Two methods for the genetic di erentiation of Lactococcus lactis ssp. lactis and cremoris based on di erences in the 16S rrna gene sequence

Transcription

1 FEMS Microbiology Letters 166 (1998) 15^20 Two methods for the genetic di erentiation of Lactococcus lactis ssp. lactis and cremoris based on di erences in the 16S rrna gene sequence Lawrence J.H. Ward *, Julie C.S. Brown, Graham P. Davey New Zealand Dairy Research Institute, Private Bag 11029, Palmerston North, New Zealand Received 9 March 1998; accepted 11 May 1998 Abstract The 16S ribosomal RNA gene sequences of Lactococcus lactis ssp. lactis and ssp. cremoris differ by 9^10 bp (depending on strain), within the first 200 bp of the sequence. These differences were used to develop two methods of genetically differentiating lactis and cremoris strains. Primers to conserved sequences in the 16S rrna gene were used in a PCR reaction to amplify fragments of the 16S rrna gene. A single base difference at position 180 of the sequence was utilised to develop a ligase chain reaction to differentiate lactis and cremoris sequences. The second method involved digestion of the amplified fragments with restriction endonucleases specific for either the lactis or cremoris sequence. Resolution of the digested fragments on an agarose gel allowed the strains to be identified as genetically lactis or cremoris. This method was used to examine lactococci isolated from raw milk. Of 31 raw milk strains examined, 21 contained the cremoris 16S rrna sequence, however, all 31 strains exhibited the phenotypic characteristics of the lactis subspecies. z 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords: Lactococcus lactis; Genetic di erentiation; 16S rrna gene; Ligase chain reaction; Raw milk isolate 1. Introduction Bacteria of the genus Lactococcus are used worldwide for the manufacture of fermented dairy products. Of particular importance are the two Lactococcus lactis subspecies, lactis and cremoris. It is necessary to di erentiate L. lactis strains as either * Corresponding author. Tel.: +64 (6) ; Fax: +64 (6) ; lawrence.ward@nzdri.org.nz lactis or cremoris because of their di erent characteristics in cheese manufacture. This di erentiation has traditionally been on the basis of phenotypic characteristics such as growth temperature, salt tolerance and the ability to hydrolyse arginine. Recently, comparison of DNA sequence data has shown genetic di erences between the two subspecies. Godon et al. [1] postulated 20^30% divergence between lactis and cremoris on the basis of DNA hybridisation stringency and reported that this was con rmed by direct comparison of 1.8-kb segments of the his operon of lactis and cremoris strains / 98 / $19.00 ß 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S (98)00305-X

2 16 L.J.H. Ward et al. / FEMS Microbiology Letters 166 (1998) 15^20 Fig. 1. A: Primers used for the LCR detection of the L. lactis ssp. cremoris 16S rrna gene. The C/G nucleotides in bold in the cremoris sequence are A/T in the lactis 16S rrna gene. The 3P terminal bases of primers CREM.1 and CREM.4R are complementary to the bases at this position. Primers CREM.2 and CREM.3R were 5P end labelled with 32 P (see text). B: Primers used for the LCR detection of the lactis 16S rrna gene. The A/T nucleotides in bold in the lactis sequence are C/G in cremoris. The 3P terminal nucleotides of primers LAC.1 and LAC.4R are complementary to the bases at this position. Di erences in the 16S ribosomal RNA sequence have been used to design speci c DNA probes to di erentiate between species of lactic acid bacteria [2]. The 16S rrna gene sequences of lactis and cremoris di er by 9^10 bp (depending on strain) within the rst 200 bp of the 16S rrna sequence. These di erences have been used in the design of Lactococcus genus-speci c and cremoris subspecies-speci c DNA probes [3]. In this paper we report two di erent methods for genetically di erentiating between the lactis and cremoris subspecies. Both methods relied on the PCR ampli cation of a 348-bp region from within the 16S rrna sequence using primers Y1 and Y2 [4] which anneal to highly conserved regions of the 16S rrna sequence. This ampli ed region contains the sequence di erences reported between lactis and cremoris [3]. Firstly, using the PCR ampli cation product, the ligase chain reaction (LCR) [5,6] was used to di erentiate a single base pair di erence between the lactis and cremoris sequences. Secondly, di erences in restriction endonuclease digest patterns of the PCR ampli cation product were used to di erentiate between the two subspecies. The second method was applied to the di erentiation of a number of raw milk isolates of Lactococcus. 2. Materials and methods 2.1. Bacterial strains Lactococcal strains were obtained from the New Zealand Dairy Research Institute (NZDRI) collection. Strains were grown in M17 broth [7] at 30³C overnight. Raw milk isolates were obtained by plating 0.1 ml of the raw milk on M17 agar plates and incubating anaerobically at 22³C and 30³C for 24 h. Strains were identi ed as Lactococcus spp. and phenotypically di erentiated as lactis or cremoris by plating single colonies on bromocresol purple agar (BCP), the di erential medium of Reddy et al. [8] Ampli cation of a 348-bp fragment of the 16S rrna gene PCR primers Y1 (5P-TGG CTC AGG ACG AAC GCT GGC GGC-3P) and Y2 (5P-CCT ACT GCT GCC TCC CGT AGG AGT-3P) [4] were used to amplify a 348-bp fragment from the 16S rrna gene. PCR conditions were 1 cycle: 94³C/3 min, 55³C/45 s, 70³C/1 min; 30 cycles: 94³C/45 s, 55³C/ 45 s, 70³C/1 min; 1 cycle: 94³C/45 s, 55³C/45 s, 70³C/5 min. Samples were prepared for PCR as described by Ward et al. [9]. PCR products were examined by electrophoresis in a 2% agarose gel.

