Sequencing of the Tyrosine Decarboxylase

Home , Decarboxylation, Tyrosine decarboxylase

2521

Journal of Food Protection, Vol. 67, No. 11, 2004, Pages 2521±2529 Copyright ᮊ, International Association for Food Protection

Sequencing of the Tyrosine Decarboxylase Cluster of Lactococcus lactis IPLA 655 and the Development of a PCR Method for Detecting Tyrosine Decarboxylating Lactic Acid Bacteria

MARIÂA FERNAÂ NDEZ, DANIEL M. LINARES, AND MIGUEL A. ALVAREZ*

Instituto de Productos LaÂcteos de Asturias, Carretera de In®esto s/n, 33300 Villaviciosa, Asturias, Spain Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021

MS 04-109: Received 12 March 2004/Accepted 9 June 2004

ABSTRACT

The enzymatic decarboxylation of tyrosine produces tyramine, the most abundant biogenic amine in dairy productsÐ especially in cheeses. The screening of lactic acid bacteria isolated from different artisanal cheeses and a number of microbial collections identi®ed 22 tyramine-producing strains belonging to different genera. The Lactococcus lactis strain IPLA 655 was selected, and the genes encoding a putative tyrosyl tRNA synthetase, a tyrosine decarboxylase (tdcA), and a tyrosine-tyramine antiporter, found together as a cluster, were sequenced. The disruption of tdcA yielded a strain unable to produce tyramine. Comparison of the L. lactis IPLA 655 tdcA gene with database tdcA sequences led to the design of two primers for use in a PCR method that identi®ed potential tyramine-producing strains. The proposed method can use puri®ed DNA, isolated colonies, milk, curd, and even cheese as a template. Molecular tools for the rapid detection of tyramine-producing bacteria at any time during the fermentation process could help prevent tyramine accumulation in fermented foods. The proposed technique could be of great use to the food industry.

Lactic acid bacteria (LAB) are widely used in a variety tures. It is therefore useful to determine which strains pro- of food fermentation processes. These gram-positive bac- duce undesirable compounds so that these are not part of teria play an important role in the quality of the ®nal prod- starter cultures. ucts because they contribute to ¯avor formation, texture Histamine and tyramine are the most studied of the BA development, and biopreservation (the prevention of spoil- because they have toxicological effects that derive from age by undesirable and pathogenic microorganisms). Some their vasoactive and psychoactive properties (44). These strains even have probiotic properties. LAB metabolism compounds are formed from histidine and tyrosine, respec- strongly in¯uences the properties of fermented foods. Ami- tively, via an enzymatic decarboxylation reaction. The his- no acid catabolism is particularly important because it helps tidine decarboxylase genes (hdcA) of different gram-posi- produce desired ¯avor molecules (methylbutanal, isovaler- tive bacterial strainsÐLactobacillus sp. strain 30a (10, 53), ate, etc.), but it can also produce off-¯avor molecules (phe- Lactobacillus buchneri (26), Clostridium perfringens (26, nol, p-cresol, etc.) and even toxic compounds such as bio- 51), Micrococcus sp. (52), and Leuconostoc oenos (11)Ð genic amines (BA). BA are organic bases of low molecular have been characterized. The biochemical properties of his- mass that are formed and degraded during the normal me- tidine decarboxylase from Lactobacillus 30a have been tabolism of microorganisms, plants, and animals. They are studied extensively by van Poejie and Snell (51) and Schelp necessary for several physiological functions in humans. et al. (43). However, although tyrosine decarboxylase However, they are toxic if high concentrations are ingested (TDC) enzymes have been well characterized in eukary- or if the detoxi®cation process is inhibited (either geneti- otes, little is known about their prokaryotic counterparts. cally or by drugs) (2). Foods likely to contain high levels The puri®cation and characterization of this enzyme has of BA include ®shery products and fermented foods such been reported only for Enterococcus faecalis (4) and Lac- as cheese, wine, beer, and cured sausages (48). tobacillus brevis IOEB 9809 (33, 37). In addition, the E. BA form via the decarboxylation of their correspond- faecalis JH2-2 gene encoding the enzyme tyrosine decar- ing amino acids through the action of enzymes produced boxylase (tdcA) has been cloned (9), and the sequence for by microorganisms present in the food (48) (e.g., Entero- L. brevis IOEB 9809 is available (32, 33). In both strains, bacteriae, Pseudomonas spp., enterococci, and some LAB the tdcA gene is part of a cluster with a gene encoding a (21)). Nonetheless, production is more related to strain than putative tyrosine-tyramine antiporter (tyrP) and a tyrosyl species. Producer strains can appear as contaminants of fermented foods, but they can also form part of starter cul- tRNA synthetase gene (tyrS). Lactococcus lactis is the most widely used starter cul- * Author for correspondence. Tel: ϩ34985893352; Fax: ϩ34985892233; ture in the manufacture of cheese, other dairy products, and E-mail: [email protected]. fermented foods. Although some strains have been identi- 2522 FERNAÂ NDEZ ET AL. J. Food Prot., Vol. 67, No. 11

