HAL Id: hal-00929265https://hal.archives-ouvertes.fr/hal-00929265

Submitted on 1 Jan 1991

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Inhibiting factors produced by lactic acid bacteria. 1.Oxygen metabolites and catabolism end-products

Jc Piard, M Desmazeaud

To cite this version:Jc Piard, M Desmazeaud. Inhibiting factors produced by lactic acid bacteria. 1. Oxygen metabolitesand catabolism end-products. Le Lait, INRA Editions, 1991, 71 (5), pp.525-541. �hal-00929265�

Lait (1991) 71, 525-541© Elsevier/INRA

525Review

Inhibiting factors produced by lactic acid bacteria.1.0xygen metabolites and catabolism end-products

Je Piard *, M Desmazeaud

INRA, Station de Recherches Laitières, Unité de Microbiologie, 78350 Jouy-en-Josas, France

(Received 29 April 1991; accepted 7 June 1991)

Summary - Lactic acid bacteria can produce a variety of antimicrobial compounds which mayaffect both the lactic acid bacteria themselves as weil as undesirable or pathogenic strains. In thisfirst section, we describe the biosynthesis, mode of action and activity spectra of these inhibitors.Metabolites of oxygen (hydrogen peroxide and free radicals) exhibit a bacteriostatic or bactericidalactivity against lactic or non-Iactic flora. When associated with the lactoperoxidase/thiocyanatesystem, hydrogen peroxide leads to the formation of inhibitory compounds which are bacteriostaticfor lactic acid bacteria and bactericidal for Gram-negative bacteria. The antimicrobial activity oforganic acids (Iactic, acetic and formic) and of pH are closely linked. It appears that thenon-dissociated fraction of these acids is the major inhibitory form. Thus, acetic acid whose pKa ishigher than that of lactic acid exhibits the highest inhibitory activity. The antimicrobial activities ofdiacetyl, acetaldehyde and of the D isomers of amino acids are also described, although their effectsare slight in usual lactic fermentations. 8acteriocins are the last type of inhibitory substanceproduced by some lactie acid bacteria. They will be described in a second section.

lactic acid bacteria / oxygen 1 acidity 1diacetyll acetaldehyde

Résumé - Facteurs inhibiteurs produits par les bactéries lactiques. 1. Métabolites de l'oxy-gène et produits du catabolisme. Les bactéries lactiques sont capables de produire une variété deproduits inhibiteurs dont les effets peuvent se répercuter sur la flore lactique el/e-même mais aussisur les flores indésirables ou pathogènes. Dans cette première partie, nous décrivons les méca-nismes d'apparition, les modes d'action et les spectres d'activité de ces inhibiteurs. Les métabolitesde l'oxygène (le peroxyde d'hydrogène et les radicaux libres) peuvent avoir des effets biologiques denature bactériostatique ou bactéricide sur la flore lactique ou non lactique. Le peroxyde d'hydrogèneassocié au système lactoperoxydase/thiocyanate, catalyse la formation de produits inhibiteurs, bac-tériostatiques pour la flore lactique et bactéricides pour les bactéries Gram-négative. Les pouvoirsantibactériens des acides organiques et du pH sont intimement liés. 1/ semble que la fraction nondissociée de ces acides soit la forme inhibitrice majeure. De ce fait l'acide acétique de pKa supérieurau pKa de l'acide lactique, a un pouvoir inhibiteur supérieur. Les activités inhibitrices du diacétyle, del'acétaldéhyde et des acides aminés de forme isomérique D sont décrites bien que leurs effetssoient mineurs dans les fermentations lactiques habituel/es. Les bactériocines produites par cer-taines bactéries lactiques seront décrites dans une seconde revue.

bactérie lactique / oxygène / acidité / diacétyle / acétaldéhyde

* Correspondence and reprillts

526

pH, diacetyl, acetaldehyde, 0 isomers ofamino acids). In the second section, wewill examine inhibitions due to small antimi-crobial molecules or to bacteriocins. Inhibi-tion resulting from inter-strain competition(Juillard et al, 1987) is not within the scopeof this literature review and so will not bediscussed.

For convenience, Lactococcus lactissubsp lactis, L lactis subsp lactis biovar di-acetylactis and L lactis subsp cremoris willbe designated L tectis, L diacetylactis andL cremoris respectively.

Je Piard,M Desmazeaud

INTRODUCTION

Aside from their fundamental roles in theformation of organoleptic and rheologicalcharacters during the transformation ofmilk to cheese and yogurt, lactic acid bac-teria are used and studied for their capaci-ty to inhibit unwanted bacteria, and thusfor increasing the shelf-life of the products.

It is known empirically that the decreaseof pH in itself is a factor of fundamental im-portance in food safety, since it leads tothe elimination of unwanted bacteria andeven pathogens. This is important sincethe ecological niches favorable to lacticacid bacteria have high water activity andare rich in nutrients, thus being very favor-able for bacteria proliferation. However,the capacity of lactic acid bacteria to effec-tively inhibit other species not only resultsfrom their ability to lower the pH, but de-pends on the nature of the organic acidsthat they produce. Also, other antimicrobi-ais such as oxygen metabolites (eg hydre-gen peroxide), bacteriocins, diacetyl, acet-aldehyde and 0 isomers of amine acidsare Iikely to enhance the inhibitory activityof lactic acid bacteria.

Renewed interest in these aspects ofthe metabolism of lactic acid bacteria hasoccurred. Industrial accidents leading tofood poisoning, notably by Listeria mono-cytogenes have provided the impetus forresearch on the potential use of lactic acidbacteria to improve food safety and evento create new fermented products.

The present review will deal with cur-rent knowledge on the inhibitory potentialof lactic acid bacteria. We will also discussthe possible benefit to food safety thatcould result from better control of thesevarious inhibitory systems.

This section deals with inhibition phe-nomena due to oxygen metabolites andend-products of catabolism (organic acids,

OXYGEN METABOLITES

Mechanism of appearance

Lactic acid bacteria are called facultativeanaerobes, or aerotolerant, and they cata-bolize sugars via fermentation. This meansthat electron transport chains do not inter-vene, but in subsequent oxidoreductionsteps nicotinamide adenine dinucleotide(NAD) participates, undergoing cycling oxi-dation and reduction. In aerobic conditions,these NAD molecules react with oxygen toform hydrogen peroxide (H202) or water asa result of the action of various enzymes(table 1).The dismutation of endogenoussuperoxide anions (02-) (table 1)can alsoadd a slight contribution to the accumula-tion of H202 by the action of a superoxidedismutase (SaD) present in most lacticacid bacteria (Britton et al, 1978) or as aresult of manganese, present in high con-centrations (20-25 mmol/I) in the cyto-plasm of bacteria lacking SaD (Archibaldand Fridovitch, 1981).

With the exception of several isolatedcases (Whittenbury, 1964; Kono and Frido-vitch, 1983), lactic acid bacteria lack cata-lase because they cannot synthesize he-moporphyrins. Thus, lactic acid bacteriacan rid themselves of hydrogen peroxide

Inhibition by lactic acid bacteria 527

Table 1. Reactions between lactic acid bacteria and oxygen or oxygen metabolites (atter Archibaldand Fridovich, 1981; Condon, 1987).Reactions des bactéries lactiques avec l'oxygène ou ses métabolites (d'après Archibald et Fridovich,1981; Condon, 1987).

