Isolation of the patC gene encoding the cystathionine ß-lyase of Lactobacillus delbrueckii subsp. bulgaricus and molecular analysis of inter-strain variability in enzyme biosynthesisa

Dominique Aubel1, Jacques Edouard Germond2, Christophe Gilbert1 and Danièle Atlan1

Unité de Microbiologie et Génétique, UMR 5122, Université Claude Bernard-Lyon 1, bât. Lwoff, 10 rue Dubois, F-69622 Villeurbanne Cedex, France1
Nestlé Research Centre, Nestec Ltd, Vers-chez-les-Blancs, CH-1000 Lausanne 26, Switzerland2

Author for correspondence: Dominique Aubel. Tel: +33 4 72 43 13 66. Fax: +33 4 72 43 26 86. e-mail: aubel{at}biomserv.univ-lyon1.fr


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
The patC gene encoding the cystathionine ß-lyase (CBL) of Lactobacillus delbrueckii subsp. bulgaricus NCDO 1489 was cloned and expressed in Escherichia coli. Overexpression of CBL complemented the methionine auxotrophy of an E. coli metC mutant, demonstrating in vivo that this enzyme functions as a CBL. However, PatC is distinguishable from the MetC CBLs by a low identity in amino acid sequence, a sensitivity to iodoacetic acid, greater thermostability and a lower substrate affinity. Homologues of patC were detected in the 13 Lb. delbrueckii strains studied, but only seven of them showed CBL activity. In constrast to CBL+ strains, all CBL-deficient strains analysed were auxotrophic for methionine. This supports the hypothesis that CBLs from lactobacilli are probably involved in methionine biosynthesis. Moreover, the results of this study suggest that post-transcriptional mechanisms account for the differences in CBL activities observed between strains of Lb. delbrueckii.

Keywords: amino acid catabolism, ß-cystathionase, PatC, lactic acid bacteria, cheesemaking

Abbreviations: CBL, cystathionine ß-lyase; DMDS, dimethyldisulfide; DMTS; dimethyltrisulfide; PLP, pyridoxal 5'-phosphate

a The GenBank accession number for the sequence determined in this work is AF423071.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Lactic acid bacteria contribute to flavour formation and texture development in dairy products and are therefore widely used in the dairy industry. Flavour development in cheese is a complex process which is initiated by proteolytic degradation of milk caseins mediated by rennet and the proteases produced by lactic acid bacteria. Subsequently, the aromatic, branched-chain and sulfur-containing amino acids are converted to volatile compounds by chemicals or enzymes which display deaminase, decarboxylase or transaminase activities, or which modify amino acid side chains. Among numerous chemicals, the volatile sulfur compounds methanethiol, dimethyldisulfide (DMDS), dimethyltrisulfide (DMTS) and methional have been found to be responsible for cheese flavour and aroma (Weimer et al., 1999 ). Indeed, these methionine breakdown products have been discovered in various cheese types (Urbach, 1993 ; Barbieri et al., 1994 ; Engels et al., 1997 ). The mechanisms by which these compounds arise have been particularly studied in the genus Lactococcus, with the ultimate goal of improving cheese flavour.

Cystathionine ß-lyase (CBL) (EC 4.4.1.8, also commonly referred to as ß-cystathionase) is a pyridoxal-5'-phosphate (PLP)-dependent enzyme which catalyses mainly the cleavage of L-cystathionine to homocysteine, pyruvate and ammonia via an {alpha}/ß-elimination reaction; that is, the enzyme splits the {alpha}C–N (deaminating) and ßC–S bonds (Clausen et al., 1997 ). Subsequently, the homocysteine is methylated by the homocysteine methyltransferase to form methionine (Weimer et al., 1999 ). However MetC, the purified CBL, from Lactococcus lactis subsp. cremoris B78 or MG1363 is also able to degrade cystathionine to cysteine, {alpha}-ketobutyrate and ammonia, or to degrade methionine ({gamma}C–S linkage) in methanethiol via an {alpha}/{gamma}-elimination reaction (Alting et al., 1995 ; Dobric et al., 2000 ). In a following reaction, the methanethiol can be oxidized to DMDS or DMTS (Parliment et al., 1982 ). However, no significant differences in the amount of DMDS formed were found between the MG1363 parental strain and the metC mutant, suggesting that other enzymes may play a more substantial role in the conversion of methionine (Fernandez et al., 2000 ). In fact, the cystathionine {gamma}-lyase (Bruinenberg et al., 1997 ), the aromatic aminotransferase AraT (Yvon et al., 1997 ; Rijnen et al., 1999 ), and the branched-chain aminotransferases AT-A and AT-B (Engels et al., 2000 ) or ILvE (Atiles et al., 2000 ) have also been reported to contribute to the conversion of methionine to flavour compounds.