TABLE 1. Strains and plasmids used in this study DNA isolation and manipulation. Escherichia coli DH5␣ was used as an intermediate host. Cloning was performed accord- Strain/plasmid Relevant characteristica Source ing to standard procedures (42). The isolation of E. coli plasmid Strain DNA and the recombinant techniques used were those described E. coli by Sambrook et al. (42). Large-scale isolation of E. coli plasmids for nucleotide sequence analysis was performed with the Plasmid INVaFЈ Invitrogen Midi Kit (Qiagen, Hilden, Germany) according to the manufac- DH5␣ (22) turer's instructions. Plasmid and chromosomal DNA of L. lactis L. lactis were isolated and transformed as described previously (15). IPLA 655 Isolated from artisanal chees- (13) Southern blot hybridization was performed at 65ЊC, and blots es. IPLA collection. were washed with 0.1ϫ SSC (i.e., 0.15 M NaCl ϩ 0.0015 M Plasmid sodium citrate) before autoradiography. The probes used for hybridization were radiolabeled with [␣-32P]dATP by nick transla- pCR2.1 Ampr Invitrogen tion. pUC18Ery Eryr, 3.8-kb pUC19 carrying Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021 the ery gene of plL253 (51) Construction of plasmids. An internal fragment of the tdc 4 pM20 Amp , 4.7-kb derivative of gene (820 bp) from L. lactis IPLA 655 was obtained directly from pCR2.1 containing a 0.8-kb genomic DNA by PCR ampli®cation with the use of the degen- fragment of tdcA from erate oligonucleotides P2 (5Ј-GAYATIATIGGIATIGGIYTIGAY L. lactis IPLA 655 This work CARG-3Ј), and P1 (5Ј-CCRTARTCIGGIATIGCRAARTCIGTRTG r pM24 Ery , 5.8-kb derivative of 3Ј), where Y is C or T, R is A or G, and I is inosine (Table 2). pM20 containing the ery The design of these oligonucleotides was based on the sequence gene of pUC18Ery This work of tdcA from L. brevis IOEB 9809 (33). The PCR product obtained was cloned in pCR2.1 (Invitrogen, Carlsbad, Calif.), gen- a Ampr, ampicillin resistant; Eryr, erythromycin resistant. erating the plasmid pM20. The erythromycin resistance gene (ery) from pUC18Ery (50) was cloned as a BamHI-HindIII fragment into pM20, previously digested with the same enzymes, to yield ®ed as tyramine producers (20), the tdcA gene has not been pM24. This was used to construct a knock-out strain. characterized. Nucleotide sequence analysis. Nucleotide sequences were Although several qualitative and quantitative methods analyzed with an ABI Prism 373 A Strech automated sequencer have been developed to determine BA production, consum- at the DNA Sequencing Department (Centro de Investigaciones er demand for better and healthier foods has increased in- BioloÂgicas, Consejo Superior de Investigaciones Cienti®cas). Se- terest in the development of rapid and sensitive methods quences were assembled and analyzed with the CLONE program for detecting BA producers in food. Molecular methods (v. 5) and compared with those in the SwissProt library (February have been used for the detection of gram-positive hista- 2004 release) with the TFASTA program. The Clustal method (25) mine-producing strains (31) and, more recently, for gram- was used for the multiple alignment of sequences. negative strains (47). This paper reports the sequencing of the tdc cluster of Oligonucleotide design and PCR reaction conditions. Primers Tdc1 and Tdc2 (Table 2) were designed by comparing genes from L. lactis IPLA 655 (13), a strain isolated from the tdcA sequences of the gram-positive bacteria E. faecalis JH2- artisanal cheese. Besides tdcA, the cluster includes a puta- 2 (9), L. brevis IOEB 9809 (33), and L. lactis IPLA 655 with the tive tyrosine-tyramine antiporter gene and an aminoacyl Clustal program. DNA was ampli®ed in an iCycler thermal cycler tRNA synthetase gene. Disruption of the tdcA gene yielded (Bio-Rad, Hercules, Calif.). The PCR conditions involved an ini- a strain unable to produce tyramine. The comparison of the tial denaturation step (95ЊC for 5 min), 35 ampli®cation cycles cluster sequence with those held in databases led to the (95ЊC for 45 s, 50ЊC for 1 min, and 72ЊC for 1 min), and a ®nal design of speci®c oligonucleotides that ampli®ed an inter- extension step at 72ЊC for 7 min. All ampli®cations were per- nal fragment of the tdcA gene of different LAB genera. To formed with the use of puRe Taq Ready-To-Go PCR beads fol- our knowledge, this is the ®rst PCR method to detect ty- lowing the manufacturer's instructions (Amersham-Biosciences, ramine-producing bacteria in foods. Buckinghamshire, UK) and 200 nM of oligonucleotides as primers. MATERIAL AND METHODS When DNA was used as a template, 10 ng was diluted in 25 ␮l of MilliQ water and used directly in the reaction. When col- Bacterial strains and media. Table 1 shows the strains and onies were used as templates, cells were resuspended in 30 ␮lof plasmids used in this study. Lactococcus and Enterococcus were MilliQ water and heated at 98ЊC for 5 min. Of this mixture, 25 routinely grown at 30ЊC in media based on M17 medium (Oxoid, ␮l was used for PCR as above. To check the detection limit of Basingstoke, Hampshire, UK) supplemented with 0.5% glucose. the PCR method, milk was inoculated with serial dilutions of an Lactobacillus strains were grown at 37ЊC in deMan Rogosa overnight culture of the producer strain L. lactis IPLA 655. Sam- Sharpe (Oxoid). In some experiments, L. lactis IPLA 655 was ples (2-␮l) were used directly in the PCR reaction. To use curd grown in skim milk (Oxoid). When appropriate, the media con- as template, 5 ml of milk, previously inoculated with 100 ␮lof tained tyrosine (2 g/liter). Bover-Cid decarboxylation broth (5), an overnight culture of L. lactis IPLA 655, was allowed to ferment containing tyrosine as the amino acid precursor and a pH indicator, for 24 h. When the milk was coagulated (after ϳ8 h of incuba- was used to screen for tyrosine decarboxylase strains. When nec- tion), 2 ␮l of curd (with no previous treatment) was used directly essary, ampicillin (50 ␮g/ml) or erythromycin (5 ␮g/ml) was add- in PCR. For the detection of tyramine-producing LAB in cheese ed to the media. samples,5gofsample was removed, placed in a stomacher bag J. Food Prot., Vol. 67, No. 11 L. LACTIS TDC CLUSTER SEQUENCING AND TYRAMINE PCR DETECTION METHOD 2523

