role of inorganic carbon in lactic acid bacteria metabolism

HAL Id: hal-00895525https://hal.archives-ouvertes.fr/hal-00895525

Submitted on 1 Jan 2004

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.

Role of inorganic carbon in lactic acid bacteriametabolism

Florence Arsène-Ploetze, Françoise Bringel

To cite this version:Florence Arsène-Ploetze, Françoise Bringel. Role of inorganic carbon in lactic acid bacteriametabolism. Le Lait, INRA Editions, 2004, 84 (1-2), pp.49-59. �10.1051/lait:2003040�. �hal-00895525�

https://hal.archives-ouvertes.fr/hal-00895525

https://hal.archives-ouvertes.fr

49Lait 84 (2004) 49–59© INRA, EDP Sciences, 2004DOI: 10.1051/lait:2003040

Original article

Role of inorganic carbon in lactic acid bacteria metabolism

Florence ARSÈNE-PLOETZE*, Françoise BRINGEL

Laboratoire de dynamique, évolution et expression de génomes de microorganismes, Université Louis-Pasteur-CNRS, FRE2326, 28 rue Goethe, 67000 Strasbourg, France

Abstract – Capnophiles are bacteria stimulated by bicarbonate and CO2, the two major forms ofinorganic carbon (IC) in physiological neutral liquids. Capnophiles are often pathogenicheterotrophs found in IC-rich ecological niches such as human cavities. Like capnophiles, thegrowth of lactic acid bacteria (LAB) such as Lactobacillus plantarum and Enterococcus faecalis isstimulated by IC. CO2 or HC are substrates in carbamoyl phosphate (CP) synthesis and othercarboxylation reactions in amino acid and nucleotide biosynthesis. When media were supplementedwith nucleotides and all the amino acids, potassium bicarbonate still stimulated L. plantarumgrowth. This suggests that IC may be involved in other aspects of L. plantarum physiology besidesits implication as a substrate in carboxylation reactions. Carbonic anhydrase (CA) catalyses thehydration of CO2 into bicarbonate. Since inorganic carbon stimulated L. plantarum growth, wesearched for CA encoding genes in LAB genomes. CA can be classified into three classes accordingto their protein relatedness: α, β and γ . A class α CA was found in the L. plantarum, Leuconostocmesenteroides, Streptococcus thermophilus, Oenococcus oeni, Enterococcus faecalis andEnterococcus faecium. These enterococci harboured a second CA encoding gene belonging to theγ class. No CA encoding gene was found in the Lactococcus lactis genome. These observations arediscussed with regard to LAB evolution and ecological niches, which are often rich in IC.

Lactic acid bacteria / Lactobacillus plantarum / carbon dioxide / bicarbonate / carbonicanhydrase

Résumé – Effet du carbone inorganique sur le métabolisme des bactéries lactiques. Lecarbone inorganique est trouvé principalement sous deux formes dans la plupart des milieuxbiologiques, le CO2 (sous forme de gaz ou dissous) et le bicarbonate (HC ). Le carboneinorganique stimule ou inhibe la croissance d’un certain nombre d’organismes. Le terme decapnophiles désigne les bactéries dont la croissance est facilitée ou nécessite des concentrations deCO2 plus élevées que celle de l’air. Ces bactéries à Gram négatif ou positif, chimio-organotropheshétérotrophes, sont souvent trouvées dans la flore commensale ou pathogène de l’homme, etprésentent un métabolisme aérobie strict ou anaérobie facultatif. Comme les capnophiles, lesbactéries lactiques sont retrouvées dans divers environnements, souvent enrichis en carboneinorganique (tractus intestinal et vaginal, végétaux en décomposition ou fermentés). Le CO2 stimulela croissance de certaines bactéries lactiques (Lactobacillus, Enterococcus faecalis) et de bactériesrelativement proches phylogénétiquement (Streptococcus pneumoniae). Mais le rôle du carboneinorganique dans le métabolisme de ces bactéries a été très peu étudié. On considérait que l’effet duCO2 chez ces hétérotrophes s’expliquait par son rôle de substrat dans les réactions de carboxylation.En analysant en détail l’effet du CO2 et du bicarbonate (HC ) sur sa croissance, nous avons puproposer que les réactions de carboxylation seules n’expliquaient pas l’effet du carbone inorganiquesur la croissance de L. plantarum, bactérie lactique que nous avons définie comme capnophile.

* Corresponding author: [email protected]

O3–

O3–

O3–

50 F. Arsène-Ploetze, F. Bringel

1. INTRODUCTION

Inorganic carbon (IC) corresponds toCO2 as a gas (CO2(g)) or dissolved in theaqueous phase (CO2(aq)), and its hydratedor dissociated forms (HC , C andH2CO3). The concentration of inorganiccarbon and the ratio of the IC species varyin different environments. In the atmos-phere, CO2 concentration (expressed in %or ppm: part per million) is much lowertoday (360 ppm) than it was 3 billion yearsago when the first bacteria, low G+C %Gram-positive bacteria, appeared [14].The CO2(g) concentration in the soil ishigher (>0.1%, 1000 ppm) than in theatmosphere [10, 36]. In aqueous phases,CO2(g) can dissolve in CO2(aq) accordingto the following equilibrium:

CO2(g) + H2O CO2(aq).

