Metabolism of lactate and sugars by dairy

propionibacteria: A review

Pascal Piveteau

To cite this version:

Pascal Piveteau. Metabolism of lactate and sugars by dairy propionibacteria: A review. LeLait, INRA Editions, 1999, 79 (1), pp.23-41. <hal-00929635>

HAL Id: hal-00929635

https://hal.archives-ouvertes.fr/hal-00929635

Submitted on 1 Jan 1999

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, estdestinee au depot et a la diffusion de documentsscientifiques de niveau recherche, publies ou non,emanant des etablissements d’enseignement et derecherche francais ou etrangers, des laboratoirespublics ou prives.

Lait (1999) 79, 23-41© InralElsevier, Paris

23

Review

Metabolism of lactate and sugars by dairypropionibacteria: A review

Pascal Piveteau *

17, résidence le Coutelet, 86370 Vivonne, France

Abstract - Dairy propionibacteria are important organisms for the manufacture of Swiss-typecheese, for the biological production of propionate and vitamin 812 and have probiotic properties. Inail the se applications, their metabolic activities play a critical role. A complete understanding ofpropionate fermentation and of the metabolic routes used is therefore necessary. Dairy propioni-bacteria have a complex metabolism and involves several cycles. Lactate or sugars utilisation yieldspyruvate which can be reduced to produce propionate via the transcarboxylase cycles, or oxidised toyield acetate and CO2• During the coupled oxidation-reduction, ATP is produced by an electrontransport system and fumarate acts as the final acceptor. Although propionibacteria are mainly anaer-obes, the electron transport system can be used in the presence of oxygen and they possess the citratecycle, but can not grow under normal atmospheric oxygen pressure. The proportions of propionateacetate and CO2 produced vary depending on the strain used and this can be explained, to sorneextent, by their relative ability to utilise pyruvate via reactions of the citrate cycle. The physico-chemical environ ment during growth affects propionic acid fermentation; it is impaired by the pres-ence of oxygen and nitrate, but fermentation at acidic pHs enhances propionate production. Fer-mentation in presence of more than one substrate is complex and still poorly understood. When bothL- and D-lactate isomers are available, L-lactate is used preferentially. Although sugar fermenta-tion is more efficient, in the presence of lactate and sugars, there is evidence that lactate is usedpreferentially. Propionic acid fermentation is affected by the utilisation of amino acids, especially aspar-tate; co-metabolism of lactate and aspartate results in a lower propionate production and a decreaseof the ratio propionate:acetate. There is evidence that the utilisation of the products of proteolysis isan important event in the ripening of Swiss-type chee se and could account for the low propi-onate:acetate ratios observed in Swiss-type cheese. © Inra/Elsevier, Paris.

propionic acid bacteria {lactate { sugar { metabolism

Résumé - Métabolisme du lactate et des sucres par les bactéries propioniques laitières: unerevue. Les bactéries propioniques laitières sont importantes pour la fabrication de fromages de typeemmental, pour la production biologique de propionate et de vitamine 812 ainsi que pour leurs pro-priétés probiotiques. Dans ces domaines, leurs activités métaboliques ont un rôle crucial. Il est donc

Oral communication at the 2nd Symposium on Propionibacteria, Cork, Ireland, June 25-27, 1998.* Correspondence and reprints. P.Piveteau@hotmail.com

24 P. Piveteau

nécessaire de comprendre de façon approfondie la fermentation propionique et les voies métabo-liques empruntées. Les bactéries propioniques possèdent un métabolisme complexe impliquant plu-sieurs cycles. Le pyruvate issu de l'utilisation du lactate et des sucres peut être réduit, ce qui conduità la production de propionate via les cycles de la transcarboxylase. Le pyruvate peut aussi être oxydéen acétate et en CO2. Au cours de ces réactions d'oxydation-réduction, un transport d'électron, danslequel l'accepteur final est le fumarate, produit de l' ATP. Bien que les bactéries propioniques soientprincipalement anaérobies, le transport d'électron est possible en présence d'oxygène et elles possèdentle cycle des acides tricarboxyliques. Elles ne peuvent cependant pas se développer sous les pres-sions partielles en oxygène atmosphériques habituelles. Les proportions de propionate, d'acétate etde CO2 produites sont variables en fonction de la souche utilisée; ceci peut être en partie expliqué parla proportion de pyruvate utilisée via les réactions du cycle des acides tricarboxyliques. L'environ-nement physico-chimique influence la fermentation propionique. Elle est inhibée par la présenced'oxygène et de nitrate alors que la production de propionate est accrue à des pH acides. La fer-mentation en présence de plusieurs substrats est complexe et mal comprise. Lorsque les deux isomèresdu lactate sont présents, le L-Iactate est utilisé de façon préférentielle. La fermentation propioniqueest affectée par l'utilisation d'acides aminés, et plus particulièrement en présence d'acide aspar-tique. L'utilisation de l'aspartate pendant la fermentation du lactate provoque une diminution de laquantité de propionate produite ainsi que du ratio propionate : acetate. Au cours de l'affinage des fro-mages de type emmental, l'utilisation des produits de la protéolyse est un événement important quipourrait être corrélé aux faibles ratios propionate : acetate observés. © Inra/Elsevier, Paris.

bactérie propionique / lactate / sucre / métabolisme

1. INTRODUCTION

Interest in propionibacteria (PAB) is notrecent. They were first isolated fromEmmental cheese by Freudenreich andJensen in 1906 [28] and play a critical rolein the ripening and quality of Swiss-typecheese by producing propionate, acetate andCO2. The acids are important fIavour com-ponents and the eyes in the chee se are pro-duced from the CO2 [30, 31, 33, 40-42,83,88, 97]. Their ability to produce largeamounts of propionate makes PAB idealcandidates for the biological production ofthis anti-fungal fatty acid, and several pro-cesses have been proposed [9]. Other indus-trial applications include the production ofvitamin B12 and their use as possible probi-otics [51].

Dairy PAB have the ability to utilise avariety of substrates for energy [30, 31].Lactate is the main energy source for pro-pionic acid fermentation in cheese, but sug-ars and other organic compounds such as

glycerol, cellulose and complex by-prod-ucts of other industries such as gluten, whey,corn steep liquor, distiller' s waste, spentpotato wash, casein hydrolysates can alsobe used. The quality of cheese can be relatedto the extent of propionic acid fermentationboth in terms of the actual amounts of pro-pionate, acetate and CO2 produced as weil asthe relative proportions of each compound.The molar ratios of propionate, acetate andCO2 produced vary depending on the sub-strate used. As early as the 19th century, thestoichiometry of the reaction resulting inthe production of propionate from lactatewas established by Fitz:

3 lactate -----.2 propionate + 1 acetate+ 1CO2 + 1H20

However, in Swiss-type cheese, thesetheoretical equations are rarely found andthe relative concentrations of propionateacetate and CO2 may be significantly dif-ferent from the expected molar ratios of2: 1: 1.

Metabolism of propionibacteria

2. CARBOHYDRA TEAND LACTATE DEGRADATION

The early literature conceming PAB hasbeen reviewed by Hettinga and Reinbold[30-32]. Ali dairy PAB, except P.freuden-reichii subsp. freudenreichii are able tometabolise lactose [30, 31, 37, 89]. Thisimplies that PAB are equipped with a lactosetransport system and must be able to hydrol-yse lactose as the first step of its utilisation.To my knowledge, there are no reports ontransport systems in PAB. ~-Galactosidaseactivity was demonstrated in P.freudenre-ichii subsp. shermanii strains by Hartleyand Vedamuthu [29]. Depending on thestrain tested, toluene-acetone treatment ofthe cells either increased or decreased theenzymatic activity, while with one strain,almost ail the activity was lost in the cell-free extract. There was sorne evidence thatthe ~-galactosidase system of P. freuden-reichii subsp. shermanii is constitutive ratherthan inducible as only slight differences in~-gal activity were observed in culturesgrown on lactate and different carbohydratesincluding lactose, glucose and galactose ormixtures of sugar and lactate. The optimalconditions for enzyme activity depended onthe strain and treatment used (untreated cells,solvent treated cells or cell-free extract); theoptimal temperature and pH ranged from32 -c to 58 -c and 7.0 to 7.5, respectively.

Hexoses are utilised via the Embden-Meyerhof-Parnas (EMP) pathway. Phos-phohexose isomerase, aldolase, triose phos-phate isomerase, glyceraldehyde 3-phosphatedehydrogenase and a pyrophosphate phos-phofructokinase are present [31, 59, 109,114]. PAB also contain two glucokinaseactivities, a po1yphosphate dependent andan ATP dependent activity both carried onthe same protein [62,63,73] and polyphos-phate was iso1ated during growth on glu-coseand lactate [10]. Another line of evi-dence for the EMP pathway was fumishedby the study of fermentations using radio-labeled glucose [108, 109, 112]. However,the presence of another pathway for the uti1-

25

isation of glucose was clearly demonstratedby the distribution of the label in the prod-ucts depending on the position of the labelon the glucose molecule, and the fact thatglucose is still utilised in the presence ofglycolytic inhibitors. PAB contain the pen-tose phosphate pathway and all the enzymesof the pathway have been detected [7,99,102, 104]. The position of the label in pro-pionate during the fermentation of labeledribose and gluconate is consistent with theutilisation of pentoses via transketolase-transaldolase conversion to fructose ô-phos-phate [108, 109]. However, under anaerobieconditions, most of the glucose is utilisedthrough the EMP pathway alone [108].