Reactions and catalysing enzymes Lactic acid bacteriain which the enzymehas been identified

NADH :H,O, oxidase

NADH + H+ + 02 > NAD+ + H202

NADH:H,O oxidase

2NADH + 2H+ + 02 > NAD+ + 2H20

pyruvate oxidase

Most lactic acid bacteria

Most lactic acid bacteria

Pyruvate + phosphate + 02 ------:> acetylphosphateTPP, FAD + CO2 + H202

L delbrueckii, L plantarum,P halophilus, S mutans

oxidase a-GP

a-Glycerophosphate + 02 > dihydroxyacetone phosphate+ H202

Lactic acid bacteria utilizingglycerol in aerobiosis(enterococci,lactobacilli,P pentosaceus)

202- + 2H+ > H202 + 02 Most lactic acid bacteria

NADH peroxidase

NADH + H+ + H202 > 2H2O Most lactic acid bacteria

Mn2+

202-+ 2 H+ > H202 + 02 Bacteria lacking SOD

TPP: thiamine pyrophosphate; FAD: flavine adenine dinucleotide; GP: glycerophosphate; SOD: superoxide dismu-tase.

SOD

formed, only by their NADH peroxidase. Itfollows that the concentration of H202formed in aerobiosis is variable and de-pends on the activities of NADH oxidaseand NADH peroxidase, catalyzing the for-mation and degradation of H202 (Anderset al, 1970).

The simultaneous presence of hydrogenperoxide and superoxide anions can bethe origin of hydroxyl radical formation ac-cording to the reaction: 02- + H202 ->

OH- + OH" + O2 (Gregory and Fridovltch,1974). Aerobic growth of lactic acid bacte-ria thus leads to the formation of 3 princi-pal derivatives of oxygen: H202• 02- andOH" which are vectors of "oxygen toxicity".

Mode of action

Although the toxicity of oxygen has beendernonstrated, its modes of action havenot been as c1earlyelucidated. Two types

528

presence of molecular oxygen may rein-force the damage, perhaps by a direct ef-fect at the level of the affected membranesites.

Bactericidal effect

Free radicals and hydrogen peroxide candamage bacterial nucleic acids, leadingto reversible or irreversible alterations.Ananthaswamy and Eisenstark (1977)demonstrated the modification of DNA byhydrogen peroxide by showing that E colimutants lacking DNA repair systems weremore sensitive to H202 than strains with in-tact repair systems. Hydrogen peroxide ap-parently causes breaks in the carbon-phosphate backbone of DNA, releasing nu-c1eotidesand preventing chromosome rep-Iication (Freese et al, 1967).

Hydroxyl radicals can attack the methylgroup of thymine, thereby damaging DNA(Byczkowski and Gessner, 1988). Activemolecular oxygen can also react with gua-nine and cause breaks in one strand ofDNA. This was shown in E coli, wheretreatment of plasmid DNA prevented trans-formation (Di Mascio et al, 1989).

Gram-negative bacteria are better pro-tected from the toxic effect of oxygen me-tabolites because of the presence of anouter Iipopolysaccharide layer. This enve-lope traps active molecular oxygen (Dahlet al, 1989). Thus, the kinetics of cell deathis a multistep process in Salmonella sp(multihit killing), while it is linear in the caseof L lactis. Lactic acid bacteria have an ad-ditional handicap in the resistance to mo-lecular oxygen because they lack mem-brane carotenoids. Active molecularoxygen thus diffuses through the peptido-glycan layer rapidly and reacts with sensi-tive membranes sites (Dahl et al, 1989).

Je Piard, M Desmazeaud

may be distinguished, one leading to abacteriostatic effect, the other being bacte-ricidal.

Bacteriostatic effect

Homolactic bacteria catabolize glucose viathe Embden-Meyerhof pathway, while het-erolactic bacteria follow the pentose phos-phate shunt. In the case of homo/actic fer-mentation, the 2 molecules of NADreduced during the formation of glyceral-dehyde-3-phosphate (G3P) are reoxidizedto favor the reduction of pyruvate to lacticacid by a lactic dehydrogenase (LDH).When the cells are in aerobic conditions,their NADH oxidase and NADH peroxi-dase systems are induced (Murphy andCondon, 1984), which compete with LDHfor available NADH. This results in a de-crease in the acidifying activity of the bac-teria. The same reasoning applies to het-erofermenting bacteria using the pentosephosphate pathway.

The oxidation of sulfhydryl compoundsmay have a bacteriostatic or bactericidaleffect, depending on whether or not the re-action is reversible when the cells return toanaerobic conditions. The denaturing ef-fect of oxygen and its metabolites on G3Pdehydrogenase has been shown (Kongand Davison, 1980) and it is probable thatoxidation of sulfhydryl to disulfide affects anumber of enzymes (lactate dehydroge-nase, atcohol dehydrogenase) or coen-zymes with sulfhydryl groups (coenzymeA) (Haugaard, 1968).

The toxicity of oxygen may result fromthe peroxidation of membrane lipids(Haugaard, 1968), which would explain theincreased membrane permeability causedby H202, O2- and OH" (Kong and Davi-son, 1980). This hypothesis was support-ed by the finding that the degree of mem-brane lipid peroxidation was related to a1055 of viability by E coli (Harley et al,1978). Once cell envelopes are damagedby the above-mentioned metabolites, the

Effect on lactic flora

The relative toxicity of H202• O2- and OH"towards bacteria is somewhat controver-

Inhibitionby lacticacidbacteria

sial. Concerning lactococci, there is an ap-parent consensus that hydrogen peroxideis the major inhibitor (Condon, 1987).Anders et al (1970) reported that 0.2mmol/l H202 was the concentration whichinhibited the growth of these bacteria by50% and that 1.5 mmol/l induced celldeath. Autoinhibition by H202, however, isnot universal among lactococci. Thus,Smart and Thomas (1987) showed that inoxygen-insensitive strains, there is an ade-quateness between the activities ofNADH:H202 oxidase and NADH peroxi-(jase,' preventing the accumulation of hy-drogen peroxide. In some strains, reduc-tion of oxygen to water by an NADH:H202oxidase enables these bacteria to remainunaffected by aerobiosis.

Although inhibition is occasionally ob-served in lactobacilli in aerobic conditions,the modes of action remain to be estab-Iished. It is generally admitted that hydro-gen peroxide is not responsible for this in-hibition (Gregory and Fridovitch, 1974;Yousten et al, 1975; Condon, 1983). Someauthors have advanced the notion that theabsence of SOD in Lactobacillus could en-able the formation of hydroxyl radicals viaa reaction between the superoxide anionand hydrogen peroxide (Gregory andFridovitch, 1974). The authors attributedthe toxicity of oxygen to these OH" radi-cals. Subsequently, however, Archibaldand Fridovitch (1981) reported that lacticacid bacteria lacking SOD had a dismuta-tion activity resulting from their high Mn2+

levels (see table 1).This system protectsthem from the superoxide anion and pre-vents the formation of hydroxyl radicals.The authors concluded that inhibition re-sulted from the combination of several fac-tors (Condon, 1983) or from the shuntingof carbohydrate metabolism towards othermetabolic pathways where energy produc-tion is slower (Murphy and Condon, 1984).

