HAL Id: hal-00929441https://hal.archives-ouvertes.fr/hal-00929441

Submitted on 1 Jan 1995

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.

Interactions between lactic and propionic acid bacteriaPg Piveteau, S Condon, Tm Cogan

To cite this version:Pg Piveteau, S Condon, Tm Cogan. Interactions between lactic and propionic acid bacteria. Le Lait,INRA Editions, 1995, 75 (4_5), pp.331-343. �hal-00929441�

Lait (1995) 75, 331-343© Elsevier/INRA

331

Original article

Interactions between lacticand propionic acid bacteria

PG Piveteau 1, S Condon 1, TM Cogan 2

1 oepartment of Microbiology, University College, Cork;e National oairy Products Centre, Teagasc, Fermoy, Ireland

5ummary - ln sorne cheeses, propionic acid bacteria (PAB) ferment lactate to propionate, acetate andCOz which are important in determining the flavour and texture of the cheese. Interactions between 14strains of lactic acid bacteria (LAB) (Lactobacillus helveticus, Lb acidophilus, Lb lactis, Streptococ-cus thermophilus and Lactococcus lactis) and 4 strains of PAB (Propionibacterium freudenreichii andP acidipropionici) were studied in whey. Stimulation or inhibition was judged by the effect on growthrate and final cell mass (00600), No inhibition of growth was found. The growth of ail 4 strains of PABwas stimulated by Lb helveticus and Str thermophilus. The degree of stimulation of the PAB by the otherLAB varied. In control and Lb helveticus RR whey, L lactate was used preferentially over 0 lactate byP freudenreichii KM. Lb helveticus RR increased the levels of amino acids and peptides in the whey.Ali of the amino acids, except threonine and cysteine, and some of the peptides were used duringsubsequent growth of P freudenreichii KM. The addition of aspartate stimulated growth of P freuden-reichii KM in control whey and reduced the amount of lactate converted to propionate, but not acetate.The stimulant(s) was stable to heating to 121°C for 15 min and eluted in several peaks alter chro-matography on Sephadex G-25. Ultrafiltration resulted in a totalloss of activity.

interaction 1 propionic acid bacteria Ilactic acid bacteria 1 whey 1 growth

Résumé - Interactions entre bactéries lactiques et bactéries propioniques. Dans certains typesde fromage, les bactéries propioniques (PAB) fermentent le lactate en propionate, acétate et COz,produits importants pour la saveur et la texture de ces fromages. Les interactions entre 14 souches debactéries lactiques (LAB) (Lactobacillus helveticus, Lb acidophilus, Lb lactis, Streptococcus ther-mophilus et Lactococcus lactis) et 4 souches de PAB (Propionibacterium freudenreichii et P acidipro-pion ici) ont été étudiées en milieu lactosérum. La stimulation ou l'inhibition étaient jugées en fonctiondu taux de croissance et de la masse cellulaire (0060oJ en fin de fermentation. Aucune inhibition de lacroissance n'a été observée. La croissance des 4 souches de PAB était stimulée par Lb helveticus etStr thermophilus. La stimulation des PAB par les autres LAB était variable. Dans le lactosérum produitpar Lb helveticus RR et dans le contrôle, l'isomère L de l'acide lactique était utilisé par la souche deP freudenreichii KM de façon préférentielle par rapport à l'isomère O. L'ajout d'aspartate stimulait la crois-sance de P freudenreichii KM dans le lactosérum controle, et réduisait la proportion de lactate convertie

332 PG Piveteau et al

en propionate sans changer celle convertie en acétate. Le(s) stimulant(s) résistaient au chauffage à121°C pendant 15 min et formait plusieurs pics après chromatographie sur Sephadex G-25. L'ultra-filtration résultait en une perte d'activité.