During growth on lactate, lactate is ini-tially oxidised to pyruvate via a NAD+ inde-pendent lactate dehydrogenase. The degra-dation of sugars also leads to the productionof pyruvate.

3. PROPIONA TE PRODUCTIONAND THE TRANSCARBOXYLASECYCLES

Propionate is the main compound pro-duced by the reduction of pyruvate by PAB.It involves several reactions [31, 98, 108,109] arranged in cycles (figure 1). The firststep is the formation of oxaloacetate by tran-scarboxylation, in which the COOH groupof methy1malonyl-CoA reacts with pyru-vate to form oxaloacetate and propionyl-CoA. Oxaloacetate is reduced to succinatevia malate and fumarate in two NADHrequiring reactions. Succinate is then con-verted to propionate via methylmalonyl-CoA intermediates (succinyl-CoA and pro-pionyl-CoA); the carboxyl group removedfrom methylmalonyl-CoA is transferred topyruvate to yield oxaloacetate, thus corn-pleting one cycle. Methylmalonyl-CoA isalso regenerated from succinyl-CoA duringpropionate production, thus creating the sec-ond of the two transcarboxylase cycles, andcan react with a new molecule of pyruvate.Because of the transcarboxylation reaction,

26 P. Piveteau

Pol _P Glucosey n ~ glucokinase

Poly-Pn-i +1Glucose-6-P ----------,

phosphobexose isomerase ~

Fructose-6-Ppyrophosphate PhospbOtroetok~' -,

PPi ~Pi .........

Fructosf-I,6-diP "-aldolase 'f '-Glyceraldehyde-3-P 'li(

glyceraldehyde J-phosphate debydrogenase t3-P-glycerate

AMP enolase~

~~Pi P-enolpyruvatepyruvate phosphate dikiDase A11' ADP osphoenol carboxytransphosphorylase

W COb~Lactate ~ Pyruvate pp.

C02~=genase '~nSCarborYlase,phosphotransace~ Acep-coA Oxaloacetate

Ik' acetyl-P .f malicdehydrogenase

acety mAse ADP r .Methylmalonyl CoA Propionyl CoAATP Citrate Malate  methylma/onyl racemase

Acetate ~ \. fumarase T methylmaJonyl mutase'If e- transfer Succinyl CoACO2 Fumarate j

o-Ketoglutarate \(umaratereductase.L ...,

~ Succinate LoA transferase

~6-P-gluconate

~C02Pentose-P

.J

PropionateFigure 1. Cycles involved in the proprionic aeid fermentation (Wood [108]). LHD: lactate dehy-drogenase; Poly-Pn: polyphosphate; PPi: pyrophosphate. For reasons of c1arity, only the pyrophos-phate dependent conversion of fructose-ô-P to fructose-I ,6--diP is shown and ATP generation by theelectron transfer system is omitted. Ali the reactions are directed towards propionate production,even though the reactions are reversible.Figure 1. Cycles de la fermentation propionique (Wood [108]). LDH : lactate déshydrogénase;Poly-Pn : polyphosphate ; PPi : pyrophosphate. Pour faciliter la lecture, seule la réaction de conver-sion du fructose-6-P en fructose-l ,6-diP nécessitant des pyrophosphates est montrée. De même,la production d'ATP par le transfert d'électron est omise. Toutes les réactions sont dirigées vers la pro-duction de propionate, bien qu'elles soient réversibles.

CO2 fixation is minimal and is only used toproduce catalytic amounts of oxaloacetatewhen the cycle is broken when for exam-pie succinate accumulates as an end product[1]. Under such circumstances, oxaloacetateis generated by condensation of CO2 with

phosphoenolpyruvate in a reaction catal-ysed by carboxytransphosphorylase; inor-ganic phosphate is converted to pyrophos-phate and the reaction does not need ATP[60,90,91]. PAB can also produce phos-phoenolpyruvate from pyruvate; this reac-

Metabolism of propionibacteria

tion is catalysed by a dikinase and requiresinorganic phosphate and ATP [56, 57,114].Because CoA is recycled and free CO2 fix-ation is minimal, propionate is producedwith a minimum loss of energy.

4. ENZYMES INVOLVEDIN PROPIONIC ACIDFERMENTATION

The enzymes involved in propionic acidfermentation (figure 1) have been reviewedby Hettinga and Reinbold [31]. Since then,a great deal of attention has been given to thetranscarboxylase. This interest is due firstlyto the central role of the transcarboxylasereaction in the propionic acid fermentation,and secondly, because it catalyses a car-boxylation without the intervention of freeCO2 which is unusual among carboxylases.Transcarboxylase has a complex structure[5, 76, 107, 113, 116]; it comprises threesubunits giving it a flexible structure. Thecentral subunit, or l2S subunit, consists of6 identical polypeptides (Mw 60 000 dal-tons); it contains 12 CoA ester binding sites.In the complete transcarboxylase complex,this l2S subunit is surrounded by 6 outer5S subunits. Each 5S subunit comprises2 identical polypeptides and contains C02+

and Zn2+. The central and outer subunits arelinked by 12 biotinyl or US subunits, eachcontaining 2 peptides of 65 and 50 aminoacids residues. The complete 26S form ofthe enzyme dissociates to a l8S form at neu-tral pH, through the loss of 3 outer subunits.The transcarboxylase reaction is divided intwo half-reactions; on the central subunit,the carboxyl group is transferred frommethylmalonyl-CoA to the biotinyl subunitby carboxylation of biotin, the second reac-tion takes place on the outer subunit wherethe carboxyl group is transferred to pyru-vate to form oxaloacetate. Neither ATP normetal ions are needed for the transcarboxy-lation reaction. Further characterisation ofthe structure and functions of the subunitshas been facilitated by cloning and expres-sion of the transcarboxylase operon in

27

E. coli. The monomer of the l2S subunit isa protein of 604 ami no acid residues andcontains regions with extensive homologywith other carboxylases [100]. By deletingamino acids from the carboxyl end of the12S subunit, Woo et al. [106] demonstratedthat this part of the molecule is involved inthe stability of the 12S subunit and in thebinding with the outer subunits. The 5S sub-unit monomer contains 519 amino acidresidues with considerable homology toyeast pyruvate carboxylase and Klebsiellapneumoniae oxaloacetate decarboxylase[76]. The Trp residues of the 5S subunit areinvolved in binding with the 1.3S subunitand with pyruvate. The structure and func-tion of several regions of the 1.3S subunithave been determined. The amino terminusis involved in the formation of the complex[39]. Biotin is attached by an amide link-age to LYS89and the penultimate Ile residueis involved in biotinylation [58]. Samois etal. [76] observed a highly conserved tetrapep-tide Ala-Met-biotynyl-Met(87_90) in mostbiotin carboxylases which is involved in thecarboxyl transfer. By using truncated 1.3Ssubunits, Shenoy et al. [84-86] deierrninedthe role of the regions around the biotinyllysine and that residues 1 to 18 are requiredfor the assembly of the complex and residues59 to 78 for activity. Finally, there is sorneevidence for the existence of a fourth proteinsubunit in the assembly of the complex [87].

Methylmalonyl-CoA mutase is anotherinterestingP AB enzyme. It requires adeno-sylcobalamin (coenzyme B12) for activityand catalyses the rearrangement of succinyl-CoA into methylmalonyl-CoA [38]. Thisenzyme has been the object of several struc-tural studies; it is an ex~ dimer [27, 119]with an apparent Mw of 150000 daltons.The cloning and sequencing of the struc-tural genes have shown that the twomono mers have a close structural homol-ogy but are different. The exand ~ subunitscontain 638 residues and 728 residuesrespectively and have Mw of 69465 and80147 daltons, respectively. The exsubunitcontains a sequence homologous to a portion

28

of the sequence coding for the methyl-malonyl-CoA binding site of transcarboxy-lase [53, 54]. The enzymatic activity is loston dissociation ofthe two subunits. Cobal-amin is covalently bound to the enzyme butdoes not stabilise it [55]. The accepted modeof action of this type of enzyme is the gen-eration of free radicals [105]. The breakingof the Co-C bond of coenzyme B 12 inducesthe change from its Co3+ form to its C02+

form and a 5' -deoxyadenosyl free radicalis generated. The radical takes a hydrogenatom from succinyl-CoA which conse-quently is converted into a radical; this newradical undergoes rearrangement of the acyl-CoA group to the position formerly occu-pied by the hydrogen atom. Finally, a hydro-gen atom from the 5' -deoxyadenosyl groupis retumed to the product-like radical, thusgenerating methylmalonyl-CoA and a 5'-deoxyadenosyl radical. The rebinding of theCo-C bond regenerates the cofactor in itsCOH form and the enzyme is ready for anew catalytic cycle. The exact mechanism ofthereaction is still under investigation [l l].