Recent data suggest that lactic acidbacteria can adapt to oxygen. Lactococci

sensitive to hydrogen peroxide and -ex-posed to a sublethal concentration of thecompound became capable of growth inthe presence of a lethal concentration ofhydrogen peroxide (Condon, 1987). Theauthor observed the simultaneous induc-tion of NADH peroxidase and, to a lesserextent, that of NADH oxidase. The sameinduction was shown when cells were sub-jected to a sublethal temperature, leadingthe author to suggest that it may be aquestion of stress proteins. These induc-tions may partiaily explain the variability ofresults presented above.

Similar phenomena have been de-scribed in other bacteria. Thus, 30 proteinsare induced in Salmonella typhimuriumwhen cells are treated with sublethal dosesof hydrogen peroxide (Morgan et al, 1986).Five of these proteins are also induced af-ter a heat shock. Other stresses, such astreatment with nalidixic acid or ethanol,lead to the synthesis of the same 30 pro-teins, in addition to others. On the otherhand, the proteins induced by hydrogenperoxide are not the same as those in-duced by aerobiosis.

Effect on non-Iactic flora

Among the non-Iactic acid bacteria thatmay be present in rnilk, it is of interest tonote that the catalase + character of un-wanted Gram-negative bacteria (Pseudo-mones, E coli) does not render them par-ticularly tolerant to oxygen metabolites(Rolfe et al, 1978). Price and Lee (1970)concluded that hydrogen peroxide was re-sponsible for the inhibition of Pseudomo-nas, Bacillus and Proteus by Lb plantarum.Dahiya and Speck (1968) reported that hy-drogen peroxide produced by Lactobacillusinhibited Staphylococcus aureus and theoptimal H202 production was obtained at5 -c.

ln conclusion, oxygen metabolites mayaffect both lactic acid flora itself as weil as

529

530

mechanism of inhibition of the system butalso stabilizes lactoperoxidase at high dilu-tions (Steele and Morrison, 1969).

Native milk does not contain hydrogenperoxide (Reiter and Hârnulv, 1984), al-though it has been observed that milk lac-tic acid flora (Iactococci and lactobacilli) at-ter milking can produce H202 in aerobicconditions, even when the milk is stored atlow temperature (Dahiya and Speck,1968). Saliva taken aseptically from hu-man salivary glands contains hydrogenperoxide (Pruitt et al, 1982).

JC Piard, M Desmazeaud

unwanted species. Various avenues of in-vestigation are promising for obtaining op-timal benefit from the inhibition of thesemetabolites. The construction of strainscontaining amplified NADH oxidase sys-tems will lead to excess production ofH202. Detailed studies on the mechanismsof resistance to these compounds (notablyvia stress proteins) should be undertakenin parallel to avoid phenomena of autoinhi-bition.

THE LACTOPEROXIDASE-THIOCYANATE-HYDROGENPEROXIDE SYSTEM OF MILK

The inhibition observed is based on the in-teraction between three compounds: lac-toperoxidase, thiocyanate and hydrogenperoxide. Lactoperoxidase is the namegiven to milk peroxidase, of which cow'smilk contains 10-30 j1g/ml (Reiter, 1985).This activity is more than sufficient to acti-vate the lactoperoxidase system (LPS),since Pruitt and Reiter (1985) showed thatthe reactions of the LPS could be cata-Iyzed by a lactoperoxidase concentrationas low as 0.5 ng/ml.

Lactoperoxidase can tolerate an acidityequivalent to pH 3 in vitro (Wright, 1958)and resists human gastric juice (Gotheforsand Marklund, 1975). It is relatively heat-stable, since it is only partially inactivatedby a short pasteurization at 74°C (Wrightand Tramer, 1958), after which there sub-sists enough residual activity for the LPSto remain effective.

The thiocyanate anion (SCN-) is widelydistributed in animal secretions and tis-sues and its concentration is a reflection ofeating habits and lifestyle of the individuals(Reiter and Hârnulv, 1984). The thiocya-nate concentration in milk varies from 1 to10 ppm and the compound has a dual rolein the LPS: it is directly involved in the

Mode of action and biological effects

ln the presence of H202, lactoperoxidasecatalyzes the oxidation of thiocyanate tonon-inhibitory end-products, such asS042-, CO2 and NH3 (Oram and Reiter,1966). In the course of this oxidation, how-ever, more or less inhibitory intermediatesform, eg hypothiocyanate (OSCN-) (Auneand Thomas, 1977) or higher oxyacids02SCN- and 03SCN- (Pruitt et al, 1982).

The major effect of these metabolitesoccurs at the level of the oxidation of sulf-hydryl . groups of metabolic enzymes(Thomas and Aune, 1978). Thus, hexoki-nases are totally inhibited, and aldolasesand 6-phosphogluconate dehydrogenaseare partially inactivated by the LPS. Lawand John (1981), however, showed thatenergy-transducing o-lactate dehydroge-nase was also inhibited by this system,even though its active site does not con-tain SH groups.

The LPS also causes lesions in thecytoplasmic membrane, causing the leak-age of potassium ions, amino acids andpolypeptides from the cell. Sugar and ami-no acid transport systems are also inhibit-ed, as are the synthesis of DNA, RNA andproteins (Reiter and Hârnulv, 1984).

As in the case of the direct toxicity ofoxygen metabolites, the catalase + charac-

Inhibition by lactic acid bacteria 531

8 CV E. coli 8 ® L. loctis

7 7

E6 6::>

u,u 5a 50;a-l 4 4

3 32 4 6 2 4 6Time (h) Time (h)

Fig 1. Effect of various concentrations of OSCN- on the viability of E coli (a) and L lactis (b) in synthe-tic medium pH 5.5. OSCN- concentrations: (0), 0; (.), 5 urnol/l: (0), 10 Jlmol/l; (.), 20 Jlmolll; (L'i), 25Jlmol/l; (Â), 35 Ilmol/l (after Marshall and Reiter, 1980; reprinted by permission of the Society for Gen-erai Microbiology).Effet de concentrations variables de OSCfIr sur la viabilité de E coli (a) et de L lactis (b) dans du mi-lieu synthétique à pH 5,5. Concentrations de OSCfIr: (0), 0; (e), 5 j.lmoll/; (0), 10 j.lmoll/; (_), 20umou'; (.1), 25 umolâ; (~), 35 j.lmol/I (d'après Marshall et Reiter, 1980; reproduit avec la permissionde la Society for General Microbiology).

ter of Gram-negative bacteria confers noadvantage on them, and is even detrimen-tai. Marshall and Reiter (1980) showedthat the LPS is bacteriostatic for L lactisand bactericidal for E coli (fig 1), whichmay be due to structural differences in thecell walls. Oram and Reiter (1966) showedthat LPS-resistant strains of lactic acidbacteria contained an enzyme which cata-Iyzed the oxidation of NADH2 in the pres-ence of an intermediate product of thiocya-nate oxidation. lt is th us possible that lacticacid bacteria can neutralize the oxidationproducts of thiocyanate or else can repairthe damage caused (Reiter and Hàrnulv,1984).