interaction / bactérie propionique / bactérie lactique / lactosérum / croissance

INTRODUCTION

Cheese is a complex biological systemwhich undergoes numerous biochemicalchanges during manufacture and ripening.ln Swiss-type cheese manufacture, ther-mophilic lactic acid bacteria (LAB), includingLaetobaeillus helvetieus, Laetobaeillus del-brueekii subsp laetis and Streptoeoeeus ther-mophilus, are used as starters and degradelactose to D and L lactate. Propionibacte-ria (PAB) then metabolize lactate to propi-onate, acetate and CO2 during the extendedripening (Hettinga and Reinbold, 1972;Langsrud and Reinbold, 1973). The propi-onic acid fermentation is critical to the finalquality of the cheese as CO2 is responsi-ble for eye formation, while propionic acidcontributes to the development of the typicalnutty f1avour (Gautier et al, 1993). The qual-ity of the cheese, therefore, depends on theextent of the lactic acid fermentation andthe propionic acid fermentation. Differentisomers of lactate are produced by thestarter LAB. Laetobaeillus delbrueekii subsplaetis produces D lactate, Streptoeoeeusthermophilus produces L lactate and Lae-tobaeillus helvetieus produces a mixture ofD and L lactate. Some strains of PAB havea preference for the L isomer (Crow, 1986a);however, the interaction between the 2 fer-mentations is not limited to lactate produc-tion and utilization. Interactions betweenLAB and PAB have been reported in co-cul-tures (Lee et al, 1976; Parker and Moon,1982; Perez Chaia et al, 1987) and insequential cultures (Hunter and Frazier,1961; Czarnocka-Roczniakowa et al, 1972).The compounds involved in these interac-tions have not been identified. The object

of this study was to characterize the inter-actions between PAB and LAB more fully.

MATERIALS AND METHODS

Organisms and medium

Three strains of Lactobacillus helveticus (RR,303,3321), 1 strain of Lactobacillus delbrueckiisubsp lactis (LL51), 1 strain of Lactobacillus eci-dophilus (LbA), 5 strains of Streptococcus ther-mophilus (1781, 1821, 1842, STB01 and STB02),4 strains of Lactococcus lactis (C1 0, ML8, E8and AM2), 1 strain of Propionibacterium acidipro-pionici (L5) and 3 strains of Propionibacteriumfreudenreichii (T, KM and H) were used. Ali strainswere from our culture collection except STB01and STB02 (Chr Hansen Laboratories, Cork, Ire-land) and LL51 (Institut Technique du Gruyère,Rennes, France). PAB were routinely transferredweekly in sodium lactate broth (SLB) and storedat 4°C. LAB were stored at -80°C and subcul-tured at least twice in 10% (w/v) sterile reconsti-tuted skim milk (RSM) before use. The composi-tion of SLB was 1% (w/v) tryptone, 1% (w/v) yeastextract, 0.5% (w/v) KH2P04 and 1% (w/v) O/Lsodium lactate, pH 6.5.

Wheyproduction

Interactions were studied by growing PAB in filter-sterilized whey obtained after growth of the LABin RSM. Assessment of the LAB/PAB interactionwas done by comparing growth rates and final00 at 600 nm in the starter whey to those in acontrol whey. A strain of LAB was incubated at42°C (thermophilic strains) or 30°C (mesophilicstrains) for 24 h in RSM. The coagulated milkwas centrifuged (8 000 g, 10 min, 4°C); the super-natant was adjusted to pH 6.0 with NaOH (2 N)

Growth of propionic acid bacteria 333

Heat treatmentand incubated for 30 min at 45°C to allow anyprecipitation to occur. This overcame subsequent

.precipitation of colloidal Casalts dlHing qrowthv- ..·of the PAB. Alter a second centrifugation (8 000 g, Starter whey, adjusted to pH 6.0, was pasteur-30 min, 4°C), the supernatant was filtered (What- ized (63.s"C, 30 min) or autoclaved at 121°C forman n01) and filter sterilized (0,45 um), Control 15 min. Alter treatment, the whey was centrifuged

(8 000 g, 10 min, 4°C) and its pH readjusted to 6.0whey was produced by acidifying RSM with 1% before filter sterilization (0,45 um),(w/v) of a 50% mixture of D and L lactic acid. Thecoagulated milk was then processed as justdescribed. Sterile wheys were stored at 4°C untilrequired.

Growth conditions

The PAB were grown in SLB for 3 d at 30°C, cen-trifuged, washed with sterile 1/4 strength Ringer'ssolution and resuspended in the original volumeof Ringer's solution. Whey was inoculated (1%,vlv) and dispensed in 10 ml tubes, which wereincubated under static conditions at 30°C. Growthwas followed by measuring aD at 600 nm. Whenthe aD was greater than 0.5, the culture wasdiluted before reading to maintain linearitybetween Oü and cell mass. Two ml of culturewas centrifuged and the supernatant frozen forfurther analysis. Purity of cultures was checked bymicroscopic examination throughout growth andby plating aerobically and anaerobically, at 30°C,at the end of fermentation on sodium lactate agar(SLA).

