Res. Microbiol. 152 (2001) 653661 2001 ditions scientiques et mdicales Elsevier SAS.

All rights reserved S0923-2508(01)01244-X/FLA

The arcABC gene cluster encoding the arginine deiminase pathway of Oenococcus oeni, and arginine induction of a CRP-like gene
Thierry Tonon, Jean-Paul Bourdineaud, Aline Lonvaud-Funel
Facult dOenologie, Unit associe INRA, Universit Victor Sgalen, Bordeaux II, Laboratoire de Biotechnologie et de Microbiologie Applique, 351, cours de la Libration, 33405 Talence cedex, France Received 20 September 2000; accepted 23 February 2001

Abstract Oenococcus oeni, the main species which induces malolactic fermentation in wine, uses arginine via the arginine deiminase (ADI) pathway. Using degenerated primers, two specic probes, one for ornithine transcarbamoylase (OTC) and the other for carbamate kinase (CK), were synthesized. These made it possible to clone and sequence a cluster containing genes encoding ADI (arcA), OTC (arcB) and CK (arcC). In addition, sequence analysis upstream of the arcA gene revealed the presence of an open reading frame (orf229) whose 3 -end was only 101 bp-distant from the start codon of the arcA gene and showed similarity with members of the FNR (regulation for fumarate and nitrate reduction) and CRP (cAMP receptor protein) family of transcriptional regulators. Moreover, a putative binding site for such regulators lies in the promoter region of the arcA gene. Induction of the arc cluster by arginine was studied rst at the enzymatic level. The activities of the three enzymes strongly increased when cells were grown in the presence of the amino acid. In addition, the inuence of arginine on gene transcription was monitored by RT-PCR (reverse transcriptase-polymerase chain reaction). Expression of the three arc genes, and particularly that of arcA, was positively affected by arginine supplementation and thus conrmed the enzymatic results. Moreover, transcription of the putative CRP-like gene orf229 was also stimulated by arginine. These data suggest that the protein encoded by orf229 could be a CRP-like regulator involved in the metabolism of O. oeni. 2001 ditions scientiques et mdicales Elsevier SAS wine / lactic acid bacteria / arginine catabolism / gene regulation

1. Introduction The degradation of arginine by the arginine deiminase (ADI) pathway is widely distributed among prokaryotic organisms [2]. The arginine deiminase pathway consists of the following three enzymes: arginine deiminase (ADI; EC 3.5.3.6), ornithine transcarbamoylase (OTC; EC 2.1.3.3), and carbamate kinase (CK; EC 2.7.2.2). A fourth gene encoding a transport protein catalyzing an electroneutral exchange between arginine and ornithine has also been identied. Arginine is converted into ornithine and carbamoyl phosphate, which serves to generate ATP using ADP. The degradation of 1 mol of arginine results in the formation of 1 mol of ATP and 2 mol of ammonia. More recently, a gene cluster encoding

Correspondence and reprints.

E-mail address: [email protected] (A. Lonvaud-Funel).

the enzymes of the ADI pathway of the homofermentative lactic acid bacteria (LAB) Lactobacillus sakei has been cloned and sequenced [19]. Oenococcus oeni, a facultative anaerobic bacterium, occurs naturally in wine and is the LAB commonly used to induce malolactic fermentation in wine. Malolactic fermentation is a process that is often encouraged after the end of alcoholic fermentation by yeast since this secondary fermentation includes deacidication of acidic wines, addition of avor complexity and increased microbiological stability [11]. In wine LAB, the metabolism of arginine via the ADI pathway has been established by studying enzymatic activities and physiological aspects [10]. It has been shown that many O. oeni strains are able to catabolize this amino acid [10]. In addition, arginine represents a potential source of energy that can be coupled to cellular growth. We have shown that arginine increases growth and viability of the O. oeni strains able to breakdown this amino acid [18]. Arginine is one of the major amino acids found in grape

