Cloning and Characterization of Two Lactobacillus casei Genes Encoding a Cystathionine Lyase^▿

Bacterial strains and growth conditions.

L. casei strains from the Agroscope Liebefeld-Posieux collection were used in this study. They were cultured anaerobically in 10 ml of MRS (12) at 30°C.

Methionine and cysteine auxotrophy.

Strains of interest were assayed for growth in a synthetic medium described by Morishita et al. (25). The medium contained l-methionine (0.1 mg ml⁻¹), l-cysteine (0.1 mg ml⁻¹), both l-methionine and l-cysteine, or neither one. Ten-milliliter volumes of the medium were inoculated with 50-μl cultures grown in MRS. The optical density at 600 nm was determined after 3 days of incubation at 30°C.

Preparation of cell extracts.

Strains were grown for 16 h in 50 ml MRS at 30°C. Bacterial cells were harvested by centrifugation, washed twice with 20 mM sodium phosphate (pH 7.4), and suspended in 500 μl of 20 mM sodium phosphate (pH 7.4). Approximately 0.4 g of glass beads (212 to 300 μm) was added. The cells were then disrupted by violent agitation in a Mini-Beadbeater-8 (Biospec Products, Inc.). After the extract had been cleared by centrifugation, metabolites were removed by using NAP columns (GE HealthCare, Uppsala, Sweden) which had been equilibrated with 20 mM sodium phosphate (pH 7.4) according to the manufacturer's instructions.

Protein determination.

Protein concentrations were determined by Bio-Rad protein assay according to the manufacturer's instructions. Bovine serum albumin was used to obtain a standard curve.

Cystathionine lyase assay.

Cystathionine lyase activity was assayed by measuring the formation of free thiol groups by spontaneous disulfide interchange with 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB) similar to the method of Uren (32). The reaction buffer contained a 200-μl solution of 50 mM buffer, 2 mM l-cystathionine, 200 μM DTNB, 20 μM pyridoxal-5′-phosphate, and enzyme solution. Depending on the pH, 2-morpholinoethanesulfonic acid, sodium phosphate, potassium phosphate, or Tris was used as a buffer. The reaction was initiated by the addition of enzyme and carried out at 37°C. The yellow coloration that developed as a consequence of enzymatic activity was recorded at 10-min intervals at 412 nm for 90 min. For high-performance liquid chromatography (HPLC) analysis, assays were performed without DTNB and stopped after 1 h of incubation at 37°C by the addition of 2 μl of 380 mM H₂SO₄ and by heating the mixture at 95°C for 5 min.

HPLC analysis.

Samples were filtered through disposable filter holders (0.45-μm pore size). Separation of the filtrate was carried out on a ion-exchange column (300 by 7.8 mm; HPX-87H Aminex; Bio-Rad) protected with a cation H+ Microguard cartridge. The operating conditions were as follows: a flow rate of 0.6 ml min⁻¹, 65°C, and detection at 210 nm. The mobile phase was 3.8 mM H₂SO₄. Pyruvate and α-ketobutyrate (both from Sigma-Aldrich) were used as standards.

DNA techniques.

Plasmid preparations, transformation, use of restriction enzymes, and ligation were carried out as described by Sambrook et al. (28) The putative malY and metC genes were amplified from the genomic DNAs of L. casei FAM18149 and FAM18168 by PCR. The primers used for the malY gene were 5′-CGGCAGCGCTAGCATGGAAGAATTGCCAGCTGATG-3′ and 5′-CCTAAGCTTAAAGTTGCGGCTGAGCCAAGT, and those used for the metC gene were 5′-CGGCAGCGCTAGCATGACCCAATTCAATACCAAACTAG-3′ and CCTAAGCTTGTCTTTATTTCTCCGCAGCCGTCAA-3′. Restriction sites are underlined. The PCR products were digested with NheI and HindIII. The pET SUMO vector (Invitrogen) was also digested with NheI and HindIII. Thus, the region encoding the SUMO fusion protein was removed from the plasmid and the PCR fragments could be cloned in the frame directly downstream of the polyhistidine region.

Sequencing was performed with the BigDye Terminator cycle sequencing kit (Applied Biosystems) and analyzed in an ABI Prism 310 genetic analyzer (Applied Biosystems).

Expression and purification of His-tagged proteins.

Recombinant E. coli BL21 was grown in 100 ml of Luria-Bertani broth containing kanamycin at a final concentration of 50 μg ml⁻¹ at 37°C on a shaker. When the optical density of the bacterial cells at 600 nm reached 0.5, induction of genes under the control of isopropyl-β-D-thiogalactopyranoside (IPTG)-inducible promoters was performed by adding 0.5 mM IPTG and lowering the incubation temperature to 27°C. After 4 h of incubation on a shaker, the bacterial cells were harvested by centrifugation, washed twice with 20 mM sodium phosphate (pH 7.4), and frozen at −20°C.