Analytical methods

Lactose, glucose, galactose, lactate, propionate,acetate and succinate were quantified by high-performance liquid chromatography (HPLC) (col-umn Aminex HPX 87 X at 60°C and eluted with0.04 N H2S04). D and L lactate were determinedwith enzymatic kits (Boehringer Mannheim).Amino acids were measured on a 120 x 4 mmcation exchange column (Nat form) using aBeckman 6300 amino acids Analyser (BeckmanInstruments L1d, High Wycombe, UK). The resultswere processed with a PC Minichrom. Peptideprofiles were determined by reverse-phase HPLC(Shimadzu HPLC system, C8 Nucleosil wide porecolumn; solvent gradient: A: 0.1 % [wlv] trifluo-roacetic acid, B: 0.1 % [wlv] trifluoroacetic acid inacetonitrile).

Ultrafiltration

Starter whey was ultrafiltered (Minitan system,Millipore) using a 10 000 cutoff polysulfone filter(10 000 NMWL plates) until a 10-fold concentra-tion of retentate was achieved.

Perchloric acid extraction

Perchloric acid was added to the starter whey toa final concentration of 2 N. The mixture was cen-trifuged (5 000 g, 10 min, 4°C); the supernatantwas neutralized with KHC03, recentrifuged andthe pH adjusted to 6.0 before filter sterilization.The original starter whey was diluted 2.5-fold bythis procedure.

Gel filtration

A 100 ml Sephadex G 25 column was equilibratedfor 24 h at room temperature with distilled water.Five ml of starter whey was applied and the col-umn eluted with distilled water at a flow rate of20 ml/h. Three ml fractions were collected andtheir A280 determined using the first fraction as ablank. Fractions forming a peak on the chro-matogram were combined, freeze-dried, dissolvedin 5 ml of distilled water, filter sterilized andassayed for growth stimulation of the PAB cul-tures in control whey alter 72 h of incubation.

RESULTS

The influence of the 14 strains of LAB onthe growth of P treudenreichii KM is shawnin figure 1. Ali strains of LAB, except Lc lac-fis E8, stimulated the growth rate and final

334 PG Piveteau et al

~ Control8\0

LbA-e0.1 RR 0.1

~ Control

-0-- 303 ~ LL51

-l!r- 33210.01 0.01

0 20 40 60 80 0 20 40 60 80

10 10c d

ControlControl

~ STBOICIO

-0-- STB020.1

--tr- 1781--tr- E8

---Q-- 1821---Q-- AM2

1842 0.0120 40 60 80 0 20 40 60 80

Tirne, h Time,h

10a

0.1

0.01 L...- __ .L-__ .l..- __ .1-_~

o

10b

Fig 1. Influence of different strains of lactic acid bacteria on the growth of Propionibacterium freuden-reichii KM in control whey. (a) Strains of Lactobacillus helveticus and Lb acidophilus; (b) Lb lactis;(c) strains of Streptococcus thermophilus; (d) strains of Lactococcus lactis.Influence de différentes bactéries lactiques sur la croissance de Propionibacterium freudenreichii KMsur lactosérum. (a) Souches de Lactobacillus helveticus et Lb acidophilus; (b) Lb lactis; (c) souches deStreptococcus thermophilus; (d) souches de Lactococcus lactis.

cell mass, but to varying extents. The largestincreases were observed with the Lacto-bacillus strains which increased growth ratesby up to 17% and more than doubled thefinal cell mass. On the other hand, Str ther-mophilus strains had only a small effect onthe growth rate (6% increase) but increasedthe cell mass significantly: a 2.2-foldincrease was observed with strain STS01.

The interactions between the other strains ofPAS and LAS varied (table 1). Each of the 4strains of PAS used were stimulated by Lbhelveticus and Str thermophilus STS01.Sorne PAS/LAS pairs showed no interac-tion (eg PAS strain T was not affected byStr thermophilus 1781 and STS02 nor bythe Lactococcus lactis strains) (table 1). Atthe end of these preliminary experiments,

Growth of propionic acid bacteria 335

Table 1. Percentage increase in final biomass production (OD600) of 4 strains of propionic acidbacteria (PAB) alter growth for 72 h in lactic acid bacteria (LAB) wheys compared with controlwheys.Pourcentage d'augmentation de la production finale de biomasse (00600) de 4 souches depropionibactéries après 72 h de croissance dans le lactosérum produit par les bactéries lactiquescomparé au lactosérum témoin.

LAB Propionibacterium

L5 T KM H

Lactobacillus helveticus RR 249 174 112 384303 173 151 128 483321 247 218 144 35

Lactobacillus acidophilus LbA 0 201 137 35

Lactobacillus delbrueckii LL51 0 51 194 0subsp lactis

Streptococcus thermophilus 1781 77 0 116 331821 0 155 98 331842 0 136 117 535TB01 166 112 126 525TB02 51 0 107 34

Lactococcus lactis C10 212 0 159 35ML8 244 0 137 60AM2 173 0 51 34

Laetobaeillus helvetieus RR and Propioni-baeterium freudenreiehii KM were selectedfor more detailed studies, because ail strainsof PAB were stimulated by Laetobaeillushelvetieus RR and ail strains of LAB, exceptLe laetis E8, stimulated P freudenreiehii KM.

