoughs pro- strains harbour¨ varying numbers and sizes of duced by commercial starter cultures are plasmids (Lonner et al. 1990, Gobbetti et al. effective in reducing the growth of fortuitous Received: 25 July 1996 1996a). DNA–DNA hybridization showed bacteria and permit a predominance of Lb. sanfrancisco which, with very few exceptions Institute of Dairy (Martinez-Anaya et al. 1990), is considered to Microbiology, Agriculture Faculty be a key lactic acid bacteria (LAB) in the of Perugia, *Corresponding author. biotechnology of baked sourdough products. Perugia, Italy

0740-0020/97/020173+13 $25.00/0/fd960083  1997 Academic Press Limited 176 M. Gobbetti and A. Corsetti

Lactobacillus sanfrancisco and its availability of specific amino acids and pep- interactions with sourdough yeasts tides excreted by yeasts. In alcoholic beverages, it has been shown that S. cerevisiae The sourdough ecosystem may be defined as may release these nitrogen compounds either a strict microbial consortium and an under- during growth (Thorne 1957, Suomalainen standing of the interactions that occur and Oura 1971) or as a consequence of an between Lb. sanfrancisco and yeasts is fun- accelerated autolysis (Lyons and Rose 1977). damental for evaluating the real role of this The liberation of amino acids by yeasts has bacterium. made possible the growth of Lb. sanfrancisco The association between Lb. sanfrancisco even in a medium initially deficient of essen- and Saccharomyces exiguus is typical in the tial amino acids (valine and isoleucine) production of San Francisco french bread (Gobbetti et al. 1994b). Berg et al. (1981) (Sugihara et al. 1970) and Panettone have identified a growth stimulant factor for (Gobbetti et al. 1994a), and the association Lb. sanfrancisco in a small peptide between Lb. sanfrancisco and Saccharomyces (Asp–Cys–Glu–Gly–Lys) contained in the cerevisiae has been studied in sourdoughs freshly-prepared yeast extracts and commer- from different countries (Spicher et al. 1982, cial yeast, liver or protein hydrolysates are Wlodarczyk 1985), in gluten-free baked goods inadequate substitutes (Sugihara and Kline (Wlodarczyk et al. 1993) and has been found 1975). Although the fermentation by S. cerev- to be optimal in a continuously operating isiae has been associated with a total sourdough fermentation system (Vollmar and decrease of amino acids in the sourdough Meuser 1992). (Collar et al. 1992, Collar and Martinez The trophic relationships that occur 1993), the release of specific peptides may between Lb. sanfrancisco and the prevailing provide a competitive advantage and contrib- sourdough yeasts (S. cerevisiae and S. ute to the stability of Lb. sanfrancisco in exiguus) have been studied in both co-cul- sourdough. tural model systems and during sourdough The consumption of soluble carbohydrate fermentation. Stimulation of Lb. sanfrancisco and the energy yield of Lb. sanfrancisco are growth has been related to an increased greatly influenced by the associated yeasts

Table 1. Source of isolation of Lactobacillus sanfrancisco (syn. Lb. brevis subsp. lindneri) References Origin Source of isolation Spicher 1959 Germany Sour rye starters Kline and Sugihara¨ 1971 California (USA) Wheat sourdough Spicher and Schroder 1978 Germany Pure culture sourdough starters Wlodarczyk 1985 Poland Bread starters Nout and Creemers-Molenaar 1986 Netherlands Wheatmeal sourdough starters Spicher 1987 Italy, Switzerland, Germany Panettone and wheat bread Sweden sourdough Galli et al. 1988 Italy Panettone, brioches, wheat and rye bread sourdough Hansen¨ et al. 1989a Denmark Rye sourdoughs Lo¨nner et al. 1990 Sweden Commercial sourdough Lonner et al. 1990 Denmark Commercial starter culture Hammes 1990 Germany Purely cultured sourdough Okada et al. 1992 Germany Rye sourdough sponge Hochstrasser et al. 1993 Switzerland Wheat sourdough Stolz et al. 1993 Germany Sourdough starters Gobbetti et al. 1994a Italy Panettone and wheat bread sourdoughs Ehrmann et al. 1994 Germany Sourdough Faid et al. 1994 Morocco Wheat sourdough Vogel¨ et al. 1994 Germany Sourdough Ganzle et al. 1995 Germany Sourdough Lactobacillus sanfrancisco a key sourdough lactic acid bacterium 177 and vary according to the type of sugars. The ose-negative yeasts or better may prevent lack of competition between Lb. sanfrancisco competitors from utilizing abundant maltose and S. exiguus for maltose is fundamental for by glucose repression thereby giving an eco- the stability of this association (Sugihara et logical advantage to Lb. sanfrancisco. The al. 1970). S. exiguus preferentially uses disappearance of S. cerevisiae from the sucrose or glucose and has a high tolerance to microbial population of sourdough during the acetic acid produced¨ by heterolactic consecutive fermentations has been related metabolism (Suihko and Makinen 1984). Due to the repression of genes involved in maltose to the faster consumption of maltose, and fermentation (Nout and Creemers-Molenaar especially glucose, by S. cerevisiae, a decrease 1986). In Lb. sanfrancisco strains, maltose in the bacterial metabolism has been utilization is very effective and is not subject observed when associated with Lb. sanfranci- to glucose repression. The maltose uptake sco in a synthetic medium (Gobbetti et al. and glucose excretion in Lb. sanfrancisco has 1994c). A stimulation of the bacterial growth been analyzed by Neubauer et al. (1994). in the co-cultures has been only observed Maltose transport occurs by a secondary using wheat flour extract supplied with 2·5 g transport system (maltose-H+ symport) and is l−1 of various soluble carbohydrates. An driven by the proton motive force (PMF). The imbalance between yeast consumption and maltose carrier protein(s) is constitutively starch hydrolysis by the flour enzymes leads expressed, and is highly dependent on pH to a rapid depletion of some soluble carbo- with an optimum at 5·6–5·2 which matches hydrates during sourdough fermentation well with the pH of sourdough. The glucose (Gobbetti et al. 1994d, Rouzaud and efflux in maltose-grown cells is catalyzed by a Martinez-Anaya 1993). However, the maltose transport system which is induced during concentration may frequently remain growth on maltose, and which can mediate between 2–5 g kg−1 in wheat doughs fer- homologous glucose–glucose exchange but mented by yeasts and LAB (Martinez-Anaya not a maltose–glucose exchange, thus indi- et al. 1993). Barber et al. (1991) have shown cating a maltose-inducible glucose uniport an accumulation of maltose because its system which is responsible for the excretion metabolism by some yeasts may not begin of the excess glucose. Lb. sanfrancisco, Lacto- until the available glucose and fructose are bacillus reuteri, and Lactobacillus fermentum depleted. Nevertheless, the microbial compe- (Vogel et al. 1994) are unique among the Lac- tition for maltose and glucose represents a tobacillaceae in that they phosphorylize critical point in wheat flours with very low maltose, and maltose phosphorylase may be concentrations of soluble carbohydrates. The considered to be a key enzyme for the pre- addition of selected carbon sources to the dominance of lactobacilli during sourdough dough has been proposed in order to enhance fermentation. the production of lactic and acetic acids by Although sucrose fermentation strains of Lb. sanfrancisco (Corsetti et al. 1994). Lb. sanfrancisco have been isolated, the Lb. sanfrancisco hydrolyzes maltose and invertase activity of yeasts liberates ferment- accumulates glucose in the phosphate buffer able glucose and enables a partial growth of (Fig. 1) in a molar ratio of about 1:1 (Stolz et sucrose- and fructose-negative strains, which al. 1993, Gobbetti et al. 1994c). The glucose- leads to a metabolic mutualism in sourdough 1-phosphate produced by maltose phos- fermentation (Gobbetti et al. 1994c). phorylase is further metabolized, whereas Lb. sanfrancisco has a positive influence glucose is not utilized but excreted outside on yeast leavening and CO2 production the cell in order to avoid excessive intracellu- (Gobbetti et al. 1995a). Even though the lar concentrations. Excretion of glucose takes inoculum size of the baker’s yeast is the place in concomitance with maltose avail- major parameter in determining the gas pro- ability, and after maltose has been depleted duction rates (Akdogan and Ozilgen 1992), the consumption of the excreted glucose analyses conducted by using a rheofermento- begins. The glucose excreted during sour- meter have shown that the associative dough fermentation may be utilized by malt- growth of S. cerevisiae and Lb. sanfrancisco 178 M. Gobbetti and A. Corsetti