A third peculiarity.of PAB is their abilityto utilise pyrophosphate instead of ATP forseveral reactions. Pyrophosphate dependentenzymes have been described in PAB [114].Phosphoenolpyruvate carboxytransphos-phorylase catalyses the fixation of CO2 withphosphoenolpyruvate to yield oxaloacetate[49,90, 91]:

phosphoenolpyruvate + CO2+ Pi <=:> oxaloacetate + PPi

In the absence of CO2, this enzyme isalso involved in the irreversible formation ofpyruvate from phosphoenolpyruvate [114]:

phosphoenolpyruvate + Pi----.pyruvate + PPi

The active enzyme exists in monomeric,dimeric and tetrameric forms and requiresMg2+, Mn2+ or C02+ for activity. The tetra-meric enzyme has been obtained in crystalform, has a molecular weight of 430 kD, asedimentation coefficient of 15.2S,and isthe most active form of the enzyme.

P. Piveteau

The second pyrophosphate dependentenzyme is pyruvate phosphate dikinasewhich catalyses the reversible phosphory-lation of pyruvate and inorganic phosphatein the presence of ATP [56, 57]:

pyruvate + ATP + Pi <=:>phosphoenolpyruvate + AMP + PPi

A third pyrophosphate enzyme ispyrophosphate phosphofructokinase [59]which is involved in the transfer of phos-phate from pyrophosphate to fructose-6-P:

Fructose-6-P + PPi <=:>fructose-l,6-diP+Pi

ATP-phosphofructokinase activity whichis common in other organisms, is low inPAB compared to the pyrophosphate depen-dent enzyme.

5. ACETATE PRODUCTIONAND THE CITRATE CYCLE

The reduction of fumarate to succinate,during propionate production, is coupled tothe oxidation of pyruvate to acetate [12,92].This oxidation, with acetyl-CoA as an inter-mediate, results in the production of equimo-lar concentrations of ATP, acetate and CO2•

Wood [108] observed large variations in theamounts of acetate and CO2 produced dur-ing the fermentation of glucose, and stud-ies involving labelled glucose showed thatthese differences only appeared at the con-version of acetyl-CoA. Ithas been proposedthat sorne acetyl-CoA may be utilisedthrough the tricarboxylic acid (TC A) cycle[108]. In this scheme, acetyl-CoA is con-verted to citrate, which in tum yields succi-nate via œ-ketoglutarate, with the direct con-sequence that less acetate is produced.Although PAB are essentially anaerobes,Bonartseva et al. [7] detected all the enzymesof the TCA cycle in P. freudenreichii subsp.shermanii and P. jensenii, grown either aer-obically or anaerobically.

Growth and metabolism of glucose andlactate by PAB under aerobic conditions has

Metabolism of propionibacteria

been reported [13, 23, 66, 67, 78]. Oxygenuptake during growth on glycerol, glucoseand lactate and oxidation of lactate wasobserved [93]. However, growth is slower inthe presence of oxygen [23]; the formationof propionate, acetate and succinate is inhib-ited and pyruvate accumulates [81, lOI].Because pyruvate accumulates, energy cannot be produced from substrate level phos-phorylation, suggesting that it must be pro-duced by oxidative phosphorylation; indeedDe Vries et al. [23] observed that, in thepresence of 02 and at limiting lactate con-centrations, part of the lactate was com-pletely oxidised to CO2 via the TCA cycle.The presence of the TCA cycle in PABcould explain differences in the fermenta-tion balances (amounts and relative pro-portions of propionate, acetate and CO2 pro-duced) observed between strains as will beillustrated later.

6. GENERATION OF ATPVIA ELECTRON TRANSPORT-COUPLED PHOSPHORYLATION

An electron transport-coupled phospho-rylation takes place, in the presence andabsence of 02' during the cou pied reduc-tion of fumarate to succinate and oxidationof pyruvate to acetate and CO2• Componentsof the respiratory chain have been identi-fied in P. freudenreichii subsp. shermaniiand P. acidipropionici and consist of dehy-drogenases [24, 79, 93, 94] and mena-quinones [80, 95, 96], which act in collect-ing the reducing equivalents from thesubstrate and distributing them to thecytochromes [2, 13, 23, 80]. The exactmechanism of electron transport is still notclear and two models, differing in the loca-tion and role of cytochrome b in the chain,have been proposed [2, 80]. Schwartz andSporkenbach [80] suggested that cyto-chromes b are not involved in the anaero-bic transport system; they suggested thatcytochromes a, b, d and 0 were alliocated inthe oxygen linked branch of the chain and

29

are the remains of a former aerobicmetabolism. The second model involvestwo types of cytochrome b. One is a lowpotential cytochrome (E'o = -20 mV) thatcouples NADH oxidation to reduction offumarate during anaerobie metabolism; thesecond type has higher potential (E'o =+ 120 mV and E'o = + 90 m'V), and alongwith cytochrome d (E' 0 = + 140 mV), takespart in an aerobic branch of the electron trans-port. Traces of cytochrome a, c and 0 werealso observed in this study. The variableability of PAB strains to grow under aerobicconditions could be related to the ratio betweenthe high and low potential cytochromes.

When lactate is used as a substrate, gen-eration of ATP only occurs in the productionof acetate and reduction of fumarate to suc-cinate but during sugar metabolism, addi-tionaI substrate-level phosphorylations occurin the formation of pyruvate. The overallfermentation balances are:

3 lactate + 3 Pi + 3 ADP ~ 2 propionate+ 1 acetate + 1 CO2 + 2 H20 + 3 ATP

1.5 glucose + 6 Pi + 6 ADP ~ 2 propionate+ 1 acetate + 1 CO2 + 2 H20 + 6 ATP

7. INFLUENCE OF THE PHYSICO-CHEMICAL ENVIRONMENT

pH greatly influences the formation ofthe end products from lactate and carbohy-drates. Although optimal growth occurs atpHs between 6 and 7, the production of pro-pionate from lactose, glucose and lactatecan be increased by growth at acidic pH [34,46, 69] whereas acetic acid production isnot significantly affected by the pH. Thehigher production of propionate at acidicpHs can be explained by the formation ofnon-growth associated propionate, but thereis no direct evidence of such a phenomenomin PAB.

Propionic acid can inhibit the growth ofpropionibacteria under acidic conditions[26], e.g. 100 mmol-Lr! propionate pre-vented the growth of P.freudenreichii subsp.

30

shermanii during glucose metabolism, atpH 5.2 as observed by Namba et al. [50].However, this concentration of propionate isvery high and would not be reached duringnormal growth.

Propionate production is affected by thelevel of oxygen present in the environment.Pritchard et al. [66] reported that P. freuden-reichii subsp. shermanii behaves as a fac-ultative anaerobe under 02 partial pressuresof 0 to 42 mm Hg, has an aerobic metabolismunder 02 partial pressure of 42 to 330 mmHg but does not grow at 550 mm Hg. Lackof growth at normal 02 atmospheric pres-sure is attributed to the inhibition ofcytochrome synthesis [23]. Although theviability of the cells is not affected, acidproduction is inhibited under partially aer-obic growth conditions. After restoration ofanaerobie conditions, acetate is produced atthe same rate as before the aeration phasebut propionate production is drasticallyreduced [35,36,81].

Lactate fermentation is also affected bythe presence ofN03-; Van Gent-Ruijters etal. [101] reported that, during anaerobiegrowth on lactate, the production of propi-onate by P. pentosaceum was drasticallydecreased, production of acetate did notchange and pyruvate accumulated. In thepresence of N03 -, CO2 was produced fromthe TCA cycle and the molar growth yieldwas increased due to ATP production viathe TCA cycle and oxidative phosphoryla-tion cou pied to nitrate reduction. N02- wasproduced from N03- and its subsequentreduction (after N03- was exhausted) gavefermentation patterns similar to thoseobserved in the absence of any inorganicelectron acceptor. However, a N02- con-centration of 9 mmol-Lr ' inhibited growth.

8. FERMENTATIONOF SUBSTRA TE MIXTURES

The utilisation of more than one substrateduring propionic acid fermentation is com-

P. Piveteau

plex and can affect the yields of propionate,acetate and CO2.