Although the sulfur-containing amino acid catabolism of lactococci is well described, data for the genus Lactobacillus are as yet very limited. Dias & Weimer (1998) reported the presence of CBL and methionine aminotransferase activities in some strains of lactobacilli. Moreover, a cystathionine {gamma}-lyase was purified from Lactobacillus fermentum DT 41, a strain contained in the natural starter used for Parmesan cheese production (Smacchi & Gobetti, 1998 ). This enzyme was found to be stable under the conditions of cheese ripening and has therefore been thought to contribute to the biosynthesis of volatile sulfur-containing compounds, although L-methionine and L-cysteine were not the optimal substrate for the purified enzyme. Recently, methionine aminotransferase activities have been detected in 30% of Lactobacillus casei and 23% of Lactobacillus plantarum strains (Amarita et al., 2001 ).

To our knowledge there are no data available on the catabolism of sulfur-containing amino acids in Lactobacillus delbrueckii species. In the present work, 14 strains of Lb. delbrueckii subsp. bulgaricus or subsp. lactis were screened for CBL activity. The corresponding patC gene was cloned from Lb. delbrueckii subsp. bulgaricus NCDO 1489, sequenced, and its distribution and expression assessed among lactic acid bacteria.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Bacterial strains and culture conditions.
Bacterial strains used in this study are listed in Table 1. All Lactobacillus strains were grown at 40 °C without shaking in Difco lactobacilli MRS broth (De Man et al., 1960 ) or in the minimal defined medium described by Poolman & Konings (1988) and modified as follows: 0·008 g tyrosine l-1, 0·4 g cysteine l-1, 0·3 g glutamic acid l-1 and 0·1% (v/v) Tween 80. Escherichia coli strains were grown at 37 °C either in Luria–Bertani medium (Miller, 1972 ) or on minimal medium M63 [0·1 M KH2PO4, 15 mM (NH4)2SO4, 9 µM FeSO4, 0·8 mM MgSO4, 3 µM thiamin chlorhydrate] supplemented with 0·2% (w/v) glucose or glycerol as the carbon source, and 0·01% (w/v) methionine when necessary. Ampicillin, chloramphenicol and erythromycin were added to final concentrations of 100, 20 and 50 µg ml-1 respectively. Genes under the control of IPTG-responsive promoters were induced for 4 h with 1 mM IPTG at an OD600 value of 0·6.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids used in this study

 
Genomic DNA extraction and Southern-blotting.
Genomic DNA of Lactobacillus strains was purified using Qiagen Genomic-tip 100/G columns according to the manufacturer’s recommendations (Qiagen). For efficient lysis of the Lactobacillus strains, the Qiagen bacterial lysis buffer B1 was supplemented with mutanolysin (10 µg ml-1). Purified genomic DNA was digested with EcoRI. Fragments were separated on an 0·8% (w/v) agarose gel by electrophoresis and transferred onto a nylon filter (Positive membrane, Qbiogene-Quantum Appligene) by the Southern transfer method (Southern, 1975 ). The DNA probe (for example, the 569 bp NcoI–SacI fragment shown in Fig. 1) was directly labelled with horseradish peroxidase using the ECL direct nucleic acid labelling system according to the manufacturer’s instructions (Amersham Biosciences). Detection of specific hybridization was done by chemiluminescence using the ECL detection system.



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 1. Genetic organization of the prtB–patC locus of Lb. delbrueckii subsp. bulgaricus NCDO 1489. Locations of primers used for gene cloning and RT-PCR experiments are shown.

 
Construction of plasmids.
Plasmids constructed in this study are listed in Table 1. A fragment including the gene of interest was amplified from genomic DNA of Lb. delbrueckii subsp. bulgaricus NCDO 1489 and Lb. delbrueckii subsp. lactis ATCC 21815 by PCR with two pairs of primers based on the sequence of the NCDO 1489 patC gene (Fig. 1): 5'-GCGGAATTCGCATGCGGAGAGAGAACG-3' (pmal2) and 5'-GCGGGATCCGCATGCCGTTCTACTGGC-3' (pmal1), or 5'-GCGGCATGCCAGAAAAGCAATAT-3' (pmal4) and 5'-GCAAGCTTCGTTCTAC TGGC-3' (pmal5) contain an EcoRI, BamHI, SphI and HindIII site, respectively (underlined). The 1·4 kb PCR fragments obtained from strains NCDO 1489 and ATCC 21815 were digested with EcoRI and BamHI, and then were cloned either in pJDC9 (Chen et al., 1987 ) or pCR-ScriptCamSK(+), resulting in plasmids pDL19 and pDL43. The 1·2 kb PCR fragment obtained from NCDO 1489 genomic DNA was cloned as a SphI–HindIII fragment into pQE-32 (Qiagen), resulting in plasmid pDL22. Subsequently, the 832 bp BglI–SacI fragment from pDL22 was replaced by the corresponding pDL43 fragment, resulting in plasmid pDL44, which encodes a hybrid PatC protein. Plasmids pDL22 and pDL44 were used for tagging the protein with six consecutive histidine residues for simple purification using nickel resin.

