Development of a PCR-RFLP assay for the identification of Lactococcus lactis ssp. lactis and cremoris

Abstract

The development of new starter culture of Lactococcus lactis for the manufacture of fermented dairy products with unique characteristics usually requires the isolation and identification of L. lactis up to subspecies level. Therefore, a rapid and specific PCR-RFLP assay has been developed. Forward and reverse primer sets were designed targeting the conserved house keeping gene htrA and yueF encoding a trypsin-like serine protease and a non-proteolytic protein from peptidase family M16, respectively, of L. lactis. Amplicons of 265 bp and 447 bp of htrA and yueF, respectively, were subjected to restriction fragment length polymorphism analysis. Restriction of the 265 bp amplicons with TaqI produced DNA bands of 90 bp and 175 bp with ssp. lactis, and 66 bp and 199 bp with ssp. cremoris. Similarly, restriction of PCR product of 447 bp size with AluI produced digested fragments of 125 bp and 322 bp with ssp. lactis, and 71 bp and 376 bp with ssp. cremoris. The designed primer sets were observed to be specific to L. lactis because other bacteria could not be amplified. The ssp. lactis and cremoris of L. lactis could be identified by restriction of PCR products of htrA and yueF with TaqI and AluI, respectively.

Introduction

Lactococcus lactis is an important member of the lactic acid bacteria (LAB) and is used in the dairy industry as a starter culture for the manufacture of fermented dairy products such as sour milk, cream, butter, fresh cheese and many varieties of cheeses, such as Cheddar, Colby, cottage cheese, cream cheese, Camembert, Roquefort and Brie (Ayad et al. 1999; Cogan 1995; Hugenholtz et al. 2000; Prodelalova et al. 2005). L. lactis influences flavor, texture and shelf-life of fermented dairy products, produces developed acidity in milk and converts milk protein into flavored components such as diacetyl, acetoin, 2,3-butanediol, lactic acid, acetic acid, aldehydes and carbon dioxide (Wouters et al. 2002). L. lactis has been further sub-divided into ssp. lactis and cremoris. In the dairy industry, L. lactis ssp. lactis is preferred for making soft cheese while ssp. cremoris is for hard cheese (Bolotin et al. 2001). However, one biovariant, diacetylactis, has been used to provide flavor in cottage cheese, cultured sour cream and cultured butter milk (Salama et al. 1995). The bio-variant differs from L. lactis ssp. lactis and cremoris in its ability to utilize citrate with the production of diacetyl, a compound with a strong buttery flavor highly prized in some dairy products (Hugenholtz et al. 2000). Furthermore, the development of new strains of L. lactis as starter culture for the manufacture of fermented dairy products with distinct characteristics can be carried out by the isolation and characterization of L. lactis from different sources.

The identification of bacterial species is based usually on a combination of phenotypic and genotypic techniques (Aquilanti et al. 2007; Temmerman et al. 2004). Moreover, DNA-based techniques offer greater distinct discriminatory power than phenotypic methods in differentiating individual subspecies and strains of bacteria. The application of molecular tools has revolutionized the identification of many bacterial species, including LAB. Many of these techniques are based on the polymerase chain reaction (PCR), which uses oligonucleotide primers to amplify targeted DNA fragments. These PCR primers may be designed to differentiate the bacteria from genus-specific to species-specific to subspecies level. The whole genome sequence analysis of L. lactis has revealed genetic differences between the ssp. lactis and cremoris. It has been postulated on the basis of DNA hybridization stringency that 20–30 % divergence has occurred between ssp. lactis and cremoris (Godon et al. 1992). Since L. lactis ssp. lactis and cremoris have the potential metabolic properties for the fermentation of milk and the manufacture of fermented dairy products, it is pertinent to identify the bacteria up to subspecies level. Earlier studies had shown the application of PCR-RFLP-based discrimination of the two subspecies of L. lactis but one or more problems were encountered (Buist et al. 1995; Nomura et al. 2002; Ward et al. 1998). Therefore, this study attempted to develop a specific PCR-RFLP assay to differentiate L. lactis ssp. lactis and cremoris isolated from dairy and non-dairy sources.

Materials and methods

Samples

Approximately 60 milk samples from cow (34), goat (8) and buffalo (17) were collected from local herds in Varanasi (Uttar Pradesh) India. Among non-dairy sources, seven samples of buffalo hair, ten from cow dung, eight from cattle fodder and ten samples of poultry feces were collected. In addition, eight samples of cheddar cheese were obtained from the experimental dairy of the National Dairy Research Institute, Karnal, (Haryana), India.