Cells were suspended in 500 μl binding buffer (20 mM sodium phosphate [pH 7.4], 150 mM NaCl, 20 mM imidazole) and disrupted as described above. The extract was cleared by centrifugation and then applied to a 1-ml HiTrap chelating HP column (GE HealthCare, Uppsala, Sweden) which had been loaded with Ni²⁺ and equilibrated with binding buffer. After the column had been washed with 8 ml binding buffer, bound proteins were eluted with 20 mM sodium phosphate (pH 7.4)-500 mM NaCl-250 mM imidazole. Afterwards, the imidazole was removed by applying the eluate to NAP columns (GE HealthCare, Uppsala, Sweden) which had been equilibrated with 20 mM sodium phosphate (pH 7.4) and 150 mM NaCl. The eluted protein fraction was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Proteins were visualized by colloidal Coomassie staining.

Detection of VSC by gas chromatography (GC)-pulsed flame photometric detection (PFPD).

The enzymatic assays were performed at a 1-ml scale as described in the cystathionine lyase assay section except that, instead of l-cystathionine, either 10 mM l-methionine or 10 mM l-cysteine was used as the substrate and incubated for 6 h.

VSC were detected as described previously (26). Due to the use of a different matrix, the following modifications were applied. The samples (1 ml) were cooled at 4°C and then transferred to a 10-ml vial with a magnetic screw cap, which had been flushed with argon. After preheating the sample (45 min, 60°C), 1 ml of the headspace was injected into the cooled injection system (CIS-4; Gerstel GmbH). The cooled injection system inlet was packed with deactivated wool (Restek GmbH) and set at −100°C. The temperature was programmed to 180°C at a rate of 12°C/s (hold time, 8 min). The temperature was held at 40°C for 1 min before increasing again. Transfer to the capillary column was performed in split mode (10:1). GC was conducted with an SPB-1 sulfur capillary column (30 m by 0.32 mm [inside diameter], 4-μm film thickness; Supelco) with helium as the carrier gas. The oven temperature was programmed from 29°C (hold time, 7 min) to 180°C at a rate of 10°C min⁻¹ (hold time, 10.5 min). Detection was performed with a PFPD model 5380 (Ol Analytical).

Nucleotide sequence accession numbers.

The nucleotide and amino acid sequences reported in this paper appear under NCBI accession numbers EF531299 and EF531300.

Cystathionine lyase activity in L. casei.

We assayed the cell lysates of 39 L. casei genotypes for cystathionine lyase activity with different buffers and different pHs. First, DTNB was used as a reagent to detect the formation of thiols during the incubation of cell extracts with l-cystathionine. We found that cystathionine lyase activity was dependent on the strain, pH, and buffer used. Only the extracts of 23 genotypes released thiols which reacted with DTNB. Considerable cystathionine lyase activity was only observed at pH 5.5. In 2-morpholinoethanesulfonic acid buffer, the activity of some genotypes was slightly reduced compared to that in phosphate buffer. At pH 6.8, the activity was greatly reduced (Fig. 1), and no activity was detected in Tris (pH 9.0) buffer (data not shown).

In order to determine whether an α,β- or an α,γ-elimination took place, enzyme reactions were performed in the absence of DTNB and then analyzed by HPLC for the presence of α-keto acids. The analysis showed that the genotypes with cystathionine lyase activity in the assay with DTNB released pyruvate and α-ketobutyrate from cystathionine (Fig. 2). Interestingly, among the apparently inactive genotypes, strains were identified which released pyruvate but not α-ketobutyrate from cystathionine. Reactions carried out at pH 6.8 contained considerably less pyruvate or α-ketobutyrate than those carried out at pH 5.5 (data not shown). Thus, the strains could be classified into three groups, (i) strains with no detectable activity, (ii) strains with low activity producing detectable amounts of pyruvate, and (iii) strains with high activity producing pyruvate and α-ketobutyrate (Table 1). From this observation, we concluded that at least two enzymes are present in L. casei which catalyze the degradation of l-cystathionine.

Methionine auxotrophy.

All of the lactobacilli tested were able to grow in the absence of methionine, but several genotypes showed no growth in the absence of cysteine (Table 1). All strains that were able to grow on methionine as the sole source of sulfur degraded cystathionine to pyruvate and α-ketobutyrate.

Cloning and overexpression of two genes encoding cystathionine lyases.

Two enzymes in E. coli are known to cleave cystathionine by α,β-elimination (16, 35). The proteins are encoded by the genes metC and malY, respectively. We used both amino acid sequences (Swissprot accession no. P06721 and P23256) to search for homologs in the draft genome sequence of L. casei ATCC 334 (genome.jgi-psf.org/mic_home.html). The MetC sequence showed homology to the product of gene 3332 of scaffold 58; MalY was shown to be homologous to the product of gene 1334 of scaffold 16. After the genome sequence had been completed and moved to the NCBI database, it was confirmed that L. casei ATCC 334 contained genes homologous to malY (LSEI_0894) and metC (LSEI_0600) (22).