The CO2(aq) concentration ([CO2]aqexpressed in mol·L–1) varies with CO2solubility as expressed by Henry’s law:

[CO2]aq = K ρCO2

where K depends on the medium and thetemperature, and ρCO2 is the partial pres-sure of CO2 in the gas phase (expressed inatm, 1 atm = 106 ppm). When the partialpressure of CO2 is kept constant, CO2hydration and dissociation in water is afunction of the pH and other factors such assalinity [35].

(1) (2)CO2(aq) + H2O H2CO3 HC + (3)H+ C + 2H+.

It is analytically difficult to distinguishbetween the species CO2(aq) and H2CO3.For this reason, the term CO2(aq) is oftenused to express the sum of CO2(aq) sensustricto and H2CO3. However, the equili-brium for reaction (1) favours CO2(aq)formation (ratio of [CO2]aq/[H2CO3]:1/1000). In most biological conditions, pHvalues are less than 8, so that the reaction(3) is inefficient, and the concentration ofcarbonate ions (C ) may be neglected.The sum of CO2 (aq) and HC concen-trations is considered to correspond to thetotal IC concentration in most biologicalsystems.

IC is used as a carbon source by autotrophs,as an electron acceptor for methanogens,and as a substrate by almost all organismsfor several carboxylation reactions such asanaplerotic reactions, amino acids or pyri-midine biosynthesis pathways. Lactic acidbacteria (LAB) are heterotrophic Gram-positive bacteria with a low G+C% content.Variable concentrations of inorganic car-bon are found in LAB ecological niches.Many fermented products contain highlevels of inorganic carbon due to CO2production by LAB or other microorga-nisms. When associated with CO2-producing

O3– O3

2–

O3–

O32–

O32–

O3–

Une stimulation de la croissance a été observée à la fois par enrichissement en CO2 dans la phasegazeuse ou par ajout de HC dans le milieu. Nous avons recherché la présence de l’anhydrasecarbonique (CA), une enzyme ubiquitaire chez de nombreux organismes, qui catalyse l’hydratationdu CO2 en HC . Trois classes, α, β et γ sont définies à partir de leur homologie de séquence. Nousavons trouvé en recherchant des séquences présentant des homologies significatives avec des CAconnues, des gènes codant pour une anhydrase carbonique de classe α chez L. plantarum,L. mesenteroides, Oenococcus oeni, Streptococcus thermophilus, E. faecalis et E. faecium, ces deuxderniers organismes possédant également une anhydrase carbonique de classe β. Aucune ORF neprésentant une homologie significative avec une anhydrase carbonique n’a été trouvée chezLactococcus lactis. Cette observation nous conduit à proposer que certaines bactéries lactiques ontévolué vers d’autres stratégies d’hydratation du carbone inorganique que celles impliquant la CA.

Bactérie lactique / Lactobacillus plantarum / gaz carbonique / bicarbonate / anhydrasecarbonique

O3–

O3–

Growth stimulation by inorganic carbon 51

eukaryotes, as in the vicinity of intestinalepithelium, LAB are also exposed to rich ICamounts. For example, in the human epithe-lial lumen, the IC concentration is estimatedto be 20.6 µmol·L–1 [7, 18]. LAB associatedwith plants also encounter daily variableexposure to IC concentrations.

Carbonic anhydrase (CA), which cata-lyses the reversible hydration of CO2 tobicarbonate (reactions (1) and (2)), is a ubi-quitous, essential enzyme. Based onsequence similarities, three classes of CAhave been found in a lot of bacteria andArchaebacteria. Some bacteria possessmore than one CA encoding gene [39, 40],which may or may not belong to the sameclass. This enzyme seems to have appearedat the beginning of life on earth [40]. Thus,the conversion of CO2 to bicarbonate isimportant for prokaryote and eukaryotesurvival and may be implied in their cen-tral metabolism.

High or low levels of IC concentrationsinduce different responses in heterotrophicbacteria. High atmospheric CO2 concen-tration inhibits microbial growth, althoughthe extent of the inhibition depends on themicroorganism and on growth conditionssuch as the culture medium (for a reviewsee [6]). CO2 alters membrane properties,probably modifies the intracellular pH, andinterferes with carboxylation reactions. As aconsequence, CO2 is effectively used infoodstuff preservation against bacterialspoilage. On the other hand, a low concen-tration of IC is necessary for the growth ofseveral organisms [28, 32–34, 45–47]. CO2alone stimulated LAB growth and in parti-cular L. plantarum growth [13]. However,to our knowledge, the effect of non-lethalinorganic carbon concentration on LABmetabolism has never been analysed indetail. Recently, Amanatidou et al. [2] stu-died the effect of CO2 and O2 enrichmentor depletion in the gas phase on the growthof several microorganisms including threeLAB: L. lactis, Leuconostoc mesenteroidesand L. plantarum. CO2 was found to stimu-late growth of L. plantarum but not that of