decreased at one third the time necessary to pathway. The ATP produced energizes the

reach the maximum production of CO2 by the transport of fructose which may enter the cell yeast. An increase in the total CO2 has also through facilitated diffusion (uniport) by the been observed when associated with S. same or modified maltose permease (Dills et exiguus M14. al. 1980). As a consequence of the maltose– fructose mediated co-fermentation, the addition of fructose to the wheat flour Co-metabolisms in Lb. sanfrancisco induces the utilization of fructose by Lb. sanfrancisco CB1 during fermentation. Com- Co-fermentations enable micro-organisms to pared with the sourdough without fructose use substrates that are otherwise non-fer- added, a greater amount of acetic acid is pro- mentable, and increase the microbial adapta- duced, as well as a lower fermentation quo- bility to difficult ecosystems. Under the influ- tient, which drops into the optimal range of ence of several ecological factors the homo- 1·5–4·0¨ (Spicher 1983). By providing fructose, and heterofermentative LAB have a great Rocken et al. (1992) increased the acetic acid aptitude for producing metabolites other production of Lactobacillus brevis (fructose- than lactic acid (Ceslovszky et al. 1992) and positive strain) to the high levels that could for co-fermentations (Romano et al. 1987, compensate the subsequent loss of acetic acid Cunha and Foster 1992) which lead to an during the freeze drying of the sourdough. increase of the energy yield. Similar changes in the sugar metabolism of A fructose-negative strain of Lb. sanfranci- obligate heterofermentative lactobacilli may sco has been shown to co-ferment fructose also be induced by aeration (Ng 1972, Condon when it is in the presence of maltose or glu- 1987). Indeed, the switch from one branch to cose (Gobbetti et al. 1995b). About two moles the other may be governed by the presence of of fructose are consumed for each mole of oxygen or other suitable electron acceptors, maltose. In comparison with the growth with such as fructose, citrate and malate (Kandler maltose alone, the main variations which 1983, Condon 1987, Stolz et al. 1993). The occur in co-fermentation are: an increase in influence of the electron acceptors on the the cell yield correlated with the additional metabolism of Lb. sanfrancisco has been amount of ATP generated by the metabolism shown in depth by Stolz et al. (1995a) (Fig. 1). of fructose, increases in production of lactic However, while the improvement in the acid and especially of acetic acid, synthesis of acetic acid concentration by controlled aer- mannitol and a decrease in ethanol amount. ation may be a suitable method in well- Lb. sanfrancisco uses maltose as an energy equipped laboratories, it cannot be done in source and fructose as an additional electron normal bakeries. acceptor, thereby obtaining an optimal A co-metabolism of maltose or glucose and growth rate (Axelsson 1993). The action of citrate has also been shown in Lb. sanfranci- fructose as electron acceptor occurs via its sco citrate-negative strains (Gobbetti and reduction to mannitol accompanied by a Corsetti 1996). The production of succinic decrease in the level of ethanol, which is nor- acid by the partial dissimilation of citrate mally produced from the degradation of malt- through¨ the succinic pathway (Radler 1975, ose (Fig. 1). This enables Lb. sanfrancisco to Lutgens and Gottschalk 1980) and a biphasic produce ATP through the acetate kinase impedance curve have been typical of this co- reaction, and in parallel, to synthesize a metabolism. During growth of Lb. sanfranci- higher level of acetic acid. A stoichiometric sco in a medium containing both maltose and acetate production (two moles of fructose glucose or other electron acceptors, the depleted per mole of acetate formed) by Lb. changes in the redox potential were consist- sanfrancisco has been first reported by Stolz ent with the formation of acetate or ethanol et al. (1995a). As shown for other co-fermen- in addition to lactate and with a biphasic tations (Dills et al. 1980, Romano et al. 1987), growth not related to the sequential utiliz- maltose crosses the membrane by PMF and ation of the various carbon sources (Stolz et generates ATP by the hexose monophosphate al. 1993, 1995a). Since pyruvate is also an Lactobacillus sanfrancisco a key sourdough lactic acid bacterium 179

Glucose Maltose H+

Fac.dif.Symport Cell membrane

Maltose H+ Cytoplasm

Glucose Glucose-1P NADH+H+ NAD+ 3 2 ATP Erythrose-4-P Erythritol 26 25 Pi ADP Glucose-6P Fructose-6-P 8 ATP NAD+ 4 + Acetyl-P Acetate NADH+H + 17 + 2 NADH+H NAD 5 + + 15,16 CO2 NADH+H 2 NAD Pi 6 Ethanol 7 ATP Acetate Xylulose-5P ADP Pi 8 17 Fructose Malate Citrate Glycerinaldehyde-3P Acetylphosphate 20 + Acetate Pi + NADH+H 18 NAD 15 19 9 CO2 + Pi Oxaloacetate NADH+H NAD+ ADP Acetaldehyde 10 21 NADH+H+ Mannitol Lactate CO2 ATP 16 NAD+ 11 22 Ethanol NAD+ H2O 12 + ADP NADH+H 13 Pyruvate ATP O2 NAD+ 14 NADH+H+ NADH+H+ 23 NAD+ Lactate H2O2 + 24 NADH+H NAD+