In either complex or defined medium,P. freudenreichii subsp. shermanii utilisesL-Iactate, preferentially over D-Iactate, ina mixture of both isomers independently ofinitial pH, initial lactate concentration andinitial ratio of the two isomers. Whenthe medium was altered to resemble a ripen-ing Swiss cheese (pH 5.3, 2.1 % NaCI,growth at 22 "C, 145 mrnol-L? L-Iactateand 55 mmol-L:" D-Iactate), the rate ofL-Iactate utilisation was also higher thanthe rate of D-Iactate utilisation. This pref-erential utilisation was explained by the factthat L-Iactate metabolism results in the pro-duction of high intracellular pyruvate con-centration which has a stronger inhibitoryeffect on D-Iactate dehydrogenase activitythan on L-Iactate dehydrogenase activity.With 10 mrnol-Lr! lactate as substrate,20 mrnol-L" pyruvate did not inhibit L-LDHactivity but caused a 75 % decrease inD-LDH activity; 40 mmol-L:" pyruvatedecreased L-LDH and D-LDH activities by6 and 81 %, respectively [16]. Similar trendsare observed with representatives ofPi freudenreichii subsp.frelldenreichii andP. acidipropionici (Piveteau, unpublished).

As mentioned earlier, higher levels ofATP are produced from similar amounts ofsugars than from lactate. As a consequence,higher growth rates and cell yields areobtained during the fermentation of lactose,glucose and galactose than during fermen-tation of lactate. More propionate andacetate, expressed per mole of pyruvate, areproduced during lactate fermentation thanduring the fermentation of sugars [3,45-47,65]. Despite this, PAB utilise lactate fasterthan sugars when lactate and carbohydratesare both available. In a mixture of lactateand either lactose, glucose or galactose,P. freudenreichii subsp. shermanii strainsuse lactate preferentially [45,65]. Liu andMoon [48] and Crow and Turner [21] foundthat sorne carbohydrate was utilised duringlactate fermentation by P. freudenreichii

Metabolism of propionibacteria

subsp. shermanii. P. acidipropionici alsoutilised at least sorne of the lactose, glucoseand galactose during lactate metabolism[65]. However, during growth in cheesewhey supplemented with fermented milkpermeates [52] and in wheys produced byacidification of milk with cultures or addi-tion of lactic acid [64,65], lactose was notutilised during metabolism of lactate. Thispreferential utilisation of lactate in mediacontaining both lactate and a sugar is appar-ently contradictory with the fact that thesugar fermentation is more efficient thanthe fermentation of lactate. The reason(s)for this is unclear; perhaps the shortermetabolic pathway from lactate than sug-ars to pyruvate is a factor.

Although lactate is the major substratemetabolised by propionibacteria in Swiss-type cheese, the products of proteolysis arealso used. The utilisation of amino acidscan affect the yields of propionate, acetateand CO2• Crow [17] studied the effect ofaspartate on lactate fermentation. In eitherdefined or complex medium, more lactatewas converted to acetate and CO2 than topropionate when aspartate was added to thegrowth medium. Aspartate utilisationresulted in the production of equimolar con-centrations of succinate, and NH3. Duringco-metabolism of 180 mmol-L:' lactate and33 mmol-L:' aspartate by three P.freuden-reichii subsp. shermanii strains in complexmedium, the ratio of propionate to acetatedecreased from 2: 1 (lactate on its own) to1.5: 1.0. The presence of either caseinhydrolysate or aspartate during fermenta-tion of acid whey results in an increase incell yield, a production of succinate pro-portional to the amount of amino acidsadded and a decrease in the molar ratio ofpropionate:acetate [64, 65]. Moreover, thepropionate yield and the ratio of propi-onate:acetate varies, depending on thesource of nitrogen used during glucosemetabolism by P. freudenreichii subsp.shermanii [69] and lactate metabolism byP. acidipropionici [46]. Altering the growth

31

conditions to resemble Swiss chee seincreased the rate of aspartate utilisation,suggesting that aspartate metabolism canbe important in the ripening chee se [17].This hypothesis is supported by results ofchee se trials using aspartase deficient vari-ants of PAB strains. In chee ses made witha strain of low aspartase activity, the ratio ofpropionate:acetate was 1.8: 1.0 and succi-nate concentration was the lowest(10.1 mmol-kg:' cheese) compared to1.5: 1.0 and 20.4 mmol-kg:' cheese, respec-tively, in chee ses made with the strain withgreater aspartase activity [22]. The amountof succinate detected at the end of warmroom ripening depends, to sorne extent, onthe conditions used during cheese making. Incommercial cheese and experimental cheesemade with Lb. helveticus, 18 to 20 mmol-kg 'of cheese were detected, while 39.4 mmol-kg 'was found in cheese made with Lb. bulgar-icus [21]. However, the residual galactosecould have been metabolised by the rela-tively high levels of non starter lactic acidbacteria in the cheeses manufactured withLb. bulgaricus which rendersany compari-son between the cheeses made with Lb. hel-veticus and with Lb. bulgaricus difficult.Sebastiani and Tschager [82] also observedthe role of aspartate metabolism in succi-nate production in Swiss-type cheese. PABare equipped with a complex intracellularpeptidase system [43, 68] and there is strongevidence that PAB can utilise short chainpeptides [4, 64, 65]. It is likely that aminoacids and peptides, produced by proteoly-sis during ripening form a pool sorne ofwhich are subsequently used by PAB; theextent of this utilisation may explain at leastsorne of the variations in the relative con-centrations of propionate, acetate, CO2 andsuccinate produced.

During co-metabolism of aspartate andlactate, the final ratio of propionate:acetatedepends on the initial aspartate:lactate ratio.Crow [17] reported that if 2 moles of aspar-tate are utilised per mole of lactate, the ratiopropionate:acetate is 1:l, while ifless thanone mole of aspartate is used per mole of

32

lactate, the ratio will be between 1:1 and2: 1. This worker also observed that in amedium containing aspartate, propionatebut not lactate, 3 moles of aspartate areutilised with 1 mole of propionate to yield3 moles of succinate, 3 moles of NH3, 1 moleof acetate and 1 mole of CO2• The effect ofaspartate metabolism on the amount of pro-pionate produced from lactate was explainedby the involvement of the TCA cycle andthe interrnediates of this cycle (malate andfumarate) were detected during aspartatemetabolism [17, 19]. Aspartate is convertedto fumarate via aspartase activity [17, 18,25, 103]. In the scheme as proposed by Crow[17], fumarate has to be reduced to succi-nate in arder to balance the oxidation-reduc-tion status of the cell; the reduction to suc-

Ammonia

P. Piveteau

cinate is therefore coupled to the oxidationof pyruvate to acetate (figure 2). Hence, theoverall consequence of this increased pool offumarate is a higher production of acetateand CO2 and a lower amount of propionateproduced.

An alternative mechanism of aspartatemetabolism was proposed by Rosner andSchink [75]. In their scheme, propionateacts as a COOH acceptor from oxaloacetatederived from aspartate and is converted tosuccinate via malate and fumarate. Thisallows the conversion of oxaloacetate topyruvate and ultimately to acetate and CO2,

thus producing reducing equivalents usedfor the reduction of aspartate to succinate(figure 3).

Succinate

Fumarate rcductescAmmonia

Aspartate-+- ~ Aspartate2 Fumarate ;;2 WAspartasc _ _ _ _

CO 2 W- - - - - - - ADP ATP2~~ Acetyl CoA -.... Acetyl-P \.4 Acetate

PDH0,"Pyruvate CoA

__ ~ 2H+ ............

......Pyruvate ~ oxaloacetate ~

MethYlmGoA ~ t! Propionyl CoA Malate

~_s_u_c1 ......inyl CoA )

. " ~Propionate 2 + Fumarate

2 lactate

Fumaratc rcductsse

......

Acetate

Propionate

Figure 2. Aspartate and lactate metabolism according to Crow [17]. PDH, pyruvate dehydrogenase.LDH lactate dehydrogenase.Figue 2. Métabolisme de l'aspartate et du lactate selon Crow [17]. PDH : pyruvate déshydrogénase.LDH : lactate déshydrogénase.

Metabolism of propionibacteria

9. VARIABILITYOF THE RATIOS AND YIELDSOF PROPIONA TE AND ACETATE

Many studies have delt with the biolog-ical production of propionate from a vari-ety of substrates present in cheap industrialby-products. The amounts of propionate andacetate produced and the ratios propi-onate:acetate varied according to the COn-ditions and strain used (table J); in general,P. acidipropionici strains have a betterpotential for propionate production thanotherdairy PAB [3,8,15,71,115]. Thestrain dependency towards propionate pro-duction is confirmed concerning the rela-tive proportions of the fermentation endproducts obtained from lactate or carbohy-drate [3, 20]. Wood [108] reported a rangeof ratios propionate:acetate and CO2: acetatefrom 2.1 to 14.7 and 1.0 to 6.3, respectivelywhen glucose was the growth substrate; such

2 Aspartate2NH/12 Fumarate

4/3 ADP----t-- 4[H]

4/3ATP~+

2 Succinate

Malate

2[H] ....