TABLE 2. Oligonucletoides used in this study Oligo- nucleotide Sequence 5Ј±3Ј Position

P1 GAYATIATIGGIATIGGIYTIGAUCARG P2 CCRTARTCIGGIATIGCRAARTCIGTRTG Tdc1 AACTATCGTATGGATATCAACG 3,791±3,813 Tdc2 TAGTCAACCATATTGAAATCTGG 4,533±4,510 Tdc3 AACTACACCGACAACACCC 3,880±3,861 Tdc4 TAAAGAAATCGAAGTACACC 4,465±4,485 Tdc5 GTAAAATGGGTCTAGGGACC 5,210±5,230 Tdc7 GCACGAGAAGGATTCTTACC 6,136±6,156 Tdc8 GCATGTCCTGGGGCATGTAG 3,149±3,129

Tdc9 CACTACATCACGATCATCTTGG 2,245±2,223 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021 Tdc10 TTCCGTACCATGGCATAATG 3,250±3,270 Tdc11 TCAATTACAGATCGGTGGGG 2,008±2,029 Tdc12 TATATGGATGATGGCCCGGGAATTG 1,658±1,633 Tdc18 ATATTCATTTTAATGCCCTCCG 1,424±1,446 Tdc19 CGTTTAACCGTGAGGCTGGG 6,632±6,651 Tdc20 CATCACGATTGCAATGATCGCAAC 6,531±6,508 Tdc21 GCGGTGTTGATCCAACTGGAGATTC 1,553±1,587 Tdc23 TTGGCTGTCACTTCTGTATGCACTG 805±781

containing 40 ml of 2% sodium citrate solution, and homogenized RESULTS with a stomacher (Lab-Blender 400, Seward, London, UK) for 1 Identi®cation of tyramine-producing strains. The min. DNA was extracted from the homogenate following the following bacteria were screened in Bover-Cid differential method described by Ogier et al. (40). Of this DNA, 20 ng was medium to identify tyramine producers: 100 L. lactis, 15 resuspended in 25 ␮l of MilliQ water and used for PCR reactions. Enterococcus spp., and 74 Lactobacillus spp. strains from Reverse PCR. A reverse PCR strategy was used to obtain the artisanal cheeses; L. brevis 3810, L. brevis 3824, and the complete tdc cluster sequence. The template used was total L. viridescens 283 from the CECT collection (ColeccioÂn DNA from L. lactis IPLA 655 digested with different restriction EspanÄola de Cultivos Tipo); L. bulgaricus 11842 from the enzymes (EcoRI, HindIII, PstI, and XbaI) and ligated under di- American Type Culture Collection; E. faecalis 1535 from luted conditions (5 ␮g/ml) to favor intramolecular annealing. The the CNRZ collection (UniteÁ de Recherches LaitieÁres, positions and sequences of the primers, deduced from previously France); L. buchneri 301, 302, 303, and 304 from the NIZO known sequences, are listed in Table 2. PCR was performed by collection (Netherlands Institute of Dairy Research); L. lac- the Expand Long Template PCR system (Roche, Mannheim, Ger- tis MG1363 (19); and L. lactis IL1403 (8). Twenty-three many) according to the manufacturer's instructions, with 2 ␮lof colonies surrounded by a clear halo were selected. The abil- the ligations as templates and an initial denaturation step (94ЊC ity of these strains to produce tyramine was checked by Њ Њ for 2 min), 35 ampli®cation cycles (94 C for 30 s, 56 C for 30 s, HPLC. Twenty-two tyramine-producing strains were con- Њ Њ and 68 C for 3 min), and a ®nal extension step (68 C for 7 min). ®rmed (L. lactis IPLA 655, 755, 855, 983, 984, 1567, and The products generated were sequenced directly. 1584; E. faecalis 148T and CNRZ 1535; E. durans L21, L37, L70, L72, 181T, 1AA46, and 1AA57; E. faecium Tyramine determination. Quantitative analysis of tyramine production was undertaken by reversed-phase high-performance 1AA62; Enterococcus 2BA62 and 2BA64; L. brevis 3810; liquid chromatography (RP-HPLC) with a Waters liquid chro- and L. curvatus VI6 and VI14). L. lactis IPLA 655 was matograph controlled by Millenium 32 software (Waters, Milford, selected for further studies. Mass.). Tyramine-producing strains and nonproducing strains Sequencing of the tyrosine decarboxylase cluster of were grown in M17 medium supplemented with tyrosine (2 g/ L. lactis IPLA 655. Two degenerate primers (P1 and P2) ϫ liter) for 24 h. The cultures were centrifuged at 8,000 g for 10 were designed on the basis of the partial sequence of the min, and the resulting supernatants were ®ltered through a 0.2- tyrosine decarboxylase protein from L. brevis IOEB 9809 ␮m Supor membrane (Pall, Ann Arbor, Mich.). The resulting sam- (33). These were used to test L. lactis IPLA 655 by PCR ples were derivatized with the use of dabsyl chloride. Separations for the presence of the tdcA gene (Table 2). An 820-bp were performed with Waters Nova-pak C column. The gradient 18 amplicon was obtained, which was cloned into pCR2.1 to and detection conditions were similar to those described by Krause et al. (30). yield the plasmid pM20. Sequence analysis of the pM20 insert revealed an open reading frame (ORF) that showed Nucleotide sequence accession number. The nucleotide se- strong similarity to the tyrosine decarboxylase genes of E. quences recorded in this study were submitted to the Experimental faecalis JH2-2 (83% identity) (9) and L. brevis IOEB 9809 Marine Biology Laboratory Nucleotide Sequence Database under (73% identity) (33) (Fig. 1). This gene was designated tdcA. the accession number AJ630043. With reverse PCR, the L. lactis genome was walked on 2524 FERNAÂ NDEZ ET AL. J. Food Prot., Vol. 67, No. 11