Influence of milk on stimulation

Control and starter wheys were producedfrom 2 skim milk powders and a sam pie ofpasteurized skim milk. The stimulation of Pfreudenreiehii KM by Lb helvetieus RRoccurred regardless of the milk in which itwas grown (data not shown).

Growth and fermentation in control whey

ln experiments carried out in control whey,lactose was never used by P freudenreiehiiKM during the fermentation (fig 2a); at theend of fermentation, the culture reached an00600 of -0.60 and the pH decreased to5.85. Similar results were obtained for theother 3 strains of PAB. Lactate utilizationwas slow du ring the first 24 h of growth afterwhich L lactate was used preferentially overo lactate (fig 2a). At the end of growth, 8.3mmol/I of L lactate and 42.1 mmol/I of 0 lac-tate remained in the whey, correspondingto an overall utilization of 82% of the L lac-tate and 12% of the 0 lactate. Propionate

336 PG Piveteau et al

150 a 100 100

7S 7S

~IOOê~

~ § 0.1 so 0.1 50

..(J so2S 2S

20 40 60

100 b

o 0.01 L..-_-L. __ ...I...-_---' __ -'o

o 20 60

80 20 40 80

40

Time,h

60

100d

10

80

1 L..-_-L. __ ...I...-_---' __ -'

o 20 40 60 80

Time,h

Fig 2. Growth, substrate utilization and product formation by Propionibacterium freudenreichii KM in(a, b) control whey and (c, d) Lactobacillus helveticus RR whey. (0) growth; (0) 0 lactate utilization;(e) L lactate utilization; ($) lactose utilization; (A) propionate production; (t.) acetate production.Croissance, utilisation des substrats et formation des produits de la fermentation par Propionibac-terium freudenreichii KM sur (a, b) lactosérum témoin et (c, d) lactosérum produit par Lactobacillus hel-veticus RR. (0) croissance; (0) utilisation du lactate 0; (e) utilisation du lactate L; ($) utilisation du lac-tose; (A) production de propionate; (t.) production d'acétate.

and acetate were only detected late in thefermentations, after 20 h, and the combinedproduction of propionate and acetate wasstoichiometrically related to the disap-pearence of lactate (fig 2b); the molar ratiosof propionate:lactate and acetate:lactatewere 0.70 and 0.24, respectively, and thecorrelation coefficients between propionateproduced and lactate used and betweenacetate produced and lactate used were0.99 and 0.95, respectively. Succinate wasnot detected.

Growth and fermentationin Lb helveticus RR whey

Figure 2c shows the growth of P freudenre-ichii KM in Lb helveticus RR whey. The cul-ture attained a final OD of -0.99, giving anincrease in biomass of 70% compared tothat in the control whey. The pH decreasedfrom 6.0 to 5.8. As in the control whey, lac-tose was not utilized. Lactate utilization wasslow for the first 24 h and again L lactatewas used preferentially over D lactate (fig

Growth of propionic acid bacteria

2c). By the end of the fermentation, 85% ofthe L lactate and 46% of the D lactate weremetabolized and 36.0 mmoVI propionate and19.0 mmol/I acetate were produced (fig 2d).The molar ratios between propionate andlactate and acetate and lactate were 0.61and 0.23, respectively; the correlation coef-ficients between propionate and lactate andacetate and lactate were 0.99 and 0.82,respectively. Succinate (13.3 mmol/I) wasdetected, but only after 71 h of fermentation.

\.5a

~.~ë-=g 0.5ev

ollc.l_o.IJIllID.....L.-Jb...l-LILL.-LL.L

\.5

b

~E

~ë-=~ 0.5

o~ Q,) l1.l l1.l -e.~ 2 '2 '5 .~ .~ "2 .~ ~.~ ~ ~ CI) '§ â ~ ê >~> < ~ .su .=:o

Fig 3. Amino acid composition of (a) control wheyand (b) Lactobacillus helveticus RR whey at (.)o h and (m) alter 41 h of growth of Propionibac-terium freudenreichii KM.Composition en acides aminés du (a) lactosérumtémoin et (b) lactosérum produit par Lactobacillushelveticus RR à (.) 0 h et (m) après 41 h decroissance de Propionibacterium freudenreichiiKM.