RNA preparation and RT-PCR.
RNA was purified from 1 ml of a fresh bacterial culture grown in MRS medium and harvested at OD600 value of 1·0. Then, bacterial cells were treated with the Rneasy Mini kit (Qiagen) according to a modified procedure. Cells were lysed with TE buffer containing 5 mg lysozyme ml-1 (45 min at 37 °C). Genomic DNA was eliminated by the action of the Rnase-free Dnase I (Qiagen) during RNA purification on the Rneasy Mini column (40 min at room temperature). RNA preparations were analysed by agarose gel electrophoresis and RNA concentrations were determined by absorbance at 260 nm.


cDNAs from patC and prtB mRNA were generated by RT-PCR using the Qiagen OneStep RT-PCR kit according to the supplier’s instructions. Two amounts of RNA (0·15 or 1·0 µg) were used for each RT-PCR assay. Primers specific to prtB or patC genes from Lb. delbrueckii subsp. bulgaricus NCDO 1489 were used for amplification (Fig. 1). Primers PRTBsens1 (5'-GACAACCCACTCACTGGACAAG-3') and PRTBrev1 (5'-CTCTTGATCGTAAGCCGGGC-3') , primers PRTBsens2 (5'-TTGACGAAGAAAGATCAGGCCG-3') and PATCrev1 (5'-AACGCCGCCAAGTTGAAGG-3') generated a 504 bp or a 3326 bp amplicon, respectively. cDNA of patC (674 bp) was generated with PATCsens1 (5'-GGGCAATTCCATCAAGTGGG-3') and PATCrev1 primers. Before any RT-PCR assays, all primers were controlled by direct PCR using a ready-to-use PCR mixture of Gibco-BRL (Invitrogen) and genomic DNAs of four studied strains as templates. In order to check DNA contamination of RNA preparations, direct PCR was also carried out with the same primers and Gibco-BRL mixture.

Total cellular extracts for enzyme assays.
Cells from 300 ml of an overnight culture of Lactobacillus strains grown in Difco MRS broth were harvested, washed with 100 mM Tris/HCl pH 8·2 and then resuspended in 3 ml of the same buffer. The cells were afterwards disrupted at 6 °C by three passages through an Aminco French pressure cell at 20000 p.s.i. (138 MPa). Intact cells and cell debris were removed by centrifugation at 12000 g for 30 min at 6 °C. The supernatant was collected and protein concentration was estimated by the method of Bradford (1976) using bovine serum albumin as the standard. The extracts were used for enzyme assays as described below and stored at -20 °C.

Detection of CBL activity by in situ staining.
CBL activity was monitored in PAGE gels with either L-cystine or L-cysteine as the substrate. Crude extracts (0·3 mg) supplemented with SDS-free sample buffer [5x buffer is composed of 150 mM Tris/HCl pH 6·8, 30% (v/v) glycerol, 0·005% (w/v) bromophenol blue] were loaded on the polyacrylamide gel. A discontinuous system consisting of stacking and separation gels was used to separate proteins. The lower separation gel consisted of 12% (w/v) polyacrylamide in 150 mM Tris/HCl pH 8·8 with 0·1% (w/v) ammonium persulfate and 0·1% (w/v) tetramethylenediamine (TEMED). The stacking gel was composed of 4·8% (w/v) polyacrylamide in 200 mM Tris/HCl pH 6·8 with polymerization reagents at the same concentration as above. The electrophoresis buffer contained 192 mM glycine, 25 mM Tris at pH 8·3. After electrophoresis, the polyacrylamide gel was incubated in 100 mM Tris/HCl buffer pH 8·2 prewarmed to 37 °C and supplemented with 10 mM L-cysteine or L-cystine, 0·5 mM Pb(NO3)2 and 0·4 mM PLP.