3 L.J.H. Ward et al. / FEMS Microbiology Letters 166 (1998) 15^20 17 Fig. 2. Partial DNA sequences of the 16S ribosomal RNA genes from L. lactis ssp. lactis (upper case) and L. lactis ssp. cremoris (lower case). The regions to which primers Y1 and Y2 anneal are boxed, as are the recognition sequences of the restriction endonucleases MboII and CfoI.

4 18 L.J.H. Ward et al. / FEMS Microbiology Letters 166 (1998) 15^20 Fig. 3. Autoradiograph of 16% polyacrylamide gels (see text) showing: A: detection of the lactis 16S rrna sequence and B: detection of the cremoris 16S rrna sequence by LCR. Lane 1: strain AM2; lane 2: strain MG1363; lane 3: strain The lower bands in each lane correspond to the labelled primers CREM.2 (28 bp) and CREM.3R (21 bp). A ligation product of 49 bp can be seen in lane A3 (lactis) and lanes B1 and 2 (cremoris) DNA sequencing of the ampli ed fragment The PCR products were passed through Wizard1 PCR prep columns (Promega) to remove unincorporated primers. The products were then sequenced using Y1 and Y2 as primers for cycle sequencing (Promega f-mol1) with direct incorporation of [ 35 S]d-ATP Ligase chain reaction assay PCR product from the ampli cation with Y1 and Y2 primers as described above was used as a template for the LCR. Four oligonucleotide primers (Fig. 1A) were designed to detect the C/G pair at position 180 of the cremoris 16S sequence (Fig. 2). Approximately 0.4 nmol of primers CREM.2 and CREM.3R were 5P end labelled with 32 P (Amersham) using T4 polynucleotide kinase (Promega). The 3P nucleotides of primers CREM.1 and CREM.4R are complementary to the C nucleotide (G for CREM.4R) at position 180 of the 16S rrna gene sequence (this nucleotide pair is A/T in the lactis 16S rrna sequence). The LCR was used to detect the presence of this nucleotide pair. Primers CREM.1, 2, 3R and 4R were diluted to approximately 10 pm. The LCR was set up in a total volume of 20 Wl, containing 3 Wl of each diluted primer, 1 Wl of template DNA (PCR product), 1 Wl ofpfu DNA ligase (Stratagene, La Jolla, CA), 2 Wl of 10UPfu DNA ligase bu er and 4 Wl of water. This mixture was overlaid with a drop of mineral oil then subjected to 1 cycle of 94³C for 5 min and 60³C for 4 min followed by 24 cycles of 94³C for 1 min and 60³C for 2 min. The reaction products were separated on a 16% polyacrylamide gel which was then exposed to X-ray lm for 6 h. For detection of the lactis 16S rrna sequence, primers CREM.1 and CREM.4R were substituted with LAC.1 and LAC.4R (Fig. 1B) Digestion of the ampli ed PCR product with restriction endonucleases The ampli ed Y1-Y2 product was microdialysed against water on Millipore type VS lters, Wm pore size and digested with restriction endonucleases MboII, CfoI, FokI and RsaI under conditions recommended by the manufacturers (Boehringer). The resulting digestion products were separated in a 4% Nu-sieve, 1% agarose gel (FMC BioProducts). 3. Results 3.1. LCR detection From the sequences reported by Salama et al. [3], there are 9 or 10 (depending on strain) base di erences between the 16S rrna gene sequences of L. lactis ssp. lactis and cremoris (Fig. 2). All cremoris sequences reported by Salama et al. [3] contained a C at position 180 whereas all lactis strains had an A at this position. Primers (Fig. 1) were designed to allow