337

Changes in amino acid compositionof whey during fermentation

The free amino acid composition of controland RR wheys are shown in figure 3. Argi-nine and phenylalanine were not detected ineither whey. Some amino acids (eg alanine,methionine, isoleucine, tyrosine, histidineand proline), which were not found in thecontrol whey, were present in RR whey. Inaddition, the concentrations of some of theother amino acids (serine, alanine, cysteine,valine and leucine) in control whey wereincreased by the growth of strain RR. Glu-tamate was not increased by the growth ofstrain RR. Ali free amino acids, except thre-onine and cysteine, were utilized during sub-sequent growth of P freudenreichii KM instarter whey. In order to verify if these dif-ferences in amino acid composition couldexplain the stimulation, growth in amino acidsupplemented control whey was studied.

The addition of 0.1 % and 1% vitam infree, acid-hydrolysed casein to control wheyresulted in increased growth rates and cellyields of strain KM (fig 4), whereas the addi-tion of 10% resulted in an inhibitory effecton growth rate but a stimulatory effect oncell yield. In addition to propionate andacetate (data not shown), succinate wasproduced during fermentation. The amountof succinate increased in direct proportion tothe casein hydrolysate added and none wasproduced in the absence of caseinhydrolysate. Several amino acids - aspar-tate (initial concentration 3.6 mmol/I), ser-ine (1.4 mmol/I), glycine (1.2 mmol/I) andalanine (1.3 mmol/I) - were utilized com-pletely during the fermentation (fig 5). Theaddition of 1.5 mmol/I aspartate to controlwhey had no influence on the growth of KM(fig 6), but higher concentrations increasedboth the growth rate and growth yield. Theamount of lactate metabolized and propi-onate and acetate produced also increasedwith increasing aspartate addition; succi-nate was produced but only in the presence

338 PG Piveteau et al

ICX) 5

4CIl

"" ~"s: 3" E

10?'

.~33 r:S ~ 2

~u

E"§.(

0.1

0.01 '-_--J'--_----L __ --'-__ ~

o 40 HO20

Time,h

Fig 4. Effect of casein hydrolysate on growth(open symbols) and succinate production (c1osedsymbols) of Propionibacterium freudenreichiiKM.(0) 0.0% casein hydrolysate; (0,.) 0.1% caseinhydrolysate; (0,.) 1.0% casein hydrolysate; (tl,.A.)10% casein hydrolysate. Succinate was not pro-duced in the absence of casein hydrolysate.Effet d'un hydrolysat de caséine sur la croissance(symboles ouverts) et sur la production de suc-cinate (symboles pleins) de PropionibacteriumIreudenreichii KM. (0) 0,0% d'hydrolysat decaséine; (0,.) 0,1% d'hydrolysat de caséine;(0,.) 1,0% d'hydrolysat de caséine; (tl,~) 10%d'hydrolysat de caséine. Le succinate n'était pasproduit en absence d'hydrolysat de caséine.

of 6 and 30 mmol/I aspartate (fig 6). Themolar ratios of propionate:lactate were 0.71,0.65, 0.59 and 0.53 in the presence of 0,1.5, 6 and 30 mmol/I aspartate, respectively.The molar ratios of acetate:lactate did notchange significantly with increasing levelsof aspartate. No effects were observed onaddition of up to 30 mmol/I each of alanine,serine or glycine, both individualiy ortogether (data not shown).

Changes in peptide compositionduring fermentation

ln addition to amino acids, RR whey alsohad higher concentrations of peptides thancontrol whey (fig 7). Most of the peptides in

Fig 5. Free amino acid composition of controlwhey supplemented with 1.0% casein hydrolysateat the beginning (.) and at the end (0) of growthof Propionibacterium freudenreichii KM.Composition en acides aminés libres du lac-tosérum témoin supplémenté avec 1,0%d'hydrolysat de caséine en début (.) et fin (71 h)(0) de fermentation par PropionibacteriumIreudenreichii KM.

10 100

E"§.(

0.1

0.01 L-_---I. __ -L __ -'-_---'

20 40 so

Time.h

Fig 6. Ellect 01 aspartate on growth (open sym-bols) and succinate production (c1osed symbols)01 Propionibacterium freudenreichii KM in con-trol whey. (0) 0.0 mmol/I aspartate; (0) 1.5 mmol/Iaspartate; (0,.) 6.0 mmol/I aspartate; (tl,.A.) 30mmol/I aspartate.Effet de l'aspartate sur la croissance (symbolesouverts) et sur la production de succinate (sym-boles pleins) de Propionibacterium freudenreichiiKM sur lactosérum témoin. (0) 0,0 mmollld'aspartate; (0) 1,5 mmolll d'aspartate; (0,.) 6,0mmolll d'aspartate; (tl,~) 30 mmolll d'aspartate.