L. lactis or L. mesenteroides. In this study,the effect of CO2 was not studied alone butin combination with O2.

For autotrophs, the role of CO2 in meta-bolism has been extensively studied (for areview [37]). For heterotrophs, in 1941Krebs described that CO2 was required inEscherichia coli for carboxylation reac-tions such as amino acid biosynthesispathways and anaplerotic reactions [22]. InE. faecalis and L. mesenteroides, someauxotrophies were suppressed whenstrains were incubated in a CO2-enrichedatmosphere [24, 25]. In L. plantarum,auxotrophies for phenylalanine, arginine,tyrosine and pyrimidines were observed inseveral strains incubated under normal airbut not when incubated in a CO2-enrichedair [4, 25, 29]. CO2 may be required inthese organisms for carboxylation reac-tions like in E. coli. However, in E. coli,the positive effect of CO2 could not be res-tricted to the carboxylation reactions [21].In this enterobacteria, cyanase catalysesthe transformation of cyanate with bicar-bonate to give NH4 and CO2, which dif-fuse out of the cell [8]. A CA deletedmutant would not grow in the presence ofcyanate because of a depletion of cellularbicarbonate. The authors attempted torelieve this growth inhibition by addingmetabolites whose synthesis are known todepend on carboxylation reactions, but thegrowth inhibition could not be completelyovercome. In L. plantarum, it was pro-posed that IC enrichment was required forefficient carbamoyl-phosphate synthetase(CPS) activity [29]. However, as demons-trated in E. coli, CO2 could be importantfor LAB metabolism besides the sensustricto carboxylation reactions.

This study focused on how CO2 stimula-tes growth in LAB. We analysed the effect ofboth increasing ρCO2 in the gas phase andaddition of bicarbonate to the medium, onthe growth of L. plantarum in severalmedia. Our data suggested that in bothcases an effect on L. plantarum metabolismwas observed. Genome databases were


analysed for the presence of carbonicanhydrase encoding genes in L. plantarumand other LAB whose genomes have beensequenced. Such genes were found in allLAB except Lc. lactis. The consequencesof these observations will be discussed.

2. MATERIALS AND METHODS

2.1. Bacterial strains and growth conditions

The CCM 1904 and NCIMB 8826 L.plantarum strains were grown on rich defi-ned medium, DLA [5], or minimal medium[26], complemented or not with arginineand uracil at a final concentration of50 µg·mL–1. Precultures were incubatedovernight at 30 °C without shaking andused to inoculate 20 mL of the samemedium in a Klett flask (total volume130 mL) to obtain an initial OD600 nm of0.05. The flasks were closed with gas-tightcorks to prevent CO2 loss during shaking.A ρCO2 of 4% (40 µatm) was obtained byinjecting through the cork 4.8 mL of pureCO2 gas with a sterile syringe. Final potas-sium bicarbonate concentrations between1 and 15 g·L–1 were obtained by adding afilter-sterilised solution of 100 g·L–1 ofpotassium bicarbonate after autoclavingand before inoculation. To adjust the pHvalue to 6.5, HCL 6N was used. Cell cultu-res were incubated at 28 °C with shakingand growth was measured using a Klett-Summerson apparatus.

2.2. Sequence analysis

Amino acid sequences deduced fromputative ORFs were compared with those ofCA in the PUBmed nucleotide databankusing either BLASTp or tBLASTn [1]. Asignificant identity was estimated if theputative protein harboured putative conser-ved domains found in one of the three classesof known CA. Sequence data for Streptococ-cus thermophilus were obtained from theUCL Life Sciences Institute (ISV) (Belgium)

website at http://www.biol.ucl.ac.be/gene/genome/. The putative ribosome bindingsite and promoter were searched for usingPatscan (http://ir2lcb.cnrs-mrs.fr/cgi-bin/patscan.cgi).

3. RESULTS

The effect of IC on L. plantarum growthwas tested. The aim of our study was to testif IC-mediated stimulation was restrictedto its role as a substrate in biosynthetic car-boxylation reactions of amino acids and ofpyrimidines. We analysed the effect ofincreasing ρCO2 and adding KHCO3 tothe medium on L. plantarum growth in richor minimal media, with different aminoacid or nucleotide compositions. First, theeffect of inorganic carbon as carbondioxide gas was tested: 0.035% (as foundin normal air), 4% and 10%. Secondly,bicarbonate ions were added directly to theliquid media, by adding different potas-sium bicarbonate concentrations. In eachcase, the pH was fixed at 6.5 before inocu-lation. L. plantarum growth was tested inliquid media in hermetically-closed flasksto prevent loss of inorganic carbon. Theeffect of inorganic carbon concentrationson two L. plantarum strains was tested.Strain NCIMB 8826, the L. plantarumwhose genome has been sequenced, requi-res CO2-enriched air to grow on definedmedia plates when arginine and pyrimidi-nes are absent [4]. The other strain, CCM1904 (equivalent to ATCC 8014), is proto-trophic for arginine and pyrimidines andgrew in normal air.