Bacterial strains

The reference strains Lactococcus lactis ssp. lactis NCDC 094, Lactococcus lactis ssp. cremoris NCDC 086, Lactobacillus acidophilus NCDC 015, Leuconostoc mesenteroides NCDC 029, Lactobacillus plantarum NCDC 021, Enterococcus faecalis NCDC 114, Enterococcus faecium NCDC 124, Bacillus subtilis NCDC 215 and Staphylococcus aureus NCDC 237 were obtained from National Collection of Dairy Cultures (NCDC), National Dairy Research Institute, Karnal (Haryana), India. ATCC strain of L. lactis ssp. cremoris (19297) was purchased from Himedia Laboratories, India. Lactococcus, Lactobacillus and Leuconostoc strains were grown in MRS broth (Himedia Laboratories, India) at 30°C for 24–48 h. Bacillus, Enterococcus and Staphylococcus strains were grown in Nutrient broth (Himedia Laboratories) and incubated at 37°C for 24 h. All strains were maintained in glycerol stocks at −20°C and sub-culturing of strains was carried out at 15-day intervals in MRS agar.

Phenotypic and molecular identification

Gram-positive cocci, homo-fermentative, catalase negative and acid-producing LAB were isolated from dairy and non-dairy sources as per the method of Mirhosaini et al. (2006). Isolates were analyzed further at the genus level. Litmus milk reduction, and methyl red reduction tests and the growth of bacterial cultures at different temperatures (10°C, 40°C and 45°C) were carried out as suggested by Corroler et al. (1998). The isolates were tested for the ability to ferment various carbohydrates (glucose, sucrose, lactose, maltose, galactose, fructose and mannitol) as described by Nomura et al. (1999). The isolates thus identified as members of the genus Lactococcus were characterized further at subspecies level by performing an arginine hydrolysis test and growth at pH 9.2 and 4 % (w/v) NaCl (Klijn et al. 1995).

Genotypic identification of phenotypically characterized strains was carried out by PCR amplification of a fragment of gadB gene followed by RFLP using AseI as described by Nomura et al. (2002). Moreover, a set of eight representative strains from each isolation sources of dairy and non-dairy origin were also identified by gene sequencing of a fragment (> 500 bp) of 16S rRNA (B27F 5′-AGA GTT TGA TCC TGG CTC AG-3′) and (U1492R 5′-GGT TAC CTT GTT ACG ACT T- 3′). GenBank accession numbers were allocated by NCBI for the sequences of 16S rRNA of a total of eight strains of L. lactis ssp. lactis.

Oligonucleotide primers for new PCR-RFLP assay

Two sets of primer pairs were designed from the nucleotide sequences encoding the serine protease and non-proteolytic protein, peptidase family M16 of L. lactis ssp. lactis IL1403 and L. lactis ssp. lactis KF147, respectively published in the NCBI database (http://www.ncbi.nlm.nih.gov/) and synthesized by Metabion (Martinstried, Germany). One of the target genes to be amplified was a fragment of the htrA gene encoding for the trypsin-like serine protease—a unique surface housekeeping gene in L. lactis. The PCR primer set was named serF (5′-CTG TCG TTT CTG TTA TGA AT-3′) and serR (5′-GTG TAT TCA TCA TAA CCA AC-3′), corresponding to positions 262 to 281 and 507 to 526, respectively, of the htrA gene and producing an amplicon of 265 bp. Another target gene to be amplified was a fragment of the yueF gene encoding the non-proteolytic protein, peptidase family M16. The forward primer pepF (5′-TCC TGA TTC ACA GTA AAT CC-3′) and reverse primer pepR (5′-GAT GAA TGG ACT TTT GGG-3′) correspond to positions 29 to 48 and 458 to 475, respectively, of the yueF gene and producing an amplicon of 447 bp. Conserved regions were identified by multiple sequence alignment of the htrA and yueF gene sequences from L. lactis ssp. lactis (GenBank accession number: AE005176.1, CP002365.1 and CP001834.1) and L. lactis ssp. cremoris (GenBank accession number: AM406671.1, CP00425.1 and CP002094.1). The selected sequences were aligned using the multiple sequence alignment program BLASTN 2.2.24+ (Altschul et al. 1997). Alignment was carried out with published NCBI databases revealed no matches other than those for the sequence of the gene encoding L. lactis htrA and yueF.

