Insertional Mutagenesis To Generate Lantibiotic Resistance in Lactococcus lactis

While the potential emergence of food spoilage and pathogenicbacteria with resistance to lantibiotics is a concern, the creationof derivatives of starter cultures and adjuncts that can growin the presence of these antimicrobials may have applicationsin food fermentations. Here a bank of Lactococcus lactis IL1403mutants was created and screened, and a number of novel geneticloci involved in lantibiotic resistance were identified.

INTRODUCTION

Top

INTRODUCTION

The development of resistance to the lantibiotic nisin has beenwidely investigated, and it is well documented that mutantscan be readily created in a laboratory setting (14). The phenotypicand physiological consequences of this resistance have beenlargely elucidated (23, 30), but the underlying genetic alterationsthat are responsible for these phenotypic changes have beenthe focus of relatively less attention. In recent years, someefforts have been made in this regard. First, a number of genes(dltA, pbp2229, hk1021, and lisK) have been associated witheither enhanced sensitivity or resistance of Listeria monocytogenesto nisin (1, 9, 13, 15), and second, transcriptomic analysisof a Lactococcus lactis IL1403 nisin-resistant mutant enabledresearchers to speculate as to how the altered expression ofcertain genes could contribute to nisin resistance (18).

Lacticin 3147, produced by L. lactis DPC3147, is a lantibioticwhich differs from nisin in that it is active via the synergisticactivity of two peptides (22). Like nisin, it exhibits a numberof traits that are highly desirable in a food biopreservative(16). Despite being one of the most extensively characterizedlantibiotics, relatively few studies have investigated the developmentof resistance to this antimicrobial. Studies so far have establishedthat low-level resistance can occur at low frequencies aftera single exposure to low concentrations of lacticin 3147 (7,17, 26) and that mutants can become moderately more resistantfollowing repeated exposure to increasing lantibiotic concentrations(17, 27). To date, the identity of the genes that contributeto lacticin 3147 resistance is unknown. Here we endeavor toidentify genetic loci involved in lantibiotic resistance development.L. lactis IL1403 was chosen as a target organism because itis highly sensitive to lacticin 3147 (17) and its genome hasbeen fully sequenced (4), making the identification of plasmidinsertion sites relatively straightforward. In addition, thedata generated could be used by the dairy industry to determinethe likelihood and mechanism of resistance development amongundesirable spoilage bacteria during food fermentations. Importantly,however, using the corollary logic, this study could also revealhow starter and desirable nonstarter lactococci could be alteredin a targeted way to facilitate their survival in the presenceof specific lantibiotics that may be added to fermented foodsto improve food quality. A mechanism of creating stable food-gradestarters or starter adjuncts that remain resistant to lantibioticscould be an important tool in the control of microorganismsin food fermentations (for a review, see reference 16).

Isolation of lacticin 3147-resistant mutants.

Top

Isolation of lacticin 3147...

A bank of pORI19 integration mutants of L. lactis IL1403 wascreated as described previously (19). Briefly, IL1403 genomicDNA was digested with Sau3A, and fragments were ligated withBamHI-digested pORI19 (RepA^–, erythromycin resistant)(19). The plasmid bank was then electroporated into Escherichiacoli EC101 cells, which can support the replication of the pORI19plasmid from a chromosomal copy of RepA. The resultant transformants(approximately 30,000) were pooled, and the plasmids were isolatedand electroporated into L. lactis IL1403 containing the helperplasmid pVE6007 (RepA⁺, temperature sensitive, chloramphenicolresistant). To select for integration of the pORI19 derivativesinto the IL1403 genome by homologous recombination, the bankof mutants was cultured at 37°C (in the presence of 5 µg/mlerythromycin) to induce loss of the temperature-sensitive pVE6007plasmid. Concentrated lacticin 3147 was obtained from the overproducerstrain MG1363 carrying pMRC01 and pOM02 as described previously(11). The activity of the preparation was assessed by assayingagainst the lacticin 3147-sensitive indicator L. lactis HP byagar well diffusion assay (24), and activity was expressed asarbitrary units per ml (AU/ml) (25). Mutants with increasedresistance to lacticin 3147 were screened by plating on GM17plates containing 110 to 250 AU/ml of lacticin 3147 for 48 hat 30°C (17). In order to identify the site of insertion,the pVE6007 plasmid (RepA⁺) was reintroduced by electroporationto rescue the inserted pORI19 containing the original clonedfragment. The inserts were amplified from five mutants, designatedBRM1 to BRM5 (BRM for bacteriocin-resistant mutant), using primersM13F and M13R (Table 1). Sequence analysis revealed the identityof the segment of strain IL1403 DNA in the respective pORI19plasmids. All mutants were re-created by pORI19 mutagenesiswith the recovered plasmid, and the site of integration wasconfirmed by PCR (Table 1).

View this table:
[in this window]
[in a new window]

TABLE 1. Primers used in this study

Characterization of disrupted loci.

Top

Characterization of disrupted...