The LPS has a broad spectrum of activi-ty, including most Gram-negative bacteriafound in milk, eg Pseudomonas and E coli(Bjërck et al, 1975), Salmonella (Purdy etal, 1983) and Shigella (table Il). Pruitt andReiter (1985) published a Iist of the biologi-cal effects caused by the LPS in bacteria.

Among undesirable Gram-positive species,Siragusa and Johnson (1989) showed thatthe LPS delayed the development ofL monocytogenes.

Field of application

ln raw milk, the LPS is a powerful inhibitorfrom hydrogen peroxide producing strains.ln addition, this factor augments the toxici-ty of H202 by a factor of about 100: activeagainst lactococci at a concentration of 1mmol/l (Anders et al, 1970), hydrogen per-oxide present at 10 nmol/I is sufficient toactivate the LPS (Thomas, 1985).

The acid resistance of the LPS and thefact that SCN- arises from raw milk or fromthe digestion of certain plants led research-ers to examine the antibacterial activity ofthe LPS in vivo. This was shawn in calveshaving absorbed raw milk enriched with Ecoli: only 1 to 10% of the strain was foundin the abomasum, the fourth (true) stom-

532 Je Piard, M Desmazeaud

Table Il. Antibacteriai activity of the lactoperoxi-dase system against various bacteria (aflerAsperger, 1986; Siragusa and Johnson, 1989).Activité antibactérienne du système lactope-roxydase contre diverses bactéries (d'aprèsAsperger, 1986; Siragusa et Johnson, 1989).

Target strain Biological activity

Bacterio- Bacteri-static cidal

Lactic acid bacteriaLactococcus sp xThermophilic lactobacilli x

Spoilage bacteriaPseudomonas f1uorescens xGram-negative bacteria xBacil/us cereus x

Pathogenic bacteriaStaphylococcus aureus xEscherichia coli xSalmonella sp xPseudomonas aeruginosa xListeria monocytogenes x

ach of ruminants, whereas if one of thecomponents of the LPS system was neu-tralized, the number of viable coliform bac-teria was maintained (Reiter and Hârnulv,1984).

The safety of this system towards eu-karyotic cells remained to be shown.Reiter and Hârnulv (1984) reported thatthe LPS did not change the respiration ofliver cells in vitro and did not cause Iysis oferythrocytes. Hanstrôrn et al (1983) evenshowed that the LPS protected humancells from hydrogen peroxide toxicity.

The LPS is thus a valuable system forsome authors, who suggested improvingthe storage properties of raw milk by add-ing traces of SCN- (12 ppm) and H202 (8

ppm) (Reiter and Hârnulv, 1984). Thispractice led to a reduction in the undesira-ble flora of raw milk, especially Pseudomo-nas sp, whose proteolytic and Iipolytic ac-tivities pose an important problem whenprocessing milk.

The LPS has been used in pharmacolo-gy by incorporating the ingredients neces-sary to activate the system in toothpastesor contact lens solutions. This system willundoubtedly be developed in the future inthe food and agriculture industry. The LPSis a very interesting system in this field,since it is bactericidal for Gram-negativebacteria and only bacteriostatic for Gram-positive species. Whenever we speak of abacteriostatic effect, it is almost certainly asaturable system. lts use to improve thestorage of raw milk thus should not havemajor effects on the subsequent growth oflactic starters. It is also to be noted that ifthe milk is pasteurized before processing,a part of the lactoperoxidase is inactivated.

END-PRODUCTS OF CATABOLISM

End-products of carbohydratecatabolism, effect of pH

The production of acids by lactic acid bac-teria contributes to energy production inthis bacterial group, as we will see below.The acidity which forms is detrimental forthe development of unwanted bacteria andthus greatly improves the hygienic qualityof dairy products. However, this aéidity canalso affect the lactic flora itself. It is to beremembered that the dairy industry mostoften uses mixed cultures of lactic acidbacteria whose composition is more orless known with certainty. A different acidi-fying activity and a different acid toleranceof the strains composing the starter maylead to the dominance of the more acid-resistant strains.

Inhibition by lactic acid bacteria 533

Bioenergetics in /actic acid bacteria

Bacteria have 2 mechanisms for generat-ing ATP:

- phosphorylation of ADP to ATP via gly-colysis intermediates;

- chemosmotic energy characterized by aproton gradient.

The latter pathway functions by thepresence of the following secondary trans-port systems:

- coupling of lactate efflux and protons (fig2) in the ratio of 2 H+/lactate (Konings,1985);

- coupling between the entry of solutesand the influx of protons at 1 or 2 Hr/solutemolecule.

Finally, the ATPase system leads to theexpulsion of protons with energy consump-tion (1 mol ATP/2H+) or the entry of pro-tons with energy production.

During glycolysis, the coupling of lactateand protons efflux creates an electric po-tential (negative in the intracellular com-partment) and a pH gradient (alkaline inthe cytoplasm), whose resultant is the pro-ton motive force (pmf) which generatesATP via the ATPase system (fig 2). This

-system offers a positive energy balance tolactic acid bacteria during a part of theirgrowth phase. Each molecule of glucoseconsumed leads to the formation of 2molecules of lactate. During their excre-tion, they enable the bacterium to extrude4 protons. When we subtract the 2 protonsresulting from glucose catabolism, the ef-flux of lactate leads to the translocation of2 protons per glucose molecule. If they areutilized by the ATPase system, they ena-ble the synthesis of 1 molecule of ATP,which is an energy gain of 50% in compari-son to the 2 molecules of ATP producedby phosphorylation during glycolysis in ho-mofermenting bacteria (Ten Brink andKonings, 1982).

Osmoregu/ation and ragu/ationof intracellu/ar pH in /actic acid bacteria

Lactic acid, the final major product of car-bohydrate catabolism by lactic acid bacte-ria is in equilibrium according to the reac-tion:

CHs-CHOH-COOH + H20CHs-CHOH-COo- + HsO+

<->

ACID

Fig 2. Scheme of metabolic energy-generatingand energy-consuming processes in lactococci(after Konings and Otto, 1983; reprinted by per-mission of Kluwer Academic Publishers). Fourpathways related to bioenergetics of the cell areshown: 1) ATP synthesis via glycolysis; 2) thecoupling of lactate and protons efflux; 3) thecoupling of protons influx or efflux with respecti-vely concomitant ATP synthesis or consump-tion; 4) the coupling of solutes and protons in-flux.Schéma du processus de synthèse et d'utilisa-tion de l'énergie chez les lactocoques (d'aprèsKonings et Otto, 1983; reproduit avec la permis-sion de Kluwer Academie Publishers). Quatrevoies liées à la bioénergie cellulaire sont repré-sentées : 1) la synthèse d'ATP par /'intermé-diaire de la glycolyse; 2) le couplage des effluxde lactate et de protons; 3) le couplage des in-flux ou des efflux de protons avec respective-ment, synthèse d'ATP ou consommation d'ATP;4) le couplage des influx de solutés et de pro-tons.