654

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

juice and wine, and the major oenological implication of this metabolism is the formation of ethyl carbamate precursors. Ethyl carbamate, also referred to as urethane, is a known animal carcinogen found in fermented foods and beverages including wine [14]. Ethyl carbamate is formed in the wine from the non-enzymatic and spontaneous reaction occurring between alcohol and excreted citrulline. This reaction is favoured upon wine storage in warm cellars. Therefore, we cloned and sequenced the genes encoding the ADI (arcA), the OTC (arcB) and the CK (arcC) of O. oeni, since no genetic datum is available on arc genes in wine LAB, and because knowledge of such information could lead to design specic probes allowing a PCR-based discrimination of arginine catabolism among wine LAB. 2. Materials and methods
2.1. Strains, plasmids and growth conditions

measuring this activity in a CK assay mixture lacking carbamoyl phosphate. CK activity was calculated by subtracting the control values (without carbamoyl phosphate) from the value obtained with complete reaction mixture. Myokinase activity was weak but present in crude extract of O. oeni. Specic activity was 0.44 U mg1 for strain IOEB 8406. The protein concentration was determined using the BCA protein assay (Pierce).
2.3. Determination of specic probes for the O. oeni arc cluster

The O. oeni strains were grown at 25 C, without shaking, in a modied MRS medium adjusted to pH 4.5 (per liter): yeast extract, 4 g; beef extract, 8 g; bactopeptone, 10 g; glucose, 5 g; KH2 PO4 , 2 g; MgSO4 7H2 O, 0.2 g; MnSO4 H2 O, 0.1 g; Tween 80, 1 mL. For measurement of the enzymatic activities, this medium was supplemented with arginine to the concentration of 5 g L1 . For total RNA isolation, induction was performed by adding 5 g L1 arginine to a culture (in MRS medium pH 5, 10 g L1 glucose, 10 g L1 fructose, 10 g L1 malic acid) previously grown at an absorbance at 600 nm (A600 ) of 0.40.5. The culture was then followed for 16 h. Escherichia coli strain XL1 Blue (Stratagene) was used for cloning procedures. The plasmid pBluescriptIIKS+ (Stratagene) was used as a general vector for cloning and sequencing.
2.2. Measurement of the enzymatic activity

To obtain probes relevant to the OTC and CK genes of O. oeni, degenerated primers were deduced from an alignment of protein sequences from several organisms. PCR was performed on ATCC 23279 genomic DNA with Taq polymerase using 35 cycles (30 s at 95 C, 1 min at 60 C, and 2 min at 72 C), preceded by 10 min at 95 C and followed by 10 min at 72 C. The genomic DNA from O. oeni ATCC 23279 was extracted as described [6]. The set of primers OTC5 (5 -YTWYTWTTYGARAARAMNWSNACNMGNACNMGN-3 ) and OTC3 (5 -YTCWCCCATNSWNACCCANACRTCNGTRTA-3 ) produced an amplicon of 564 bp. The two other degenerated oligonucleotides CK5 (5 -CAYGGNAAYGGNCCNCARGTNGGN-3 ) and CK3 (5 -NGCRAARTCYTTRTCDATNACNGC-3 ) gave a 499-bp fragment. These amplied products were cloned in pKS+ linearized by EcoRV, and sequenced. Similarities between these fragments and the L. sakei arcB and arcC genes [19] then conrmed that we had amplied the internal portion of the homologous genes from O. oeni.
2.4. Southern blot analysis

Cells were harvested when A600 reached 0.91. The activities of ADI pathway enzymes were measured as described [10]. Very little citrulline is synthesized nonenzymatically under the standard assay conditions for ADI and OTC. Interference by myokinase (adenylate kinase) activity was estimated by

O. oeni ATCC 23279 genomic DNA was transferred onto a Hybond-N+ membrane (Amersham) under vacuum. Hybridizations (overnight at 65 C) were performed with probes labeled with digoxigeninlabelled-11-dUTP using the DIG-DNA labeling and detection kit (Roche). The detection was carried out by chemiluminescence using an anti-digoxigenin antibody and CDP-starTM (Roche) as recommended by the manufacturer.
2.5. Construction of genomic mini-libraries

10 g of genomic DNA were subjected to HindIII + EcoRI, and BspXI restrictions. For the double re-

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

655

striction, the resulting 23 kb fragments were ligated to doubly restricted pKS+ . For the single restriction, the DNA fragments between 3 to 4 kb were ligated to pKS+ cleaved by BspXI. XL1 blue E. coli strain was transformed with the ligation reaction by electroporation in a PulserTM Controller (Biorad). After blue-white selection, the recombinant plasmids were extracted by an alkaline lysis method, and the inserted fragments analyzed by agarose gel electrophoresis. Then the DNA sequences were determined by the dideoxy chain termination reaction.
2.6. RNA isolation and RT-PCR analysis

Table I. Specic activities of the ADI pathway enzymes in cellfree extracts of O. oeni.