Propionyl CoAoxaloac::e=ta=te:...- jf ~cO2 -----,

Pyruvate Methylmalonyl CoA CoA

2[H]+cal t.,,_. 1

Aœ~~coA S~~

Acetate-PADP~

ATP ---,

Acetate

Aspartate

NH4+4Fumarate

+

33

great differences between strains can beaccounted for by their relative ability tometabolise pyruvate through the TCA cycle.In this event, the acetate yield is mostlyaffected, and the decrease in acetate pro-duced explains the high ratios found.Depending on the substrate and the mediumused, succinic and pyruvic acid are producedat various levels during the fermentation oflactate and carbohydrate. P. acidipropioniciformed 18 to 38 mmol-Lr! succinate and17 mmol-Lr! to 62 mmol-L:' pyruvate dur-ing the fermentation of III mmol·L-1 lac-tose or 222 mmol-L:' glucose, but loweramounts (4 mmol-L'" succinate, no pyru-vate) were detected during the metabolismof 444 mmol-L-1 lactate [34,46,47]. Crow[19,20] detected 3 to 5 mmol-L:' succinateand pyruvate during the fermentation of155 mmol-L" lactate and 145 mmol-Lr! lac-tose. Pyruvate was detected during growth ofP.freudenreichii subsp. shermanii [52] and

Propionate

Succinate

Figure 3. Pathway of aspartate plus propionate fermentation according to Rosner and Schink [75].Figure 3. Voie de l'utilisation de l'aspartate en présence de propionate selon Rosner et Schink [75].

t;>+-

Table I. Comparison of the relative amounts of propionate and acetate produced by various strains of propionibacteria.Tableau I. Comparaison des quantités relatives de propionate et d'acétate produites par diverses souches de bactéries propioniques.

Strain used Medium Conditions Propionate Acetate Molar ratio Recovery (%) Authors

P. freudenreichii whey lactose batch fermentation pH 7.0 19 g·L-1 5 g·L-1 3.8/1.0 102 Bodie et al. [6]subsp. shermanii PS209

PS209 + Lb. casei TA 101 whey lactose batch fermentation 30 g·L-1 10 g-L-1 3.0/1.0 86

P. freudenreichii lactose complex steady state continuous 1.6/1.0 to 79 Schutz et al. [77]subsp. shermanii medium pH 4.5 to 7.5 1.9/1.0

:-0P. freudenreichii 20271 lactose complex steady state continuous 2.4 g·L-1 1.0 g·L-1 1.9/1.0 "'é

+ Lb. acidophilus 20079 medium pH5.8 <'(0

ôl:OC

P. freudenreichii 20271 lactose complex steady state continuous 1.0 g·L-1 0.8 g·L-1 0.9/1.0+ Le. Lactis medium pH 5.8

P. freudenreichii subsp. Lb. Helveticus agitated immobilised cells 6.6 g·L-1 4.1 g·L-1 1.6/1.0 137 Champagne et al. [14]shermanii fermented whey static immobilised cells 7.3 g·L-1 2.6 g·L-1 2.8/1.0 140

P. acidipropionici lactose rich nitrogen batch pH 7.12 2.111.0 59 Hsu and Yang [34]ATCC4875 complex medium pH6.15 2.5/1.0 6

pH5.5 3.3/1.0 61lactose low nitrogen batch pH 7.03 3.0/1/0 72complex medium pH 6.12 2.7/1.0 72

pH5.5 3.4/1.0 73

Table I. (Continued) / Tableau I. (Suite).

P. acidipropionici complex medium batchpH 6.6ATCC4875 lactate 18 g-L-1 7 g·L-1 2.1/1.0 90 Lewis and Yang [46,47]

lactose 15 g-L"! 5 g-L-1 2.4/1.0 71glucose 16g·L-l 4.5 g·L-1 2.9/1.0 77lactate packed bed pH 7.0 1.7/1.0 nc

packed bed pH 5.2 1.9/1.0 nc3::~

P. freudenreichii subsp. lactate whey based batch 0.1 to 1.5 % 0.1 to 0.6 % 0.8/1.0 to 3.211.0 83-85 Marcoux et al. [52] ~cr-0

shermanii B 123 medium immobilised cells (3rd re-use) 8.3 to 9.4 g-L:' 3.7 g-L:' 1.8/1.0 to 2.0/1 122 m'a0

35g·L-l 9 g.L:'...,

P. acidipropionici lactate whey permeate sequential fermentation 3.1/1.0 80 Colomban et al. [15] "0..,ATCC4965 based medium 0

"0o':;:l

0:P. acidipropionici lactose whey steady state 24.2 to 27.4 g·L-19.7 to 10 g-L:' 2.0/1.0 to 98-107 Yang et al. [117] ~

()

ATCC4875 permeate continuous immobilised cells 2.2/1.0 <ii:J.~

P. acidipropionici P9 glucose complex batch pH 6.9 2.7/1.0 nc Ozadali et al. [61]mediumcorn steep liquor batch pH 6.9 3.0/1.0 nccomplex medium

ne: the percentage recovery could not be calculated [rom the data presented.ne : le pourcentage de récupération ne pouvait pas être calculé à partir des données obtenues.

voV1

36

P. acidipropionici [117] in whey-basedmedia. Succinate, but not pyruvate was pro-duced during propionate production fromwhey lactose by immobilised cel!s ofP. acidipropionici [118].

The variability of the propionate:acetateratio by different strains is also found inSwiss-type cheese. A range of ratio from1.1:1.0 to 2.2: 1.0 was reported by Langsrudand Reinbold [42]; Crow and Turner [21]reported ratios of 0.7: 1.0 for an experimen-tal Emmental cheese made with Lb. bul-garicus, 0.95: 1.0 and 1.0:1.0 for two exper-imental cheeses made with Lb. helveticusand 1.5: 1.0 for a commercial Emmentalcheese. The large variations in the metabolicactivity of Propionibacterium strains werefurther confirmed in experimental smallscale Swiss-type chee se manufacture where19 PAB strains could be arranged into fourclusters depending on their ability to pro-duce propionate and acetate [72].

10. INFLUENCE OF MINORFERMENTA nON PATHWA YS

Another factor affecting propionate andacetate production is the formation of newcarbohydrates [32]. Trehalose is producedduring glucose [109-111] and lactate [74]fermentation. Three polysaccharides two oflow Mw (-350-450 and -200-300) and oneofhigh Mw (> 5000) containing methylpen-tose, glucose and galactose were detectedduring growth of P. acidipropionici at highlactose concentration and was related to alow propionate:acetate ratio « 2: 1) [20].The production of a polysaccharide ofMw < 5 800 was also described during lac-tose fermentation by P. acidipropionici inwhey-based medium, and its concentrationwas proportional to the C:N ratio [70]; itwas composed of glucose, galactose, man-nose and two methylpentoses (rhamnoseand fucose), but a lower proportion of methyl-pentose was present compared to thepolysaccharide studied by Crow [20]. Againthe differences are probably due to the dif-

P. Piveteau

ferences in strain and growth medium used.The effect of the polysaccharide on the pro-pionate and acetate production can beattributed to a direct loss of C for synthe-sis, but more importantly to the effect of thereduced components, the methylpentoses,which decrease the oxidation-reduction sta-tus of the cel!; the overal! result is anincreased production of oxidised products,i.e. acetate, to maintain the oxidation-reduc-tion balance and thus a decrease in propi-onate production.

The production of diacetyl by dairy PABwas observed during growth in milk by Leeet al. [44]. Only 5 out of 24 strains producedthe dicarbonyl, but the pathway is not clear.Citrate is a likely substrate as diacetyl pro-duction followed its utilisation. Citrate util-isation and diacetyl production wereincreased at pH < 6.25. Addition of citrate tothe milk increased diacetyl production.Diacetyl was detected during fermentation ofpyruvate, citrate and glucose; it is thereforepossible that diacetyl production wouldaffect propionate, acetate and CO2' but littleinformation is available,

11. FATE OF CO2 DURINGPROPIONIC FERMENTA nON

Because of its critical role in eye forma-tion in Swiss-type cheese, it is worth con-sidering the factors which determine the netamount of CO2 produced during fermenta-tion of lactate and sugars. In the classicalpathway of propionic acid fermentation,1 mole of CO2 is produced per 3 moles oflactate or 1.5 moles of glucose. However,the net quantity of CO2 produced can bedepleted by fixation; CO2 can condense withphosphoenol pyruvate (figure 1), by thereaction catalysed by the carboxytranspho-sphorylase:

phosphoenolpyruvate + CO2 + PPi---7 oxaloacetate + Pi

The relevance of CO2 fixation dependson the growth conditions and is especially

Metabolism of propionibacteria

important when the quantity of CO2 is high[90, 91, 108, 110, Ill]. In acidic conditions,CO2 is less soluble in water and its avail-ability for fixation increases. Such a situationwould arise when sugars are used. CO2 fix-ation leads to production of equimolar con-centrations of succinate and reduces theamount of propionate produced:

3 lactate -7 (2-x) propionate + 1 acetate+ (I-x) CO2 + (x) succinate

On the other hand, involvement of theTCA cycle could lead to extra productionof CO2. Considering only this cycle, 7 molesof lactate would be metabolised into 6 molesof propionate and 3 moles of CO2, whilethe hypothetical joint use of the glyoxylateenzymes (from citrate to isocitrate, glyoxy-late and malate) with the TCA cycle, andfollowed by propionate production gives atheoretical 7 moles of lactate being con-verted to 5 moles of propionate, 2 moles ofCO2 and 1 mole of succinate.