534 Je Piard, M Desmazeaud

Since its pKa is 3.86 and the initial intra-cellular pH is close to neutrality, most lac-tic acid is ionized. The bacteria must th usface 2 situations: eliminate the proton sup-ply in order to maintain intracellular pH(pHc) at a value consistent with vital meta-bolic processes, and eliminate excess lac-tate in order to restore osmotic balance.

Konings and Otto (1983) determinedthe factors governing these regulations incultures grown at uncontrolled pH. 8acteri-al growth is characterized by the presenceof a lactate gradient (àlact) which causesa pH gradient (~pH) as a result of H+ et-flux/lactate coupling "(fig 3). This intervenesat the level of the pmf, which simultane-ously increases. As cell growth continues,the external lactate concentration increas-es, ~Iact and thus ~pH and ~p decrease.This drop in the pmf leads to decreases inboth the absorption of nutrients and H+/lactate stoichiometry. The cell is th usforced to gradually tap its energy reservesto eliminate H+ with the ATPase system.Cell death occurs when the cell no longerhas energy available to control its pHc.

ln cultures run at controlled pH, there isno pH gradient and the pmf reflects onlythe difference in membrane potential, ie itis constant during growth and decreaseswhen the sugar in the medium is con-sumed. Note than in a controlled pH cul-ture, there is no cell death, and so a pmfwill reappear if the bacteria once againhave an available carbon source (Koningsand Otto, 1983).

The absence of a proton gradient incontrolled pH cultures means that the pmfwill have a lower value during the expo-nential phase of growth. As a result, thegrowth rate will be lower than in the caseof a culture at uncontrolled pH. This phe-nomenon has apparently not been report-ed, probably because it is transitory andgrowth rate calculations are generallymade over the entire exponential phase ofgrowth.

-Effects of end-productsof sugar catabolism on different tlora

Effect on lactic acid bacteria

At uncontrolled pH, acidification and thedecrease of intracellular pH can lead to de-creased activity of metabolic enzymes,even their denaturation, as reported eg byAccolas et al (1980) for the hexokinase of

200

120T

160 Ls EÉ <,

~ C'

80 "-'" 20u .SCi c:

LL ë"-080 Ci

40 ..u

40

120 240 360 480Time (min)

Fig 3. Evolution of proton motive force and ofthe lactate gradient in L cremoris during growthin batch cultures on complex medium with lac-tose (2 g/I) as sole energy source. (e), protonmotive force; (0), membrane potential; (to), lac-tate gradient; (A) pH gradient; (---), cell proteinconcentration (after Konings and Otto, 1983; re-printed by permission of Kluwer Academie Pub-Iishers).Évolution de la force proton motrice et du gra-dient de lactate au cours de la croissance de Lcremoris en culture discontinue dans un milieucomplexe avec le lactose (2 g/I) comme seulesource d'énergie. (e), force proton motrice; (0),potentiel membranaire; (,1), gradient de lactate;("'), gradient de pH; (---) concentration de pro-téines cellulaires (d'après Konings et Otto, 1983;reproduit avec la permission de Kluwer Acade-mic Publishers).

Inhibition by lactic acid bacteria 535

S thermophilus. The tolerance of metabolicenzymes and of transport systems to acidpH is variable depending on the strain,which could explain the variable acid toler-ance in lactic acid bacteria which in mixedcultures leads to phenomena of domi-nance. Thus, Accolas et al (1977) reportedthat at uncontrol/ed pH, L bulgaricus con-tinued to grow after S thermophilusstopped growing, while at control/ed pH,both strains stopped at the same time (Juil-lard et al, 1987).

Homofermenting lactic acid bacteriaproduce lactic acid from glucose in almoststoichiometric quantities. When the bacte-ria are nutritional/y deprived, however,their metabolism also yields formic andacetic acids and ethanol (Cogan et al,1989). Heterofermenting bacteria degradeglucose to lactic acid, ethanol, CO2 and insome cases acetic acid. The cel/ mem-brane is impermeable to ionized hydrophil-ic acid, while non-ionized hydrophobie ac-ids diffuse passively (Kashket, 1987).Weak organic acids, such as acetic acid(pKa = 4.74) will thus be present at thesame concentration on both sides of themembrane. The intracellular medium is ba-sic in comparison to the external mediumand will favor the ionization of these mole-cules and thus an increase in their intracel-lular concentration. This causes not onlyan acidification of the cytoplasm, but alsothe creation of inhibitions, especial/y of en-zymes, by salt excesses. Final/y, Kashket(1987) reported that the organic acids andalcohols produced could act as protono-phores in the membrane and accentuatethe anarchie entry of protons. There areconsiderable knowledge gaps in this field:although lactic acid bacteria have been ex-tensively studied during their growth, rela-tively little is known about their behavior instationary or declining phase.

Effect on non-Iactic flora

There are very few non-Iactic bacteria ca-pable of growing at pH values lower thanthe threshold pH of lactic acid bacteria(table III). Good lactic acidification shouldthus inhibit the growth of E coli, Pseudo-monas, Salmonella and Clostridium.

Although it is very easy to determine thepH, this parameter is not a reliable indica-tor for presuming whether or not microor-ganisms wil/ develop. The pH is only a par-tial illustration of the concentration of weakacids in a medium, since a large part of

Table III. Limit pH values allowing the initiationof growth of various microorganisms (afterGray and Killinger, 1966; Stadhouders andLangeveld, 1966; Asperger, 1986).Valeurs de pH limites permettant l'initiationde la croissance de divers microorganismes(d'après Gray et Kil/inger, 1966; Stadhouders etLangeveld, 1966; Asperger, 1986).

Gram-negativebacteria

LimitpH

Escherichia coliKlebsiella pneumoniaeProteus vulgarisPseudomonas aeruginosaSalmonella paratyphiSalmonella typhiVibrioparahaemolyticus

4.44.44.45.64.5

4.0-4.54.8

Gram-positive

Bacillus cereusClostridium botulinumEnterococcus spLactobacil/us spMicrococcus spStaphylococcus aureusLactococcus lactisBrevibacterium linensListeria monocytogenes

4.94.74.8

3.8-4.45.64.0

4.3-4.85.65.5

536

6.0 and 5.6 in the case of acetate, formateand lactate, respectively, and attributed themajor inhibitory power to acetate. In termsof spore germination, the same authors cit-ed, in order, formate, lactate and acetatewhich inhibited spore germination by 50%at 0.1 molli and at pH 4.4, 4.3 and 4.2, re-spectively. The cumulated effects of the 3acids were not studied, however, eventhough they may be produced simultane-ously in certain conditions of the growth oflactococci (Cogan et al, 1989).

Concerning other sporulated bacteria,especially Clostridium tyrobutyricum whichis responsible for the late blowing ofcheeses, a review by Blocher and Busta(1983) indicated that vegetative cells wereinhibited at pH 4.65, while spore germina-tion was affected at pH 4.6.