Genomic DNA extraction

The genomic DNA of test and reference strains was extracted as per the method of Ho et al. (1991). The band of extracted genomic DNA of test and reference strains was observed using 0.8 % (w/v) agarose gel electrophoresis.

PCR conditions

The PCR was carried out using reagents obtained from Bangalore Genei, India. A reaction mixture (25 μL) containing 10–15 ng bacterial genomic DNA, 2.5 μl 10× Taq buffer, 1.5 mM of MgCl₂, 10 pmol of each primer, 100 μM dNTP mixture and 2.5 U Taq DNA polymerase was prepared for each PCR reaction. Amplification was performed in a PCR thermal cycler (BioRad laboratories India). The amplification reaction involved initial denaturation at 94°C for 4 min, 35 cycles consisting of denaturation at 94°C for 1 min, annealing at 51°C for 1 min for both the primer sets, extension at 72°C for 1 min and a final extension step consisting of 72°C for 10 min. A volume of 10 μl of each PCR product was analyzed by 1.8 % (w/v) agarose (G-Biosciences, Maryland Heights, MO) gel electrophoresis in 1 × TAE buffer (pH 8.0) at 65 V for 1 h using horizontal gel electrophoresis (Tarsons Products, India). The gel was visualized using a gel documentation system (Biorad laboratories, India).

Restriction fragment length polymorphism

The digestion of PCR products was carried out using a total of seven fast-digest restriction endonuclease enzymes (AluI, BamHI, HindIII, HinfI, PvuII, SacI and TaqI; Fermentas, Hanover, MD) in appropriate buffered solution as recommended by the manufacturer. The restriction fragments were separated on a 2.5 % (w/v) agarose gel.

Results

Isolation and identification of Lactococcus lactis

During the isolation of L. lactis from a total of 98 dairy and non-dairy samples, 82 (84 %) isolates were identified as LAB, which were characterized as Gram-positive cocci arranged as chains/double/single, catalase-negative, homo-fermentative and with the ability to reduce litmus milk to a white curd with pinkish band. The yield of LAB was 95 % (56/59) and 67 % (26/39) from dairy and non-dairy samples, respectively (Table 1). A total of 14 isolates of L. lactis (24 %, 14/59) could be isolated from dairy products, and 7 (18 %, 7/39) from non-dairy products. The phenotypic characterization of 21 isolates of L. lactis by a battery of test substrates is presented in Table 2. A genotypic identification by PCR amplification of a fragment of gadB gene from genomic DNA followed by restriction digestion with AseI is presented in Fig. 1, which depicts the amplification of a 600 bp fragment of the gadB gene and digestion by AseI into two fragments of 410 bp and 190 bp by test and reference strain L. lactis ssp. lactis NCDC 094. However, amplification of a 560 bp fragment with no digestion by AseI was observed only with the reference strains L. lactis ssp. cremoris ATCC 19297 and L. lactis ssp. cremoris NCDC 086. Further, the maximum sequence similarity of a fragment of 16S rRNA from a total of eight strains with the GenBank reference strain L. lactis ssp. lactis KF147 is shown in Table 3. All 21 strains tested exhibited typical phenotypic and genotypic characters of L. lactis ssp. lactis. However, none of strains tested was phenotypically and genotypically identified as L. lactis ssp. cremoris.

Table 1 Isolation of Lactococcus lactis ssp. lactis from dairy and non-dairy samples. LAB Lactic acid bacteria

Table 2 Phenotypic characteristics of isolates from dairy and non-dairy samples

Fig. 1

PCR amplification of a fragment of the glutamate decarboxylase (gadB) gene and gadB–AseI of Lactococcus lactis ssp. lactis and cremoris. Lane 1–5, 9–10, 12 and 15–16 = test isolates (600 bp), 7 = E. faecalis NCDC 114, 11 = Leuconostoc mesenteroides NCDC 029,13 = B. subtilis NCDC 215, 6 and 14 = L. lactis ssp. lactis NCDC 094, 19 = L. lactis ssp. cremoris ATCC 19297, 20 = L. lactis ssp. cremoris NCDC 086 (560 bp), 8 = 100 bp DNA ladder, 17 & 22 = 50 bp DNA ladder, 18 = 1 kb DNA ladder, 23–27 = gad B-AseI in test isolates 21 = gad B-AseI in L. lactis ssp. cremoris ATCC 19297 and 28 = gad B-AseI in L. lactis ssp. lactis NCDC 094