CBL assay and kinetic constant determination.
The assay used for determination of kinetic parameters was based on the formation of free thiol groups by spontaneous disulfide interchange with Ellman’s reagent [5,5'-dithiobis-(2-nitrobenzoic acid)] or DTNB according to the method described by Uren (1987). Assay mixtures contained 0·2 mM DTNB, 0·1 mM PLP, various substrate concentrations (L-cystine or L-cystathionine dissolved in water and 0·04 M HCl, respectively) ranging from 0·5 to 4 mM and purified His-tagged PatCNCDO 1489 in a final volume of 1 ml Tris/HCl 100 mM pH 8. The increase in absorbance at 412 nm was measured in a Uvikon 930 spectrophotometer at 37 °C; it was recorded at 0·5 min intervals for 10 min. A molar absorption coefficient for the aryl mercaptide of 13200 l mol-1 cm-1. was used to calculate the enzyme activity. One enzyme unit represents the formation of 1 µmol aryl mercaptan min-1. The Lineweaver–Burk plot (1/v against 1/s) was employed to calculate Km and Vmax values. The values reported represent a mean of three independent experiments±SD.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Isolation and sequence analysis of the patC gene from Lb. delbrueckii subsp. bulgaricus NCDO 1489
The DNA region located downstream from the prtB gene encoding the cell-surface proteinase from strain NCDO 1489 (Gilbert et al., 1996 ) was isolated by inverse PCR, and the DNA sequence was determined on both strands (Fig. 1). A computer-assisted search for protein-coding regions revealed one ORF of 1170 bp which encodes a protein of 390 amino acids with a predicted molecular mass of 43·9 kDa. This ORF was preceded by a potential ribosome-binding site (GAGG) ending 8 bp upstream of the ATG start codon and by several sequences showing high homology to the Lb. delbrueckii consensus promoter (Matern et al., 1994 ). The putative -35 (TGATTT) and -10 (TTACTT) boxes with a spacing of 18 nucleotides and ending 39 bp upstream of the ribosome-binding site seemed to fit better with the consensus promoter for two major reasons: (i) the nucleotides TGC were found upstream of the -10 region; (ii) a potential upstream (UP)-element, characterized by oligo-A and -T stretches (A and T content of 78·2 mol%), was detected 4 nucleotides upstream of the -35 region. This UP-element was shown to stimulate transcription in bacteria and has been found upstream of numerous promoters of Lb. delbrueckii genes (Matern et al., 1994 ). The GC content of this ORF is 52 mol%, which is in good agreement with the characteristic mol% GC content of Lb. delbrueckii subsp. bulgaricus (Kilpper-Balz et al., 1982 ).

A search for protein homology revealed significant similarity of the deduced amino acid sequence to the potential amino transferase (PatB) protein (37·3% identity) of Bacillus subtilis and the MalY protein (31·4% identity) of E. coli (Fig. 2). Whereas the biological function of PatB is still unknown (Trach & Hoch, 1993 ), MalY has been described as a pleiotropic protein acting as a PLP-dependent CBL and as a repressor of the maltose regulon by interaction with MalT, the central transcription activator of the maltose regulon (Zdych et al., 1995 ; Clausen et al., 2000 ). Considering the significant identity of the deduced amino acid sequence with PatB from B. subtilis (Fig. 2), the ORF located immediately downstream of prtB gene from Lb. delbrueckii subsp. bulgaricus NCDO 1489 was named patC. The DNA sequence of the Lb. delbrueckii subsp. bulgaricus NCDO 1489 patC gene has been assigned GenBank number AF423071.



View larger version (69K):
[in this window]
[in a new window]
 
Fig. 2. Alignment of the PatC amino acid sequence from Lb. delbrueckii subsp. bulgaricus NCDO 1489 (PatCNCDO 1489) with PatB from Bacillus subtilis (Swissprot Q08432) and MalY of E. coli (P23256). Boxes indicate identical amino acid residues over the three proteins; conserved amino acids with either PatB or MalY are indicated by a dash above the sequence. Amino acids of PatC from ATCC 21815 and CNRZ 250 which differ from NCDO 1489 are indicated underneath the PatCNCDO 1489 sequence: white letters on a black background and in bold indicate substitutions in PatCATCC 21815 and PatCCNRZ 250, respectively. Letters shown in bold and underlined indicate substitutions that were found in PatCATCC 21815 as well as in PatCCNRZ 250. The potential binding site of PLP is shown by a black triangle; amino acids that might be involved in PLP stabilization are indicated by a + symbol (Clausen et al., 2000 ). Other residues essential for the MalY CBL activity are shown by an asterisk. The lines numbered I through III correspond to regions described by Clausen et al. (2000) as being involved in the interaction between MalY and MalT.

 
PatC and CBL activity in vivo
Amino acid sequence alignments (Fig. 2) showed that the Lys-233 residue by which the cofactor PLP is covalently attached to MalY, as well as 7 out of 11 residues which had been previously reported to stabilize the cofactor, were conserved in PatC (Clausen et al., 2000 ). Moreover, the Tyr-121 and Asp-201 residues, essential for the MalY CBL activity, were also conserved in PatC. All these data suggested that PatC might represent a PLP-dependent enzyme with CBL activity.

To test our hypothesis, plasmid pDL19 (harbouring the 1·4 kb PCR fragment encompassing patC from NCDO 1489 and its 217 bp upstream region) was constructed and introduced into E. coli strain CAG18475, which contains a mutation in the CBL-encoding gene (metC) essential in the methionine biosynthetic pathway (Singer et al., 1989 ). E. coli strain CAG18475 harbouring plasmid pDL19 was able to grow on minimal medium M63 supplemented with glucose or glycerol as the carbon source, whereas the same strain carrying the parental vector pJDC9 still required methionine to grow on the same medium. Thus, we clearly demonstrated that the overproduction of PatC, as with MalY (Zydch et al., 1995), complements the methionine requirement of a metC mutant and that PatC functions in vivo as a CBL. Moreover, we confirmed that the patC gene of strain NCDO 1489 is expressed under its own promoter since the fragment cloned in pDL19 is flanked by efficient transcriptional terminators which prevent any interference by vector promoters (Chen & Morrison, 1987 ).