Metabolism of amino acids, and espe-cially aspartate would yield extra CO2 pro-duction during lactate utilisation:

3 aspartate + 1 propionate -7 1 acetate+ 113CO2 + 3 succinate + 3 NH4 +

The overall quantity of CO2 producedcan be explained by the relative involve-ment of each of these processes.

12. CONCLUSION

The metabolism of propionibacteria iscomplex and involves several cycles notvery weIl understood yet. Although the tran-scarboxylase cycle is the major pathwayleading to the production of propionate,acetate and CO2, the TCA cycle can be usedas another route for the dissimilation of lac-tate and sugars. The presence of alternativepathways can partly explain the discrepan-cies from the theoretical molar ratios of endproducts during lactate and sugar metabolismwhich are found. Hence, CO2 can be pro-duced from lactate, according to the Fitz

37

equation, and from amino acids, but it canalso be utilised by fixation with phospho-enolpyruvate, and lead to succinate pro-duction. Moreover, the physico-chemicalcomposition of the medium, utilisation ofextra substrates such as amino acids or shortchain peptides and production of polysac-charides can affect the propionic acid fer-mentation by reducing the amount of pro-pionate produced compared to acetate. Theability to use those alternative pathways forlactate and sugar dissimilation could explainthe great variability observed between dif-ferent Propionibacterium strains.

REFERENCES

[1] Allen S.H.G., Kellenneyer R.W., Stjernholm R.L.,Wood H.G., Purification and properties ofenzymes involved in the propionic acid fer-mentation, J. Bacteriol. 87 (1964) 171-187.

[2] Asmundson RV., Pritchard G.G., Multiple fonnsof cytochrome b in the electron transport sys-tem of Propionibacterium shermanii, Arch.Microbiol. 136 (1983) 285-290.

[3] Babuchowski A., Hammond E.G., Glatz B.A.,Survey of propionibacteria for ability to pro-duce propionic and acetic acids, J. Food Pro-tect. 56 (1993) 493-296.

[4] Baer A., Ryba L, Influence of casein proteolysisby starter bacteria, rennet and plasmin on thegrowth of propionibacteria in Swiss-type cheese,Lait 75 (1995) 391-400.

[5] Berger M., Wood H.G., Purification of the sub-units of transèarboxylase by affinity chro-matographyon avidin-Sepharose, J. Biol. Chem.250 (1975) 927-933.

[6] Bodie E.A., Anderson T.M., Goodman N.,Schwartz R.D., Propionic acid fermentation ofultra-high-temperature sterilized whey usingmono- and mixed-cultures, Appl. Microbiol.Biotechnol. 25 (1987) 434-437.

[7] Bonartseva G.A., Krainova a.A., Vorob'eva L.L,Aerobic metabolism of propionibacteria, Micro-biology 42 (1973) 520-525.

[8] Boyaval P., Corre C; Continuous fermentationof sweet whey permeate for propionic acid pro-duction in a CSTR with UF recycle, Biotech-nol. Lett. 9 (1987) 801-806.

[9] Boyaval P., Corre C; Production of propionicacid, Lait 75 (1995) 453-461.

[10] Clark J.E., Beegen H., Wood H.G., Isolation ofintact chains of polyphosphate from Propioni-bacterium shermanii grown on glucose and lac-tate, J. Bacteriol. 168 (1986) 1212-1219.

38

[1 1] Calafat A.M., Taoka S., Puckell lM., Semerad C,Yan H., Luo L., Chen H., Banerjee R., Mazilli L.G.,Structural and electronic similarity but func-tional difference in methylmalonyl-CoA mutasebetween coenzyme B 12 and the analog2' ,5' -dideozyadenosylcobalamin, Biochem-istry 34 (1995) 14125-14130.

[12J Castberg H.B., Morris H.A., The pyruvate oxi-dizing system of Propionibacteriumfreudenre-ichii subsp. shermanii, Milchwissenschaft 9(1978) 541-544.

[13] Chaix P., Fromageot C; Intluence de l'oxydede carbone sur la respiration de Propionibac-terium pentosaceum, Bull. Soc. Chim. Biol. 24(1942) 1128-1131.

[ 14] Champagne CP., Baillargeon-Côté C, GouletJ.,Whey fermentation by immobilized cells of Pro-pionibacterium shermanii, J. Appl. Bacteriol.66 (1989) 175-184.

115] Colomban A., Roger L., Boyaval P., Produc-tion of propionic acid from whey permeate bysequential fermentation, ultrafiltration and cellrecycling, Biotechnol. Bioeng. 42 (1993)1091-1098.

[16] Crow V.L., Utilization of lactate isomers byPropionibacteriumfreudenreichii subsp. sher-manii: regulatory role for intracellular pyruvate,Appl. Environ. Microbiol. 52 (1986) 352-358.

[17] Crow V.L., Metabolism of aspartate by Propi-onibacterium freudenreichii subsp. shermanii:effect on lactate fermentation, Appl. Environ.Microbiol. 52 (1986) 359-365.

[18J Crow V.L., Properties of alanine dehydroge-nase and aspartase from Propionibacteriumfreudenreichii subsp. shermanii, Appl. Environ.Microbiol. 53 (1987) 1885-1892.

[19] Crow V.L., Citrate cycle intermediates in themetabolism of aspartate and lactate by Propi-onibacterium freudenreichii subsp. shermanii,Appl. Environ. Microbiol. 53 ( 1987) 2600-2602.

[20] Crow V.L., Polysaccharide production by Pro-pionibacteria during lactose fermentation, Appl.Environ. Microbiol. 54 (1988) 1892: 1895.

[21] Crow V.L., Turner K,W., The effect of succi-nate production on other fermentation productsin Swiss-type cheese, N. Z. J. Dairy Sci. Tech-nol. 21 (1986) 217-227.

[22) Crow V.L., Martley F.G., Delacroix A., Isolationand properties of aspartase-deficient variants ofPropionibacterium freudenreichii subsp. sher-manii and their use in the manufacture of Swiss-type cheese, N. Z. J. Dairy Sci. Technol. 23(1988) 75:85.

[23] De Vries W., van Wijck-Kapteijn W.M.C.,Stouthamer A.H., Influence of oxygen ongrowth, cytochrome synthesis and fermentationpattern in propionic acid bacteria, J. Gen, Micro-biol. 71 (1972) 515-524.

P. Piveteau

[24] De Vries W., van Wijck-Kapteijn W.M.C.,Stouthamer A.H., Generation of ATP duringcytochrome-linked anaerobie electron transportin propionic acid bacteria, J. Gen. Microbiol.76 (1973) 31-41.

[25J Elfolk N., Studies on aspartase 1. Quantitativeseparation of aspartase from bacterial cells andits partial purification, Acta Chem. Scand. 7(1953) 824-830.

[26] Foschino R., Galli A., Ponticelli G.,Volonterio G.,Propionic bacteria activity in different culturecondition, Ann. Microbiol. Enzymo!. 38 (1988)207-222.

[27] Francalanci F., Davis N.K., Fuller J.Q., Murfit D.,Leadlay P.F., The subunit structure of methyl-malonyl-CoA mutase from P. shermanii,Biochem. J. 236 (1986) 489-494.

[28] Freudenreich E., Orla Jensen O., Über die imEmmentalerkase stattfindende propionsaüre-garung, Zbl. Bakteriol. II Abt. Band XVII (1906)34-52.

[29] Hartley J.c., Vedamuthu E.R., [3-galactosidaseof Propionibacrerium shermanii, Appl, Micro-biol. 29 (1975) 74-80.

[30] Hettinga D.H., Reinbold G.W., The PropionicAcid Bacteria - a review. 1 Growth, J. MilkFood Technol. 35 (1972) 295-301.

[31] Hettinga D.H., Reinbold G.W., The PropionicAcid Bacteria - a review. II Metabolism, J. MilkFood Technol. 35 (1972) 358-372.

[32J Hettinga D.H., Reinbold G.W. The PropionicAcid Bacteria - a review. 1II Miscellaneousmetabolic activities, J. Milk Food Technol. 35(1972) 436-447.

[33] Hettinga D.H., Reinbold G.W., Split defect ofSwiss cheese. Il. Effect of low temperatures onthe metabolic activity of Propionibacterium,J. Milk Food Technol. 38 (1975) 31-35.

[34] Hsu S.T., Yang S.T., Propionic acid fermentationof lactose by Propionibacterium acidipropi-on ici: effects of pH, Biotechnol. Bioeng. 38(1991) 571-578.

[35] Ibragimova S.I., Shul'Govskaya E.M., Growthof P. shermanii culture under various aerationconditions, Mikrobiologiya 48 (1979) 668-671.

[36] Ibragimova S.I., Shul'Govskaya E.M., Rabot-nova I.L., Intluence of aeration on the retardationof growth of P. shermanii at various pH, Mikro-biologiya 48 (1979) 845-848.