Growth of propionic acid bacteria

the RR whey were utilized during subse-quent growth of P freudenreichiiKM, regard-less of their hydrophobicities (fig 8). Somepeptides were higher in concentration after

1 1 1 Iii 1 i 1 i i10 15 ~ ~ M JS ~ ~ ~ ~ ~ ~ ~ ~

Fig 7. Peptide profiles of (a) control whey beforegrowth of Propionibacterium freudenreichii KM,(b) Lactobacillus helveticus RR whey beforegrowth of Propionibacterium freudenreichii KMand (c) Lactobacillus helveticus RR whey afterultrafiltration.Profiles peptidiques du (a) lactosérum témoinavant croissance de Propionibacterium freuden-reichii KM, (b) lactosérum produit par Lacto-bacillus helveticus RR, avant croissance de Pro-pionibacterium freudenreichii KM et (c) lactosérumde Lactobacillus helveticus RR ayant subi uneultrafiltration.

339

growth of the PAB, which may be due toproduction of peptides by the PAB duringgrowth or to hydrolysis of some peptidesinto smaller ones, which elute at differenttimes, but which are not used or used onlyto a limited extent by the PAB.

Ultrafiltration

<s.: Ultrafiltration of RR whey was used in anattempt to determine whether the stirnu-lant(s) was a high or low molecular masscompound(s). After ultrafiltration of RRwhey, KM did not grow in either the reten-tate, permeate or in the recombined per-meate plus retentate. Ultrafiltration resultedin a dramatic reduction of the peptides ofthe recombined permeate and retentate (fig7); some peptides were missing and somewere present but at a much lower concen-tration.

Influence of heat treatmenton the stimulant(s)

Pasteurization and autoclaving of RR wheyresulted in a precipitate which was removedby centrifugation before filtering and filtersterilizing the supernatant. These heat treat-ments did not affect the rates of growth ofstrain KM, but a slightly reduced final cellmass was obtained in the case of auto-claved whey (data not shown).

Effect of perchloric acid extractionon the stimulant(s)

Treatment with perchloric acid precipitatedthe proteins from the whey. This procedurecaused a 2.5-fold dilution of the whey, butstimulation was still observed in it comparedwith a control whey which had also beendiluted 2.5-fold (data not shown).

340 PG Piveteau et al

Partial isolation of the stimulant(s)by gel filtration

The gel-filtration profile of RR whey, afterseparation on Sephadex G25, is shown infigure 9. Eleven different peaks wereobtained and peaks 1 to 4 contained most ofthe stimulatory activity. These results sug-gest that several stimulants ,are involvedwith molecular masses of less than 5 000Da.

DISCUSSION

Earlier workers (Hunter and Frazier, 1961;Czarnocka-Roczniakowa et al, 1972; Lee

o 10 15 20 25 30 35 40 45 50 55 60 65 70 75

Time, min

Fig 8. Elleet 01growth 01 Propionibacterium freudenreichii KM on the peptide prolile 01 Lactobacillushelveticus RR whey. Solid line, belore growth; broken line, alter growth.Modifications du profile peptidique du lactosérum de Laetobaeillus helvetieus RR, induites par la crois-sance de Propionibacterium Ireudenreichii KM. Lignes pleines, avant croissance ..lignes pointillées, aprèscroissance.

5

et al, 1976; Liu and Moon, 1982; Parker andMoon, 1982; Perez Chaia et al, 1987)observed stimulation of some strains of PABby some strains of LAB. The present studyconfirms the ability of several starter LABto stimulate the growth of PAB in wheys, inthat 13 of 14 strains of LAB tested stimu-lated growth of 1 or more of 4 strains ofPAB. The stimulatory activity of Lb aci-dophilus and Lb lactis depended on thestrain of PAB used. Similar results wereobtained by Liu and Moon (1982) andParker and Moon (1982). The stimulatoryactivity of Lactococcus lactis depended, onthe one hand, on the strain used (strain E8had no effect) and on the other hand, onthe PAB strain used (no stimulation by anyof the 4 strains tested was observed with

Growth of propionic acid bacteria

PAB strain T). This variation of the stimu-lation according to the PAB strain used wasnoted by Czarnocka-Roczniakowa et al(1972), but the stimulation was low. The lat-ter workers also found that Str thermophilusT149 stimulated the growth of 3 strains ofPAB. In the present study, ail 5 strains ofStr thermophilus stimulated PAB strains KMand H. However, for PAB strains L5 and T,the stimulation depended on the strain of

06 4

0.5

0.-1

Il~ 0.3 Il

.i.