Although PatC and MalY share some features, they differ in their number of subunits and their capacity to repress the maltose regulon of E. coli. The native enzyme MalY consists of two identical subunits (Clausen et al., 2000 ), whereas PatC from Lb. delbrueckii subsp. bulgaricus NCDO 1489 is apparently a homotetramer (data not shown) as for CBL from numerous bacteria (Alting et al., 1995 ; Belfaiza et al., 1986 ). In contrast to MalY, PatC overproduction did not repress the expression of a malK'–'lacZ fusion (data not shown). So, PatC is not able to control the regulation of the maltose regulon (malT-dependent genes). This is not surprising, since most amino acid residues involved in the interaction with the transcriptional activator MalT are not conserved in PatC, as shown in Fig. 2 (Clausen et al., 2000 ).

Purification of PatC and biochemical features
In order to produce large amounts of PatC, the patC gene from NCDO 1489 was amplified from genomic DNA using primers pmal4 and pmal5 (Fig. 1) and subsequently cloned in-frame with a His6-tag coding sequence (pDL22). Moreover, this hybrid patC gene was under control of the IPTG-inductible T5 coliphage promoter. The His-tagged PatC protein was purified under native conditions by affinity chromatography on nickel-nitrilotriacetic acid (Ni-NTA) resin and its purity was confirmed by SDS-PAGE. Indeed, a single band running at the expected size of 45 kDa was observed following staining with Coomassie blue as shown in Fig. 3(a, lane 2).



View larger version (57K):
[in this window]
[in a new window]
 
Fig. 3. SDS-PAGE (a) and non-denaturing PAGE (b) analysis of purified His-tagged PatC proteins with Coomassie blue staining and in situ staining for CBL activity. Lanes: 1, purified hybrid PatCNCDO 1489/ATCC 21815; 2, purified PatCNCDO 1489; M, molecular mass markers (Gibco-BRL, BenchMark prestained protein ladder). The PatC bands are arrowed.

 
CBL activity could easily be monitored in non-denaturing gels using either L-cysteine or L-cystine as the substrate; both compounds have a ßC–S bond. Hydrogen sulfide resulting from L-cysteine or L-cystine cleavage precipitated with Pb(NO3)2 to an insoluble PbS product forming brown to black bands on the polyacrylamide gel. Electrophoresis under non-denaturing conditions of purified His-tagged PatC revealed a single dark brown band (Fig. 3b, lane 2). This suggests that the 13 additional amino acids at the N-terminus of PatC probably do not interfere with enzyme activity. Suppression of the methionine auxotrophy in a metC mutant harbouring pDL22 corroborates our hypothesis.

Previous results (Aubel et al., 2002 ) have shown that like most CBLs (Alting et al., 1995 ; Smacchi & Gobetti, 1998 ; Bruinenberg et al., 1997 ; Gentry-Weeks et al., 1993 ), PatC is a PLP-dependent enzyme that is sensitive to hydroxylamine. PatC is also sensitive to the thiol-reactive agent iodoacetic acid and, in this respect, is similar to the CBL of Bordetella avium (Gentry-Weeks et al., 1993 ). Nevertheless, the two enzymes differ in their thermostability (Gentry-Weeks et al., 1993 ; Aubel et al., 2002 ) and substrate affinity. Indeed under the conditions used (see Methods), the PatC enzyme exhibited slightly higher affinity for L-cystine (Km 1·88±0·37 mM; Vmax 48·69±0·95 U mg-1) than for L-cystathionine (Km 5·13±0·42 mM; Vmax 13·17±1·01U mg-1). Both of these Km values were higher than those reported for the MetC enzyme of B. avium (Km 0·070–0·077 mM for L-cystathionine and L-cystine, respectively; Gentry-Weeks et al., 1993 ) and E. coli (Km 0·04–0·25 mM; Dwivedi et al., 1982 ) or Salmonella (Km 0·22–0·80 mM; Guggenheim, 1971 ). On the other hand, the affinity of PatC and MalY of E. coli (Km 1·70±0·10 mM; Zdych et al., 1995 ) for L-cystine was quite similar.

Distribution of the patC gene among lactic acid bacteria
An ORF with 97·7 and 98·4% identity to patC was detected downstream of the prtB genes of Lb. delbrueckii subsp. lactis CNRZ 250 and ATCC 21815, respectively, which suggested that the locus encompassing both genes was common to most Lb. delbrueckii strains. In order to determine whether the patC gene was specific to Lb. delbrueckii species or widely distributed in lactic acid bacteria, Southern hybridizations were performed. The patC probe hybridized with a single DNA fragment in all eight Lb. delbrueckii subsp. bulgaricus and all five subsp. lactis strains studied, as shown in Fig. 4(a, b). However, under the conditions used, no hybridizing band was detected in the genome of Lactobacillus helveticus (four strains), Lb. casei (one strain) or Lc. lactis (three strains) (data not shown). The absence of hybridation with the Lc. lactis genome is not surprising, because the identity between PatC and MetC from Lc. lactis (Fernandez et al., 2000 ) is very low (18·3%). Our results suggest that the patC gene is only present in Lb. delbrueckii species.