Finally, it is ta be noted that the sensitiv-ity of bacteria to acids depends on varia-tions which in turn depends on the simulta-neous action of other factors (salt levels,activity of water, redox potential, heat treat-ment, etc). For example, Kleter et al (1984)showed the synergy between acidity andthe NaCI concentration in the inhibition ofC tyrobutyricum.

ln conclusion, it is important to note thehigh inhibitory capacity of organic acids inthe molecular (non-ionized) form. lt is con-sistent to imagine several improvementsfor the more efficient utilizatiôn of this sys-tem. Improved acid and salt tolerance ofthe enzymes of lactic acid bacteria wouldenable the lactic flora to more efficientlycompete with unwanted bacteria. As in thecase of the induced resistance to hydrogenperoxide, a more thorough understandingof the stress proteins of lactic acid bacteriacould indicate interesting prospects in thisfield. It would also be of interest to designlactic acid bacteria which at the propertime would switch their catabolism toheterofermenting pathways in order toprofit from the high inhibitory power of ace-tic acid at low pH.

Je Piard, M Desmazeaud

these compounds may be non-dissociatedas a function of pH. We saw above, how-ever, that in the case of weak organic ac-ids, it is precisely this non-dissociated frac-tion that diffuses across the membraneand is more or less ionized depending onintracellular pH. This type of molecule ex-erts a shadowy effect which cannot be de-termined by only measuring the pH of themedium. In this context, the study byChung and Goepfert (1970) is interesting,since it showed that salmonellas are inhib-ited at pH values < 4.4 for lactic acid and5.4 for acetic acid. In light of the respectivepKa values of 3.86 and 4.75 for the 2 acidsand of their dissociation constant, inhibi-tion apparently occurs when the non-ionized fraction reaches about 20%. Dalyet al (1972) showed that L diacetylactiscouId inhibit a wide range of unwantedbacteria in dairy products. Using S aureusas the model strain, the authors estab-Iished that the inhibitory activities of aceticand lactic acids increased as the pH de-creased. Adams and Hall (1988) investi-gated the individual and cumulated effectsof lactic and acetic acids towards E coliand Salmonella enteritidis. They confirmedthat the molecular form of these 2 acids isthe toxic factor for the bacteria (fig 4).They attributed the higher toxicity of aceticacid to the fact that its pKa is higher thanthat of lactic acid. The authors alsoshowed that the 2 acids acted synergisti-cally in weakly-buffered media: lactic aciddecreased the pH of the medium, therebyincreasing the toxicity of acetic acid.

Several research groups have exam-ined the inhibitory capacity of lactic acidbacteria in terms not only of the growth ofsporulated bacteria, but also towards thespores of these bacteria. Thus, Wong andChen (1988) showed that Bacillus cereusis first progressively inhibited and thenkilled when cocultured with various lacticacid bacteria. They showed that growthcompletely stopped at pH values of 6.1,

Inhibition by lactic acid bacteria 537

120

100

75

Q) 50'§J:: 25i~'"u 120;;:'u 100Q) ~a.C/)

750

50

25

a 2 4 8 16 32 63 126251500 a 2 4 8 16 32 63 126251500Total ocid (mmol/I)

Fig 4. Effect of total acid concentration on microbial specific growth rate (r:I) and % dissociation ofacid ("'). (a) E coli and acetic acid, (b) E coli and lactic acid, (c) Salmonella enteritidis and acetic acld,(d) Salmonella enteritidis and lactic acid. Specifie growth rate units are expressed on an arbitraryscale (atter Adams and Hall, 1988; reprinted by permission of International Journal of Food Scienceand Technology).Effet de la concentration totale d'acide sur le taux de croissance bactérien spécifique (0) et sur le %de dissociation de l'acide ("). (a) E coli et acide acétique, (b) E coli et acide lactique, (c) Salmonellaenteritidis et acide acétique, (d) Salmonella enteritidis et acide lactique. Les unités de taux de crois-sance spécifique sont exprimées sur une échelle arbitraire (d'après Adams et Hall, 1988; reproduitavec la permission de International Journal of Food Science and Technology).

Diacetyl

L diacetylactis and Leuconostoc sp canuse citrate present in milk (== 8.3 mmol/l).Among the products formed during thismetabolism (Cogan, 1980), diacetyl is im-portant since it is the major aromatic con-stituent of butter and fresh cheeses.

The antibacterial properties of diacetylhave been shown in a large number ofstudies involving different microorganisms.Jay (1982) published a thorough table ofthe inhibitory capacity of diacetyl, showingthat lactic acid bacteria are insensitive to

diacetyl up to the dose of 350 llg/ml.Gram-negative bacteria, on the otherhand, begin to be inhibited at 100 llg/mland are almost totally inhibited at 200 1191ml. The author indicated that diacetyl is le-thal for Gram-negative bacteria and onlybacteriostatic for Gram-positive bacteria.

The mean diacetyl concentration pro-duced by L diacetylactis in pure culture orin coculture with Leuconostoc is 4 llg/ml(Cogan, 1980) and so there is little chancethat the compound is antibacterial in lacticfermentations, but it could act synergisti-cally with other factors (Gilliland, 1985). In

538 Je Piard. M Desmazeaud

light of the value of the broad spectrum ofantibacterial action of diacetyl, especiallyagainst Gram-negative bacteria, Jay(1982) suggested its use in manufacturingprocesses in the food and agriculture in-dustry, as an aseptic agent for surfacesand utensils. Its volatility would avoid con-taminating the flavor of the manufacturedproducts. In addition, work is currently un-der way to construct hyperproducing dia-cetyi bacteria. The use of these strainscould reinforce the antagonist role of dia-cetyI. Uncertainties remain, however, con-cerning the mutagenic potential of diacetyl(Jay, 1982).

Acetaldehyde

Acetaldehyde production in certain dairyproducts and notably in yogurt, where it isthe principal aromatic constituent, is attrib-uted principally to L bulgaricus. This spe-cies contains a threonine aldolase whichc1eaves threonine into acetaldehyde andglycine (Accolas et al, 1980). The flora ofyogurt (L bulgaricus and S thermophilus)cannot rnetabolize acetaldehyde and so itaccumulates in the product at a concentra-tion of", 25 ppm (Accolas et al, 1980).

The possible inhibition of lactic acidbacteria by acetaldehyde is relativelyundocumented, limited to the work ofKulshrestha and Marth (1975), whoshowed that the lactic acid bacteria tested(L lactis, L citrovorum, S thermophilus)were very slightly inhibited at acetalde-hyde doses starting at 10 ppm, reaching3-10% inhibition at 100 ppm. The growthof undesirable bacteria in dairy products,eg Staphylococcus aureus, Salmonella ty-phimurium or E coli decreased at acetalde-hyde concentrations of 10-100 ppm. It isalso to be noted that cell division of Ecoli is inhibited by 10-3 mol/I (44 ppm) ac-etaldehyde (Egyud, 1967). These data areto be compared with the level of 25 ppm of

acetaldehyde produced in yogurt, suggest-ing that the compound could play a role,undoubtedly minor but not negligible, in an-tagonism among different flora.