FIGURE 1. Deduced amino acid sequence for the internal fragment of the tdcA gene in E. faecalis JH2-2 (9), L. brevis IOEB 9809 (32), and L. lactis IPLA 655. The boxes represent the VHVDAAY motif and the pyridoxal phosphate attachment site. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021

either side of this fragment. Upstream of tdcA, a complete tdcA, another ORF (tyrP) was found that shared 84 and ORF was found that shared 78 and 72% identity, respec- 66% identity, respectively, with the antiporters of E. fae- tively, with the tyrS of E. faecalis JH2-2 and L. brevis calis JH2-2 and L. brevis IOEB 9809. The deduced amino IOEB 9809. The analysis of its sequence revealed that the acid sequence of this protein (determined with the Sosui encoded tyrosyl tRNA synthetase (TyrS) belongs to the program, Tokyo University of Agriculture & Technology, class I aminoacyl tRNA synthetases characterized by HIGH Japan) revealed a membrane protein structure with 11 trans- and KMSKS motifs (16). The HIGH motif is perfectly con- membrane helices. This protein might be involved in ty- served in L. lactis IPLA 655 TyrS, and the KMSKS motif rosine-tyramine exchange. is represented by the KFGKT sequence, as in E. coli (1), Disruption of the tdcA gene. For the construction of Bacillus subtilis (23), and E. faecalis (9). Downstream of a tdcA knock-out strain, the erythromycin resistance gene (ery) from pUC18Ery (Table 1) was cloned in pM20 (a plasmid containing an internal fragment of the tdcA gene from L. lactis IPLA 655), generating plasmid pM24. This plasmid, which is unable to replicate in gram-positive bacteria, was introduced by electroporation into L. lactis IPLA 655. Candidate single-crossover mutants at the tdcA locus were isolated as erythromycin-resistant colonies. Plasmid integration that disrupted tdcA was con®rmed by Southern blot and PCR analyses (data not shown). One colony was used for further analysis and designated L. lactis IPLA 655 tdcA. The ability of this strain to decarboxylate tyrosine was compared with that of the wild-type strain with the use of Bover-Cid medium supplemented with this amino acid (5). No tyrosine decarboxylation halo was seen for the knock- out strain. To determine whether it could produce tyramine, the supernatants of cultures grown for 24 h in M17 with tyrosine were analyzed by HPLC. Tyramine was not detected (Fig. 2). This suggests that the disruption of the tdcA gene incapacitates tyramine production. The growth of L. lactis IPLA 655 and the knock-out strain followed similar patterns, although a slight difference in the ®nal pH of the medium was observed when tyrosine was present. The knock-out strain produced a lower pH than the wild type (Fig. 3), suggesting that tyrosine decarboxylase might play a role in pH control. More experiments are required to investigate the potential role of the tdcA gene in the control of intracellular pH. FIGURE 2. Chromatographic analysis of supernatants from (1) L. lactis IPLA 655 and (2) L. lactis IPLA 655 tdcA cultures. Ty- Design of speci®c oligonucleotides for detecting the ramine peak is indicated by a black arrow. tdcA gene by PCR. Primers TDC1 and TDC2 (Table 2) J. Food Prot., Vol. 67, No. 11 L. LACTIS TDC CLUSTER SEQUENCING AND TYRAMINE PCR DETECTION METHOD 2525

FIGURE 3. (A) Growth of L. lactis IPLA 655 (squares) and L. lactis IPLA 655 tdcA (circles) in M17 with tyrosine. (B) pH var- iation of the media. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021

were designed to detect tyramine-producing strains by PCR. technique for determining the presence of tyramine-produc- Reactions were initially performed with the use of total ing strains during food fermentation was then evaluated. DNA from 17 different LAB strains identi®ed as tyramine The ®rst assay tested the sensitivity of the detection method producers by Bover-Cid plates and HPLC. The analysis of in milk under laboratory conditions. Milk was inoculated the PCR products showed one single fragment of 720 bp. with serial dilutions of an overnight culture of the producer No PCR product was detected when DNA from 14 non- strain L. lactis IPLA 655. PCR products were obtained with producing strains was used as a template in a negative con- a detection limit of 104 CFU/ml. trol. PCR reactions were also performed with curd as the The deduced amino acid sequences of the PCR prod- template. Samples were taken at 0, 1, 2, 4, 6, 8, and 24 h ucts showed 70 to 90% similarity (Fig. 4). All contained after milk inoculation with L. lactis IPLA 655 and used for the active-site peptide from pyridoxal phosphate (27). Sim- PCR. Products of the expected size were obtained for all ilarity with other TDC sequences from the databases ranged the samples assayed, indicating that the proposed method from 70% (L. brevis IOEB 9809) to 80% (E. faecalis JH2- detects tyramine-producing strains in the early stages of the 2). fermentation process. The internal fragment of tdcA was successfully ampli- To assess whether this technique could detect the pres- ®ed when colonies were used directly as templates in the ence of tyramine-producing strains in commercial samples, PCR reaction. The 22 tyramine-producing strains identi®ed, tests were performed on a market cheese (90 days of rip- as well as 14 nonproducing strains belonging to different ening) in which the presence of tyramine had been con- genera, were grown on a plated medium, and an isolated ®rmed by HPLC (600 mg/kg). DNA was isolated from the colony of each was used in PCR. The 720-bp PCR fragment cheese for these assays as described in the section ``Oli- was detected for all the producer strains. No PCR product gonucleotide design and PCR reaction conditions.'' The was obtained with nonproducer strains. 720-bp PCR fragment was again obtained. These results indicate that this PCR method is a powerful, sensitive tool Detection of tyramine-producing strains in milk, that easily and rapidly detects tyramine-producing strains fermented milk, and cheese. The usefulness of the PCR in foods. 2526 FERNAÂ NDEZ ET AL. J. Food Prot., Vol. 67, No. 11