[37] Janicka 1., Malizewska M., Pedziwilk F., Vti-lization of lactose and production of corrinoidsin selected strains of propionic acid bacteria incheese whey and casein media, Acta Microbiol.Polinica 25 (1976) 205-210.

[38] Kellermeyer R.W., Allen S.H.G., Stjernholm R.,Wood H.G., Methylmalonyl isomerase. IV.Purification and properties of the enzyme frompropionibacteria, J. Biol. Chem. 239 (1964)2562-2569.

Metabolism of propionibacteria 39

[39] Kumar a.K., Babler C.R., Merrifield R.B., [54] Marsh N., McKie N., Davis N.K., Leadlay P.P.,Wood H.a., The amino acid sequences of the Cloning and structural characterisation of thebiotinyl subunit essential for the association of genes coding for adenosylcobalamin-dependenttranscarboxylase, J. Biol. Chem. 257 (1982) methylmalonyl-CoA mutase from Propioni-13828-13834. bacterium shermanii, Biochem. J. 260 (1989)

[40] Langsrud T., Reinbold a.w., Flavor develop- 345-352.

ment and microbiology of Swiss cheese - [55] Marsh E.N., Harding S.E., Methylmalonyl-CoAa review.1. Milk quality and treatments, J. Milk mutase from P. shermanii: characterisation ofFood Technol. 36 (1973) 487-490. the cobalarnin-inhibited form and subunit-eofac-

[41] Langsrud T., Reinbold a.w., Flavor develop- tors interactions studied by analytical ultracen-

ment and microbiology of Swiss chee se - trifugation, Biochem. J. 290 (1993) 551-555.

a review. II. Starters, manufacturing processes [56] Milner Y., Wood H.a., Isolation of a pyrophos-and procedures, J. Milk Food Technol. 36 (1973) phoryl form ofpyruvate posphate dikinase from531-542. propionibacteria, Proc. Nat. Acad. Sei, USA 69

[42] Langsrud T., Reinbold a.w., Flavor develop- (1972) 2463-2468.

ment and microbiology of Swiss cheese - [57] Milner Y., Wood H.a., Steady state anda review. III. Ripening and flavor production, exchange kinetics of pyruvate phosphate diki-J. Milk Food Technol. 36 (1973) 593-{j09. nase from Propionibacterium shermanii, J. Biol.

[43] Langsrud T., Sorhaug T., Vegarud a.E., Pro- Chem. 251 (1976) 7920-7928.

tein degradation and amino acid metabolism by [58] MurtifL., Samois D., Mutagenesis affecting thepropionibacteria, Lait 75 (1995) 325-330. carboxylterminus of the biotinyl subunit of tran-

[44] Lee l.H., Vedamuthu E.R., Washam c.J., scarboxylase, 1. Biol. Chem. 262 (1987)

Diacetyl production by Propionibacterium sher- 11813-11816.

manii in milk cultures, Cano J. Microbiol. 16 [59] O'Brien W.E., Bowien S., Wood H.a., Isola-(1970) 1231-1242. tion and characterization of a pyrophosphate-

[45] Lee l.H., Fredrickson A.a., Tsuchiya H.M., dependent phosphofructokinase from Propi-

Diauxic growth of Propionibacterium shermanii, onibacterium shermanii, J. Biol. Chem. 250

Appl. Microbiol. 28 (1974) 831-835. (1975) 8690-8695.

[46] Lewis V.P., Yang S.T., Propionic acid fermen- [60] O'Brien W.E., Wood H.a., Carboxytransphos-

tation by Propionibacterium acidipropionici: phorylase. VII. Ligand mediated interaction of

effect of growth substrate, Appl. Microbiol. subunits as possible control mechanism in pro-pionibacteria, J. Biol. Chem. 249 (1974)Biotechnol. 37 (1992) 437-442. 4917-4925.

[47] Lewis V.P., Yang S.T., Continuous propionic [61] Ozadali F., G1atz B.A., G1atz C.E., Fed-batchacid fermentation by immobilized Propioni- fermentation with and without on line extrac-bacterium acidipropionici in a novel packed- tion for propionic and acetic acid production bybed bioreactor, Biotechnol. Bioeng. 40 (1992) P. acidipropionici, Appl. Microbiol. Biotech-465-474. nol. 44 (1996) 710-716.

[48] Liu J.A.P., Moon N.1., Commensalistic inter- [62] Pépin C.A., Wood H.a., Polyphosphate glu-action between Lactobacillus acidophilus and cokinase from P. shermanii: kinetics and demon-Propionibacterium shermanii, Appl. Environ. stration that the mechanism involves both pro-Microbiol. 44 (1982) 715-722. cessive and non processive type reactions,

[49] Lochmüller H., Wood H.a., Davis J.1., Phos- J. Biol. Chem. 261 (1986) 4476-4480:phoenolpyruvate carboxytransphosphorylase. [63] Phillips N.F.B., HornP.J., Wood H.a., TheII. Crystallisation and properties, J. Biol. Chem. polyphosphate- and ATP-dependent glucoki-241 (1966) 5678-5691. nase from Propionibacterium shermanii: both

[50] Namba A., Nukada R., Nagai S., Inhibition by activities are catalysed by the same protein,acetic and propionic acids of the growth of Pro- Arch. Biochem. Biophys. 300 (1993) 309-319.pionibacterium shermanii, J. Ferment. Technol. [64] Piveteau P., Condon S., Cogan T.M., Interac-61 (1983) 551-556. tions between lactic and propionic acid bacteria,

[51] Mantere-Alhonen S., Propionibacteria used as Lait 75 (1995) 331-343.probiotics. A review, Lait 75 (1995) 447-452. [65] Piveteau P., Stimulation of propionic acid bac-

[52] Marcoux V., Beaulieu Y., Champagne c.P., teria by lactic acid bacteria, Ph.D. thesis,Goulet J., Production of Propionibacterium National University ofireland, 1997.freudenreichii subsp. shermanii in whey-based [66] Pritchard a.a., Wimpenny J.W.T., Morris H.A.,media, J. Ferment. Bioeng. 74 (1992) 95-99. Lewis M.W.A., Hughes D.E., Effects of oxy-

[53] Marsh E.N., Harding S.E., Leadlay P.P., Methyl- gen on Propionibacterium shermanii grown inmalonyl-CoA mutase studies by analytical ultra- continuous culture, J. Gen. Microbiol. 102centrifugation, Biochem. J. 260 (1989) 353-358. (1977) 223-233.

40 P. Piveteau

[67J Pritchard G.G., Asmundson R.V., Aerobic elec- [81] Schwartz AC., Mertens B., Voss K.W., Hahn H.,tron transport in Propionibacterium shermanii. Inhibition of acetate and propionate formationEffects of cyan ide, Arch. Microbiol. 126 ( 1980) upon aeration of resting cells of the anaerobie167-173. Propionibacterium shermanii: evidence of

[68] Quelen L., Dupuis C; Boyaval P., Proline-spe- the Pasteur reaction, Z. Mikrobiol. 16 (1976)cific activities of Propionibacterium freuden- 123-131.reichii subsp. shermanii, J. Dairy Sei. 62 (1995) [82] Sebastiani H., Tschager E., Succinatbildung661-666. durch propionsaurebakterien-eine ursache der

[69] Quesada-Chanto A, Schroeder A.G., Schmid- nachgârung von Emmentaler? DMZ Lebensm.Meyer A.C., Lopez J.A., Silveira M.M., Jonas R., Milchwirtsch. 114 (1993) 76-80.Organic acid production by P. shermanii: effect [83] Shaw R.H., Sherman J.M., The production ofof pH, temperature and vitamin-nitrogen source, volatile fatty acids and carbon dioxide by pro-J. Biosciences 52 (1997) 193-196. pionic acid bacteria with special reference to

[70] Racine M., Dumont 1., Champagne c.P., their action in cheese, J. Dairy Sei. 6 (1923)Morin A., Production and characterisation of 303-309.the polysaccharide from P. acidipropionici on [84] Shenoy B.C., Xie Y., Park V.L., Kumar G.K.,whey-based media, J. Appl. Bacteriol. 71 (1991) Beegen H., Wood H.G., Samois D., The impor-233-238. tance of methionine residues for the catalysis

[71] Rehberger J.L., Glatz B.A., Response of cul- of the biotin enzyme transcarboxylase, J. Biol.tures of Propionibacterium to acid and low pH: Chem. 263 (1992) 6461-6464.

tolerance and inhibition, J. Food Protec!. 61 [85] Shenoy B.C., Kumar G.K., SamoIs D., Dissec-(1998) 211-216. tion of the biotinyl subunit oftranscarboxylase

[72] Richoux R., Kerjean J.R., Caractérisation tech- into regions essential for activity and assembly,nologique de souches pures de bactéries pro- J. Biol. Chem. 268 (1993) 2232-2238.pioniques : test en minifabrication de fromages [86] Shenoy B.C., SamoIs D., Kumar G.K., The con-à pâte cuite, Lait 75 (1995) 45-59. served methionines of the 1.3S biotinyl subunit