Table 3 Identification of a group of phenotypically characterized strains of L. lactis ssp. lactis by sequencing of a portion (> 500 bp) of the universal 16S rRNA

Characterization of L. lactis ssp. lactis/cremoris by new genotypic method

The PCR amplification of chromosomal DNA of the identified strains, positive reference strains L. lactis ssp. lactis NCDC 094, L. lactis ssp. cremoris ATCC 19297, L. lactis ssp. cremoris NCDC 086 and other reference strains Lactobacillus acidophilus NCDC 015, Leuconostoc mesenteroides NCDC 029, Lactobacillus plantarum NCDC 021, E. faecalis NCDC 114, E. faecium NCDC 124, S. aureus NCDC 237 and B. subtilis NCDC 215 using the designed primer sets targeting a fragment of nucleotide sequence encoding for serine protease (htrA) and non-proteolytic protein, peptidase family M16 (yueF) resulted in the amplicons of 265 bp (Fig. 2) and 447 bp (Fig. 3). Amplification was obtained with the chromosomal DNA of the test and reference strains L. lactis ssp. lactis NCDC 094, L. lactis ssp. cremoris ATCC 19297 and L. lactis ssp. cremoris NCDC 086. However, no amplicon was generated with the chromosomal DNA of Lactobacillus acidophilus NCDC 015, Leuconostoc mesenteroides NCDC 029, Lactobacillus plantarum NCDC 021, E. faecalis NCDC 114, E. faecium NCDC 124, S. aureus NCDC 237 or B. subtilis NCDC 215. Therefore, our results confirmed the specificity of primer sets (serF/R and pepF/R) under the described PCR conditions for the identification of L. lactis ssp. lactis/cremoris.

Fig. 2

PCR-RFLP of a fragment of the serine protease (htr A) gene of L. lactis ssp. lactis and cremoris with TaqI. Lane 1–3, 9 and 25–28 = test isolates, 4 and 17–19 = L. lactis ssp. cremoris ATCC 19297, 5 = Lactobacillus acidophilus NCDC 015, 6, 16 and 24 = L. lactis ssp. lactis NCDC 094, 7 = Leuconostoc mesenteroides NCDC 029, 8 = Lactobacillus plantarum NCDC 021, 10,13 and 20–23 = L. lactis ssp. cremoris NCDC 086, 11 = E. faecalis NCDC 114, 12 = E. faecium NCDC 124, 14 = S. aureus NCDC 237, 15 = B. subtilis NCDC 215, M1 = 100 bp DNA ladder and M2 = 50 bp DNA ladder

Fig. 3

PCR-RFLP of a fragment of non-proteolytic protein, peptidase family M16 (yueF) gene of L. lactis ssp. lactis and cremoris with AluI. Lane 1–3, 11–13 and 18–21 = test isolates, 4, 9 and 22–25 = L. lactis ssp. cremoris ATCC 19297, 5, 10 and 17 = L. lactis ssp. lactis NCDC 094, 6 = E. faecalis NCDC 114, 7 = E. faecium NCDC 124, 8 and 26–28 = L. lactis ssp. cremoris NCDC 086, 14 = Lactobacillus acidophilus NCDC 015, 15 = Leuconostoc mesenteroides NCDC 029, 16 = Lactobacillus plantarum NCDC 021, M1 = 50 bp DNA ladder and M2 = 1 Kb DNA ladder

PCR-RFLP of serine protease and non-proteolytic protein, peptidase family M16 gene fragment

The PCR-RFLP patterns of the 265 bp and 447 bp amplicons encoding the serine protease (htrA) and non-proteolytic protein, peptidase family M16 (yueF), respectively, of the identified test and reference strains L. lactis ssp. lactis NCDC 094, L. lactis ssp. cremoris ATCC 19297 and L. lactis ssp. cremoris NCDC 086 with seven restriction endonucleases (AluI, BamHI, HindIII, HinfI, PvuII, SacI and TaqI) are presented in Tables 4 and 5. The amplification of PCR products and restriction of DNA bands are depicted in Figs. 2 and 3. The digestion of PCR product of 265 bp and 447 bp with TaqI and AluI, respectively, revealed discriminatory restriction patterns between ssp. lactis and cremoris. The restriction of 265 bp amplicons with TaqI produced DNA bands measuring 90 bp and 175 bp with test strains and L. lactis ssp. lactis NCDC 094, while 66 bp and 199 bp were obtained with L. lactis ssp. cremoris ATCC 19297 and L. lactis ssp. cremoris NCDC 086. Similarly, the restriction of amplicons of 447 bp with AluI produced DNA bands measuring 125 bp and 322 bp with ssp. lactis, and 71 bp and 376 bp with L. lactis ssp. cremoris ATCC 19297 and L. lactis ssp. cremoris NCDC 086. Thus, it was observed that ssp. lactis and cremoris of L. lactis could be identified by restriction of the PCR products of gene htrA and yueF with the single restriction enzymes TaqI and AluI, respectively.