Strain Enzyme IOEB 8406 ADIb OTCb CKc

Specic activities +arg arg +arg/arga 0.14 0.04 0.01 0.01 14 46 5 0.25 0.07 184 22 3 1 0.3 22

a Ratio of specic activities in the presence and absence of

arginine. b Specic activities are expressed in mol citrulline min1 (mg protein)1 . c Specic activities are expressed in mol ATP min1 (mg protein)1 .

3. Results RNA preparation was carried out using the StrataPrep Total RNA Miniprep kit (Stratagene) which includes a DNase (RNase-free) treatment. Bacterial cells were disrupted with glass beads (0.1 mm diameter) by shaking on a vortex. The quality of the RNA samples was checked on a 1% (w/v) agarose gel, and the concentration was determined by measuring the absorbance at 260 nm. Synthesis of cDNA was carried out with the ProSTARTM First-Strand RTPCR Kit (Stratagene). Reverse transcriptase reaction used random primers of the kit with 1015 g of total RNA, in a nal volume of 50 L. Three L of the rst strand cDNA synthesis reaction was used as template for PCR amplication. Thirty-ve cycles were carried out, each consisting of 30 s of denaturation at 95 C, 30 s of annealing at 48 C (for the 23S rRNA gene) or 60 C (for the arcA, arcB and arcC genes), and 30 s of enzymatic primer extension at 72 C. PCR fragments were visualized on a 1% (w/v) agarose gel. Negative controls included PCR with either water instead of RT mix or 3 L of crude RNA. None of the negative controls resulted in DNA amplication. Genomic DNA was used as a positive control. Five primer pairs were designed to amplify fragments of arcA, arcB, arcC and orf229 genes. The last primer set corresponds to the 23S rRNA gene [13] which was used as a reference since 23S rRNA is a housekeeping gene and therefore cannot be induced by arginine. The RT-PCRs were transferred onto a Hybond-N+ membrane under vacuum. Hybridizations (overnight at 65 C) were performed with digoxigenin-labelled probes prepared from the relevant cloned arc genes.
3.1. Specic activities of ADI pathway enzymes

A screening of the O. oeni strains of the collection of the Faculty of Oenology of Bordeaux revealed that among the numerous strains which could use this amino acid, the IOEB 8406 proved to be very efcient since it could consume up to 60% of the added arginine [17]. Then, this strain was cultured in MRS broth containing 5 g L1 glucose, which was supplemented with 5 g L1 arginine to test the possible induction of the ADI pathway by this amino acid. Cells were harvested at the end of the exponential growth phase, when arginine was not completely degraded. Table I shows the specic activities of the three enzymes involved in the ADI pathway of strain O. oeni IOEB 8406. Without arginine, ADI and CK activity were hardly detectable. When this amino acid was added to the medium, the activities were much higher, since ADI and CK became detectable and OTC was increased 184-fold.
3.2. Cloning of the O. oeni arc cluster

By using degenerated primers, two specic probes of the OTC and CK of O. oeni were synthesized. These two probes were used in Southern blot experiments. A membrane carrying the restricted genomic DNA from strain O. oeni ATCC 23279 was hybridized either with the OTC or the CK probe (gure 1A). A 2214 bp Hind III-EcoRI fragment was revealed by both probes and was cloned. The insert within the resulting plasmid pTT1 contains at the 5 end a truncated orf which showed similarities with the OTC of L. sakei, and a complete ORF encoding

656

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

Figure 2. Sequence comparison between O. oeni ADI protein (the product of the arcA gene) and L. sakei and P. aeruginosa ADI proteins. Identical amino acids are indicated by asterisks and similar by dots. Highly conserved residues found in other ADI sequences appear in bold characters. Figure 1. (A) Southern hybridization with either the OTC or the CK probe of the O. oeni ATCC 23279 genomic DNA digested by BspXI (lane 1), EcoRI (lane 2), HindIII (lane 3), BspXI + HindIII (lane 4), HindIII + EcoRI (lane 5), and BspXI + EcoRI (lane 6). Numbers to the left are sizes of the marker DNA fragments in kilobases. (B) Structure and restriction map of the 5624 bp BspXIEcoRI fragment, and sequence of the promoter region upstream arcA. Restriction sites for EcoRI (E), HindIII (H), and BspXI (B) are shown. In arcB gene, the restriction site for HindIII is localized at nucleotide 134, and restriction sites for BspXI at nucleotide 337 and 692 respectively. The positions of orf86, orf119, orf229, arcA, arcB and arcC are indicated by boxes. The sequence spanning the promoter region upstream arcA is shown in detail (positions 1900 to 2041 in GenBank accession no. AF124851). Bold underlined letters indicate the two putative halves of the CRP binding site matching the consensus. Bold letters highlight the putative 10 promoter region. The ribosomal binding site is indicated by underlined italics. The translational start site of arcA is shown in bold underlined italics.