FIGURE 4. Alignment of partial tyrosine decarboxylase amino acid sequences of LAB tyramine producers. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021

DISCUSSION ramine. However, several factors are involved in tyramine Tyramine and histamine are the most common BA in production, including the growth stage of bacteria (18), pH cheese. Although a PCR method for the detection of his- (7), and the presence of tyrosine in the medium (38). There- tamine-producing strains was introduced by Le Jeune et al. fore, a method that could detect potential tyramine produc- (31), no equivalent method for the detection of tyramine- tion without having to consider the culture conditions producing strains was ever developed. The techniques cur- would be useful: the proposed PCR method appears to offer rently used for this are based on either growth in a differ- such an alternative. ential medium (5) or the chromatographic detection of ty- Comparison with information held in databases showed J. Food Prot., Vol. 67, No. 11 L. LACTIS TDC CLUSTER SEQUENCING AND TYRAMINE PCR DETECTION METHOD 2527 that the tyrosine decarboxylation clusters of L. lactis IPLA however, to con®rm the potential role of the tyrosine de- 655, E. faecalis JH2-2 (9), and L. brevis IOEB 9809 (32) carboxylase operon in acid tolerance. have the same genetic organization: a tyrS gene followed Another ORF, tyrP, was located downstream of tdcA. by tdcA and a tyrP gene. Its deduced amino acid sequence revealed a membrane pro- The L. lactis IPLA 655 tyrS gene shared 78% identity tein structure with 11 transmembrane helices. This protein with the E. faecalis JH2-2 tyrS gene, and 72% with that of could be involved in tyrosine-tyramine exchange. L. brevis IOEB 9809. E. faecalis V583, whose genome has Comparison of the tdcA gene from different tyramine- been fully sequenced (41), shows two tyrosine aminoacyl producing strains allowed the design of two oligonucleo- tRNA synthetase genes (tyrS-1 and tyrS-2). tyrS-1 is part tides that ampli®ed an internal fragment of the tdcA genes of a tyrosine decarboxylase cluster and shares 81.3% iden- from LAB by PCR. All the strains used in this study that tity with the L. lactis IPLA 655 tyrS gene, compared with gave a positive PCR result were able to produce tyramine. 51% with tyrS-2. In addition to their essential catalytic roles Comparison of the amplicon sequences con®rmed the spec-