2 H2O

Figure 1. Metabolic key reactions of Lactobacillus sanfrancisco: metabolism of maltose and fate of potential electron acceptors upon maltose fermentation. The enzymes involved are given in the following list: (1) maltosephosphorylase; (2) phosphoglucomutase; (3) hexokinase; (4–8) enzymes of the phosphogluconate pathway; (8) phosphoketolase; (9–14) enzymes of the Embden–Meyerhof glycolytic pathway; (15) phosphotransacetylase; (16) alcohol dehydrogenase; (17) acetate kinase; (18) mannitol dehydrogenase; (19) malolactic enzyme; (20) citrate lyase; (21) oxaloacetate-decarboxylase; (22) lactate dehydrogenase; (23) NADH-H2O2 oxidase; (24) NADH-peroxidase; (25) glucose phosphate isomerase; (26) erythritol dehydrogenase and erythrose-4-P-phosphotransferase. Reproduced by permission of Stolz et al. (1995b). 180 M. Gobbetti and A. Corsetti

intermediate of the citrate metabolism of in the concentration of aliphatic, dicarboxylic LAB (Cogan 1987), in cells co-metabolizing and hydroxy amino acid groups which, for the citrate, pyruvate coming from two pathways most part, are stimulatory for the bacterial is used to reoxidize NADH/NADPH (Fig. 1). growth (Spicher and Nierle 1984, Gobbetti et This leads the cell to save acetaldehyde from al. 1994e). The synthesis of D-alanine and D- reduction to ethanol and the precursor acetyl- glutamic acid isomers has also been observed phosphate is diverted to acetic acid, the pro- by Lb. sanfrancisco and other sourdough duction of acetic acid being greater than can starters. Cellular excretion of D-forms or be accounted for in terms of maltose fermen- enzymatic isomerization of L-amino acids tation. A sourdough started with Lb. sanfran- may explain the biotic generation (Gobbetti cisco CB1 and with an addition of citrate to et al. 1994e). the dough was characterized by an appropri- The peptide hydrolase and proteinase sys- ate quotient of fermentation. tems of Lb. sanfrancisco have been characterized and their subcellular distribution localized (Gobbetti et al. 1996b). Compared Proteolytic activity of Lb. with the activities of other sourdough LAB, sanfrancisco Lb. sanfrancisco strains show the highest aminopeptidase, dipeptidase, tripeptidase Accumulation of amino acids during sour- and iminopeptidase activities. Most pepti- dough fermentation has been determined dases are contained in the cytoplasm except (Collar et al. 1991). Although some authors for dipeptidase which is located in the mem- have shown a proteolysis by flour enzymes brane, and di- and tri-peptidase have a (Kratochvil and Holas 1984, 1988, Spicher higher overall specific activity than the other and Nierle 1988), the increase of the amino peptidases. Because wheat flour contains acid content has been related to the spon- considerable amounts of low molecular taneous microflora, and especially, to the weight peptide (Quaglia 1984), the di- and presence of various LAB strains (Spicher and tri-peptidases may be more important than Nierle 1984, 1988, Collar and Martinez 1993, other proteolytic activities for the predomi- Mascaros et al. 1994). To a lesser extent, nance of LAB in the sourdough. Although amino acids may also be derived from other iminopeptidase activity is limited when com- metabolic routes (Okuara and Harada 1971) pared with those from strong proteolytic bac- and from the cell mass of micro-organisms teria, it may represent a pre-requisite due to (Rothenbuehler et al. 1982, Gobbetti et al. the high content of proline residues in the 1994e). The sourdough microflora require- gluten (Quaglia 1984). Endopeptidase and ments (Spicher and Nierle 1988, Collar et al. proteinase activities have been shown to be 1991), the environmental factors (Collar and pronounced and distributed in the cytoplasm Martinez 1993, Mascaros et al. 1994) and and cell wall–envelope of Lb. sanfrancisco, yeasts which, as a balance between release respectively. and consumption, lead to a depletion (Collar A cell-envelope 58 kDa proteinase, a cyto- and Martinez 1993) are the main parameters plasmic 67 kDa dipeptidase and 75 kDA ami- which influence the kinetics of amino acids. nopeptidase from Lb. sanfrancisco CB1 have Peptides and amino acids also play an essen- been purified to homogeneity and charac- tial role as important flavour precursors of terized (Table 2) (Gobbetti et al. 1996c). The the baked sourdough products (Spicher proteinase is a serine enzyme most active at 1983). pH 7·0 and 40°C which, when compared with The use of Lb. sanfrancisco in sourdough the activity of the cell-envelope proteinase fermentation has been related to a consider- from Lactobacillus delbrueckii subsp. bulgar-

able increase of the total concentration of free icus B397 and with the PIII proteinase from amino acids (Spicher and Nierle 1984, Lactococcus lactis subsp. lactis SK11, is