insert was identical with the 5 -end sequence of the pTT1 insert. The large 5624 bp BspXI-EcoRI fragment contains a truncated ORF at the BspXI end and ve complete ORFs, all transcribed in the same direction (gure 1B).
3.3. DNA sequence

a 313-aa sequence similar to the CK of L. sakei. In addition, a 3618-bp BspXI fragment revealed by the OTC probe was ligated in plasmid pKS+ to yield plasmid pTT2. The sequence of the 3 -end of the pTT2

The GenBank accession number of the reported sequence is AF124851. Its analysis revealed that the last three complete ORFs of the 5624-bp fragment were homologous to the L. sakei arcA, arcB and arcC genes [19]. These were therefore named the O. oeni arcA, arcB and arcC genes and encode respectively for ADI (415 aa, 47.1 kDa), OTC (351 aa, 39.8 kDa) and CK (313 aa, 34 kDa). The length of the intergenic regions are 101 bp between orf229 and arcA, only 16 bp between arcA and arcB, and 109 bp between arcB and arcC. Potent ribosome binding sites are found upstream arcA and arcC genes, but not arcB. Putative 35 and 10 promoter regions are found

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

657

Figure 3. Sequence comparison between O. oeni OTC protein (the product of the arcB gene) and L. sakei and P. aeruginosa OTC proteins. Identical amino acids are indicated by asterisks and similar by dots. Highly conserved residues found in other OTC sequences appear in bold characters.

Figure 4. Sequence comparison between O. oeni CK protein (the product of the arcC gene) and L. sakei and P. aeruginosa CK proteins. Identical amino acids are indicated by asterisks and similar by dots.

upstream arcA and arcC genes, but not arcB. By using probes specic to arcA and arcC genes, we could amplify from retro-transcribed RNAs a 1300-bp DNA fragment covering the end of arcA and the beginning of arcC genes, and encompassing the arcB gene (gure 6C). This suggests that the arc cluster is organized in an operon structure with a large transcript containing the three arc genes. Comparison with enzymes isolated from other bacteria was performed. The highest identities values were obtained with sequences of Gram-positive bacteria, mainly with enzymes from the other LAB L. sakei [19]. Indeed, identity values of 60, 71, and 64 % were found for ADI, OTC, and CK respectively, between O. oeni and L. sakei corresponding sequences (gures 2, 3 and 4). At the 3 end of the HindIII-EcoRI fragment, no evidence was found for the presence of a homolog of the arcD gene. However, two complete and one incomplete orf were found upstream arcA. The incomplete orf86 (gure 1B) was found at the 5 end of the 5624-bp fragment. For the orf119, no similarity could be found by screening of databases. Interestingly, analysis of the deduced amino acid sequence of orf229 revealed a weak homology with sequences of some related transcription regulators of the CRP (cAMP receptor protein, also named CAP for catabolite gene activator protein) and FNR (fumarate and nitrate reduction regulator) families. Members of the CRP-FNR family control a variety of physiological functions in many bacteria. This deduced amino acid sequence was aligned with the protein sequence of CRP of E. coli [1] and Flp (FNR like protein) of Lactobacillus casei [9] (gure 5). Similarities between these proteins are weak (17% identical residues between ORF229p and Ec-CRP, 18% identical residues between ORF229p and LcFLP), but extend over the entire length of the proteins. Another sequence alignment indicated that ORF229p has sequence identities of 17% (47% similarity) with FlpA and 19% (42% similarity) with FlpB of Lactococcus lactis [8] (not shown). However, the similarity is higher in the C-terminal region that contains the helix-turn-helix motif. This DNA binding domain features the highly conserved Arg200, Glu-201 and Arg-205 residues, the numbering of amino acid refers to the ORF229p sequence, which are essential for binding of the protein to the CRP box [4]. Moreover, analysis of the promoter region of the O. oeni arcA gene made it possible to pinpoint the DNA motif TGTGA-N6-TCACT which matches the

