Novel sucrose transposons from plant strains of Lactococcus lactis

Authors


*Corresponding author. Tel.: +64 (6) 356 8080; Fax: +64 (6) 354 6731, E-mail address: wkelly@hort.cri.nz

Abstract

Lactococcus lactis strains isolated from vegetable products transferred the ability to ferment sucrose in conjugation experiments with the recipient strain L. lactis MG1614. Nisin production and sucrose fermentation were transferred together from two strains, but transfer also occurred from several other strains which did not produce nisin. Pulsed-field gel electrophoresis analysis showed that all transconjugants had acquired large chromosomal insertions at two main sites. Nisin–sucrose transconjugants had gained inserts of 70 kb, while those that fermented sucrose without nisin production contained inserts of between 50 and 110 kb. Transconjugants from one donor had acquired a separate insertion of 55 kb which correlated with enhanced bacteriophage resistance, but contained neither nisin nor sucrose fermentation genes.

1Introduction

It is well established that biosynthesis of the bacteriocin nisin and the ability to ferment sucrose are genetically linked as part of a 70-kb conjugative transposon in Lactococcus lactis subsp. lactis[1–4]. This transposon is able to integrate itself into the lactococcal chromosome with one preferred site of insertion in strain MG1614 [3]. Resistance to small isometric bacteriophage [1,5], production of N5-(carboxyethyl)ornithine synthase [6], and altered peptidoglycan composition [7] have also been associated with the nisin–sucrose element. A non-lactococcal origin of this transposon is supported by G+C percent and codon usage data [8,9], and differences in phenotype, conjugative ability, nisin type and transposon architecture suggest that a heterogeneous group of nisin–sucrose transposons exists [9].

We have recently described nisin producing strains of L. lactis isolated from fruit and vegetable products [10], and several other strains which fermented sucrose but did not produce nisin were isolated from the same environment. This paper describes novel conjugative elements from these strains in which the genes for sucrose metabolism are not linked with those for nisin production.

2Materials and methods

2.1Bacterial strains and media

The lactococci used in this study are shown in Table 1, and were grown at 28°C in M17 broth supplemented with 0.5% (w/v) glucose, or 0.5% sucrose. Indicator agars [11] containing lactose, raffinose, or sucrose (SIA) were used to differentiate lactococci.

Table 1. Lactococcus strains used and frequencies of sucrose metabolism transfer
  1. aRaf, raffinose; Lac, lactose; Suc, sucrose; Nis, nisin production; Bac, bacteriocin production, Sm, streptomycin.

  2. bSuc+ transconjugants per donor cell.

  3. cThese strains are closely related but are distinguishable by slight differences in PFGE and plasmid profiles [10].

StrainPhenotypeaTransfer frequencyb
MG1614Raf, Lac, Suc, Smr 
KF5cRaf+, Lac+, Suc+, Bac+1×10−6
KF31cRaf+, Lac+, Suc+, Bac+1×10−6
KF152cRaf+, Lac+, Suc+6×10−4
KF165Raf+, Lac+, Suc+, Nis+4×10−7
KF169Raf+, Lac, Suc+3×10−5
KF201Raf, Lac+, Suc+1×10−4
KF225Raf+, Lac+, Suc+5×10−5
KF241cRaf+, Lac+, Suc+, Bac+1×10−6
KF292Raf+, Lac, Suc+, Nis+2×10−6

2.2Conjugation

Conjugal matings were set up between sucrose fermenting strains and the plasmid-free recipient strain MG1614 [12] as previously described [10]. Selection was on SIA supplemented with 600 μg ml−1 streptomycin.

2.3Pulsed-field gel electrophoresis (PFGE)

Suc+ transconjugants and the MG1614 recipient strain were prepared for PFGE analysis as described previously [10]. DNA was digested for 16 h with SmaI, or I-CeuI (New England Biolabs, Beverly, MA, USA). Gels were run using either a CHEF DR II pulsed-field gel apparatus (Bio-Rad Laboratories, Richmond, CA, USA), or with a FIGE Mapper electrophoresis system (Bio-Rad). DNA fragments are referred to by the numbering used on the physical and genetic map of strain MG1363 [13], the parent of strain MG1614.

2.4DNA methodology

Pulsed-field gels were hybridized with probes for the nisin structural gene, and the sucrose hydrolase gene constructed by amplification of lactococcal DNA with primers already described [14,15]. Sequence data on the nisin gene cluster [16] were used to prepare primers to the nisT gene.

2.5Bacteriophage sensitivity

Changes in bacteriophage sensitivity resulting from the inserted DNA were determined by measuring the efficiency of plating of small isometric (φsk1, φ712, φ742, φ1738 and φ1760) and prolate (φc2) lactococcal bacteriophage on the recipient and transconjugant strains.

3Results

3.1Transfer of sucrose metabolism

The sucrose fermenting phenotype was transferred to MG1614 from 9 of 38 Suc+ donor strains at frequencies between 10−4 and 10−7 CFU per donor cell (Table 1). Transfer occurred from 2 of 22 nisin producing strains and 7 of 16 non-producers. Acquisition of sucrose fermentation could not be correlated with any change in plasmid pattern in the transconjugants (data not shown).

3.2PFGE and sites of DNA insertion

The PFGE patterns for SmaI digests of genomic DNA from the donor and recipient strains have been reported previously [10]. The Suc+ transconjugants showed PFGE patterns closely similar to the MG1614 recipient strain, but with inserts of different sizes and at different sites (Fig. 1A and Table 2). Correlation between the inserted DNA and the phenotypic changes that occurred was established using probes to the nisA, nisT and sucrose hydrolase genes (Fig. 1B,C).