Gobbetti et al. 1994e). When compared with characterized by lower activity on αs1- and β- the unstarted sourdough, the sourdough fer- caseins and by a higher capacity to produce mented by Lb. sanfrancisco showed increases peptides from the gliadin. A higher substrate Lactobacillus sanfrancisco a key sourdough lactic acid bacterium 181 specificity for vegetable proteins than for acids (propionic, isobutyric, butyric, α-methyl caseins and a great adaptability to the sour- n-butyric, isovaleric, and valeric acids) gener- dough environment have been assumed. Both ated by the activity of microbial enzymes also dipeptidase and aminopeptidase have opti- contributed to the titratable¨ acidity and flav- mum activity at pH 7·5 and 35–30°C, respect- our (Galal et al. 1978). Lonner and Preve- ° ively. The dipeptidase is a metalloenzyme Akesson (1988) studying the properties of which shows high affinity for peptides con- LAB in sourdough have found that Lb. san- taining hydrophobic amino acids but has no francisco is the heterofermentative bacteria activity on tri- or larger peptides showing with the best properties. several features in common with another A characterization of sourdough LAB dipeptidase isolated from Lb. delbrueckii based on nine parameters of acidification subsp. bulgaricus B14 (Wohlrab and rates was done (Gobbetti et al. 1995c, d). Lb. Bockelmann 1992). After glutamic acid and sanfrancisco strains isolated from sour- proline, the hydrophobic amino acids are the doughs are characterized by the highest major amino acid residues contained in the variability among the heterofermentative gluten of wheat flour (Carnovale and Miuccio species and differ markedly from the collec- 1981). Aminopeptidase is also inhibited by tion cultures. They are characterized by a metal-chelating agents, shows a broad N-ter- rather long latency phase, high maximum- minal hydrolytic activity including di- and acidification rate and by tolerance to acidity. tri-peptides and, according to the classifi- These pre-requisites lead to an elevated pro- cation proposed by Tan et al. (1993), has been duction of lactic and acetic acids (3·48–3·70 g defined as an aminopeptidase type N. All the kg−1 and 0·38–0·44 g kg−1, respectively). characterized enzymes from Lb. sanfrancisco Acetic acid in heterofermentative cultures CB1 maintain relatively high activity at the accounts for about 16–29% by weight of the pH (4·0–5·0) and temperature (30–35°C) of total lactic acid (Lund et al. 1989) but this sourdough fermentation. percentage varies according to the firmness of the dough and the fermentation temperature (Salovaara and Valjakka 1987, Onno Acidification rates and production of and Roussel 1994). The screening of Lb. san- volatile compounds by Lb. francisco cultures showed strains which dif- sanfrancisco fered completely due to their fastest acidification. The total lactic acid produced by Lb. Heterolactic metabolism by means of fermen- sanfrancisco approaches those of a weak tation quotient mainly influences the flavour homofermentative strain such as Lactobacil- of the various leavened baked products lus alimentarius (Gobbetti et al. 1995d). ¨ ° (Spicher 1983, Lonner and Preve-Akesson The flavour of leavened baked goods is 1989). Acetic acid is assumed to act as a influenced by the raw materials (McWilliams flavour enhancer and, together with lactic and McKey 1969, Hansen and Hansen acid, is a catalyst during¨ the Maillard type- 1994a), sourdough fermentation (Hansen et reaction (Seibel and Brummer 1991). Minor al. 1989a, Lund et al. 1989), proofing, baking

Table 2. Characteristics of the purified proteinase, dipeptidase and aminopeptidase from Lactobacillus sanfrancisco CB1 Proteinase Dipeptidase Aminopeptidase Molecular mass 58 kDa 67 kDa 75 kDa Type of enzyme serine-proteinase metalloenzyme metalloenzyme pH optimum 7·0 7·5 7·5 Relative activity at higher than 70% higher than 70% higher than 70% pH 4·0–5·0 Temperature optimum 40°C35°C30°C Substrate specificity Gliadin Leu–Leu Leu–pNA 182 M. Gobbetti and A. Corsetti

and by the type of starters (Hansen et al. other homo- or heterofermentative LAB 1989b). The volatile flavour compounds of the and/or S. exiguus are characterized by a bal- wheat bread crust have been identified by anced profile, the sourdoughs produced with Schieberle and Grosch (1985) which also the association Lb. sanfrancisco–S. cerevisiae determined the differences between the 141 contained higher concentrations of the bread crumb and crust and the variations of yeast fermentation products (1-propanol, 2- these potent odorants during storage of methyl-1-propanol and 3-methyl-1-butanol) wheat bread (Schieberle and Grosch 1992, and a lower amount of the bacterial com- 1994). Even though the greatest amounts of pounds (Gobbetti et al. 1995e, Damiani et al. aroma substances are formed during baking 1996). An activation of the yeast metabolism (Spicher 1983), sourdough fermentation is in the presence of the homofermentative LAB essential for achieving an acceptable flavour was demonstrated by Hansen et al. (1989b), because chemically acidified breads failed in but the same effect, probably due to the com- sensory quality (Rothe and Ruttloff 1983). bination of bacterial acidification and proteol- The importance of the individual starters ysis (Levesque 1991, Gobbetti et al. 1994e), used and of the microbial interactions on the may be attributed to Lb. sanfrancisco. production of volatile compounds has been Hansen and Hansen (1994b) used the associ- carefully considered (Hansen and Hansen ation of Lb. sanfrancisco, Lactobacillus plan- 1994b). tarum and S. cerevisiae to guarantee an equi- Sourdoughs fermented with the heterofer- librated aroma in wheat sourdough breads. mentative lactobacilli had a higher titratable acidity, higher acetic acid content, lower yeast count and lower content of yeast fer- Antimicrobial activity and genetics mentation products than those fermented of Lb. sanfrancisco with homofermentative lactobacilli (Hansen et al. 1989a). While ethanol and ethylacetate Spicher and Mastik (1988) first reported inhi- were produced in the highest amounts in bition of the typical bacterial flora of flour by sourdoughs fermented with Lb. sanfrancisco, sourdough LAB cultures. Homofermentative ethyl-n-propanoate, butyl-acetate and n-pen- LAB have greater inhibitory effects than het- tyl acetate were only produced in the sour- erofermentative against coliforms (Barber doughs started with yeasts (Hansen and and Baguena 1989). A mixed culture pre-fer- Hansen 1994b). Lb. sanfrancisco strains ment of lactic and propionic acid bacteria has show a wide and homogeneous profile of vol- been used in breadmaking in order to pro- atiles which differ greatly from those of the duce propionic acid for its antimicrobial other heterolactic species (Damiani et al. properties (Javanainen and Linko 1993). 1996). It may be defined as unique among the Larsen et al. (1993) purified and charac- sourdough LAB and irreplaceable in sour- terized the bavaricin A from Lactobacillus dough production. Ethylacetate (less than bavaricus¨ MI401 isolated from sourdoughs with the other species), alcohols (ethanol, 1- and Ganzle et al. (1995) recently purified a propanol, 2-methyl-1-pentanol, 1-heptanol bacteriocin from Lb. reuteri LTH 2584 which and 1-octanol), aldehydes (3-methyl-1-but- differed from the bactericins of meat-associ- anal, heptanal, trans-2-heptanal, octanal and ated lactobacilli, such as sakacin P, or anti- nonanal) and acetic acid are the main com- biotics, such as nisin. A screening among diff- pounds produced by Lb. sanfrancisco. Some erent strains of Lb. sanfrancisco has shown of these compounds, and in particular 3- inhibition of Bacillus subtilis but not against methyl-1-butanal and nonanal (Schieberle sourdough yeasts and moulds (Corsetti et al. and Grosch 1994), have been considered as 1996). Two main groups of Lb. sanfrancisco potent odorants in sourdough rye and wheat strains were differentiated for their inhibi- breads (Hansen et al. 1989b). tory spectrum. According to the classification The microbial interactions also affect the proposed by Jack et al. (1995), a bacteriocin- volatile synthesis. While sourdoughs started like inhibitory substance (BLIS C57) which is with the association of Lb. sanfrancisco and heat-stable (100°C for 20 min), insensitive to Lactobacillus sanfrancisco a key sourdough lactic acid bacterium 183 lipase and α-amylase, of a protein nature, and the Amy gene was expressed (Gobbetti et with an inhibitory spectrum centered about al. 1996a). Lb. sanfrancisco CB1 was also LAB, with a bactericidal or bacteriolytic transformed with pGKV210, pNZ12 and mode of action and with a chromosomally pMG36e by electroporation. Amylase activity located encoding gene has been isolated from was extracellular and retained for at least Lb. sanfrancisco C57. With the exception of 140 generations. The transformant degraded the Lactobacillus fructivorans strains, all the starch despite the presence of maltose, other strains from sourdough, as well as most regardless of its concentration and had simi- of the dairy LAB strains, are inhibited by lar or slightly greater growth than the wild BLIS C57. Listeria monocytogenes is also sen- type on maltose. The expression of the α- sitive. A narrower spectrum of inhibition has amylase activity in Lb. sanfrancisco could generally been observed from bacteriocins potentiate the fermentation ability of this produced by lactobacilli (Klaenhammer strain, reduce the competition between LAB 1993). The production of anti-bacterial sub- and yeasts for the soluble carbohydrates of stances by Lb. sanfrancisco may be related to the flour and have a positive role in reducing its predominance and may contribute to the stailing of baked sourdough products. stability of sourdough products. Basic knowledge about the genetics of sourdough LAB must increase because to Conclusions date¨ very few studies have been conducted. Lonner et al. (1990) have examined the plas- Lb. sanfrancisco is physiologically well- mid contents of LAB isolated from different adapted to the sourdough system (utilization types of sourdoughs, including Lb. sanfranci- of specific amino acids and peptides, presence sco, and compared them with the plasmid of maltose phosphorylase, pronounced proteo- contents of culture collection strains. No plas- lytic activity and synthesis of antimicrobial mids or one plasmid of varying size (ca. compounds), physiologically related to the 4·2–38 kbp) have been identified in the Lb. sourdough yeasts (very strict trophic sanfrancisco strains from the collection, relationships with yeasts are based on funda- while strains isolated from sourdoughs have mental metabolisms) and biotechnologically a high number of plasmids. Wild-type indispensable for producing typical baked strains, other than the inoculated strains, sourdough products (the acidification rate gradually become dominant during sour- and the volatile compound productions dif- dough fermentation. They possess a high fered greatly from the other sourdough LAB). number of plasmids which are assumed to be However, its performances must still be related to the carbohydrate fermentations. improved by selection of strain or, better, of On the contrary, a ca. 17·0 kbp cryptic plas- balanced starter associations, by conditioning mid unrelated to maltose fermentation, bac- the sourdough ecosystem (e.g. use of teriocin production or to the proteolytic additional electron acceptors such as fructose activity has constantly been identified or citrate) and probably, in the future, by (Gobbetti et al. 1995b, 1996b, Corsetti et al. using engineered strains. 1996) in Lb. sanfrancisco strains. A rapid method for reliable and simul- taneous identification of different LAB in fer- Acknowledgements mented food, including Lb. sanfrancisco, has been developed by Ehrmann et al. (1994). The research was supported by the National Sixteen-S and 23-S rRNA-targeted, species- Research Council of Italy, Special Project specific oligonucleotides have been applied as RAISA, Sub-project 4, Paper No. 3090. capture probes in a non-radioactive reverse dot-blot hybridization. pC194Amy, a construct containing an References amylase encoding gene, has been introduced in Lb. sanfrancisco CB1 by electroporation Akdogan, H. and Ozilgen, M. (1992) Kinetics of 184 M. Gobbetti and A. Corsetti