658

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

lactobacilli proteins, and R plays a key role during binding of the regulator to DNA by interacting with the GC base pairs in the TGTGA motif of the CRP half site [15]. These data suggest that the sequence protein deduced from orf229 is closer to CRP proteins than to FLP proteins. This conclusion was supported by the construction of a phylogenetic tree considering the ORF229p and members of the FNR, FLP and CRP proteins (gure 5B).
3.4. Induction of ADI pathway genes by arginine

Figure 5. (A) Alignment of the amino acid sequence deduced from orf229 (ORF229p) with that of the cAMP receptor protein of E. coli [1] (Ec-CRP), and Flp of L. casei. [9] (Lc-FLP). The domains determined by the Motifnder program are boxed. Region I: hypothetical cAMP binding domain. Region II: putatif DNA binding domain. Residues thought to interact directly with DNA are in bold characters. Identical amino acids are indicated by asterisks and similar ones by dots. (B) Phylogenetic relationships between members of FNR, CRP, FLP proteins and ORF229p established by parsimony analysis. Sequences alignment was performed using the ClustalW program and the tree was constructed using the Protpars program and TreeView software.

Assays to demonstrate induction by arginine of the three putative genes of the ADI pathway of O. oeni were carried out. Moreover, orf229 was also considered since it codes for a protein which looks like an effector involved in the regulation of several metabolic pathways. Total RNA was extracted from cells harvested at mid-logarithmic growth phase after incubation for 16 h in the presence of 10 g L1 arginine. The RT-PCR technique was used to study transcriptional induction (gure 6A). Controls were conducted to verify that no DNA was extracted with RNA. A control PCR with the extracted RNAs was done with specic primer sets related to the 23S rRNA and arcB genes. No amplication was noted for either the 23S rRNA gene (gure 6B) or the arcB gene (data not shown). Concerning the study of the ADI pathway genes and that of the orf229 gene, the expression of the 23S rRNA gene was monitored as a constitutive expression control since its level of expression did not vary upon arginine supplementation. A basal level of expression was observed for orf229 and the putative arcA, arcB and arcC genes without arginine addition (gure 6A). However, a large increase in expression was observed in the presence of arginine for the four genes, in the strain IOEB 8406 (gure 6A).
3.5. The IOEB 8413 strain which lacks the arc cluster is unable to degrade arginine

CRP binding domain mentioned above [7, 16], except the last base (gure 1B). A cyclic nucleotide binding domain was identied too in the N-terminal region of the two Gram-positive sequences (gure 5A), and no typical cysteine residues of the FNR proteins family were found in the sequence of the ORF229p. In addition one of the main differences with FLP of L. casei is in the DNA binding domain, since the RE side chains of ORF229p are replaced by PE in the

There is no known transformation method for O. oeni. Thus, knockout constructions and complementations are impossible. To make a direct link between these putative arc genes and arginine catabolism, we sought for a strain impeded in arginine consumption. During the screening of the O. oeni IOEB strain collection [17], the IOEB 8413 strain proved to be unable to catabolize arginine. Then, we carried out compared PCRs and hybridizations between strains IOEB

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

659

Figure 6. RT-PCR analysis of arc genes expression from an O. oeni strain. (A) Left panels: RT-PCR patterns obtained with cDNA coming from RNA of O. oeni IOEB 8406 extracted after 16 h of induction with 5 g L1 arginine (+) or without arginine (). The gene-specic primers that were used are indicated on the left part of panels. Right panels: the DNA contained in the agarose gels was transferred to a Nylon membrane and hybridized with labeled probes prepared from the relevant cloned arc genes. (B) Controls include PCR with 23S rRNA specic primers on water (H2 O, negative control), RNA extracted from cultures with or without arginine (RNA+ and RNA, negative controls) and genomic DNA instead of RNA (DNA, positive control). (C) Left panel: amplication of a 1300-bp arcA,B,C segment using primers specic to arc A and arcC genes with cDNA coming from RNA of O. oeni IOEB 8406 extracted after 16 h of induction with 5 g L1 arginine (cDNA lane). In a control experiment RNA extracted from cultures with arginine was used with the same arcA- and arcC specic primers (RNA+ lane). Right panel: the DNA contained in the agarose gel was transferred to a Nylon membrane and hybridized with a labeled arcB probe prepared from the cloned arcB gene. Lanes M: molecular size marker (100 bp DNA ladder, Promega).

