Garvieacin Q, a Novel Class II Bacteriocin from Lactococcus garvieae BCC 43578

Abstract

Lactococcus garvieae BCC 43578 produces a novel class II bacteriocin, garvieacin Q (GarQ), 70 amino acids in length and containing a 20-amino-acid N-terminal leader peptide. It is cleaved at the Gly-Gly site to generate the mature GarQ (5,339 Da), which is especially inhibitory against Listeria monocytogenes ATCC 19115 and other L. garvieae strains.

TEXT

Ribosomally synthesized antimicrobial polypeptides of bacterial origin, designated peptide bacteriocins, play an important role in bacterial competition, allowing bacteriocin-producing bacteria considerable survival advantages. Most bacteriocins from lactic acid bacteria (LAB) are small, heat-stable, cationic, amphiphilic, membrane-permeabilizing polypeptides (11, 12, 20). LAB bacteriocins are categorized into two main classes (5), class I (lanthionine-containing lantibiotics) and class II (non-lanthionine-containing bacteriocins), which has 4 subclasses, namely, class IIa (pediocin-like like), IIb (two peptide), IIc (circular), and IId (nonpediocin linear one peptide).

Lactococcus is one of the most important LAB genera, which is widely found in fermentation food and in the environment (3, 4). A large number of strains of Lactococcus spp. have been found to produce bacteriocins. Lactococcal bacteriocins are represented in all the classes and subclasses (28). To date, only two bacteriocins from Lactococcus garvieae have been reported: garviecin L1-5 (2.5 kDa) from L. garvieae L1-5, isolated from bovine milk (26), and garvicin ML (6,022 Da) from L. garvieae DCC43, isolated from the intestinal content of mallard duck (Anas platyrhynchos) (2, 23).

Here, we describe the isolation and purification of a novel class II bacteriocin, garvieacin Q (GarQ), obtained from culture supernatant of L. garvieae BCC 43578. A genomic DNA fragment containing garQ was cloned and sequenced, allowing the identification of garQ's open reading frame, as well as some of the adjacent DNA sequences.

Purification and molecular characterization of GarQ from L. garvieae BCC 43578.

Of 30 strains of L. garvieae isolated from nham (local fermented pork sausage), only L. garvieae BCC 43578 exhibited antilisterial activity using the spot-on-lawn method as described previously (18), with Listeria monocytogenes ATCC 19115 as the indicator strain. Thus, its bacteriocin (GarQ) was purified and characterized. Culture supernatant obtained from the 18-h culture grown at 30°C in MRS broth produced 1.6 × 10⁶ activity units (AU)/liter with a specific activity of 57 AU/mg protein. Purification of GarQ was performed using a sequential series of chromatographies (Amberlite XAD-16, SP-Sepharose, and reverse-phase high-performance liquid chromatography [HPLC]), resulting in a final yield and fold purification of GarQ of 0.13% and 12,754-fold (731,429 AU/mg protein), respectively (Table 1).

Table 1.

Summary of the purification steps of garvieacin Q

Step	Volume (ml)	Total activity (AU)a	Total protein (mg)b	Sp act (AU/mg)	Purification (fold)	Yield (%)
Culture supernatant	1,000	1,600,000	27,900	57	1	100
Hydrophobic chromatography (Amberlite XAD-16)	100	640,000	1,665	384	7	0.4
Cation-exchange column chromatography (SP-Sepharose Fast Flow)	10	256,000	71	3,606	63	0.16
Reverse-phase HPLC (Resource RPC column)	1	204,800	0.28	731,429	12,754	0.13

^aBacteriocin activity was measured using the spot-on-lawn method against the indicator strain L. monocytogenes ATCC 19115 and was defined as the reciprocal of the highest dilution causing a clear zone of growth inhibition in the indicator lawn and expressed in activity units (AU) per milliliter of a bacteriocin preparation.

^bProtein concentration was measured by the method of Lowry et al. (15).

Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) analysis of this purified GarQ indicated that a monoisotopic peak ([M+H]⁺) of the bacteriocin was present, suggesting that the molecular mass of GarQ is 5,339 Da (Fig. 1). N-terminal amino acid sequence determination by Edman degradation revealed the following unambiguous first 40 amino acid residues: EYHLMNGANGYLTRVNGKYVYRVTKDPVSAVFGVISNGWG. The calculated molecular mass of this polypeptide (4,434 Da) indicated that about 9 or 10 amino acid residues remained to be sequenced.

Fig 1.

Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) profile of purified garvieacin Q. A monoisotopic peak ([M+H]⁺) of m/z 5,340 is evident.

Cloning of garQ.

Degenerate primers LG1F (5′ GARTAYCAYYTIATGAAYGGIGCIAAYGG-3′) and LG2R (5′-CCAICCRTTISWDATIACICCRAAIAC-3′) [R = (A/G), Y = (C/T), I = inosine, S = (C/G), W = (A/T), D = (A/T/G)], designed from the amino acid sequence obtained, were used to amplify a genomic DNA fragment containing a part of GarQ. The 120-bp PCR amplicon was cloned using the pGEM-T vector system (Promega, Madison, WI) into Escherichia coli JM109. Plasmid inserts from transformants were sequenced (Macrogen, Seoul, South Korea), and the sequences were used to design a new set of primers, LGBF1 (5′-TCGAGTTACAAAGGATCCCGTT-3′) and LGBR1 (5′-CTCTCGTTAAATAACCGTTCGCC-3′), to amplify the complete garQ by means of inverse PCR of L. garvieae BCC 43578 HindIII-digested genomic DNA (21). This generated an amplicon of 6.2 kb, which then was inserted into the pGEM-T vector for subsequent sequencing by the primer-walking procedure. The amplicon contained three complete and two partial putative open reading frames (ORFs) (Table 2). Four ORFs were oriented in the same direction and one in the opposite (Fig. 2).

Table 2.

Identities of the putative proteins encoded by orf1 to orf5 present in the 6.2-kb HindIII fragment

ORF	Position in nucleotide sequence		Gene					Protein characteristicsa	Accession no.
	5′	3′	Designation	G+C content (%)	Length (amino acids)	Molecular mass (Da)	pI
1	1835	4009	orf1	29.04	724	83,787	5.5	Cation transport ATPase; 722 aa; 1e−92; 33%/51%; Lactobacillus delbrueckii subsp. abulgaricus ATCC BAA-365 (YP_813123); 17	AEN79390
2	4426	4671	garQ	42.21	70	7,609	6.1	Hypothetical protein llmg_0854; 140 aa; 1e−12; 75%/81%; Lactococcus lactis subsp. cremoris MG1363 (YP_001032182); 27	AEN79392
3	4695	5009	garI	28.12	104	11,647	9.8	Orf2; 103 aa; 1e−31; 58%/75%; Lactobacillus casei BL23 (YP_001986031); 19	AEN79391
4	5135	6298	garT	35.15	388	44,070	9.6	Lactococcin A ABC transporter ATP binding and permease protein; 562 aa; 2e−169; 71%/86%; Lactococcus lactis subsp. lactis Il1403 (NP_266235); 1	AEN79393
5	1	1203	orf5	27.5	401	47,199	6.6	Span lang=PT-BR>Hypothetical protein SGO_1244; 466 aa; 6e−120; 55%/69%; Streptococcus gordonii strain Challis substrain CH1 (YP_001450528); 25	AEN79394

^aClosest relative; length (amino acids [aa]); e value; level of amino acid identity/level of amino acid similarity; microorganism (accession no.); reference.

Fig 2.

Orientation of orf1 to orf5 in the garQ-containing 6.2-kb HindIII fragment. Arrowheads indicate direction of transcription.