microbial growth, gas production, and dough Damiani, P., Gobbetti, M., Cossignani, L., Cor- volume increase during leavening. Enzyme setti, A., Simonetti, M. S. and Rossi, J. (1996) Microbial. Technol. 14, 141–143. The sourdough microflora. Characterization of Axelsson, L. T. (1993) Lactic acid bacteria: classifi- hetero- and homofermentative lactic acid bac- cation and physiology. In Lactic acid bacteria teria, yeasts and their interactions on the (Eds Salminen, S. and von Wright, A.) pp. basis of the volatile compounds produced. Leb- 1–63, New York, Dekker. ensm. Wiss. u. Technol. 29, 63–70. Barber, S. and Baguena, R. (1989) Microflora of Dills, S. S., Apperson, A., Schmidt, M. R. and the sourdough of wheat flour bread. XI. Saier, M. H. (1980) Carbohydrate transport in Changes during fermentation in the microflora bacteria. Microbiol. Rev. 44, 385–418. of sour doughs prepared by a multi-stage pro- Ehrmann, M., Ludwig, W. and Schleifer, K. H. cess and of bread doughs thereof. Rev. Agro- (1994) Reverse dot blot hybridization: a useful quim. Tecnol. Aliment. 29, 478–491. method for the direct identification of lactic Barber, S., Baguena, R., Benedito de Barber, C. acid bacteria in fermented food. FEMS Micro- and Martinez-Anaya, M. A. (1991) Evolution of biol. Lett. 117, 143–150. biochemical and rheological characteristics Faid, M., Boraam, F., Zyani, I. and Larpent, J. P. and breadmaking quality during a multistage (1994) Characterization of sourdough bread wheat sour dough process. Z. Lebensm. Unters. ferments made in the laboratory by traditional Forsch. 192, 46–52. methods. Z. Lebensm. Unters. Forsch. 198, Berg, R. W., Sandine, W. E. and Anderson, A. W. 287–291. (1981) Identification of a growth stimulant for Galal, A. M., Johnson, J. A. and Varriano-Mar- Lactobacillus sanfrancisco. Appl. Environ. ston, E. (1978) Lactic and volatile (C2–C5) Microbiol. 42, 786–788. organic acids of San Francisco sourdough Carnovale, E. and Miuccio, F. (1981) Tabelle di French bread. Cereal Chem. 55, 461–468. Composizione Degli Alimenti, Rome, Italy, Isti- Galli, A., Franzetti, L. and Fortina, M. G. (1988) tuto Nazionale della Nutrizione. Isolation and identification of sourdough Ceslovszky, J., Wolf, G. and Hammes, W. P. (1992) microflora. Microbiol.-Aliments-Nutr. 6, Production of formate, acetate, and succinate ¨ 345–351. by anaerobic fermentation of Lactobacillus Ganzle, M., Hertel, C. and Hammes, W. P. (1995) pentosus in the presence of citrate. Appl. Mic- Antimicrobial activity of lactobacilli isolated robiol. Biotechnol. 32, 94–97 . from sour dough. In Proceedings of the Lactic Cogan, T. M. (1987) Co-metabolism of citrate and acid bacteria Conference, Cork, Ireland. glucose by Leuconostoc ssp: effects on growth, Gobbetti, M., Corsetti, A., Rossi, J., La Rosa, F. substrates and products. J. Appl. Bacteriol. 63, and De Vincenzi, S. (1994a) Identification and 551–558. clustering of lactic acid bacteria and yeasts Collar, C. Mascaros, A. F., Prieto, J. A. and Bened- from wheat sourdoughs of central Italy. Ital J. ito de Barber, C. (1991) Changes in free amino Food Sci. 1, 85–94. acids during fermentation of wheat doughs Gobbetti, M., Corsetti, A. and Rossi, J. (1994b) started with pure culture of lactic acid bac- The sourdough microflora. Interactions teria. Cereal. Chem. 68, 66–72. between lactic acid bacteria and yeasts: metab- Collar, C., Mascaros, A. F. and Benedito de Bar- olism of amino acids. World J. Microbiol. ber, C. (1992) Amino acid metabolism by Biotechnol. 10, 275–279. yeasts and lactic acid bacteria during bread Gobbetti, M., Corsetti, A. and Rossi, J. (1994c) The dough fermentation. J. Food Sci. 57, sourdough microflora. Interactions between 1423–1427. lactic acid bacteria and yeasts: metabolism of Collar, C. and Martinez, C. S. (1993) Amino acids carbohydrates. Appl. Microbiol. Biotechnol. 41, profiles of fermenting wheat sour doughs. J. 456–460. Food Sci. 58, 1324–1328. Gobbetti, M., Corsetti, A. and Rossi, J. (1994d) Condon, S. (1987) Responses of lactic acid bacteria The sourdough microflora. Evolution of soluble to oxygen. FEMS Microbiol. Rev. 46, 269–280. carbohydrates during the sourdough fermen- Corsetti, A., Gobbetti, M. and Rossi, J. (1994) tation. Microbiol.-Aliments-Nutr. 12, 9–15. Sourdough fermentation in presence of added Gobbetti, M., Simonetti, M. S., Rossi, J., Cossig- soluble carbohydrates. Microbiol.-Aliments- nani, L., Corsetti, A. and Damiani, P. (1994e) Nutr. 12, 377–385. Free D- and L-amino acid evolution during Corsetti, A., Gobbetti, M. and Smacchi, E. (1996) sourdough fermentation and baking. J. Food Antibacterial activity of sourdough lactic acid Sci. 59, 881–884. bacteria: isolation of a bacteriocin-like inhibi- Gobbetti, M., Corsetti, A. and Rossi, J. (1995a) tory substance from Lactobacillus sanfrancisco Interactions between lactic acid bacteria and C57. Food Microbiol. 13, 447–456. yeasts in sour dough using a rheofermento- Cunha, M. V. and Foster, M. A. (1992) Sugar-gly- meter. World J. Microbiol. Biotechnol. 11, cerol cofermentations in Lactobacilli: the fate 625–630. of lactate. J. Bacteriol. 174, 1013–1019. Gobbetti, M., Corsetti, A. and Rossi, J. (1995b) Lactobacillus sanfrancisco a key sourdough lactic acid bacterium 185