D-isomers of amino acids

Gilliland and Speck (1968) reported that acompound present in the supernatant of alactic starter culture inhibited the samestarter. The compound was identified as D-leucine, inhibitory at the concentration of 1mg/ml. Teeri (1954) showed that several D-amino acids inhibited lactobacilli.

The mechanism of these inhibitions hasnot been elucidated. The initial idea wasthat D-amino acids compete with L-forms,essential for the growth of lactic acid bac-teria. Another hypothesis is that D-aminoacids are accidentally incorporated in pro-teins, rendering them inactive (Gillilandand Speck, 1968).

The conditions for the appearance of D-amino acids are unknown. Even thoughsorne bacterial antibiotics contain D-aminoacids, there is to our knowledge no proofof the presence of a racemase in lacticacid bacteria. This subject requires furtherstudy, ail the more so since unwantedbacteria such as Pseudomonas are alsoapparently sensitive to D-amino acids(Daniels, 1966).

CONCLUSION

The first part of this review has examinedthe wide range of inhibitory substances inlactic acid bacteria resulting from the me-tabolism of oxygen and from other catabol-ic pathways. These inhibitors are only verypartially used in the fight against undesira-ble flora. Future applications are neverthe-less promising.

We have seen that the lactoperoxidase-thiocyanate-hydrogen peroxide system is a

Inhibition by lactic acid bacteria 539

powerful inhibitor, capable of exerting abactericidal effect on a number of patho-gens. This system ls understood to a suffi-cient extent to be used and optimized atthe pilot scale for the preservation of milk.More detailed studies on the mechanism ofresistance to this system by lactic acidbacteria are necessary to envision its usein lactic fermentations.

Inhibitions resulting from oxygen metab-olism remain poorly elucidated. There ex-ists a wide variability of sensitivity to thesecompounds among lactic acid bacteria andthis variability is of value wh en studyingthe mechanisms involved in the resistanceto these metabolites. Stress proteins areapparently involved and more detailedknowledge of these phenomena couldhave repercussions on other aspects, egthe heat resistance of lactic acid bacteria.

Organic acids are the major products ofsugar catabolism. lt is important to notethat it is the molecular, or non-ionized,form of these acids which exerts the great-est inhibition. Technological advantagescould arise from these observations.

ACKNOWLEDGMENTS

We would Iike to express our gratitude to Dr JPAccolas for his insight and guidance in the criti-cal reading of the manuscript. We also thankJ Gallé for preparing the figures.

REFERENCES

Accolas JP, Bloquel R, Didienne R, Régnier J(1977) Propriétés acidifiantes des bactérieslactiques thermophiles en relation avec lafabrication du yoghourt. Lait 57, 1-23

Accolas JP, Hemme D, Desmazeaud MJ, Vas-sal L, Bouillanne C, Veaux M (1980) Les le-vains lactiques thermophi/es : propriétés etcomportement en technologie laitière. Unerevue. Lait 60, 487-524

Adams MR, Hall CJ (1988) Growth inhibition offood borne pathogens by lactic and acetic ac-ids and their mixtures. Int J Food Sei Technol23,287-292

Ananthaswamy HN, Eisenstark A (1977) Repairof hydrogen peroxide-induced single-strandbreaks in Escherichia coli deoxyribonucleicacid. J Bacterio/130, 187-191

Anders RF, Hogg DM, Jago GR (1970) Forma-tion of hydrogen peroxide by group N strepto-cocci and its effect on their growth and me-tabolism. Appl Microbio/19, 608-612

Archibald FS, Fridovich 1(1981) Manganese, su-peroxide dismutase and oxygen tolerance insome lactic acid bacteria. J Bacteriol 146,928-936

Asperger H (1986) Wirkungen vonMilchsaurebakterien auf andere mikroorga-nismen. Oesterr Milchwirts 41, suppl 1 to is-sue 4,1-22

Aune TM, Thomas EL (1977) Accumulation ofhypothiocyanite ion during peroxidase-catalyzed oxidation of thiocyanate ion. Eur JBiochem 80, 209-214

Bjërck L, Rosen CG, Marshall VM, Reiter B(1975) Antibacterial activity of the lactoperox-idase system in milk against pseudomonadsand other Gram-negative bacteria. Appl Mi-crobio/30, 199-204

Blocher JC, Busta FF (1983) Bacterial spore re-sistance to acids. Food Technol37, nov, 87-99

Britton L, Malinowski OP, Fridovitch 1(1978) Su-peroxide dismutase and oxygen metabolismin Streptococcus faecalis and comparisonswith other organisms. J Bacteriol 134, 229-236

Byczkowski J, Gessner T (1988) Biological roleof superoxide ion-radical. Int J Biochem 20,569-580

Chung KC, Goepfert JM (1970) Growth of Sal-monel/a at low pH. J Food Sci 35, 326-328

Cogan TM (1980) Les levains lactiques méso-phi/es. Une revue. Lait 60, 397-425

Cogan JF, Walsh D, Condon S (1989) Impact ofaeration on the metabolic end-productsformed from glucose and galactose by Strep-tococcus lactis. J Appl Bacteriol66, 77-84

Condon S (1983) Aerobic metabolism of lacticacid bacteria. Ir J Food Sei Technol7, 15-25

540

tured mammalian cells against hydrogen per-oxide toxicity. Med Biol 61, 268-274

Harley JB, Santangelo GM, Rasmussen H,Goldfine H (1978) Dependence of Escheri-chia coli hyperbaric oxygen toxicity on the llp-id acyl chain composition. J Bacteriol 134,808-820

Haugaard N (1968) Cellular mechanisms of oxy-gen toxicity. Physiol Rev 48, 311-373

Jay JM (1982) Antimicrobial activity of diacetyl.Appl Environ Microbio/44, 525-532

Juillard V, Spinnier HE, Desmazeaud MJ, Boqui-en CY (1987) Phenomènes de coopération etd'inhibition entre les bactéries lactiques utili·sées en industrie laitière. Lait67, 149-172

Kashket ER (1987) Bioenergetics of lactic acidbacteria: cytoplasmic pH and osmotolerance.FEMS Microbiol Rev 46, 233-244

Kleter G, Lammers WL, Vos EA (1984) The in-fluence of pH and concentration of lactic acidand NaCI on the growth of Clostridum tyrobu-tyricum in cheese. Neth Milk Dairy J 38, 31-41

Kong S, Davison AJ (1980) The role of interac-tions between O2, H2, 'OH, e- and O2- in freeradical damage to biological systems. ArchBiochem Biophys 204, 13-29

Konings WN (1985) Generation of metabolic en-ergy by end product efflux. Trends BiochemSci10, 317-319

Konings WN, Otto R (1983) Energy transductionand solute transport in streptococci. AntonieLeeuwenhoek 49, 247-257

Kono Y, Fridovich 1 (1983) Isolation and charac-terization of the pseudocatalase of Lactoba-cillus plantarum. J Biol Chem 258, 6015-6019

Kulshrestha DC, Marth EH (1975) Sorne volatileand nonvolatile compounds associated withmilk and their effects on certain bacteria. Areview. J Milk Food Techno/38, 604-620