The deduced amino acid sequence of orf2 matched that of GarQ obtained from the Edman sequencing and thus was designated garQ (Table 2). The N-terminal amino acid (Glu) of the 40-mer polypeptide fragment matched the 21st amino acid deduced from the sequence of garQ, indicating that this bacteriocin is initially synthesized in a preform of 70 amino acids (Fig. 3). The putative promoter and terminator sequences and ribosome binding site were readily identified. The deduced amino acid sequence of pre-GarQ possesses a glycine-glycine cleavage site, resulting in a mature GarQ of 50 amino acid residues with a theoretical molecular mass of 5,340 Da and a pI of 9.73. Database search showed that GarQ shares only 42% identity with bacteriocin BacSJ2-8 (GenBank accession no. YP_002720034) from Lactobacillus paracasei subsp. paracasei (14, 16) and 32% identity with lactococcin A (M63675) from Lactococcus lactis subsp. cremoris LMG 2130 (10) but has the highest homology with the hypothetical protein llmg_0854 (YP_001032182) from L. lactis subsp. cremoris MG1363 (27), with 75.1% identity (Fig. 4b). The leader sequence of GarQ has the highest homology (48%) with that of lactococcin A (Fig. 4a).

Fig 3.

Nucleotide and deduced amino acid sequences of the region containing garQ (orf2). The putative −35 and −10 (Pribnow box) promoter sequences and ribosome binding site (RBS) are underlined. The vertical arrow indicates the Gly-Gly cleavage site of the presequence. The stop codon is indicated by the asterisk.

Fig 4.

Amino acid alignments of leader peptides (a) and mature peptides (b) of garvieacin Q with those of other reported class II bacteriocins. The sequences were aligned using the ClustalW2 program. Asterisks, dots, and double dots show identical, weakly similar, and highly similar amino acid residues, respectively. Dashes represent spaces introduced to maximize the alignment. The accession numbers are AEN79392 for garvieacin Q, YP_002720034 for BacSJ2-8, and M63675 for lactococcin A.

orf1 encodes a polypeptide with the highest sequence similarity (51%) to that of Enterococcus casseliflavus EC20 transport ATPase, and orf3 has the highest similarity to a putative immunity protein homologous to the EntA_immunity protein family (Pfam accession no. PF0895; http://pfam.sanger.ac.uk) (6) (designated garI) (Table 2). The deduced amino acid sequence encoded by the partial orf4 has significant sequence similarity to the ABC transporter superfamily (Pfam accession no. PF00005) and was designated garT. This type of ABC transporter is believed to cleave the N-terminal leader peptide from the bacteriocin preform at the Gly-Gly site and to transport the resulting mature polypeptide across the plasma membrane (7, 9, 24). The incomplete orf5 contains a sequence similar to the Abi group (Pfam accession no. PF07751), which is involved in Lactococcus spp. bacteriophage resistance (8, 13).

Cytotoxicity test and antibacterial properties of purified GarQ.

Purified GarQ, up to 1 mg ml⁻¹, was not cytotoxic to Vero (African green monkey kidney), HepG2 (human liver hepatocarcinoma), and Caco-2 (human colon adenocarcinoma) cell lines using an MTT [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide] assay (22) (Table 3). In the tests of GarQ's antibacterial activity against various indicator strains (Table 4), the MIC is defined as the minimum GarQ concentration that yielded a clear zone of growth inhibition in the indicator lawn. GarQ had a wide activity range, inhibitory against the food-borne pathogenic bacterium L. monocytogenes ATCC 19115 and various L. garvieae strains, as well as L. lactis, Enterococcus, and Pediococcus pentosaceus but not E. coli or Salmonella. Its spectrum of antimicrobial activity included many LAB, as well as strains of potentially pathogenic Enterococcus faecalis, L. garvieae, and L. monocytogenes, similar to garviecin L1-5 (26) and garvicin ML of L. garvieae (2, 23). GarQ retained its activity over a wide pH range (pH 2 to 8), even following treatment at 100°C for 15 min. However, heating at 121°C for 15 min led to a complete loss of antibacterial activity. It is worth noting that the target cell specificity of GarQ displayed a high degree of similarity to the target cell specificities of other L. garvieae bacteriocins, despite the marked differences in GarQ's molecular mass and structure compared with those of the other bacteriocins. In addition, GarQ lacks the signature YGNGVXC motif.