3.1. Inorganic carbon stimulated L. plantarum growth in the absence of arginine and pyrimidines

We measured the effect of a CO2-enrichedatmosphere on the growth of NCIMB 8826and CCM 1904 in DLA liquid medium.This medium contains all the amino acidsexcept arginine and pyrimidines. As expec-ted from its phenotype on agar-plates, in theabsence of arginine and pyrimidines (ura-cil), NCIMB 8826 grew only when the


atmosphere was enriched with CO2. Agrowth rate of 0.41 h–1 was obtained whenair was supplemented with 4% CO2. In theDLA media, growth of the prototroph CCM1904 occurred at all the tested gas phaseCO2 concentrations. With a ρCO2 of0.035% as found in normal air, a specificgrowth rate of 0.33 h–1 was obtained. With4% or 10% ρCO2, the growth rate was0.41 h–1. Increasing the CO2 tension signi-ficantly stimulated the growth of both L.plantarum strains. The effect of variableKHCO3 concentration (from 0 to 15 g·L–1)on strain CCM 1904 growth, in the samemedium (in the absence of arginine and ura-cil) was measured to determine the growthrate, the lag phase and the maximal yield(Tab. I). In defined rich media, DLA (Fig. 1and Tab. I), and in minimal media, MM(data not shown), without arginine and pyri-midines, CCM 1904 growth depended onbicarbonate concentration. The best yield atthe stationary phase and the optimal growthrate were observed in DLA with bicarbo-nate concentration of 2 g·L–1 (Fig. 1). Inminimal medium, optimal growth was at2 g·L–1 KHCO3 (data not shown). To test ifthe bicarbonate-mediated growth stimula-tion in L. plantarum was pH-dependent, weanalysed the effect of bicarbonate on L.plantarum growth rates at different pHbetween 5 and 7 in DLA. The tested pH hadno effect on bicarbonate growth stimulation(data not shown).

3.2. Addition of potassium bicarbo-nate stimulated L. plantarum growth in media supplemented with arginine and uracil

IC stimulation of L. plantarum growthmay be explained by the requirement ofCO2 as a substrate for the carboxylationreaction of the CPS (EC 6.3.5.5) involvedin both arginine and pyrimidine biosynthe-sis pathways [29]. To test this hypothesis,we analysed the prototroph CCM 1904’sgrowth with respect to IC supply, in thepresence of arginine (A) and uracil (U) indifferent media: minimal medium, MM, or

rich defined medium, DLA. Bicarbonateconcentrations had no effect on the lagphase (data not shown). Growth stimula-tion depended on bicarbonate concentra-tion: an optimal effect was obtained at 1 to2 g·L–1 of bicarbonate in minimal medium,complemented with A (data not shown),and at 8 g·L–1 in DLA complemented withA and U (Fig. 1 and Tab. I). When the cul-tures were performed in the open air andnot in hermetically-closed flasks, growthwas not stimulated because bicarbonatewas transformed into CO2 which subse-quently diffused into the air. Thus, even inDLA supplemented with A and U, and inthe presence of all the 20 common aminoacids and pyrimidines, bicarbonate stimu-lated L. plantarum growth. The Ymaxobtained in the stationary phase was redu-ced at 8 g·L–1 of KHCO3 as compared withculture with less than 8 g·L–1 of KHCO3.This was observed in DLA, independent of

Figure 1. Effect of the KHCO3 concentrationon the growth rate (µ) in L. plantarum CCM1904. The growth rate (µ) obtained in theabsence of KHCO3 was used as a reference toexpress the relative µ (µ obtained in thepresence of KHCO3 divided by µ obtained inthe absence of KHCO3). Experiments wereperformed in defined rich medium DLA. Opencircles: in the presence of arginine andpyrimidines; closed squares: in the absence ofarginine and pyrimidines. The data representthe mean value of at least 4 independentexperiments.


A and U’s presence. Additional experi-ments such as viability testing would benecessary to confirm this observation.

3.3. The carbonic anhydrase enzyme in LAB

Carbonic anhydrase catalyses the rever-sible hydration of CO2 to bicarbonate [39].Since growth was stimulated in response toIC, the presence of such an enzyme islikely in L. plantarum. Indeed, a proteincalled Lp2736, homologous to carbonicanhydrase and encoded by the cah genewas found in the L. plantarum WCFS1genome [15]. In front of the cah gene, aconsensus for a good putative ribosomebinding site and a promoter were found(data not shown). This gene encoded a 211amino acid-long protein belonging to the αclass of known carbonic anhydrase andshared more than 47% similarity with theCA of O. oeni, Vibrio parahaemolyticus,Xanthomonas campestris, and Bacillushalodurans (for the most homologous pro-teins). α class CA are not often found inbacteria. Therefore, to test if α class CAwere commonly found in LAB, we sear-ched for putative CA proteins with signifi-cant identity with known CA in LABwhose genomes have been sequenced:

E. faecalis, E. faecium, L. lactis subsp. lactis,Lactobacillus gasseri, L. mesenteroidessubsp. mesenteroides ATCC 8293, Oeno-coccus oeni MCW and S. thermophilus andin Bifidobacteria (Bifidobacterium longumDJO10A and Bifidobacterium longum bio-var longum). ORFs with deduced aminoacid sequences showing identity with theL. plantarum cah-encoded protein werefound in L. mesenteroides subsp. mesente-roides (Lmes0773), O. oeni (Ooen0510),E. faecalis (EF1711), E. faecium (Efae2400)and S. thermophilus (C79-115). TheseORFs shared 29, 32, 34, 25 and 31% iden-tity, respectively with the L. plantarumCA. The genetic contexts of the putativeCA encoding gene were compared in thedifferent LAB (Fig. 2). We found that CAgenes were linked to genes with differentfunctions, such as regulators, transporters,catabolic or anabolic enzymes and pro-teases. In L. plantarum, the cah gene wasfound in the vicinity of a cell surfacehydrolase encoding gene (Lp2737) (Fig. 2).Only in the two Enterococcus strains,which are closely related phylogenetically,did we detect synteny. In these cases, thecah gene was linked to a gene encoding aLysR family regulator (Fig. 2). In all theother LAB tested, no gene clusters in thevicinity of the CA-related genes were con-served even when they belonged to thesame class of CA.