Figure 1.

A: PFGE patterns of SmaI-digested genomic DNA. Lanes: 1, MG1614; 2, KF1113; 3, KF1116; 4, KF1141; 5, KF1142; 6, KF1144; 7, KF1148; 8, KF1119; 9, KF1131; 10, KF1111; 11, KF1112; 12, KF1112-1. Gel run for 21 h at 200 V with the pulse time ramped from 1 to 30 s. B and C: Southern blots of (A) hybridized with sequences amplified using primers for the lactococcal sucrose hydrolase gene (B), and the nisT gene (C). Hybridization with the nisA probe gave identical results to nisT (data not shown).

Table 2.  Properties of transconjugants
  1. aSI, small isometric phage; P, prolate phage.

  2. bThe inserted DNA and the sucrose phenotype are unstable in these strains.

StrainDonorInsert sizeInsert sitePhage resistancea
    SucNisSIP
KF1113KF16570 kbSm2/Sm5B++
KF1116KF16570 kbSm5B+++
KF1119KF20155 kbSm3(+)b
KF1141KF20155 kbSm5B++
KF1142KF20155 kbSm2++
KF1144KF20155 kbSm2/Sm5B+
KF1148KF201110 kbSm5B(+)b+
KF1131KF15250 kbSm5B++
KF1111KF3150 kbSm2(+)b+
KF1112KF3150 and 55 kbSm2 and Sm12(+)b++
KF1112-1KF3155 kbSm12++

3.3Properties of transconjugants

Suc+ transconjugants were unable to ferment lactose or raffinose, but did show differences in nisin production and phage resistance when compared to the MG1614 recipient strain (Table 2). Strains KF1113 and KF1116 show insertion of a 70-kb nisin–sucrose element into the MG1614 chromosome at the same sites that have previously been described with nisin-producing transconjugants [1–3]. Most transconjugants from the KF201 donor show insertion at the same sites as the nisin–sucrose element but differ in that the inserted fragment is approximately 15 kb smaller and does not contain the nisin genes. In KF1119 insertion was into SmaI fragment Sm3 which is not recorded as a site for the nisin–sucrose element, but sucrose fermentation was not stable in this strain. The 110-kb insert in KF1148 is also unstable and is probably the result of insertion of two copies of the transposon.

The donor strains KF31 and KF152 are closely related [10], and their transconjugants also show inserts (of approximately 50 kb) into the same SmaI fragments as the nisin–sucrose elements. However, with these donors insertion into fragment Sm2 was unstable and the Suc+ phenotype was rapidly lost during subculturing. The SmaI digest DNA from KF1112 and its sucrose negative derivative (KF1112-1) also showed a new band at 52 kb which did not hybridize with the nisin or the sucrose probes. These strains showed an insertion into the 38-kb fragment Sm12, and there was also additional DNA at 42 kb (Fig. 2A). Digestion with I-CeuI showed that this insertion increased the 80-kb Ce4 fragment to approximately 135 kb, whereas the main site for insertion of sucrose or nisin–sucrose elements is into the Ce2 fragment (Fig. 2B). Strains containing this extra fragment showed greater phage resistance, including resistance to prolate phage, than did the other transconjugants. The other strains related to KF31 (Table 1) all gave transconjugants which harbored this additional 55-kb fragment (data not shown).

Figure 2.

PFGE patterns of digested genomic DNA from MG1614 and transconjugants. A: FIGE of SmaI-digested DNA. Gel run at 180 V (forward direction) and 120 V (reverse direction) for 20 h with the pulse time ramped from 0.1 to 2.0 s. Lanes: 1, MG1614; 2, KF1112-1. B: PFGE of I-CeuI-digested DNA. Gel run for 20 h with the pulse time ramped from 5 to 60 s. Lanes: 1, MG1614; 2, KF1112-1; 3, KF1116; 4, KF1141; 5, KF1148.

4Discussion

This is the first report of conjugative transposon-like elements that encode sucrose metabolism without nisin production in L. lactis, and follows the report of a large conjugative sucrose transposon in enterobacteria [17]. The results suggest that a family of related sucrose transposons exists in Lactococcus, and that those described here represent a new sub-class of these elements. It is likely that the nisin–sucrose elements have arisen by the integration of nisin genes into a genetic element encoding sucrose, conjugal transfer and transposition genes as has previously been proposed [2]. The conjugative element from these plant strains show the same site specificity as the documented nisin–sucrose transposons, but are smaller and lack the nisin genes. The 15–20 kb difference in size between the transposons with and without the nisin genes is in agreement with the size quoted for the complete nisin gene cluster [18].

The results highlight the untapped genomic diversity of L. lactis as many of the donor strains were also able to transfer a second conjugative element encoding raffinose metabolism that inserted at different sites in the MG1614 chromosome [15]. Three different chromosomal segments each of ∼50 kb can be transferred to specific sites on the MG1614 chromosome during matings with the strains similar to KF31. One of these insertions, the 55-kb fragment in KF1112-1 could be correlated with enhanced phage resistance. Use of these transferable elements to modify the properties of industrial lactococcal strains is being investigated.

Acknowledgments

The authors wish to thank Marie Timmins and Joella Pinkney for technical assistance. This research was partially funded by the New Zealand Foundation for Research, Science and Technology.

Ancillary