02

0.1

~ R 12 16 20 2-1- 28 "~2 36 40 ...J-l. -48 52 .'Ï6 60 f).l. 61

Fraction number

0.8

Ilo.s

§

" OA

0.2

b 2

1

•••••

43 =

1'7'

!lë 1 1

'§ 5 6 9 10 :~)t7 8u ~~~, .;.

li! .i 'i

,. ~t:~?

i.'l li.,:::::i1i1i: ,•• 11!1

}j~ m~~1

o 1 2 3 ~ 5 6 7 8 9 10 Il

Peak number

Fig 9. Separation of Lactobacillus helveticus RRwhey on Sephadex G-25. (a) Chromatogram; (b)biomass (A600) of Propionibacterium freudenreichiiKM in control whey supplemented with mate rialfrom the different fractions.Séparation du lactosérum de Lactobacilius hel-veticus RR sur Sephadex G-25. (a) Chro-matogramme; (b) biomasse (A600) de Propioni-bacterium freudenreichii KM dans le lactosérumtémoin et dans le lactosérum témoin supplémentéavec les différentes fractions.

341

Str thermophilus used. In the present work,Lb helveticus strains were the most consis-tent stimulators, in that the 3 strains usedstimulated the 4 PAB strains studied. Suchstimulation has also been reported previ-ously by Czarnocka-Roczniakowa et al(1972) and Perez Chaia et al (1987).

The more detailed study of the stimula-tion of P freudenreichii KM by Lb helveticusRR showed that the increase in growth rateand cell yield coincided with an increasedconversion of lactate to propionate andacetate (fig 2). L lactate was used fasterthan D lactate in both starter and controlwhey, in agreement with previous data ofCrow (1986a). Ali strains of PAB used inthis study metabolize lactose, but lactosewas not utilized during growth in whey con-taining lactate which was still present whengrowth ceased. Marcoux et al (1992)observed lactose utilization by PAB in whey-based media containing lactose, lactate andyeast extract, but only when most of the lac-tate had been metabolized.

Amino acids are metabolized by PAB inthe presence of lactate (fig 3). In a mixtureof amino acids (vitamin-free, acid-hydrol-ysed casein), aspartate, serine, glycine, ala-nine and glutamic acid were mainly metab-olized (fig 5). These results are in agreementwith those of Brendehaug and Langsrud(1985). Crow (1986b) showed that aspar-tate, alanine and serine were used by PAB,but aspartate appeared to be the only onereadily metabolized in a Swiss-type cheeseenvironment. Of the 4 main free amino acidsused by PAB in this study, aspartate wasthe only one found to stimulate strain KM(fig 6). However, the concentration of freeaspartate was less than 0.1 mmol/I in starterand control wheys. In control whey, freeaspartate concentrations greater than 15times the concentration found in the starterwhey were needed to stimulate the growthof KM. Succinate was detected duringgrowth in control whey supplemented withaspartate or casein hydrolysate. This coin-

342 PG Piveteau et al

cided with a decrease in the ratio of propi-onate:acetate formed from lactate. Crow(1986b) reported similar findings. In the pre-sent study, the am ount of succinate detectedat the end of fermentation was 5 mmol/Igreater than the amount of aspartate used.Crow (1986b) proposed that the addition alsuccinate was formed from pyruvate in apathway involving CO2 fixation and oxaloac-etate production. CO2 was not measuredin the present study.

The stimulatory compound(s) was resis-tant to pasteurization and autoclaving andstable to precipitation with perchloric acid.These results agree with those of Hunterand Frazier (1961) but contrast with those ofCzarnocka-Roczniakowa et al (1972), whoreported a decrease of stimulatory activityafter pasteurization. Attempts were madeto determine the molecular mass. Ultrafil-tration was useless because of absorption ofthe stimulant(s) on the membrane. Gel fil-tration with Sephadex G25 suggests thatthe molecular mass is less than 5 000 Da,since most of the activity eluted after thevoid volume. These results also suggestthat several compounds are involved sincestimulatory activity was present in severalpeaks. However, since water was used aseluent, nonspecific adsorbtion of one par-ticular compound to the column may resultin the compound eluting in several peaks.Such interactions can be reduced by elu-tion with NaCI which increases the ionicstrength. However, NaCI could not be usedhere because of the inhibitory activity ofsalt on the growth of PAB (Hettinga andReinbold,1972).