In some bacteria, more than one puta-tive encoding CA gene was detected, andtwo out of three known classes of CA werefound. Therefore, we analysed if CA enco-ding genes belonging to other classes werefound in LAB. Using the E. coli paaYgene, two genes of the γ class CA werefound in E. faecalis (EF2918) and E. fae-cium (Efae1549). Using the E. coli cynTgene or the Oceanobacillus iheyensiOB1097 locus, two genes of the β class CAwere found in Bifidobacterium longumNCC2705 (icfA) and DJO10A. Finally, inthe Lc. lactis subsp. lactis completegenome and in the L. gasseri partially-sequenced genome, we found no genespresenting significant homology with the

Table I. Effect of bicarbonate on L. plantarumgrowth in DLA medium complemented or notwith arginine and uracil.

Medium [KHCO3]

(g·L–1)

µ

(h-1)

Ymax *

DLA

DLA+AU

028028

0.34 ± 0.050.44 ± 0.060.40 ± 0.050.46 ± 0.040.49 ± 0.030.56 ± 0.04

100107 ± 1

93 ± 4100100 ± 1

91 ± 5

Specific growth rate (µ) is the mean value of atleast four independent experiments. *: maxi-mum population density (Ymax) is the Klett unitpercentage obtained at stationnary phase as com-pared with the 0 g·L–1 of [KHCO3] condition.


Fig

ure

2. G

enom

ic o

rgan

isat

ion

of c

arbo

nic

anhy

dras

e en

codi

ng g

enes

in

6 la

ctic

aci

d ba

cter

ia. T

he c

arbo

nic

anhy

dras

e (C

A)

is r

epre

sent

ed b

y a

grey

rec

tang

le. T

he a

rrow

s in

dica

te th

e tr

ansc

ript

ion

orie

ntat

ion

of th

e di

ffer

ent g

enes

sch

emat

ised

by

rect

angl

es. T

he C

A c

lass

wer

e de

term

ined

by

sequ

ence

hom

olog

y (s

ee M

ater

ials

and

met

hods

).


three classes of CA encoding genes from avariety of origins (L. plantarum class α cahgene; E. faecium class γ CA encodinggene; E. coli cynT, cah, or caiE; the O.iheyensis OB1097 locus; or the B. longumβ class CA encoding genes). Indeed, theabsence of CA-related genes in the L. gas-seri genome should be confirmed when itsgenome is completely sequenced.

4. DISCUSSION

The aim of our work was to study theeffect of non-lethal IC concentration on L.plantarum metabolism in order to unders-tand better how IC stimulates growth inLAB. Our results showed that either gasphase CO2 enrichment or addition of bicar-bonate increased the growth rate of L.plantarum in rich defined medium or inminimal medium, whether or not arginineor pyrimidines were present: 20% in thepresence of arginine and pyrimidines and30% in the absence of arginine and pyrimi-dines. Heterotroph bacteria were designa-ted as capnophilic bacteria on the basis oftheir CO2 growth stimulation on solidmedium. Such bacteria (for example, Cap-nocytophaga and Actinobacillus) arepathogens found in CO2-rich human cavi-ties [3, 12, 38, 41]. On the other hand, hete-rotrophs showing a reduced lag phase inthe presence of IC but no change in thegrowth rate, were not defined as capnophi-les (case of E. coli) [32–34]. On the basisof the observed IC growth stimulation, wepropose to define L. plantarum as a capno-philic bacterium.

4.1. How does inorganic carbon stimulate L. plantarum growth?

CO2 is required for carboxylation reac-tions since mutants of CPS encoding genesin L. plantarum required CO2-enrichedatmosphere to grow [30]. High-CO2-requi-ring natural auxotrophs are prevalent innatural isolates of L. plantarum withoutarginine and pyrimidine supplements [4].This suggests that inorganic carbon isrequired as a substrate of CPS. However,

in our experiments, growth stimulationwas still observed in a defined richmedium containing all the amino acids andnucleotides. This suggests that IC plays abroader role than simply an intermediate inamino acid or nucleotide biosynthesispathways. In LAB, decarboxylation reac-tions are important for pH homeostasis andproton motive force energy storage [19,20]. Decarboxylation of histidine to hista-mine in Lactobacillus buchneri [27] or glu-tamate to γ -aminobutyrate in Lactobacillussp. strain E1 [11] led to proton extrusion. Itis still not clear if both species HC andCO2(aq) (comprising H2CO3) are equallyrequired for growth stimulation. However,growth stimulation was observed at pH 5,when CO2(aq) is the most abundant spe-cies, and at pH 7, when HC is the mostabundant species. Based on these observa-tions, carbonic anhydrase may be involvedin controlling the amounts of both speciesat various pH, since this enzyme convertsCO2 to HC .