Table 3.

Cytotoxicity of garvieacin Q tested against Vero, HepG2, and Caco-2 cells

Cell lines	% Survival with indicated concn (mg ml−1) of purified garvieacin Qa
	1	0.5	0.25	0.125
Vero	94 ± 8	95 ± 7	95 ± 4	101 ± 7
HepG2	97 ± 4	102 ± 6	90 ± 2	96 ± 3
Caco-2	89 ± 7	93 ± 4	99 ± 5	95 ± 6

^aPercent survival compared to that of untreated control using the MTT assay.

Table 4.

Antimicrobial spectrum of garvieacin Q^a

Indicator strain	MIC (μM)
Bacillus coagulans JCM 2257	0.2048
Enterococcus faecalis JCM 5803	NA
Enterococcus faecium D48S52	3.2775
Enterococcus faecium JCM 5804	NA
Escherichia coli ATCC 25922	NA
Lactobacillus plantarum BCC 9546	0.0512
Lactobacillus sakei JCM 1157	0.0256
Lactococcus garvieae BCC 7247	0.0512
Lactococcus garvieae BCC 7251	0.1024
Lactococcus garvieae BCC 7255	0.0512
Lactococcus garvieae BCC 7266	1.6388
Lactococcus garvieae BCC 7268	0.0256
Lactococcus garvieae BCC 7282	0.0256
Lactococcus garvieae BCC 7322	0.0512
Lactococcus garvieae BCC 7338	0.1024
Lactococcus garvieae BCC 7339	0.1024
Lactococcus garvieae BCC 7344	0.0256
Lactococcus garvieae BCC 7352	0.0256
Lactococcus garvieae BCC 7379	0.8194
Lactococcus garvieae BCC 7380	0.0256
Lactococcus garvieae BCC 7389	0.1024
Lactococcus garvieae BCC 7417	1.6388
Lactococcus garvieae BCC 7433	0.0512
Lactococcus garvieae BCC 7446	0.0256
Lactococcus garvieae BCC 7448	3.2775
Lactococcus garvieae BCC 7374	0.2048
Lactococcus lactis ATCC 19435	NA
Lactococcus lactis NCDO 497	NA
Listeria monocytogenes ATCC 19115	0.2048
Listeria ivanovii ATCC 700402	NA
Listeria ivanovii DMST 9012	0.1024
Micrococcus luteus NBRC 12708	NA
Pediococcus acidilactici BCC 9545	NA
Pediococcus pentosaceus BCC 3749	0.1024
Pediococcus pentosaceus JCM 5885	NA
Salmonella Typhimurium ATCC 13311	NA
Staphylococcus aureus ATCC 6538	NA
Staphylococcus xylosus BCC 3710	NA

^aActivity was determined by the spot-on-lawn method and defined as the lowest concentration causing a clear zone of growth inhibition in the indicator lawn. ATCC, American Type Culture Collection, Rockville, MD; JCM, Japan Collection of Microorganisms, Wako, Japan; NCDO, National Collection of Dairy Organisms, Reading, United Kingdom; BCC, BIOTEC Culture Collection, National Science and Technology Development Agency, Ministry of Science and Technology, Bangkok, Thailand; DMST, Department of Medical Sciences, Ministry of Public Health, Thailand; NA, no activity.

In summary, this study describes the identification of a novel bacteriocin, GarQ, from L. garvieae BCC 43578 culture medium. In addition, the complete garQ, together with three other ORFs within the vicinity, was cloned and sequenced, enabling the full-length sequence of GarQ to be elucidated. Based on the genetic organization of its gene, its bactericidal activity in the absence of other peptides, and the absence of any modified amino acids, GarQ was designated as belonging to the one-peptide class IId bacteriocins.

Nucleotide sequence accession number.

The DNA sequence obtained has been deposited in the GenBank database under accession number JN605800.