Maltose-fructose co-fermentation by Lacto- tic acid bacteria and sourdough yeasts. Z. Leb- bacillus brevis subsp. lindneri CB1 fructose- ensm. Unters. Forsch. 198, 202–209. negative strain. Appl. Microbiol. Biotechnol. Hochstrasser, R. E., Ehret, A., Geiges, O. and 42, 939–944. Schmidt-Lorenz, W. (1993) Microbiology of Gobbetti, M., Corsetti, A. and De Vincenzi, S. dough preparation. III. The microflora of eight (1995c) The sourdough microflora. Charac- bakeries’ wheat doughs produced directly and terization of heterofermentative lactic acid using wheat starter or sourdough. Mitt. Gebi- bacteria based on acidification kinetics and ete Lebensm. Hyg. 84, 581–621. impedance test. Ital. J. Food Sci. 2, 103–111. Jack, R. W., Tagg, J. R. and Ray, B. (1995) Bac- Gobbetti, M., Corsetti, A. and De Vincenzi, S. teriocins of Gram-positive bacteria. Microbiol. (1995d) The sourdough microflora. Charac- Rev. 59, 171–200. terization of homofermentative lactic acid bac- Javanainen, P. and Linko, Y. Y. (1993) Mixed cul- teria based on acidification kinetics and ture pre-ferment of lactic and propionic acid impedance test. Ital. J. Food Sci. 2, 91–101. bacteria for improved wheat bread shelf-life. J. Gobbetti, M., Simonetti, M. S., Corsetti, A., Santi- Cereal Sci. 18, 75–88. nelli, F., Rossi, J. and Damiani, P. (1995e) Vol- Kandler, O. (1983) Carbohydrate metabolism in atile compound and organic acid productions lactic acid bacteria. A. van Leeuwenhoek 49, by mixed wheat sour dough starters: influence 209–224. of fermentation parameters and dynamics dur- Kandler, O. and Weiss, N. (1986) Genus Lacto- ing baking. Food Microbiol. 12, 497–507. bacillus Beijerinck 1901, 212. In Bergey’s Man- Gobbetti, M., Corsetti, A., Morelli, L. and Elli, M. ual of Systematic Bacteriology 2 (Eds Sneath, α (1996a) Expression of -amylase gene from P. H. A., Mair, N. S., Sharpe, M. E. and Holt, J. Bacillus stearothermophilus in Lactobacillus G.) pp. 1208–1234, Baltimore USA, Williams sanfrancisco. Biotechnol. Lett. 18, 969–974. and Wilkins. Gobbetti, M., Smacchi, E., Fox, P., Stepaniak, L. Klaenhammer, T. R. (1993) Genetics of bacterioc- and Corsetti, A. (1996b) The sourdough micro- ins produced by lactic acid bacteria. FEMS flora. Cellular localization and characteriz- Microbiol. Rev. 12, 39–86. ation of proteolytic enzymes in lactic acid bac- Kline, L. and Sugihara, T. F. (1971) Microorgan- teria. Lebensm. Wiss. u. Technol. 29, 561–569. isms of the San Francisco sour dough bread Gobbetti, M., Smacchi, E. and Corsetti, A. (1996c) process. II. Isolation and characterization of The proteolytic system of Lactobacillus san- undescribed bacterial species responsible for francisco CB1: purification and characteriz- souring activity. Appl. Microbiol. 21, 459–465. ation of a proteinase, dipeptidase and amino- Kratochvil, J. and Holas, J. (1984) Studies on pro- peptidase. Appl. Environ. Microbiol. 62, teolysis in rye sour dough. Getreide Mehl Broth 3220–3226. 38, 330–332. Gobbetti, M. and Corsetti, A. (1996) Co-metabolism of citrate and maltose by Lactobacillus Kratochvil, J. and Holas, J. (1988) Use of a cereal brevis subsp. lindneri CB1 citrate-negative protein substrate for characterization of pro- strain: effect on growth, end-products and teolytic enzymes in rye sourdough. Getreide sourdough fermentation. Z. Lebensm. Unters. Mehl Broth 42, 166–169. Forsch. 203, 82–87. Larsen, A. G., Vogensen, F. K. and Josephsen, J. Hammes, W. P. (1990) Bacterial starter cultures (1993) Antimicrobial activity of lactic acid bac- in food production. Food Biotechnol. 4, teria isolated from sourdough: purification and 383–397. characterization of bavaricin A, a bacteriocin Hammes, W. P. and Vogel, R. F. (1995) The genus produced by Lactobacillus bavaricus MI401. J. Lactobacillus.InThe Genera of Lactic Acid Appl. Bacteriol. 75, 113–122. ` Bacteria 2 (Eds Wood, B. J. B. and Holzapfel, Levesque, C. (1991) Etude des potentialitesdela` W. H.) pp. 19–54, London, Blackie Academic & souche S47 de Saccharomyces´ ` cerevisae a pro- Professional. duire´ des alcools superieurs´ a partir´ de com- Hansen, A° ., Lund, B. and Lewis, M. J. (1989a) poses organiques identifies et des precurseurs ¨ de la farine. PhD Thesis. Nantes. Flavour production and acidification of sour- ° doughs in relation to starter culture and fer- Lonner, C. and Preve-Akesson, K. (1988) Acidifi- mentation temperature. Lebensm. Wiss. u. cation properties of lactic acid bacteria in rye Technol. 22, 145–149. ¨ sour doughs. Food Microbiol. 5, 43–58. ° Hansen, A° ., Lund, B. and Lewis, M. J. (1989b) Lonner, C. and Preve-Akesson, K. (1989) Effects of Flavour of sourdough rye bread crumb. Leb- lactic acid bacteria on the properties of sour ensm. Wiss. u. Technol. 22, 141–144. ¨ dough bread. Food Microbiol. 6, 19–35. ´ Hansen, A° . and Hansen, B. (1994a) Influence of Lonner, C., Preve-A° kesson, K. and Ahrne,S. wheat flour type on the production of flavour (1990) Plasmid contents of lactic acid bacteria compounds in wheat sourdoughs. J. Cereal Sci. isolated from different types of sour doughs. 19, 185–190. Curr. Microbiol. 20, 201–207. Hansen, A° . and Hansen, B. (1994b) Volatile com- Lund, B., Hansen, A° . and Lewis, M. J. (1989) The pounds in wheat sourdoughs produced by lac- influence of dough yield on acidification and 186 M. Gobbetti and A. Corsetti