View larger version (41K):
[in this window]
[in a new window]
 
Fig. 4. Hybridization of genomic DNA from Lb. delbrueckii subsp. bulgaricus (a) and Lb. delbrueckii subsp. lactis (b) strains with the Lb. delbrueckii subsp. bulgaricus NCDO 1489 patC gene. An internal 569 bp DNA fragment of the patC gene was hybridized to EcoRI-digested genomic DNA (a) of NCDO 1489 (lane 1), Lfi5.H7 (lane 2), LB4 (lane 3), CNRZ 1187 (lane 4), CNRZ 416 (lane 5), Lfi5 (lane 6), NZDRI 5005 (lane 7), CNRZ 397 (lane 8) and LT1 (lane 9), or (b) of NCDO 1489 (lane 1), Lb. delbrueckii subsp. lactis CNRZ 250 (lane 2), CNRZ 245 (lane 3), LL44 (lane 4), N141 (lane 5) and ATCC 21815 (lane 6). Arrows indicate the sizes of the hybridizing fragments.

 
CBL biosynthesis and methionine auxotrophy among Lb. delbrueckii species
As shown above, sequences homologous to patC of NCDO 1489 have been detected in 13 Lb. delbrueckii subsp. bulgaricus or subsp. lactis strains. However, CBL activity could only be detected in cellular extracts of six (CBL+) out of eight Lb. delbrueckii subsp. bulgaricus strains and in a single strain of Lb. delbrueckii subsp. lactis (CNRZ 250) with the in situ enzymic test on native PAGE (Aubel et al., 2002 ; Table 2). Enzyme activity was also assayed by determining the formation of free thiol groups (Uren, 1987 ). The specific activity of the crude extract of NCDO 1489 (control strain) towards the substrate L-cystine was 76x10-4 µmol min-1 mg-1. In contrast, the values were eight- to tenfold weaker for strains ATCC 21815 and CNRZ 397, previously described as CBL- strains (Aubel et al., 2002 ). Therefore, the in situ enzymic test and the quantitative CBL assay yielded similar results. Whereas CBL+ strains were able to grow in a minimal defined medium without methionine, all CBL- strains were auxotrophic for methionine (Table 2). These results suggest that PatC is probably involved in the methionine biosynthesis pathway of Lb. delbrueckii strains.


View this table:
[in this window]
[in a new window]
 
Table 2. Relationships between CBL activity and methionine auxotrophy in Lb. delbrueckii strains

 
The mechanism(s) that might be involved in the lack of CBL activity in some strains have been investigated and, therefore, the patC genes of two other Lb. delbrueckii strains, CNRZ 250 (CBL+) and ATCC 21815 (CBL-), were sequenced (data not shown). The PatC amino acid sequence from NCDO 1489 (PatCNCDO 1489) differs from PatCATCC 21815 and PatCCNRZ 250 at 7 positions (Fig. 2). The substitutions in positions 23, 46, 257 and 327 do not interfere with the enzyme activity since they were found in PatCATCC 21815 (CBL) as well as PatCCNRZ 250 (CBL+). On the other hand the three substitutions Val-100, Leu-158 and Lys-186, which are only detected in PatCATCC 21815 (CBL-), may account for the lack of CBL activity in this strain. In order to test this hypothesis, we constructed a hybrid PatCNCDO 1489/ATCC 21815 protein (plasmid pDL44); as a consequence the Met-100, Pro-158 and Glu-186 of PatCNCDO 1489 were respectively substituted by Val, Leu and Lys. The purified hybrid PatC displayed similar enzyme activity to PatCNCDO 1489, as shown in Fig. 3(b), suggesting that the lack of CBL activity in strain ATCC 21815 was not due to these amino acid substitutions. Consequently, PatCATCC 21815 is enzymically active; this was confirmed by the observation that the overexpression of PatCATCC 21815 complemented the methionine auxotrophy of the E. coli metC mutant. Therefore, it could be supposed that the transcription of the patC gene may be altered in CBL- strains of lactobacilli compared to CBL+ strains. To check our hypothesis, total RNA was extracted from two CBL+ strains (NCDO 1489 and CNRZ 250) and two CBL- strains (CNRZ 397 and ATCC 21815) belonging to the bulgaricus or lactis subspecies, respectively. No cDNA was obtained by RT-PCR using PRTBsens2 and PATCrev1 primers (Fig. 1). These results indicate that prtB and patC do not form an operon in Lb. delbrueckii. This conclusion is in agreement with the presence of a CBL activity in LFi5.H7, a spontaneous lacZprtB deletion mutant (Germond et al., 1995 ). Moreover, no ORF was detected immediately downstream of the patC gene. Thus, expression of patC is quite different from that of the metC gene that encodes the CBL from Lc. lactis, which is co-transcribed with the downstream cysK gene encoding a putative cysteine synthase (Fernandez et al., 2000 ). On the other hand, patC and prtB cDNAs of the expected size were generated by RT-PCR from all RNA extracts studied (Fig. 5), demonstrating that the patC gene was transcribed in CBL- as well as in CBL+ strains. All these data suggested that unknown post-transcriptional mechanisms account for the differences in CBL activities.