5 the LCR detection of this base in the respective subspecies. L. lactis ssp. cremoris strain AM2 and ssp. lactis strain 1404 were used to demonstrate the speci c LCR products (Fig. 3). Strain MG1363 exhibits the phenotypic characteristics of the lactis subspecies, however, it has recently been shown to be genetically cremoris. Fig. 3 shows that MG1363 gives a cremoris-speci c LCR product. The results obtained were con rmed by DNA sequencing of the ampli ed fragment (data not shown) Restriction endonuclease sites in the lactis and cremoris 16S rrna genes L.J.H. Ward et al. / FEMS Microbiology Letters 166 (1998) 15^20 19 The partial 16S rrna sequence of L. lactis ssp. lactis and ssp. cremoris were analysed for restriction sites using the Genetics Computer Group (GCG) sequence analysis package [10]. In the region ampli- ed by the primers Y1 and Y2, L. lactis ssp. lactis had sites for HaeII, HhaI (CfoI), FokI and RsaI (two sites 7 bp apart) which were not in the corresponding region of the cremoris sequence. The L. lactis ssp. cremoris 16S rrna gene has two MboII sites in the ampli ed region. In contrast, there are no MboII sites in the corresponding region of the lactis 16S rrna sequence. The location of these restriction sites is shown in Fig. 2. The restriction enzymes MboII, CfoI, FokI and Fig. 5. Nu-sieve gel (4%) of CfoI and MboII digestion products of the Y1-Y2 ampli ed fragment from seven raw milk lactococcal isolates A^G. Lane 1: BRL 1-kb ladder; lanes 2, 4, 6, 8, 10, 12, 14: MboII digests; lanes 3, 5, 7, 9, 11, 13, 15: CfoI digests. RsaI were used to digest the Y1-Y2 ampli ed product from lactis and cremoris strains. The cremoris PCR product is only digested with MboII. The reduction of the 348-bp fragment to fragments of 23, 65 and 257 bp can be seen in Fig. 4 lane 2 (the smallest fragment is not readily visualised), whereas no reduction in size is observed on digestion with CfoI, FokI or RsaI (Fig. 4, lanes 3^5). The Y1-Y2 ampli ed fragment from lactis is not digested with MboII (Fig. 4, lane 6), but the reduction in size of the 348-bp lactis fragment can be seen when the enzymes CfoI, FokI and RsaI are used (Fig. 4, lanes 7^ 9). Digestion with CfoI gives two products of 299 bp and 49 bp (Fig. 4, lane 7), FokI gives two fragments of 87 bp and 261 bp ((Fig. 4, lane 8). RsaI has two sites in the lactis sequence, 7 bp apart. The fragments visible on the gel were 61 bp and 280 bp (Fig. 4, lane 9). For routine analyses, the enzymes MboII and CfoI were used Examination of raw milk strains Fig. 4. Nu-sieve agarose gel (4%) of digestion products of the Y1-Y2 ampli ed fragment. Lanes 1 and 10: BRL 1-kb ladder; lanes 2^5: ampli ed fragment from L. lactis ssp. cremoris strain AM2 digested with MboII (lane 2), CfoI (lane 3), FokI (lane 4) and RsaI (lane 5); lanes 6^9: ampli ed fragment from L. lactis ssp. lactis strain 1404 digested with MboII (lane 6), CfoI (lane 7), FokI (lane 8) and RsaI (lane 9). Thirty-one lactococcal strains were isolated from raw milk samples. The partial 16S rrna gene fragment was ampli ed from these strains using primers Y1 and Y2 and digested using CfoI and MboII as described above. Of the 31 strains examined, 21 were digested with MboII but not CfoI, indicating that they contained the 16S rrna sequence of the cre-