in protein biosynthesis, aminoacyl tRNA synthetases par- i®city of the PCR method; all were very similar to that Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021 ticipate in other functions, such as the regulation of gene previously described for tdcA genes (Fig. 4). No ampli®- expression (45). The presence of a second tyrosyl tRNA cation product was obtained with any of the nonproducer synthetase gene (tyrS-2) in the genome of the tyramine- strains, suggesting that these have no tdc genes. In fact, the producing strain E. faecalis V583 suggests that tyrS-1 could negative LAB strains whose genomes have been sequenced play a role in the regulation of tyrosine decarboxylation. (4, 29) do not have these genes. Similar results have been Although further experiments are needed to investigate the obtained for the histidine decarboxylase and lysine decar- role of tyrosyl tRNA synthetase, this enzyme might sense boxylase genes (14, 47), which suggests that a lack of ty- the intracellular tyrosine pool. In enteric bacteria, the ex- rosine decarboxylase activity is more likely the result of a pression of biosynthetic histidine genes is regulated by his- large chromosomal deletion than a minor mutation or dis- tidine availability via histidyl tRNA synthetase (17). The ruption of the TDC gene, as demonstrated for lysine de- concentration of the substrate amino acid has to be a key carboxylase activity (14). The ability to produce tyramine regulating factor, something already described for histidine might also be explained by the horizontal transference of decarboxylase (10), glutamate decarboxylase (24, 54), and the tdc cluster. The similar genetic organization of different lysine decarboxylase (39). tdc clusters and strong sequence identity support this idea. A similar organization has also been described for oth- The proposed PCR method is, to our knowledge, the er decarboxylase operons, such as those of lysine decar- ®rst to detect whether a LAB strain can produce tyramine boxylase in E. coli (36), glutamate decarboxylase in E. coli (making its use as a food fermentation starter undesirable). (46) and Listeria monocytogenes (12), arginine decarbox- Because it is unnecessary to purify the DNA of the strains ylase in L. lactis IL1403 (3), and ornithine decarboxylase to be tested (an isolated colony can be used as a template), in E. coli (28). However, the presence of an aminoacyl the proposed method is fast and easy. Several authors in- tRNA synthetase gene as part of the cluster has only been dicated that LAB produce HPLC-detectable tyramine con- described for the arginine decarboxylase operon in L. lactis centrations when present in cheeses at densities of 107 IL1403 (3). CFU/g or greater (34), but this PCR method is more sen- Analysis of the deduced amino acid sequence of the L. sitive, allowing tyramine-producing strains to be detected lactis IPLA 655 tdcA gene revealed its strong similarity to at 104 CFU/ml. Current techniques based on tyramine de- those of the pyridoxal phosphate±dependent decarboxylas- tection require the presence of its precursor amino acid. In es, as well as the presence of the VHVDAAY motif. This milk, or in the early ripening stages of fermented milks, the similarity suggests that the protein belongs to the group II tyrosine concentration necessary to detect tyramine is not decarboxylases. The tdcA gene seems to be involved in ty- often reached. The proposed PCR technique, however, rosine decarboxylation because the knock-out strain, when avoids this problem and offers the possibility of detecting grown in the same conditions as the wild type, did not potential tyramine-producing strains in milk at any point in produce tyramine, although no differences were observed milk fermentation and in ®nal products such as cheese. This between the growth curves of both strains. However, the technique could be useful to the food industry. ®nal pH attained was a little lower in the knock-out culture than in the wild-type culture (Fig. 3). This difference might ACKNOWLEDGMENTS be explained by the pH stress protection afforded by de- We thank Victoria BascaraÂn and Teresa Delgado for their help with carboxylase activity. Amino acid decarboxylation, in com- the chromatography experiments. This work was supported by European bination with amino acid±BA antiport, has been proposed Union research grant no. QRLT-2001-02388. D. M. Linares is the recipient as a method of maintaining pH homeostasis. In E. coli and of a fellowship from the Spanish Ministry of Science and Technology. Shigella ¯exneri, the glutamate-␥-aminobutyrate antiporter, REFERENCES together with decarboxylase activity, confers acid resistance in media supplemented with glutamate (24, 54). In Strep- 1. Barker, D. G., C. J. Bruton, and G. Winter. 1982. The tyrosyl-tRNA tococcus thermophilus and L. lactis, glutamate decarbox- synthetase from Escherichia coli. Complete nucleotide sequence of the structutal gene. FEBS Lett. 150:419±423. ylase plays an important role in acid stress tolerance (49). 2. Bodmer, S., C. Imark, and M. KneubuÈhl. 1999. Biogenic amines in Similar effects have been described for the arginine de- foods: histamine and food processing. In¯am. Res. 48:296±300. aminase system (6, 35). Further experiments are necessary, 3. Bolotin, A., P. Wincker, S. Mauger, O. Jaillon, K. Malarme, J. Weis- 2528 FERNAÂ NDEZ ET AL. J. Food Prot., Vol. 67, No. 11