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 5. Amplification of cDNAs of patC and prtB from CBL+ or CBL- strains of Lb. delbrueckii. RT-PCR was carried out with primers PATCsens1 and PATCrev1 (lanes 2–5) or primers PRTBsens1 and PRTBrev1 (lanes 6–9) using 0·15 µg RNA from Lb. delbrueckii subsp. bulgaricus NCDO 1489 (lanes 2 and 6) or CNRZ 397 (lanes 3 and 7), and Lb. delbrueckii subsp. lactis CNRZ 250 (lanes 4 and 8) or ATCC 21815 (lanes 5 and 9). Product length was controlled by reference to components of 1 kb ladder (0·8 µg, Gibco-BRL, Invitrogen) (lane 1).

 
In conclusion, like most lactic acid bacteria, Lb. delbrueckii synthesizes a CBL, which we have named PatC. The sequence and biochemical properties of this enzyme are different from those of other published CBLs (MetC) involved in methionine metabolism. Among Lb. delbrueckii species, the biosynthesis level of PatC is variable and strain dependent. Further studies are necessary to elucidate mechanisms controlling PatC biosynthesis. Methionine is a precursor of several aroma compounds, and a screening of CBL+ strains for cheesemaking could enhance the development of flavours.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Alting, A. C., Engels, W. J. M., van Schalkwijk, S. & Exterkate, F. A. (1995). Purification and characterization of cystathionine ß-lyase from Lactococcus lactis subsp. cremoris B78 and its possible role in flavor development in cheese. Appl Environ Microbiol 61, 4037-4042.[Abstract]

Amarita, F., Requena, T., Taborda, G., Amigo, L. & Pelaez, C. (2001). Lactobacillus casei and Lactobacillus plantarum initiate catabolism of methionine by transamination. J Appl Microbiol 90, 971-978.[Medline]

Atiles, M. W., Dudley, E. G. & Steele, J. L. (2000). Gene cloning, sequencing, and inactivation of the branched-chain aminotransferase of Lactococcus lactis LM0230. Appl Environ Microbiol 66, 2325-2329.[Abstract/Free Full Text]

Aubel, D., Germond, J.-E., Gilbert, C. & Atlan, D. (2002). L’activité ß-cystathionase au sein de l’espèce Lactobacillus delbrueckii. Sci Aliments (in press).

Barbieri, G., Bolzoni, L., Careri, M., Manglia, A., Parolari, G., Spagnoli, S. & Virgili, R. (1994). Study of the volatile fraction of Parmesan cheese. J Agric Food Chem 42, 1170-1176.

Belfaiza, J., Parsot, C., Martel, A., la Tour, C. B., Margarita, D., Cohen, G. N. & Saint-Girons, I. (1986). Evolution in biosynthetic pathways: two enzymes catalyzing consecutive steps in methionine biosynthesis originate from a common ancestor and possess a similar regulatory region. Proc Natl Acad Sci USA 83, 867-871.[Abstract]

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248-254.[Medline]

Bruinenberg, P. G., de Roo, G. & Limsowtin, G. K. Y. (1997). Purification and characterization of cystathionine {gamma}-lyase from Lactococcus lactis subsp. cremoris SK11: possible role in flavor compound formation during cheese maturation. Appl Environ Microbiol 63, 561-566.[Abstract]

Chen, J.-D. & Morrison, D. A. (1987). Cloning of Streptococcus pneumoniae DNA fragments in Escherichia coli requires vectors protected by strong transcriptional terminators. Gene 55, 179-187.[Medline]

Clausen, T., Laber, B. & Messerschmidt, A. (1997). Mode of action of cystathionine ß-lyase. Biol Chem 378, 321-326.[Medline]

Clausen, T., Schlegel, A., Peist, R. & 7 other authors (2000). X-ray structure of MalY from Escherichia coli: a pyridoxal 5'-phosphate-dependent enzyme acting as a modulator in mal gene expression. EMBO J 19, 831–842.[Abstract/Free Full Text]

De Man, J. C., Rogosa, M. & Sharpe, M. E. (1960). A medium for the cultivation of lactobacilli. J Appl Bacteriol 23, 130-135.

Dias, B. & Weimer, B. (1998). Conversion of methionine to thiols by lactococci, lactobacilli, and brevibacteria. Appl Environ Microbiol 64, 3320-3326.[Abstract/Free Full Text]

Dobric, N., Limsowtin, G. K. Y., Hillier, A. J., Dudman, N. P. & Davidson, B. E. (2000). Identification and characterization of a cystathionine ß/{gamma}-lyase from Lactococcus lactis ssp. cremoris MG1363. FEMS Microbiol Lett 182, 249-254.[Medline]

Dwivedi, C. M., Ragin, R. C. & Uren, J. R. (1982). Cloning, purification, and characterization of ß-cystathionase from Escherichia coli. Biochemistry 21, 3064-3069.[Medline]

Engels, W. J. M., Dekker, R., de Jong, C., Neeter, R. & Visser, S. (1997). A comparative study of volatile compounds in the water-soluble fraction of various types of ripened cheese. Int Dairy J 7, 255-263.