4.2. Carbonic anhydrase is not ubiquitous in LAB and may play different physiological roles

Analysis of LAB complete genomesrevealed the presence of putative carbonicanhydrase (CA) encoding genes in fiveLAB genera: Leuconostoc, Oenococcus,Enterococcus, Streptococcus, and Lacto-bacillus, but not in Lactococcus. CA iswidespread in organisms, which suggeststhat this enzyme plays a fundamental rolein prokaryotes and in eucaryotes. Very lowamino acid identity between different CAsuggests that this enzyme is ancient [39,40]. Low amino acid identity was alsoobserved between CA from LAB (25 to35% of identity). A single copy of a CAencoding gene of the α class was found inL. plantarum, L. mesenteroides, S. thermo-philus and O. oeni and of the β class inBifidobacteria. Two copies of CA enco-ding genes, one homologous to the α classand the other to the γ class, were found inE. faecalis and in E. faecium. The α class

O3–

O3–

O3–


CA enzymes are widespread in mammalsand unicellular green algae [39, 40]. Thus,even though only a few prokaryotes harbourthe α class of CA, this class of CA wasalways found in LAB harbouring at leastone CA. α class CA of Neisseria or Anae-baena contain a peptide signal, suggestingthat this protein is secreted or associatedwith the inner membrane [39], perhaps forthe transport of inorganic carbon. A pep-tide signal was found in CA from S. ther-mophilus, but not from other LAB. Thedifferent classes of CA in LAB may playdifferent roles, as suggested in othermicroorganisms by Smith et al. [39],although the physiological role of the CAenzyme is not clear in microorganisms. CAcould be involved in CO2 or HC trans-port and modulate enzymatic reactions byvariation of its substrate or product con-centrations (CO2 or HC ). In E. coli,cyanate degradation would require CAactivity to prevent HC depletion duringthis process [8, 9]. In some pathogens suchas Salmonella typhymurium, this enzymewould be important for bacterial survivalin the host [39]. The β and γ classes of CAare involved in the carbon concentratingmechanism (ccm) in photosynthetic bacte-ria. Genes homologous to the carboxy-some components were found in two LABstudied: E. faecalis and E. faecium (datanot shown). In S. thyphimurium [17], thecarboxysome-component encoding genesare probably involved in ethanolaminecatabolism, which suggests a similar rolein enterococci. The genetic context ofthese genes may help to determine the roleof these CA in LAB. However, we did notobserve any conservation in gene organisa-tion around CA encoding genes in LAB.This may reflect the fact that CA is anancient enzyme [40].

Unlike the other LAB tested, in Lc. lac-tis no CA encoding gene was found. Inother bacteria such as Borrelia burgdoerferi,Chlamydia trachomatis or Mycoplasmapneumoniae, no ORF with identity withknown classes of CA was found [39]. Thiswould suggest that CA is not an ubiquitous

enzyme. Smith et al. [39, 40] suggestedthat CA could play a role in removing CO2produced by decarboxylation, by conver-sion to HC to drive the decarboxylationreaction forward. Since these decarboxyla-tion reactions are important in Lc. lactis[19, 20], we were surprised not to find anyCA encoding gene in this bacterium. It ispossible that in Lc. lactis the sequence ofthe CA gene has diverged from the otherknown CA so that we did not detect thecorresponding gene. CA may also not beneeded for Lc. lactis survival. In fact, whileLc. lactis is mostly found in milk products,the other LAB studied are found in morediverse environments where CA expres-sion may be essential. In Lc. lactis a differentstrategy for CO2/bicarbonate interconver-sion may have evolved. Enzymatic dosagewould be necessary to confirm the absenceof CA in Lc. lactis.

4.3. Is there CO2-regulated gene expression in LAB?

The growth stimulation by IC in com-plete media suggested that CO2-regulationwould exist in L. plantarum. CO2-media-ted gene regulation was observed in hete-rotrophs [42] such as Pseudoalteromonas([43], unknown function); the cad operonin E. coli [44]; and virulence genes inpathogenic bacteria such as Bacillus anthra-cis, Actinobacillus actinomycetemcomitans,Staphylococcus aureus, Streptococcus pyo-genes, and enteropathogenic E. coli [16,42]. In Cyanobacteria, LysR-type regula-tors are known to regulate gene expressionin response to CO2, including genes enco-ding the carboxysome components [23,31]. We found that in E. faecalis, two LysRregulators are clustered with a γ class CAencoding gene, perhaps in an operon. ThisCA may be CO2-regulated, as observed incyanobacteria. A homologue of this gene(Lp1857) was found in L. plantarum. Acombination of proteomic and geneticexperiments would be required to identifyCO2-regulated genes in LAB.

O3–

O3–

O3–

O3–


Acknowledgements: Sequence data for Strep-tococcus thermophilus were obtained from theUCL Life Sciences Institute (ISV) website athttp://www.biol.ucl.ac.be/gene/genome/.Sequencing of Streptococcus thermophilus wassupported by the Walloon Region (BIOVALgrant #9813866).

REFERENCES

[1] Altschul S.F., Madden T.L., Schaffer A.A.,Zhang J., Zhang Z., Miller W., Lipman D.J.,Gapped BLAST and PSI-BLAST: a newgeneration of protein database search pro-grams, Nucl. Acids Res. 25 (1997) 3389–3402.