Figure 7. An O. oeni strain devoid of the arc genes is unable to consume arginine. (A) PCRs (left panels) were carried out with the cognate gene-specic primers. Lanes M: molecular size marker; lanes (): control PCR without genomic DNA; lanes 1: IOEB 8406 strain genomic DNA was added to the PCR mix; lanes 2: IOEB 8413 strain genomic DNA was added to the PCR mix. Then, the DNA contained in the agarose gels was transferred to a Nylon membrane and hybridized with labeled probes prepared from the relevant cloned arc genes (right panels). (B) PCRs were carried out with the arcA gene specic primers. Three O. oeni strains that degrade and consume arginine were tested (lanes 1 to 3) along with the three O. oeni strains of the IOEB collection that are unable to degrade arginine (lanes 4 to 6). Lane 1: ATCC 23279; lane 2: IOEB 8406; lane 3: ATCC 23277; lane 4: IOEB 8413; lane 5: IOEB 8908; lane 6: IOEB 9614.

660

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

8406 and IOEB 8413 (gure 7A). Results showed that the arginine user 8406 strain allowed PCR amplication of orf229 and arcA, B, C gene-derived fragments from its genomic DNA (lanes 1). These fragments could hybridise with specic labelled probes prepared from the relevant cloned genes. In contrast, it was not possible to amplify or hybridise arc gene fragments from genomic DNA extracted from the arginine-decient 8413 strain (lanes 2). Thus, the 8413 strain cannot degrade arginine likely because it lacks the putative arc genes. Two other O. oeni strains that consume and degrade arginine (ATCC 23279 and ATCC 23277) were tested positively for the arcA gene amplication (gure 7B). For the two other strains of the IOEB collection that do not degrade or consume arginine (IOEB 8908 and IOEB 9614) faint bands and smears were amplied. This indicates that at least one of the two primers cannot efciently hybridise to the genomic DNA and therefore this pinpoints differences in the arcA gene sequence that could account for the observed lack of arginine consumption. In summary, the high identity values among ADI, OTC and CK protein sequences (including those of O. oeni) between various bacterias, the arginine induction of the O. oeni arc gene expression paralleling that of the enzymatic activities, and the lack of arginine degradation by a strain devoid of the putative arc genes, strongly suggests that these putative arc genes are genuine arc genes controlling arginine catabolism. 4. Discussion By cloning and sequencing the arc genes encoding the arginine deiminase pathway of O. oeni, we showed that these genes are clustered on the genome of this species. They appear in the same order as in L. sakei [19], the only other LAB for which the corresponding loci have been sequenced. Incubation with arginine allowed transcriptional stimulation of the orf229 gene and of the three genes of the ADI pathway. This is in keeping with the stimulation of arcABC genes by arginine observed in L. sakei [19]. However, unlike the L. sakei arc operon (repressed by as little as 0.1 g L1 glucose) [19], the O. oeni arc cluster expression is not submitted to catabolite repression since RNAs were prepared from cultures containing 10 g L1 glucose. The deduced amino acid sequence of the orf229 showed signicant similarity with proteins of the

CRP-FNR family. To our knowledge, no such protein is known in O. oeni. A p gene has recently been isolated from the LAB L. casei [9] and L. lactis [8]. The p genes of these last two species belong to operons that are specically involved in the oxidative stress response. However, no inuence of anaerobiosis or oxidative stress was observed on the arginine stimulation of the O. oeni arc cluster (data not shown). Involvement of proteins of the CRP-FNR family regulators have already been found in anaerobic arginine utilization. In P. aeruginosa, ANR (arginine and nitrate reduction regulator) positively regulates expression of the arcDABC operon [5]. Moreover, a motif homologous to the E. coli FNR box was identied 700 bp upstream from the R. etli arcA gene [3], and an imperfect palindromic sequence similar to the E. coli CRP and B. subtilis FNR consensus sequence has been found between the ArgR binding site and the site covered by RNA polymerase upstream from the arc promoter of B. licheniformis [12]. In conclusion, the location of orf229 upstream the arcA gene, together with the induction of its expression by arginine and the presence of a typical CRP binding site lying in the promoter region of arcA, suggest the following working hypothesis: the ORF229p factor might bind to this consensus site and therefore might positively regulate arginine catabolism in O. oeni. Acknowledgements We thank Prof. Prez-Martnez for having sent us the sequence of the gene cluster encoding ADI pathway enzymes of L. sakei as a personal communication. T.T. is supported by a grant from the Ministre de lEducation Nationale, de la Recherche et de la Technologie. References
[1] Cossart P., Gicquel-Sanzey B., Cloning and sequence of the crp gene of Escherichia coli K 12, Nucleic Acids Res. 10 (1982) 13631378. [2] Cunin R., Glansdorff N., Pierard A., Stalon V., Biosynthesis and metabolism of arginine in bacteria, Microbiol. Rev. 50 (1986) 314352. [3] DHooghe I., Vander Wauven C., Michiels J., Tricot C., de Wilde P., Wanderleyden J., Stalon V., The arginine deiminase pathway in Rhizobium etli : DNA sequence analysis