6 20 L.J.H. Ward et al. / FEMS Microbiology Letters 166 (1998) 15^20 moris subspecies. The remaining 10 isolates showed digestion products with CfoI but not MboII, indicating a lactis genotype. The results of digestion of seven strains are shown in Fig. 5. All 31 isolates fermented arginine (white colonies on BCP plates). 4. Discussion Di erences in the 16S rrna gene sequence of L. lactis ssp. lactis and cremoris have been previously used to di erentiate between the subspecies [2,3,11,12]. These workers used oligonucleotide probes and DNA hybridisation methodology for subspecies di erentiation. We report two additional methods of genetically di erentiating the L. lactis ssp. lactis and cremoris subspecies using a PCR ampli ed 16S rrna gene fragment followed by either LCR or restriction endonuclease digestion. Of the two methods, LCR is highly sensitive, and because of the ampli cation steps, has the potential to permit detection of a low level of a particular sequence (i.e. subspecies) in a mixed population. The LCR assay using the designed primers was able to discriminate between the single base di erence at position 180 (A in lactis vs C in cremoris). Although this is a sensitive assay, a single base mutation in the sequence at the base being detected would a ect the analysis. Although this is also true of the restriction endonuclease digestion method, where a single base change would alter an enzyme site, more than one enzyme can be used to con rm the result. Additionally, in the case of the lactis sequence, the enzymes FokI (which has two lactis-speci c bases in its recognition sequence) or RsaI (which has two sites in the sequence, 7 bp apart) could be used. This would eliminate the possibility of single base changes giving false results. In agreement with our results, it has previously been shown that lactococci isolated from raw milk are predominantly phenotypically lactis [13]. However, Salama et al. [11] and Klijn et al. [12] recently isolated wild-type strains which, although phenotypically lactis, were genetically cremoris. Our results con rm these observations and although only a small sample of strains were examined, demonstrate that lactococci with the cremoris genotype can be isolated from raw milk. Acknowledgments We would like to thank Marie Timmins for excellent technical assistance. References [1] Godon, J.-J., Delorme, C., Dusko Ehrlich, S. and Renault, P. (1992) Divergence of genomic sequences between Lactococcus lactis subsp lactis and Lactococcus lactis subsp cremoris. Appl. Environ. Microbiol. 58, 4045^4047. [2] Klijn, N., Weerkamp, A.H. and de Vos, W.M. (1991) Identi- cation of mesophilic lactic acid bacteria by using polymerase chain reaction-ampli ed viable regions of 16S rrna and speci c DNA probes. Appl. Environ. Microbiol. 57, 3390^ [3] Salama, M., Sandine, W. and Giovannoni, S.J. (1991) Development and application of oligonucleotide probes for identi- cation of Lactococcus lactis subsp cremoris. Appl. Environ. Microbiol. 57, 1313^1318. [4] Young, J.P.W., Downer, H.L. and Eardly, B.D. (1991) Phylogeny of the phototrophic Rhizobium strain BTAil by polymerase chain reaction-based sequencing of a 16S rrna gene segment. J. Bacteriol. 173, 2271^2277. [5] Barany, F. (1991) Genetic disease detection and DNA ampli- cation using cloned thermostable ligase. Proc. Natl. Acad. Sci. USA 88, 189^193. [6] Barany, F. (1991) The ligase chain reaction (LCR) in a PCR world. PCR Methods Applic. 1, 5^16. [7] Terzaghi, B.E. and Sandine, W.E. (1975) Improved medium for lactic streptococci and their bacteriophages. Appl. Microbiol. 29, 807^813. [8] Reddy, M.S., Vedamuthu, E.R. and Reinbold, G.W. (1971) A di erential broth for separating the lactic streptococci. J. Milk Food Technol. 34, 43^45. [9] Ward, L.J.H., Brown, J.C.S. and Davey, G.P. (1995) Detection of dairy Leuconostoc strains using the polymerase chain reaction. Lett. Appl. Microbiol. 20, 204^208. [10] Devereux, J., Haeberli, P. and Smithier, O. (1984) A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12, 387^395. [11] Salama, M., Sandine, W. and Giovannoni, S.J. (1993) Isolation of Lactococcus lactis subsp cremoris from nature by colony hybridization with rrna probes. Appl. Environ. Microbiol. 59, 3941^3945. [12] Klijn, N., Weerkamp, A.H. and de Vos, W.M. (1995) Detection and characterization of lactose-utilizing Lactococcus spp in natural ecosystems. Appl. Environ. Microbiol. 61, 788^ 792. [13] Heap, H.A., Limsowtin, G.K.Y. and Lawrence, R.C. (1978) Contribution of Streptococcus lactis strains in raw milk to phage infection in commercial cheese factories. N.Z. J. Dairy Sci. Technol. 13, 16^22.