Law BA, John P (1981) Effect of the lactoperoxi-dase bactericidal system on the formation ofthe electrico-chemical proton gradient inE coli. FEMS Microbiol Lett 10, 67-70

Marshall WME, Reiter B (1980) Comparison ofthe antibacterial activity of the hypothiocya-nite anion towards Streptococcus lactis andEscherichia coli. J Gen Microbiol 120, 513-516

JC Piard, M Desmazeaud

Condon S (1987) Responses of lactic acid bac-teria to oxygen. FEMS Microbiol Rev 46,269-280

Dahiya RS, Speck ML (1968) Hydrogen perox-ide formation by lactobacilli and its effect onStaphylococcus aureus. J Dairy Sci 51,1568-1572

Dahl TA, Midden WR, Hartman PE (1989) Com-parison of killing of Gram-negative andGram-positive bacteria by pure singlet oxy-gen. J Bacterio/171 , 2188-2194

Daly C, Sandine WE, Elliker PR (1972) Interac-tions of food starter cultures and food bornepathogens: Streptococcus diacetylactis ver-sus food pathogens. J Milk Food Techno/35,349-357

Daniels HJ (1966) Inhibition of growth of Pseu-domonas denitrificans by amino acids. Can JMicrobio/12,1095-1098

Di Mascio P, Wefers H, Do-Thi HP, LafleurMVM, Sies H (1989) Singlet molecular oxy-gen causes loss of biological activity in plas-mid and bacteriophage DNA and inducessingle-strand breaks. Biochim Biophys Acta1007,151-157

Egyud LG (1967) Studies on cell division, the ef-fect of aldehydes, ketons, and a-keto-aldehydes on the proliferation of Escherichiacoli. Curr Med Biol 1, 14-20

Freese EG, Gerson J, Taber H, Rhaese HJ,Freese E (1967) Inactivating DNA alterationsinduced by peroxides and peroxide-producing agents. Mutat Res 4,517-531

Gilliland SE (1985) Concentrated starter cul-tures. In: Bacterial Starter Cultures for Food(Gilliland SE, ed) CRC Press, 145-157

Gilliland SE, Speck ML (1968) D-Leucine as anauto-inhibitor of lactic streptococci. J DairySei 51, 1573-1578

Gothefors L, Marklund S (1975) Lactoperoxi-dase activity in human milk and in saliva ofnewborn infants. Infect Immun 11, 1210-1215

Gray ML, Killinger AH (1966) Listeria monocyte-genes and Listeria infections. Bacteriol Rev30,309-382

Gregory EM, Fridovich 1 (1974) Oxygen meta-bolism in Lactobacillus plantarum. J Bacteriol117,166-169

Hanstrëm L, Johansson A, Carlsson J (1983)Lactoperoxidase and thiocyanate protect cul-

Inhibition by lactic acid bacteria 541

Morgan RB, Christman MF, Jacobson FS, StorzG, Ames BN (1986) Hydrogen peroxide-inducible proteins in Salmonella typhimuriumoverlap with heat shock and other stress pro-teins. Proc Nat! Acad Sei USA 83, 8059-8063

Murphy MG, Condon S (1984) Comparison ofaerobic and anaerobie growth of Lactobacil-lus plantarum in a glucose medium. Arch Mi-crobio/138, 49-53

Oram JD, Reiter B (1966) The inhibition of strep-tococci by lactoperoxidase, thiocyanate andhydrogen peroxide. Biochem J 100, 373-381

Priee RJ, Lee JS (1970) Inhibition of Pseudomo-nas species by hydrogen peroxide producinglactobacilli. J Milk Food Techno/33, 13-18

Pruitt KM, Reiter B (1985) Biochemistry of per-oxidase system: antimicrobial effects. In: TheLactoperoxidase System (Pruitt KM, TenovuoJ, eds) Immunology Series, M Dekker Inc,NY, vol 27, 143-178

Pruitt KM, Tenovuo J, Andrews RW, McKane T(1982) Lactoperoxidase-catalysed oxidationof thiocyanate: polarographic study of the oxi-dation products. Biochemistry 21, 562-567

Purdy MA, Tenovuo J, Pruitt KM, White WE(1983) Effect of growth phase and cell enve-lope structure on susceptibility of Salmonellatyphimurium to the lactoperoxidase thiocya-nate hydrogen peroxide system. Infect Im-mun39,1187-1195

Reiter B (1985) The lactoperoxydase system ofbovine milk. In: The Lactoperoxidase System(Pruitt KM, Tenovuo J, eds) Immunology Se-ries, M Dekker Inc, NY, vol 27, 123-141

Reiter B, Hârnulv G (1984) Lactoperoxidase an-tibacterial system: natural occurrence, biolog-ical functions and practical applications.J Food Prot47, 724-732

Rolfe RD, Hentges DJ, Campbell BJ, Barett JT(1978) Factors related to the oxygen toler-ance of anaerobic bacteria. Appl Environ Mi-crobio/36, 306-313

Siragusa GR, Johnson MG (1989) Inhibition ofListeria monocytogenes growth by the lacto-

peroxidase-thiocyanate-H202 antimicrobialsystem. Appl Environ Microbiol 55, 2802·2805

Smart JB, Thomas TD (1987) Effect of oxygenon lactose metabolism in lactic streptococci.Appl Environ Microbio/53, 533-541

Stadhouders J, Langeveld LPM (1966) The mi-croflora of the surface of cheese. Factors af-fecting its composition. 17th Int Dairy Gongr,Vol D, 577-584

Steele WF, Morrison M (1969) Antistreptococcalactivity of lactoperoxidase. J Bacteriol 97,635-639

Teeri AE (1954) Effect of D-amino acids ongrowth of lactobacilli. J Bacterio/67, 686-689

Ten Brink B, Konings WN (1982) Electrochemi·cal proton gradient and lactate concentrationgradient in Streptococcus cremoris ceIlsgrown in batch cultures. J Bacterio/152, 682-686

Thomas EL (1985) Bacterial hydrogen peroxideproduction. In: The Lactoperoxidase System(Pruitt KM, Tenovuo J, eds) ImmunologicalSeries, M Dekker Inc, NY, vol 27, 179-202

Thomas EL, Aune TM (1978) Lactoperoxidase,peroxide, thiocyanate antimicrobial system:correlation of sulfhydryl oxidation with antimi-crobial action. Infect Immun 20, 456-463

Whittenbury R (1964) Hydrogen peroxide forma-tion and catalase activity in the lactic acidbacteria. J Gen Microbio/35, 13-26

Wong HC, Chen YL (1988) Effects of lactic acidbacteria and organic acids on growth andgermination of Bacillus cereus. Appl EnvironMicrobiol54, 2179-2184

Wright RC, Tramer J (1958) Factors influencingthe activity of cheese starters. The role ofmilk peroxidase. J Dairy Res 25, 104-118

Yousten AA, Johnson JL, Salin M (1975) Oxy-gen metabolism of catalase negative and cat-alase-positive strains of Lactobacillus plantar-um. J Bacterio/123, 242-247