Table 4 PCR-RFLP patterns of a 265 bp fragment encoding the serine protease (htrA) of L. lactis ssp. lactis and cremoris

Table 5 PCR-RFLP patterns of a 447 bp fragment encoding the non-proteolytic protein, peptidase family M16 (yueF) of L. lactis ssp. lactis and cremoris

Discussion

During discrimination o L. lactis ssp. lactis and cremoris by PCR-RFLP, the PCR product of 447 bp from both subspecies was subjected to a battery of restriction endonucleases. It was observed that AluI was able to generate 125 bp and 322 bp fragments from ssp. lactis while ssp. cremoris yielded 71 bp and 376 bp fragments. Two more endonucleases (PvuII and TaqI) could restrict only the amplicons of ssp. lactis. Another conserved housekeeping gene, htrA encoding for trypsin like serine protease, was targeted to discriminate the two subspecies of L. lactis by restriction analysis of 265 bp amplicons with endonucleases. The restriction of amplicons with TaqI yielded fragments of 90 bp, 175 bp from ssp. lactis and 66 bp, 199 bp from ssp. cremoris. Further, AluI and HindIII could discriminate ssp. cremoris and ssp. lactis, respectively. Other endonucleases such as BamHI, HinfI and SacI could not act on the amplicons of yueF and htrA and therefore were not useful for the discrimination of subspecies of L. lactis. Larger restriction fragments were observed with yueF-AluI as compared to htrA-TaqI and therefore were better visualized in agarose gels.

Earlier studies had already discriminated the two subspecies of L. lactis using PCR-RFLP assay. Buist et al. (1995) designed a set of primers (PALA-4 and PALA-14) targeting the L. lactis N-acetylmuramidase (acmA) gene, which differentiated ssp. lactis from ssp. cremoris. However, these results could not be confirmed by Prodelalova et al. (2005) with the primers designed by Buist et al. (1995).

Ward et al. (1998) designed a primer pair from the conserved sequence of 16S rRNA for the discrimination of ssp. lactis and cremoris. But the smaller size of fragments, measuring 23 bp, 65 bp and 257 bp in 16S rRNA-MboII, was difficult to visualize in agarose gels. Later, Nomura et al. (2002) also designed a primer set targeting the glutamate decarboxylase gene (gadB) to discriminate between two subspecies of L. lactis. The amplicons produced by ssp. lactis and cremoris were 600 bp and 560 bp, respectively, which were also difficult to discriminate easily. Therefore, this method was found best suitable for species level identification. Further, upon restriction digestion of gadB amplicons by AseI, two fragments of 410 bp and 190 bp were generated in ssp. lactis but there was no restriction site in ssp. cremoris. However, the restriction site could be observed with one strain L. lactis ssp. cremoris MG1363 (Nomura et al. 2002).

Therefore, it may be concluded from the present study that amplicons targeting yueF and htrA subjected to restriction by AluI and TaqI, respectively, are good enough to discriminate L. lactis ssp. lactis and cremoris and hence PCR-RFLP assay could be used routinely for the identification of subspecies of L. lactis from dairy and non-dairy sources. Further, the detection of L. lactis ssp. lactis from dairy and non-dairy samples confirms their survival in dairy as well as non-dairy environments such as poultry farms and on cattle (Klijn et al. 1995). However, no detection of ssp. cremoris strains in the tested samples indicated its rare occurrence or survival. Our study also supports the observation that L. lactis strains present in starter cultures and in spontaneous milk fermentations may originate from cattle and their surroundings. However, it was observed that the number of lactococci in non-dairy environments was significantly lower than in dairy environments. Therefore, the survival of most starter cultures could be difficult in non-dairy environments.