[73] Robinson N.A., Wood H.G., Polyphosphate of transcarboxylase: effect of mutations on con-kinase from P. shermanii: demonstration that formation and activity, Arch. Biochem. Bio-the synthesis and utilisation of poly P is by a phys. 304 (1993) 359-366.processed mechanism, J. Biol. Chem. 261 (1986) [87] Shenoy RC., Xie Y., Sha D., SamoIs D., Iden-4481-4485. tification and characterisation of a factor which

[74] Rolin D.B., Girard F., de Certaines J.D., Boyaval P., is essential for assembly of transcarboxylase,I3C-NMR of lactate metabolism in P. freuden- Biochemistry 32 (1993) 10750-10756.reichii subsp. shermanii, Appl. Microbiol. [88] Sherman J.M., The cause of eyes and charac-Biotechnol. 44 (1995) 210-217. teristic f1avor in Emmental or Swiss cheese,

[75] Rosner B., Schink B., Propionate acts as car- J. Bacteriol. 6 (1921) 379-392.boxylic group acceptor in aspartate fermenta- [89] Sherman J.M., Shaw R.H., The propionic acidtion by Propionibacteriumfreudenreichii, Arch. fermentation of lactose, J. Biol. Chem. 56 (1923)Microbiol. 155 (1990) 46-51. 695-700.

[76] SamoIs D., Thornton c.G., Murtif V.L., [90] Siu P.M.L., Wood H.G., Stjernholm R.L., Fix-Kumar G.K., Haase F.C., Wood H.G., Evolu- ation of CO2 by phosphoenolpyruvic carboxy-tionary conservation among biotin enzymes, transphosphorylase, J. Biol. Chem. 236 (1961)J. Biol. Chem. 263 (1988) 6461-6464. PC21-22.

[771 Schütz M., Sievers M., Klemm K., Metaboli- [91] Siu P.M.L., Wood H.G., Phosphoenolpyruviccally controlled continuous fermentation of lac- carboxytransphosphorylase, a CO2 fixationtose to propionate, acetate and carbon dioxide, enzyme from propionic acid bacteria, J. Biol.Milchwissenschaft 42 (1987) 71-74. Chem. 237 (1962) 3044-3051.

[78] Schwartz A.C., Anaerobiosis and oxygen con- [92] Smart J.B., Pritchard G.G., Control ofpyruvatesumption of sorne strains of Propionibacterium kinase activity during glycolysis and glucogen-and a modified method for comparing the oxy- esis in Propionibacterium shermanii, J. Gen.gen sensitivity ofvarious anaerobes, Z. Mikro- Microbiol. 128 (1982) 167-176.biol. 13 (1973) 681-691. . [93] Sone N., The redox reactions in propionic acid

[79] Schwatz A.C., Krause AE., Partial purification fermentation. 1. Occurrence and nature of anand properties of NADH dehydrogenase from electron tranfer system in PropionibacteriumPropionibacterium shermanii, Z. Milkrobiol. arabinosum, J. Biochem. 71 (1972) 931-940.15 (1975) 99-110. [94] Sone N., The redox reactions in propionic acid

[80] Schwartz A.C., Sporkenbach J., The electron fermentation. III. Enzymatic properties of NAD-transport system of the anaerobie Propionibac- independent glycerol-phosphate dehydrogenaseterium shermanii, Arch. Microbiol. 102 (1975) from Propionibacterium arabinosum, J. Biochem.261-273. 74 (1973) 297-305.

Metabolism of propionibacteria

[95] Sone N., Isolation ofa novel menaquinone withparti y hydrogenated side chain from Propioni-bacterium arabinosum, J. Biochem. 76 (1974)133-136.

[96] Sone N., The redox reactions in propionic acidfermentation. IV. Participation of menaquinonein the electron transfer system in Propionibac-terium arabinosum, J. Biochem. 76 (1974)137-145.

[97] Steffen C; Eberhard P., Bosset J.O., Rüegg M.,3 Swiss-type varieties, in: Fox P.F. (Ed.),Cheese: Chemistry, Physics and Microbiology,vol 2, 2nd ed., Elsevier Applied Science, Lon-don, 1993.

[98J Stjernholm R., Wood H.G., Glycerol dissimi-lation and occurrence of a symmetrical three-carbon intermediate in propionic acid fermen-tation, J. Biol. Chem. 235 (1960) 2757-2761.

[99] Stjernholm R., Flanders F., Metabolism of D-ribose-f-I-C and 14C-Iabeled D-gluconate in anenzyme system of the genus Propionibacterium,J. Bacteriol, 84 (1962) 563-568.

[100] Thornton c.G., Kumar G.K., Haase F.C., PhillipsN.F.B., Woo S.B., Park V.M., Magner W.J.,Shenoy B.e., Wood H.G., Samois D., Primarystructure of the monomer of the 12S subunit oftranscarboxylase as deduced from DNA andcharacterisation of the products expressed inEscherich ia coli, J. Bacteriol. 175 (1993)5301-5308.

[101] Van Gent-Ruijters M.L.W., De Vries W.,Stouthamer A.W., Influence of nitrate on fer-mentation pattern, molar growth yields and syn-thesis of cytochrorne b in P. pentosaceum,J. Gen. Microbiol. 88 (1975) 36-48.

[102] Vandernark P.J., Fukui G.M., An enzymaticstudy of the utilisation of gluconic acid byP. pentosaceum, 1. Bacteriol. 72 (1956) 610-614.

[103] Virtanen A.I., Tatanen J., Die enzymatischespaltung und synthese der asparaginsaüre,Biochem. Z. 250 (1932) 193-211.

[104] Volk W.A., The enzymatic formation ofD ara-binose 5-phosphate from L-arabinose and ATPby P. pentosaceum, J. Biol. Chem. 234 (1959)1931-1936.

[105] Wolfe K., Michenfelder M., Kônig A., Hull W.E.,Rétey J., On the rnechanism of action of methyl-malonyl-CoA mutase, change of the steric courseon isotope substitution, Eur. J. Biochem. 156(1986) 545-554.

[106] Woo S.B., Shenoy RC., Wood H.G., Magner WJ.,Kumar G.K. Beegen H., Samois D., Effect ofdeletion from the carboxylterminus of the 12S

41

subunit on activity of transcarboxylase, J. Biol.Chem. 268 (1993) 16413-16419.

[107] Wood H.G., The anatomy of transcarboxylaseand the role of its subunits, Crit. Rev. Biochem.7:2 (1979) 143-160.

[108] Wood H.G., Metabolic cycles in the fermenta-tion by propionic acid bacteria, Curr. TopicsCell. Regul. 18 (1981) 255-287.

[109] Wood H.G., Werkman CH; Mechanism of glu-cose dissimilation by the propionic acid bacte-ria, Biochem. J. 30 (1936) 618-623.

[110] Wood H.G., Werkman c.a., The utilization ofCOz by the propionic acid bacteria, Biochem.J. 32 (1938) 1262-1271.

[III] Wood H.G., Werkman C.H., The relationshipof bacterial utilization of COz to succinic acidformation, Biochem. J. 34 (1940) 129-138.

[112] Wood RG., Stjernholm R., Leaver F.W., Themetabolism of labeled glucose by propionic acidbacteria, J. Bacteriol. 70 (1955) 510-520.

[113] Wood H.G., ChiaoJ.P., Poto E.M., A new largeform of transcarboxylase with six outer subunitsand twelve biotinyl carboxyl carrier subunits,J. Biol. Chem. 252 (1977) 1490-1499.

[114] Wood H.G., O'Brien W.E., Michaels G., Prop-erties of carboxytransphosphorylase, pyruvatephosphate dikinase, pyrophosphate-phospho-fructokinase and pyrophosphate-acetate kinaseand their roles in the metabolism of inorganicpyrophosphate, Adv. Enzymol. 45 (1977)85-155.

[115] Woskow S.A., Glatz B.A., Propionic acid pro-duction by a propionic acid tolerant strain ofP. acidipropionici in batch and semicontinuousfermentation, Appl. Environ. Microbiol. 57(1991) 2821-2828.

[116] Wrigley N.G., Chiao J.P., Wood H.G., Electronmicroscopy of the large forrn of transcarboxylasewith six attached subunits, J. Biol. Chem. 252(1977) 1500-1504.

[117] Yang S.T., Zhu H., Li Y., Continuous propi-onate production from whey permeate using anovel tïbrous bed bioreactor, Biotechnol. Bio-eng. 43 (1994) 1124-1130.

[1 18] Yang S.T., Huang Y., Hong G., A novel recyclebatch immobilized cell bioreactor for propionateproduction from whey lactose, Biotechnol. Bio-eng.45 (1995)379-386.

[119] Zagalak B., Rétey J., Studies on methylmalonyl-CoA mutase from P. shermanii, Eur. J. Biochem.44 (1974) 529-535.