senbach, S. D. Ehrlich, and A. Sorokin. 2001. The complete genome 25. Higgins, D. G., and P. M. Sharp. 1989. Fast and sensitive multiple sequence of the lactic acid bacterium Lactococcus lactis ssp lactis sequence alignments on a microcomputer. Comput. Appl. Biosci. 5: IL 1403. Genome Res. 11:731±753. 151±153. 4. Bùrrensen, T., N. K. Klausen, L. M. Larsen, and H. Sùrensen. 1989. 26. Huynh, Q. K., and E. E. Snell. 1985. Pyruvoyl-dependent histidine Puri®cation and charcterization of tyrosine decarboxylase and aro- decarboxylase. Preparation and amino acid sequences of the chains matic-L-amino acid decarboxylase. Biochim. Biophys. Acta 993:108± of histidine decarboxylase from Clostridium perfringens and Lac- 115. tobacillus buchneri. J. Biol. Chem. 260:2798±2803. 5. Bover-Cid, S., and W. H. Holzapfel. 1999. Improved screening pro- 27. Jackson, F. R. 1990. Prokaryotic and eukaryotic pyridoxal dependent cedure for biogenic amine production by lactic acid bacteria. Int. J. decarboxylases are homologous. J. Mol. Evol. 31:325±329. Food Microbiol. 53:33±41. 28. Kashiwagi, K., S. Miyamoto, F. Suzuki, H. Kobayashi, and K. Igar- 6. Casiano-ColoÂn, A., and R. E. Marquis. 1988. Role of the arginine ashi. 1992. Excretion of putrescine by the putrescine-ornithine an- deiminase system in protecting oral bacteria and an enzymatic basis tiporter encoded by potE gene of Escherichia coli. Proc. Natl. Acad. for acid tolerance. Appl. Environ. Microbiol. 54:1318±1324. Sci. USA 89:4529±4533. 7. Chander, H., V. H. Batish, S. Babu, and R. S. Singh. 1989. Factors 29. Kleerebezem, M., J. Boekhorst, R. van Kranenburg, D. Molenaar, O. affecting amine production by a selected strain of Lactobacillus bul- P. Kuipers, R. Leer, R. Tarchini, S. A. Peters, H. M. Sandbrink, M. garicus. J. Food Sci. 54:940±942. W. Fiers, W. Stiekema, R. M. Lankhorst, P. A. Bron, S. M. Hoffer, Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021 8. Chopin, A., M. C. Chopin, A. Moillo-Batt, and P. Langella. 1984. M. N. Groot, R. Kerkhoven, M. de Vries, B. Ursing, W. M. de Vos, Two plasmid-determined restriction and modi®cation systems in and R. J. Siezen. 2003. Complete genome sequence of Lactobacillus Streptococcus lactis. Plasmid 11:260±263. plantarum WCFS1. Proc. Natl. Acad. Sci. USA 100:1990±1995. 9. Connil, N., Y. L. Breton, X. Dousset, Y. Auffray, A. RinceÂ, and H. 30. Krause, I., A. Bockhardt, H. Neckermann, T. Henle, and H. Klos- PreÂvost. 2002. Identi®cation of the Enterococcus faecalis tyrosine termeyer. 1995. Simultaneous determination of amino acids and bio- decarboxylase operon involved in tyramine production. Appl. Envi- genic amines by reversed-phase high performance liquid chromatog- ron. Microbiol. 68:3537±3544. raphy of the dabsyl derivatives. J. Chromatogr. A 715:67±79. 10. Copeland, W. C., J. D. Domena, and J. D. Robertus. 1989. The mo- 31. Le Jeune, C., A. Lonvaud-Funel, B. Ten Brink, H. Hofstra, and J. lecular cloning, sequence and expression of the hdcB gene from M. B. M. Van der Vossen. 1995. Development of a detection system Lactobacillus 30a. Gene 85:259±265. for histamine decarboxylating lactic acid bacteria based on DNA 11. Coton, E., G. C. Rollan, and A. Lonvaud-Funel. 1998. Histidine probes, PCR and activities test. J. Appl. Bacteriol. 78:316±326. carboxylase of Leuconostoc oenos 9204: puri®cation, kinetic prop- 32. Lucas, P., J. Landete, M. Coton, E. Coton, and A. Lonvaud-Funel. erties, cloning and nucleotide sequence of the hdc gene. J. Appl. 2003. The tyrosine decarboxylase operon of Lactobacillus brevis Microbiol. 84:143±151. IOEB 9809: characterization and conservation in tyramine producing 12. Cotter, P. D., C. G. Gahan, and C. Hill. 2001. A glutamate decar- bacteria. FEMS Microbiol. Lett. 229:65±71. boxylase system protects Listeria monocytogenes in gastric ¯uid. 33. Lucas, P., and A. Lonvaud-Funel. 2002. Puri®cation and partial gene Mol. Microbiol. 40:465±475. sequence of the tyrosine decarboxylase of Lactobacillus brevis IOEB 13. Cuesta, P. 1995. Development of a starter culture for the artisanal 9809. FEMS Microbiol. Lett. 211:85±89. cheese ``afuega'l pitu.'' Ph.D. thesis. University of Oviedo, Spain. 34. Marino, M., M. Maifreni, S. Moret, and G. Rondinini. 2000. The 14. Day, W. A., R. E. FernaÂndez, and A. T. Maurelli. 2001. Pathoadap- capacity of Enterobacteriaceae species to produce biogenic amines tive mutations that enhance virulence: genetic organization of the in cheese. Lett. Appl. Microbiol. 31:169±173. cadA regions of Shigella ssp. Infect. Immun. 69:7471±7480. 35. Marquis, R. E., G. R. Bender, D. R. Murray, and A. Wong. 1987. 15. de Vos, W. M., P. Vos, H. Dehaard, and I. Boerritger. 1989. Cloning Arginine deiminase system and bacterial adaptation to acid environ- and expression of the Lactococcus lactis ssp cremoris SK11 gene ments. Appl. Environ. Microbiol. 53:198±200. encoding an extracellular serine proteinase. Gene 85:169±176. 36. Meng, S. Y., and G. N. Bennett. 1992. Nucleotide sequence of the 16. Eriani, G., M. Delarue, O. Poch, J. Gangloff, and D. Moras. 1990. Escherichia coli cad operon: a system for neutralization of low ex- Partition of tRNA synthetases into two classes based on mutually tracellular pH. J. Bacteriol. 174:2659±2669. exclusive sets of sequence motifs. Nature 347:203±206. 37. Moreno-Arribas, M. V., and A. Lonvaud-Funel. 2001. Puri®cation 17. Francklyn, C., J. Adams, and J. Augustine. 1998. Catalytic defects and characterization of tyrosine decarboxylase of Lactobacillus brev- in mutants of class II histidyl-tRNA synthetase form Salmonella ty- is IOEB 9809 isolated from wine. FEMS Microbiol. Lett. 195:103± phimurium previously linked to decreased controlled of histidine bio- 107. synthesis regulation. J. Mol. Biol. 280:847±858. 38. Moreno-Arribas, V., S. Torlois, A. Joyeux, A. Bertrand, and A. Lon- 18. Gardini, F., M. Martuscelli, M. C. Caruso, F. Galgano, M. A. Cru- vaud-Funel. 2000. Isolation, properties and behaviour of tyramine- dele, F. Favati, M. E. Guerzoni, and G. Suzzi. 2001. Effects of pH, producing lactic acid bacteria from wine. J. Appl. Microbiol. 88: temperature and NaCl concentration on the growth kinetics, proteo- 584±593. lytic activity and biogenic amine production of Enterococcus fae- 39. Neely, M. N., and E. R. Olson. 1996. Kinetic of expression of the calis. Int. J. Food. Microbiol. 64:105±117. Escherichia coli cad operon as a function of pH and lysine. J. Bac- 19. Gasson, M. J. 1983. Plasmid complements of Streptococcus lactis teriol. 178:5522±5528. NCDO 712 and other lactic streptococci after protoplast-induced cur- 40. Ogier, J. C., O. Son, A. Gruss, P. Tailliez, and A. Delacroix-Buchet. ing. J. Bacteriol. 175:7045±7055. 2002. Identi®cation of the bacterial micro¯ora in dairy products by 20. Gonzalez de Llano, D., P. Cuesta, and A. Rodriguez. 1998. Biogenic temporal temperature gradient gel electrophoresis. Appl. Environ. amine production by wild lactococcal and leuconostoc strains. Lett. Microbiol. 68:3691±3701. Appl. Microbiol. 26:270±274. 41. Paulsen, I., L. Banerjei, G. S. A. Myers, K. E. Nelson, R. Seshadri, 21. HalaÂsz, A., A. BaraÂth, L. Simon-Sarkadi, and W. H. Holzapfel. 1994. T. D. Read, D. E. Fouts, J. A. Eisen, S. R. Gill, J. F. Heidelberg, H. Biogenic amines and their production by microorganisms in food. Tettelin, R. J. Dodson, L. Umayam, L. Brinkac, M. Beanan, S. Trends Food Sci. Technol. 5:42±49. Daugherty, R. T. DeBoy, S. Durkin, J. Kolonay, R. Madupu, W. 22. Hanahan, D. 1983. Studies on transformation of Escherichia coli on Nelson, J. Vamathevan, B. Tran, J. Upton, T. Hansen, J. Shetty, H. plasmids. J. Mol. Biol. 5:557±580. Khouri, T. Utterback, D. Radune, K. A. Ketchum, B. A. Dougherty, 23. Henkin, T. M., B. L. Glass, and F. J. Grundy. 1992. Analysis of the and C. M. Fraser. 2003. Role of mobile DNA in the evolution of Bacillus subtilis tyrS gene: conservation of a regulatory sequence in vancomycin-resistant Enterococcus faecalis. Science 299:2071± multiple tRNA synthetase genes. J. Bacteriol. 174:1299±1306. 2074. 24. Hersh, B. M., F. T. Farooq, D. N. Barstad, D. L. Blankenhorn, and 42. Sambrook, H., E. F. Fritsch, and T. Maniatis. 1989. Molecular clon- J. L. Slonczewski. 1996. A glutamate-dependent acid resistance gene ing: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, in Escherichia coli. J. Bacteriol. 178:3978±3981. Cold Spring Harbor, N.Y. J. Food Prot., Vol. 67, No. 11 L. LACTIS TDC CLUSTER SEQUENCING AND TYRAMINE PCR DETECTION METHOD 2529