Engels, W. J. M., Alting, A. C., Arntz, M. M. T., Gruppen, H., Voragen, A. G. J., Smit, G. & Visser, S. (2000). Partial purification and characterization of two aminotransferases from Lactococcus lactis subsp. cremoris B78 involved in the catabolism of methionine and branched-chain amino acids. Int Dairy J 10, 443-452.

Fernandez, M., van Doesburg, W., Rutten, G. A. M., Marugg, J. D., Alting, A. C., van Kranenburg, R. & Kuipers, O. P. (2000). Molecular and functional analyses of the metC gene of Lactococcus lactis, encoding cystathionine ß-lyase. Appl Environ Microbiol 66, 42-48.[Abstract/Free Full Text]

Gentry-Weeks, C. R., Keith, J. M. & Thompson, J. (1993). Toxicity of Bordetella avium ß-cystathionase toward MC3T3-E1 osteogenic cells. J Biol Chem 268, 7298-7314.[Abstract/Free Full Text]

Germond, J.-E., Lapierre, L., Delley, M. & Mollet, B. (1995). A new mobile genetic element in Lactobacillus delbrueckii subsp. bulgaricus. Mol Gen Genet 248, 407-416.[Medline]

Gilbert, C., Atlan, D., Blanc, B., Portalier, R., Germond, J. E., Lapierre, L. & Mollet, B. (1996). A new cell surface proteinase: sequencing and analysis of the prtB gene from Lactobacillus delbrueckii subsp. bulgaricus. J Bacteriol 178, 3059-3065.[Abstract]

Guggenheim, S. (1971). ß-Cystathionase (Salmonella). Methods Enzymol 17B, 439-442.

Kilpper-Balz, R., Fisher, G. & Schleifer, K. (1982). Nucleic acid hybridation of group N and D streptococci. Curr Microbiol 7, 245-250.

Matern, H. T., Klein, J. R., Henrich, B. & Plapp, R. (1994). Determination and comparison of Lactobacillus delbrueckii ssp. lactis DSM7290 promoter sequences. FEMS Microbiol Lett 122, 121-128.[Medline]

Miller, J. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Parliment, T. H., Kolor, M. G. & Rizzo, D. J. (1982). Volatile components of Limburger cheese. J Agric Food Chem 30, 1006-1008.

Poolman, B. & Konings, W. N. (1998). Relation of growth of Streptococcus lactis and Streptococcus cremoris to amino acid transport. J Bacteriol 170, 700-707.

Rijnen, L., Bonneau, S. & Yvon, M. (1999). Genetic characterization of the major lactococcal aromatic aminotransferase and its involvement in conversion of amino acids to aroma compounds. Appl Environ Microbiol 65, 4873-4880.[Abstract/Free Full Text]

Singer, M., Baker, T. A., Schnitzler, G. & 7 other authors (1989). A collection of strains containing genetically linked alternating antibiotic resistance elements for genetic mapping of Escherichia coli. Microbiol Rev 53, 1–24.

Smacchi, E. & Gobetti, M. (1998). Purification and characterization of cystathionine {gamma}-lyase from Lactobacillus fermentum DT 41. FEMS Microbiol Lett 166, 197-202.[Medline]

Southern, E. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98, 503-517.[Medline]

Trach, K. A. & Hoch, J. A. (1993). Multisensory activation of the phosphorelay initiating sporulation in Bacillus subtilis: identification and sequence of the protein kinase of the alternate pathway. Mol Microbiol 8, 69-79.[Medline]

Urbach, G. (1993). Relations between cheese flavour and chemical composition. Int Dairy J 3, 389-422.

Uren, J. R. (1987). Cystathionine ß-lyase from Escherichia coli. Methods Enzymol 143, 483-486.[Medline]

Weimer, B., Seefeldt, K. & Dias, B. (1999). Sulfur metabolism in bacteria associated with cheese. Antonie Leeuwenhoek 76, 247-261.[Medline]

Yvon, M., Thirouin, S., Rijnen, L., Fromentier, D. & Grippon, J.-C. (1997). An aminotransferase from Lactococcus lactis initiates conversion of amino acids to cheese flavor compounds. Appl Environ Microbiol 63, 414-419.[Abstract]

Zdych, E., Peist, R., Reidl, J. & Boos, W. (1995). MalY of Escherichia coli is an enzyme with the activity of a ßC-S lyase (cystathionase). J Bacteriol 177, 5035-5039.[Abstract]

Received 20 December 2001; revised 15 March 2002; accepted 8 April 2002.