production of volatiles in sourdoughs. Leb- ting, C. V. and Whiting, G. C.) pp. 17–27, Lon- ¨ ensm. Wiss. u. Technol. 22, 150–153. ¨ don, Academic Press. Lutgens, M. and Gottschalk, G. (1980) Why a co- Rocken, W., Rick, M. and Reinkemeier, M. (1992) substrate is required for anaerobic growth of Controlled production of acetic acid in wheat Escherichia coli on citrate. J. Gen. Microbiol. sour doughs. Z. Lebensm. Unters. Forsch. 195, 119, 63–70. 259–263. Lyons, T. P. and Rose, A. H. (1977) Whisky. In Romano, A. E., Brino, G., Peterkofsky, A. and Alcoholic Beverages (Ed Rose, A. H.) pp. Reizer, J. (1987) Regulation of β-galactoside 635–692, London, Academic Press. transport and accumulation in heteroferment- Martinez-Anaya, M. A., Pitarch, B., Bayarri, P. ative lactic acid bacteria. J. Appl. Bacteriol. and Benedito de Barber, C. (1990) Microflora 169, 5589–5596. of the sourdoughs of wheat flour bread. X. Rothe, M. and Ruttloff, H. (1983) Aroma retention Interactions between yeasts and lactic acid in modern bread production. Die Nahrung 27, bacteria in wheat doughs and their effects on 505–512. bread quality. Cereal Chem. 67, 85–91. Rothenbuehler, E., Amado, R. and Solms, J. (1982) Martinez-Anaya, M. A., Pitarch, B. and Benedito Isolation and identification of amino acid de Barber, C. (1993) Biochemical character- derivatives from yeast. J. Agric. Food Chem. istics and breadmaking performance of freeze- 30, 439–442. dried wheat sour dough starters. Z. Lebensm. Rouzard, O. and Martinez-Anaya, M. A. (1993) Unters. Forsch. 196, 360–365. Effect of processing conditions on oligosacchar- Mascaros, A. F., Martinez, C. S. and Collar, C. ide profile of wheat sourdoughs. Z. Lebensm. (1994) Metabolism of yeasts and lactic acid Unters. Forsch. 197, 434–439. bacteria during dough fermentation relating Salovaara, H. and Valjakka, T. (1987) The effect of functional characteristics of fermented doughs. fermentation temperature, flour type, and Rev. Esp. Cienc. Tecnol. Aliment. 34, 623–642. starter on the properties of sour wheat bread. McWilliams, M. and Mckey, A. C. (1969) Wheat Int. J. Food Sci. Technol. 22, 591–597. flavor components. J. Food Sci. 34, 493–496. Schieberle, P. and Grosch, W. (1985) Identification Nelson, L. S., Sriranganathan, N., Dyryee, F. L., of the volatile flavour compounds of wheat Seidler, R. J., Sandine, W. E. and Elliker, P. R. bread crust—comparison with rye bread crust. (1971) Studies on sourdough Lactobacillus Z. Lebensm. Unters. Forsch. 180, 474–478. species. Bacteriological-Proceedings 4. Schieberle, P. and Grosch, W. (1992) Changes in Neubauer, H., Glaasker, E., Hammes, W. P., Pool- the concentrations of potent crust odorants man, B. and Konings, W. N. (1994) Mechanism during storage of white bread. Flavour Fragr- of maltose uptake and glucose excretion in ance J. 7, 213–218. Lactobacillus sanfrancisco. J. Bacteriol. 176, Schieberle, P. and Grosch, W. (1994) Potent odor- 3007–3012. ants of rye bread crust—differences from the Ng, H. (1972) Factors affecting organic acid pro- crumb and from wheat bread crust. Z. Leb- duction by sourdough (San Francisco) bacteria. ensm. Unters. Forsch.¨ 198, 292–296. Appl. Microbiol. 23, 1153–1159. Seibel, W. and Brummer, J. M. (1991) The sour- Nout, M. J. R. and Creemers-Molenaar, (1986) dough process for bread in Germany. Cereal Microbiological properties of some wheatmeal Food World 36, 299–304. sourdough starters. Chem. Mikrobiol. Technol. Spicher, G. (1959) Die mikroflora¨ des sauerteiges. Lebensm. 10, 162–167. I. Mitt.: Untersuchungen uber¨ die art¨ der in Onno, B. and Roussel, Ph. (1994) Technologie et sauerteigen¨ anzutreffenden stabchenformigen microbiologie´ de la panification au levain. In milchsaurebakterien (Genus Lactobacillus Bacteries lactiques 2 (Eds de Roissart, H. and Beijerinck). Zbl. Bakt.¨ II. Abt. 113, 80–106. Luquet, F. M.) pp. 293–321, F-38410 Uriage, Spicher, G. and Schroder, R. (1978) The microflora Lorica. of sour dough. IV. Communication: Bacterial Okada, S., Ishikawa, M., Yoshida, I., Uchimura, composition of sourdough starters. Genus T., Ohara, N. and Kozaki, M. (1992) Identifi- Lactobacillus Beijerinck. Z. Lebensm. Unters. cation and characteristics of lactic acid bac- Forsch. 167, 342–354. teria isolated from sour dough sponges. Biosci. Spicher, G., Rabe, E., Sommer, R. and Stephan, H. Biotech. Biochem. 56, 572–575. (1982) XV. Communication: on the behaviour Okuhara, M. and Harada, T. (1971) Formation of of heterofermentative sourdough bacteria and N-acetyl-L-alanine and N-acetylglycine from yeasts in mixed culture. Z. Lebensm. Unters. glucose by Candida tropicalis OH23. Biochim. Forsch. 174, 222–227. Biophys. Acta 244, 16–22. Spicher, G. (1983). Baked Goods. In Biotechnology Quaglia, G. (1984) Scienza e Tecnologia della Pan- 5 (Eds Rehm, H. J. and Reed, G.) pp. 1–80, ificazione, 2nd edn. Pinerolo, Italy, Chiriotti. Weinheim, Verlag Chemie. Radler, F. (1975) The metabolism of organic acids Spicher, G. and Nierle, W. (1984) The microflora of by lactic acid bacteria. In Lactic Acid Bacteria sourdough. XVIII. Communication: The pro- in Beverages and Food (Eds Carr, J. G., Cut- tein degrading capabilities of the lactic acid Lactobacillus sanfrancisco a key sourdough lactic acid bacterium 187