43. Schelp, E., S. A. Worley, F. Monzingo, S. Ernst, and J. D. Robertus. P. Rattray, and G. Smit. 2002. Glutamate decarboxylase of Stepto- 2001. pH-induced structural changes regulate histidine decarboxyl- coccus thermophilus, G-75. Abstr. 7th Symposium on Lactic Acid ase activity in Lactobacillus 30a. J. Mol. Biol. 306:727±732. Bacteria. Genetics, metabolism and applications, 2002. Federation of 44. Silla Santos, M. H. 1996. Biogenic amines: their importance in food. European Microbiological Societies and Netherlands Society for Mi- Int. J. Food Microbiol. 29:213±231. crobiology, Egmond aan Zee, The Netherlands. 45. Sissler, M., C. Delorme, J. Bond, S. D. Ehrlich, P. Renault, and C. 50. van Kranenburg, R., J. D. Marugg, I. I. van Swam, N. J. Willem, Francklyn. 1999. An aminoacyl-tRNA synthetase paralog with a cat- and W. M. de Vos. 1997. Molecular characterization of the plasmid- alytic role in histidine biosynthesis. Proc. Natl. Acad. Sci. USA 96: encoded eps cluster essential for exopolysaccharide biosynthesis in 8985±8990. Lactococcus lactis. Mol. Microbiol. 24:387±397. 46. Smith, D. K., T. Kassam, B. Singh, and J. F. Elliott. 1992. Esche- 51. van Poelje, P. D., and E. E. Snell. 1990. Cloning, sequencing, ex- richia coli has two homologous glutamate decarboxylase genes that pression, and site-directed mutagenesis of the gene from Clostridium map to distinct loci. J. Bacteriol. 174:5820±5826. perfringens encoding pyruvoyl-dependent histidine decarboxylase. 47. Takahashi, H., B. Kimura, M. Yoshikawa, and T. Fujii. 2003. Cloning Biochemistry 29:132±139. and sequencing of the histidine decarboxylase genes of gram-nega- 52. van Poelje, P. D., and E. E. Snell. 1990. Pyruvoyl-dependent en- tive, histamine-producing bacteria and their application in detection zymes. Annu. Rev. Biochem. 59:29±59. and identi®cation of these organisms in ®sh. Appl. Environ. Micro- 53. Vanderslice, P., W. C. Copeland, and J. D. Robertus. 1986. Cloning Downloaded from http://meridian.allenpress.com/jfp/article-pdf/67/11/2521/1674402/0362-028x-67_11_2521.pdf by guest on 01 October 2021 biol. 69:2568±2579. and nucleotide sequence of wild type and a mutant histidine decar- 48. Ten Brick, B., C. Damink, H. M. L. J. Joosten, and J. H. J. Huis in't boxylase from Lactobacillus 30a. J. Biol. Chem. 261:15,186±15,191. Veld. 1990. Occurrence and formation of biologically active amines 54. Waterman, S. R., and P. L. Small. 1996. Identi®cation of ␴S-depen- in foods. Int. J. Food Microbiol. 11:73±84. dent genes associated with the stationary-phase acid resistance phe- 49. van Hylckama, J. E. T., N. Leferink, Z. Kruijswijk, M. L. Jensen, F. notype of Shigella ¯exneri. Mol. Microbiol. 21:925±940.