We cloned both genes from L. casei FAM18149 and FAM18168. The amino acid sequence deduced from LSEI_0894 of FAM18168 was identical to that of FAM18149 and showed 99% identity with the sequence of ATCC 334. Furthermore, the amino acid sequences from LSEI_0600 of FAM18168 and FAM18149 were identical and were 99% identical to that of ATCC 334. A search for protein homology revealed the similarity of LSEI_0894 to PatB (34%) of B. subtilis, PatC (33%) of L. delbrueckii, Lcd (31%) of S. anginosus, YtjE (31%) of L. lactis, and MalY (29%) of E. coli. LSEI_0600 showed similarity to MetC of L. lactis (46%) and L. helveticus (44%), methionine γ-lyase (35%) of B. linens, and MetB (33%) and MetC (25%) of E. coli. For clarity, the gene names malY for LSEI_0894 and metC for LSEI_0600 are used in this report.

In order to produce large amounts of protein, the genes were cloned in frame with a His₆ tag coding sequence (pET-SUMO). The region encoding the SUMO protein was omitted since we considered that a large, prolonged N terminus would impair the quaternary structure of the enzymes. Due to the construct, the N terminus of each protein was nevertheless prolonged by 21 amino acid residues (MGSSHHHHHHGSGLVPRGSAS). After purification by affinity chromatography, the apparent molecular mass of the His-tagged MalY protein was 47 kDa (theoretical molecular mass of 46.9 kDa) and that of MetC was 43 kDa (theoretically, 42.7 kDa).

Cystathionine lyase activity was determined at different pHs. We found that the MalY protein showed the highest activity at pH 6.0, whereas the MetC protein had an optimum pH of 9.0. At pH 5.5, a pH usually found in 4-month-old Gruyère cheese, the residual activities of MalY and MetC were 55 and 15%, respectively. To ensure that a cystathionine lyase of E. coli was not copurified by affinity chromatography, the SUMO-chloramphenicol acetyltransferase fusion protein encoded by the unmodified plasmid was isolated under the same conditions. No cystathionine lyase activity was detected in the purified SUMO-chloramphenicol acetyltransferase preparation (data not shown).

After incubation at 56°C for 30 min, the enzyme preparations were no longer active. After several weeks of storage at 4°C or −20°C, both enzyme preparations still showed considerable cystathionine lyase activity. The addition of 4% NaCl completely inhibited the activity of MetC but not that of MalY.

The formation of α-keto acids from l-cystathionine was determined by HPLC (Fig. 3 and 4). The results showed that the product of the malY gene degraded cystathionine to pyruvate, whereas the product of the metC gene released pyruvate and α-ketobutyrate. At pH 9.0, MetC produced more α-ketobutyrate than pyruvate.

To determine substrate specificity, assay mixtures containing l-cysteine, l-methionine, dl-homocysteine, l-serine, O-succinyl-l-homoserine, l-leucine, l-isoleucine, l-valine, l-phenylalanine, l-aspartate, l-glutamate, and l-lysine were prepared. It was found that MalY formed pyruvate from l-cysteine (Fig. 3B) and l-serine at pH 5.5 and pH 6.8. The latter reaction was clearly observed after a prolonged incubation time (data not shown). Reaction vials containing cysteine as a substrate emitted a strong hydrogen sulfide odor after incubation.

The MetC protein formed α-ketobutyrate and succinate from O-succinyl-l-homoserine (Fig. 4C) at pH 5.5, pH 6.8, and pH 9.0. After incubating O-succinyl-l-homoserine and l-cysteine together with MetC, the formation of α-ketobutyrate was no longer observed (Fig. 4D). This indicated that the α-ketobutyrate moiety was used for the synthesis of l-cystathionine. Reaction vials containing MetC and methionine had a cabbage-like odor after a prolonged incubation, which indicated the release of methanethiol. However, the formation of α-ketobutyrate was only detected by HPLC analysis after 20 h of incubation at pH 9.0 but not at pH 6.8. When DTNB was added, the solution turned yellow, indicating the presence of a thiol-containing group. At pH 9.0, a more intense color formation was observed than at pH 6.8. Control experiments with methionine alone or with MalY and methionine did not turn yellow (data not shown). We concluded that MetC was able to liberate methanethiol from methionine.

In order to confirm the release of methanethiol from methionine, we used GC coupled to flame photometric detection, which is selective for phosphorus and sulfur. The headspace of an enzyme assay carried out with methionine in the absence of enzyme contained very small amounts of methanethiol, dimethyl disulfide, and DMTS. Whereas the addition of the MalY protein had no effect on the content of VSC, the MetC protein clearly released methanethiol, dimethyl disulfide, and DMTS from methionine (Fig. 5). Furthermore, it was confirmed that MalY released hydrogen sulfide from l-cysteine, as mentioned above (data not shown).