[2] Amanatidou A., Smid E.J., Gorris L.G.,Effect of elevated oxygen and carbondioxide on the surface growth of vegetable-associated micro-organisms, J. Appl. Micro-biol. 86 (1999) 429–438.

[3] Bolmstrom A., Karlsson A., Influence ofCO2 incubation on quinolone activityagainst Streptococcus pneumoniae and Hae-mophilus influenzae, Diagn. Microbiol.Infect. Dis. 42 (2002) 65–69.

[4] Bringel F., Hubert J.C., Extent of geneticlesions of the arginine and pyrimidine bio-synthetic pathways in Lactobacillus planta-rum, L. paraplantarum, L. pentosus andL. casei: prevalence of CO2 dependent auxo-trophs and characterization of deficient arggenes in L. plantarum, Appl. Environ.Microbiol. 69 (2003) 2674–2683.

[5] Bringel F., Frey L., Boivin S., Hubert J.C.,Arginine biosynthesis and regulation in Lac-tobacillus plantarum: the carA gene and theargCJBDF cluster are divergently transcri-bed, J. Bacteriol. 179 (1997) 2697–2706.

[6] Dixon N.M., Kell D.B., The inhibition byCO2 of the growth and metabolism of micro-organisms, J. Appl. Bacteriol. 67 (1989)109–136.

[7] Groeneveld A.B., Kolkman J.J., Splanchnictonometry: a review of physiology, metho-dology, and clinical applications, J. Crit.Care 9 (1994) 198–210.

[8] Guilloton M.B., Korte J.J., Lamblin A.F.,Fuchs J.A., Anderson P.M., Carbonic anhy-drase in Escherichia coli. A product of thecyn operon, J. Biol. Chem. 267 (1992) 3731–3734.

[9] Guilloton M.B., Lamblin A.F., Kozliak E.I.,Gerami-Nejad M., Tu C., Silverman D.,Anderson P.M., Fuchs J.A., A physiologicalrole for cyanate-induced carbonic anhydrasein Escherichia coli, J. Bacteriol. 175 (1993)1443–1451.

[10] Hamada Y., Tanaka T., Dynamics of carbondioxide in soil profiles based on long-termfield observation, Hydrol. Process 15 (2001)1829–1845.

[11] Higuchi T., Hayashi H., Abe K., Exchangeof glutamate and gamma-aminobutyrate in aLactobacillus strain, J. Bacteriol. 179 (1997)3362–3364.

[12] Hofstad T., The classification and identifica-tion of the anaerobic Gram-positive cocci,Scand. J. Infect. Dis. Suppl. 46 (1985) 14–17.

[13] Kandler O., Weiss N., Regular, nonsporingGram-positive rods, in: Sneath P.H.A., MairN.S., Sharpe M.E., Holt J.G. (Eds.), Bergey’smanual of systematic bacteriology, Vol. 2,Williams and Wilkins, Baltimore, MD,USA, 1986, pp. 1208–1260.

[14] Kasting J.F., Earth’s early atmosphere,Science 259 (1993) 920–926.

[15] Kleerebezem M., Boekhorst J., van KranenburgR., Molenaar D., Kuipers O.P., Leer R.,Tarchini R., Peters S.A., Sandbrink H.M.,Fiers M.W., Stiekema W., Lankhorst R.M.,Bron P.A., Hoffer S.M., Groot M.N.,Kerkhoven R., de Vries M., Ursing B., de VosW.M., Siezen R.J., Complete genomesequence of Lactobacillus plantarum WCFS1,Proc. Natl. Acad. Sci. USA 100 (2003) 1990–1995.

[16] Koehler T.M., Bacillus anthracis geneticsand virulence gene regulation, Curr. Top.Microbiol. Immunol. 271 (2002) 143–164.

[17] Kofoid E., Rappleye C., Stojiljkovic I., RothJ., The 17-gene ethanolamine (eut) operon ofSalmonella typhimurium encodes fivehomologues of carboxysome shell proteins,J. Bacteriol. 181 (1999) 5317–5329.

[18] Kolkman J.J., Otte J.A., Groeneveld A.B.,Gastrointestinal luminal PCO2 tonometry:an update on physiology, methodology andclinical applications, Brit. J. Anaesth. 84(2000) 74–86.

[19] Konings W.N., The cell membrane and thestruggle for life of lactic acid bacteria, Anto-nie Van Leeuwenhoek 82 (2002) 3–27.

[20] Konings W.N., Lolkema J.S., Bolhuis H.,van Veen H.W., Poolman B., Driessen A.J.,The role of transport processes in survival oflactic acid bacteria. Energy transduction andmultidrug resistance, Antonie Van Leeuwen-hoek 71 (1997) 117–128.

[21] Kozliak E.I., Fuchs J.A., Guilloton M.B.,Anderson P.M., Role of bicarbonate/CO2 inthe inhibition of Escherichia coli growth bycyanate, J. Bacteriol. 177 (1995) 3213–3219.

[22] Krebs H.A., Carbon dioxide assimilation inheterotrophic organisms, Nature 147 (1941)560–563.