The production of high levels of aminoacids and peptides by LAB in milk andcheese is not unusual (Accolas et al, 1980;Desmazeaud, 1983). Lb helveticus RR pro-duced free amino acids and peptides, sorneof which were utilized during growth of strainKM. PAB have intracellular (Perez Chaia etal, 1990; El-Soda et al, 1992) and extracel-lular peptide hydrolase activities (Langsrud

et al, 1977). Peptide utilization has beenreported du ring Swiss-type chee se ripeningwhere late fermentation was linked to anincreased breakdown of sorne peptides withmolecular masses around 3 000 (Blanc etal, 1979). Although the concentration of freeaspartate in starter whey was too low toaccount for the stimulation observed, it ispossible that aspartate is responsible for thestimulation as a constituent of a peptide.

ACKNOWLEDGMENTS

This research was partly funded by the EuropeanUnion through the Structural Fund Programme.PGP thanks Teagasc for the award of a stu-dentship.

REFERENCES

Accolas JP. Hemme D, Desmazeaud MJ, Vassal L,Bouillanne C, Veaux M (1980) Les levains lactiquesthermophiles : propriétés et comportement en tech-nologie laitière. Lait 60,487-524

Blanc B, Ruegg M, Baer A, Casey M, Lukesch A (1979)Essais comparatifs dans le fromage d'emmentalavec et sans fermentation secondaire. IV. Com-paraison biochimique et physico-chimique. SehweizMi/ehw Forsehung 8,27-36

Brendehaug J, Langsrud T (1985) Amino acidmetabolism in prapionibacteria: resting cells exper-iments with four strains. J Dairy Sei 68,281-289

Crow VL (1986a) Utilization of lactate isomers by Pro-pionibaeterium freudenreiehii subsp shermanii: reg-ulatory raie of intracellular pyruvate. App/ EnvironMierobio/52, 352-358

Crow VL (1986b) Metabolism of aspartate by Propioni-baeterium freudenreiehii subsp shermanii: effect onlactate fermentation. App/ Environ Mierobio/52, 359-365

Czarnocka-Roczniakowa B, Jaworski J, Kornacka D(1972) Effet stimulant de la microflore lactique surl'activité biochimique des bactéries prapioniques.Lait 52, 193-202

Desmazeaud M (1983) L'état des connaissances enmatière de nutrition des bactéries lactiques. Lait 63,267-316

El-Soda M, Ziada N, Ezzat N (1992) The intracellularpeptide-hydrolase system of Propionibaeterium.Mierobios 72,65-74

Growth of propionic acid bacteria

Gautier M, Lortal S, Boyaval P, Girard F, Lemée R, deCarvalho AF, Dupuis C (1993) Les bactéries propi-oniques laitières. Lai/73, 257-263

Hettinga OH, Reinbold GW (1972) The propionic acidbacteria, a review. J Milk Food Technol 35, 295-301,358-372,436-447

Hunter JE, Frazier WC (1961) Gas production by associ-ated Swiss cheese bacteria. J Dairy Sci44, 2176-2186

Langsrud T, Reinbold GW (1973) Flavor developmentand microbiology of Swiss cheese: a review. J MilkFood Techno/36, 531-542, 593-609

Langsrud T, Reinbold GW, Hammond EG (1977) Prolineproduction by Propionibacterium shermanii P59.J Dairy Sci 60, 16-23

Lee IH, Fredrickson AG, Tsuchiya HM (1976) Oynamicsof mixed cultures of Lactobacillus plantarum andPropionibacterium shermanii. Biotechnol Bioeng 18,513-526

343

Liu JAP, Moon J (1982) Commensalistic interactionbetween Lactobacillus acidophilus and Propioni-bacterium shermanii subsp shermanii. Appl EnvironMicrobio/44,715-722

Marcoux V, Beaulieu Y, Champagne CP, Goulet J (1992)Production of Propionibacterium freudenreichii subspshermanii in whey-based media. J Ferment Bioeng74,95-99

Parker JA, Moon NJ (1982) Interactions of Lactobacillusand Propionibacterium in mixed cultures. J FoodProt 45, 326-330

Perez Chaia A, Ruiz Hoigado AP, Oliver G (1987) Inter-action between Lactobacillus tietveticus and Propi-onibacterium freudenreichii subsp shermanii. Micro-bief Aliment Nutr 5,325-331

Perez Chaia A, Ruiz Hoigado AP, Oliver G (1990) Pep-tide hydrolases of propionibacteria: effect of pH andtemperature. J Food Prot 53,237-240