bacteria of sourdough. Z. Lebensm. Unters. acetate, propionate and sorbate by Saccharo- Forsch. 178, 389–392. myces cerevisiae and Torulopsis holmii. Food Spicher, G. (1987) The microflora of sourdough. Microbiol. 1, 105–110. XXII. Communication: the lactobacillus spec- Suomalainen, H. and Oura, E. (1971) Yeast ies of wheat sourdough. Z. Lebensm. Unters. nutrition and solute uptake. In The Yeasts Forsch. 184, 300–303. (Eds Rose, A. H. and Harrison, J. S.) pp. 3–59, Spicher, G. and Mastik, G. (1988) Interactions London, Academic Press. between the lactobacilli of sourdough and flour Tan, P. S. T., Poolman, B. and Konings, W. N. microflora. Getreide, Mehl Brot 42, 338–342. (1993) Proteolytic enzymes of Lactococcus Spicher, G. and Nierle, W. (1988) Proteolytic lactis. J. Dairy Res. 60, 269–286. activity of sourdough bacteria. Appl. Microbiol. Thorne, R. S. (1957) Brewers’ yeast. In Yeasts (Ed Biotechnol. 28, 487–492. Roman, W.) pp. 79–113, The Hague, DR. W. Sriranganathan, N., Seidler, R. J., Sandine, W. E. Junk. ¨ and Elliker, P. R. (1973) Cytological and deox- Vogel, R. F., Bocker, G., Stolz, P., Ehrmann, M., yribonucleic acid-deoxyribonucleic acid Fanta, D., Ludwig, W., Pot, B., Kersters, K., hybridization studies on Lactobacillus isolates Shleifer, K. H. and Hammes, W. P. (1994) from San Francisco sourdough. Appl. Micro- Identification of Lactobacilli from sourdough biol. 25,¨461–470. and description of Lactobacillus pontis sp. nov. Stolz, P., Bocker, G., Vogel, R. F. and Hammes, W. Int. J. System. Bacteriol. 44, 223–229. P. (1993) Utilization of maltose and glucose by Vollmar, A. and Meuser, F. (1992) Influence of lactobacilli isolated from sourdough. FEMS starter cultures consisting of lactic acid bac- Microbiol. Letters 109, 237–242. ¨ teria and yeasts on the performance of a con- Stolz, P., Bocker, G., Hammes, W. P. and Vogel, R. tinuous sourdough fermenter. Cereal Chem. F. (1995a) Utilization of electron acceptors by 69, 20–27. lactobacilli isolated from sourdough. I. Lacto- bacillus sanfrancisco. Z. Lebensm. Unters. Weiss, N. and Schillinger, U. (1984) Lactobacillus Forsch. 201, 91–96. sanfrancisco sp. nov., nom. rev. System Appl. Stolz, P., Vogel, R. F. and Hammes, W. P. (1995b) Microbiol. 5, 230–232. Utilization of electron acceptors by lactobacilli Wlodarczyk, M. (1985) Associated cultures of lac- isolated from sourdough. II. Lactobacillus tic acid bacteria and yeasts in the industrial pontis, L. reuteri, L. amylovorus and L. fer- production of bread. Acta Alimentaria Polonica mentum. Z. Lebensm. Unters. Forsch. 201, XI, 345–359. 402–410. Wlodarczyk, M., Jezynska, B. and Warzywoda, A. Sugihara, T. F., Kline, L. and McCready, L. B. (1993) Associated cultures of lactic acid (1970) Nature of the San Francisco sour dough bacteria and yeast as starters in gluten-free French bread process. II. Microbiological baker’s leavens. Polish J. Food Nutr. Sci. 2/43 aspects. Bakers Dig. 44, 51–56. (1), 83–91. Sugihara, T. F. and Kline, L. (1975) Further stud- Wohlrab, Y. and Bockelmann, W. (1992) Purifi- ies on a growth medium for Lactobacillus san- cation and characterization of a dipeptidase francisco. J. Milk¨ Food Technol. 38, 667–672. from Lactobacillus delbrueckii subsp. bulgar- Suihko, M. L. and Makinen, V. (1984) Tolerance of icus. Int. Dairy J. 2, 354–361.