T. Tonon et al. / Res. Microbiol. 152 (2001) 653661

and functional study of the arcABC genes, J. Bacteriol. 179 (1997) 74037409. Dong Q., Ebright R.H., DNA binding specicity and sequence of Xanthomonas campestris catabolite gene activator protein-like protein, J. Bacteriol. 174 (1992) 54575461. Galimand M., Gamper M., Zimmermann A., Haas D., Positive FNR-like control of anaerobic arginine degradation and nitrate respiration in Pseudomonas aeruginosa, J. Bacteriol. 173 (1991) 15981606. Gasson M.J., Davies F.L., Conjugal transfer of drug resistant plasmid pAM 1 in lactic Streptococci , FEMS Microbiol. Lett. 7 (1980) 5153. Gaston K., Kolb A., Busby S., Binding of the Escherichia coli cyclic AMP receptor protein to DNA fragments containing consensus nucleotide sequences, Biochem. J. 261 (1989) 649653. Gostick D.O., Grifn H.G., Shearman C.A., Scott C., Green J., Gasson M.J., Guest J.R., Two operons that encode FNR-like proteins in Lactococcus lactis , Mol. Microbiol. 31 (1999) 15231535. Irvine A.S., Guest J.R., Lactobacillus casei contains a member of the CRP-FNR family, Nucleic Acids Res. 21 (1993) 753. Liu S.-Q., Pritchard G.G., Hardman M.J., Pilone G.J., Occurrence of arginine deiminase pathway enzymes in arginine catabolism by wine lactic acid bacteria, Appl. Environ. Microbiol. 61 (1995) 310316. Lonvaud-Funel A., Lactic acid bacteria in the quality improvement and depreciation of wine, Antonie van Leeuwenhoek 76 (1999) 317331.

661

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12] Maghnouj A., de Sousa Cabral T.F., Stalon V., Van der Wauven C., The arcABDC gene cluster, encoding the arginine deiminase pathway of Bacillus licheniformis, and its activation by the arginine repressor ArgR, J. Bacteriol. 180 (1998) 64686475. [13] Martinez-Murcia A.J., Harland N.M., Collins M.D., Phylogenetic analysis of some leuconostocs and related organisms as determined from large-subunit rRNA gene sequences: assessment of congruence of small- and large-subunit rRNA derived trees, J. Appl. Bacteriol. 74 (1993) 532541. [14] Ough C.S., Ethyl carbamate in fermented beverages and foods, J. Agri. Food Chem. 24 (1976) 323328. [15] Schultz S., Shields G., Steitz T., Crystal structure of a CAPDNA complex: the DNA is bent by 90 degrees, Science 253 (1991) 10011007. [16] Spiro S., Gaston K.L., Bell A.I., Roberts R.E., Busby S.J.W., Guest J.R., Interconversion of the DNA-binding specicities of two related transcription regulators, CRP and FNR, Mol. Microbiol. 4 (1991) 18311838. [17] Tonon T., Etude du catabolisme de larginine par les bactries lactiques du vin. Aspects physiologiques, enzymatiques et gntiques chez Oenococcus oeni, PhD thesis (2000). [18] Tonon T., Lonvaud-Funel A., Metabolism of arginine and its positive effect on growth and revival of Oenococcus oeni , J. Appl. Microbiol. 89 (2000) 526531. [19] Zniga M., Champomier-Vergs M., Zagorec M., PrezMartinez G., Structural and functional analysis of the gene encoding the enzymes of the arginine deiminase pathway of Lactobacillus sake, J. Bacteriol. 180 (1998) 41544159.