[23] Kusian B., Bowien B., Organization andregulation of cbb CO2 assimilation genes inautotrophic bacteria, FEMS Microbiol. Rev.21 (1997) 135–155.

[24] Lascelles J., Cross M.J., Woods D.D., Thefolic acid and serine nutrition of Leuconos-toc mesenteroides P60 (Streptococcus equi-nus P60), J. Gen. Microbiol. 10 (1954) 267–284.

[25] Lyman C.M., Moseley O., Wood S., ButlerB., Hale F., Some chemical factors whichinfluence the amino acid requirements of thelactic acid bacteria, J. Biol. Chem. 167(1947) 177–187.

[26] Masson A., Kammerer B., Hubert J.C.,Selection and biochemical studies of pyrimi-dine-requiring mutants of Lactobacillusplantarum, J. Appl. Bacteriol. 77 (1994) 88–95.

[27] Molenaar D., Bosscher J.S., ten Brink B.,Driessen A.J., Konings W.N., Generation ofa proton motive force by histidine decar-boxylation and electrogenic histidine/hista-mine antiport in Lactobacillus buchneri, J.Bacteriol. 175 (1993) 2864–2870.

[28] Neidhardt F.C., Bloch P.L., Smith D.F., Cul-ture medium for enterobacteria, J. Bacteriol.119 (1974) 736–747.

[29] Nicoloff H., Hubert J.C., Bringel F., In Lac-tobacillus plantarum, carbamoyl phosphateis synthesized by two carbamoyl-phosphatesynthetases (CPS): carbon dioxide differen-tiates the arginine-repressed from the pyri-midine-regulated CPS, J. Bacteriol. 182(2000) 3416–3422.

[30] Nicoloff H., Hubert J.C., Bringel F., Carba-moyl-phosphate synthetase (CPS) in lacticacid bacteria and other Gram-positive bacte-ria, Lait 81 (2001) 151–159.

[31] Omata T., Gohta S., Takahashi Y., HaranoY., Maeda S., Involvement of a CbbR homo-log in low CO2-induced activation of thebicarbonate transporter operon in cyanobac-teria, J. Bacteriol. 183 (2001) 1891–1898.

[32] Repaske R., Clayton M.A., Control ofEscherichia coli growth by CO2, J. Bacte-riol. 135 (1978) 1162–1164.

[33] Repaske R., Ambrose C.A., Repaske A.C.,De Lacy M.L., Bicarbonate requirement forelimination of the lag period of Hydrogeno-monas eutropha, J. Bacteriol. 107 (1971)712–717.

[34] Repaske R., Repaske A.C., Mayer R.D.,Carbon dioxide control of lag period and

growth of Streptococcus sanguis, J. Bacte-riol. 117 (1974) 652–659.

[35] Schumpe A., Quicker G., Deckwer W.D.,Gas solubilities in microbial culture media,Adv. Biochem. Eng. 24 (1982) 1–38.

[36] Sheppard S.K., Lloyd D., Direct mass spec-trometric measurement of gases in soilmonoliths, J. Microbiol. Methods 50 (2002)175–188.

[37] Shively J.M., van Keulen G., Meijer W.G.,Something from almost nothing: carbondioxide fixation in chemoautotrophs, Annu.Rev. Microbiol. 52 (1998) 191–230.

[38] Slots J., Salient biochemical characters ofActinobacillus actinomycetemcomitans, Arch.Microbiol. 131 (1982) 60–67.

[39] Smith K.S., Ferry J.G., Prokaryotic carbonicanhydrases, FEMS Microbiol. Rev. 24(2000) 335–366.

[40] Smith K.S., Jakubzick C., Whittam T.S.,Ferry J.G., Carbonic anhydrase is an ancientenzyme widespread in prokaryotes, Proc.Natl. Acad. Sci. USA 96 (1999) 15184–15189.

[41] Socransky S.S., Holt S.C., Leadbetter E.R.,Tanner A.C., Savitt E., Hammond B.F.,Capnocytophaga: new genus of Gram-negative gliding bacteria. III. Physiologicalcharacterization, Arch. Microbiol. 122(1979) 29–33.

[42] Stretton S., Goodman A.E., Carbon dioxideas a regulator of gene expression in microor-ganisms, Antonie Van Leeuwenhoek 73(1998) 79–85.

[43] Stretton S., Marshall K.C., Dawes I.W.,Goodman A.E., Characterisation of carbondioxide-inducible genes of the marine bacte-rium, Pseudomonas sp. S91, FEMS Micro-biol. Lett. 140 (1996) 37–42.

[44] Takayama M., Ohyama T., Igarashi K.,Kobayashi H., Escherichia coli cad operonfunctions as a supplier of carbon dioxide,Mol. Microbiol. 11 (1994) 913–918.

[45] Talley R.S., Baugh C.L., Effects of bicarbo-nate on growth of Neisseria gonorrhoeae:replacement of gaseous CO2 atmosphere,Appl. Microbiol. 29 (1975) 469–471.

[46] Valley G., Rettger L.F., The influence ofcarbon dioxide on bacteria, J. Bacteriol. 14(1927) 101–137.

[47] Walker H.H., Carbon dioxide as a factoraffecting the lag in bacterial growth, Science76 (1932) 602–604.

role of inorganic carbon in lactic acid bacteria metabolism

Documents