In mutant strain BRM1, pORI19 inserted near the end of the mleSgene (Fig. 1). mleS encodes a malate-inducible malolactic enzymeand is followed by mleP, which encodes the associated malatepermease. This operon has been previously characterized in L.lactis, and the two genes are thought to be cotranscribed (2,3). Therefore, it is likely that the pORI19 insertion has apolar effect on mleP. In mutant strain BRM2, the disrupted geneis yjjC, which encodes a putative ATP binding domain of an ABCtransporter and is followed by a putative permease-encodinggene, yjjD. YjjD has 12 hydrophobic domains characteristic ofa secondary membrane transporter (28). Interestingly, transcriptionof yjjCD was also found to be reduced in a spontaneous lacticin3147-resistant mutant (unpublished data). A homologue of yjjC,tnrB2 in Streptomyces longisporoflavus (51% identity), has beenfound to be important in resistance to the polyether ionophoreantibiotic tetronasin (20), and bioinformatics analysis revealeda possible role in lantibiotic immunity (www.ncbi.nlm.nih.gov/structure/cdd/cdd.shtml).API 50CH-based carbohydrate utilization analysis (bioMerieux)carried out in line with the manufacturer's instructions revealedthat the yjjC mutant was impaired compared to the parental controlwith respect to maltose utilization (after 24 h) and gluconateutilization (after 48 h), indicating that resistance was a consequenceof an altered metabolic pathway. In mutant strain BRM3, thepORI19 plasmid has inserted within the coding region of theymcF gene encoding a putative peptidoglycan-bound protein, theC terminus of which shows homology to a cell wall surface anchorfamily of proteins. ymcF is followed by an insertion element(IS905) in the same orientation, but the genes were not foundto be cotranscribed (not shown). Strain BRM4 had a plasmid insertionin an operon encoding phage-like proteins, and in BRM5, pORI19disrupts an insertion sequence (IS981). Curiously, the IS981sequence is found at 10 different locations (tra981A [the tragene that is found in copy A of IS981] to tra981J) throughoutthe IL1403 genome. Thus, it was necessary to determine whichof these loci had been disrupted. A PCR-based strategy establishedthat the plasmid had inserted in tra981I (data not shown).

View larger version (28K):
[in this window]
[in a new window]

FIG. 1. Genes disrupted in the L. lactis IL1403 genome. Black arrowheads point to the sites of pORI19 insertion in IL1403 lacticin 3147-resistant mutants. White regions in BRM2 and BRM3 mutant strains indicate the positions of deletions made through splicing by overlap extension PCR. Putative terminators are indicated by lollipops, and operons proven experimentally are represented by broken-line arrows.

Lantibiotic resistance.

Top

Lantibiotic resistance.

The growth of L. lactis IL1403 and BRM1 to BRM5 strains in GM17broth in the presence of various levels of lantibiotics lacticin3147 and nisin was analyzed (optical density at 600 nm). Themajority of the mutants (pORI:19:yjjC, pORI:19:ymcF, pORI:19:pi322,and pORI:19:mleS) were not dramatically more resistant to lacticin3147 but did show statistically significant (P < 0.05; logphase) increased growth in the presence of low levels of lacticin3147 (25 AU/ml) (Fig. 2A). pORI::tra981I was unique howeverin that it demonstrated an ability to grow even in the presenceof 50 AU/ml lacticin 3147 (Fig. 2B). Analysis of the culturesin the presence of nisin demonstrated that the growth of pORI19::yjjCand pORI19::ymcF was significantly greater than that of thewild-type IL1403 in the presence of 4 µg nisin powder/ml(Sigma; data not shown). To confirm the association betweenmutagenesis of yjjC and ymcF and nisin resistance, additionalmutants in which these genes were deleted in a nonpolar food-grademanner were created as described previously (10) (Table 1).It was established that the deletion mutants (i.e., ${Delta}$ yjjC and ${Delta}$ ymcF) behaved identically to their respective pORI19 counterparts,indicating that the phenotypes observed in the correspondingpORI19 mutants were not due to polar effects (Fig. 2C). A double ${Delta}$ yjjC ${Delta}$ ymcF mutant was also generated to assess whether nisinresistance could be further enhanced. It was established that,although this mutant grew at levels of nisin that preventedgrowth of the parental strain, it was not more tolerant thanthe corresponding single mutants and, in fact, displayed a relativelyextended lag phase. Thus, while under certain circumstancesin the future, it may be possible to combine different mutationswith a view to enhancing the lantibiotic tolerance of starterstrains, this was not the case in this instance. The three remainingmutants did not show significantly increased resistance to nisinas determined by growth analysis (data not shown).

View larger version (14K):
[in this window]
[in a new window]

FIG. 2. Growth of L. lactis IL1403 pORI19 mutants in the presence of 25 AU/ml of lacticin 3147 (A) and growth of pORI19::tra981I in 50 AU/ml of lacticin 3147 (B) at 30°C. Symbols: ${blacksquare}$ , IL1403; ${square}$ , pORI19::mleS; •, pORI19::yjjC; ${circ}$ , pORI19::ymcF; ${blacktriangleup}$ , pORI19::pi322; ${Delta}$ , pORI19::tra981I. (C) Growth of IL1403 ( ${blacksquare}$ ), IL1403 ${Delta}$ yjjC (•), IL1403 ${Delta}$ ymcF ( ${circ}$ ) and IL1403 ${Delta}$ yjjC ${Delta}$ ymcF (x) in 4 µg/ml of nisin powder (Sigma).

Of the five pORI19 mutants, disruption of the tra981I gene resultedin the greatest resistance to lacticin 3147. It is not apparentwhy integration of the vector into 1 of the 10 IS981 sites resultsin this phenotype. A study on IS elements in L. lactis IL1403reported a high frequency of transposition of IS981 under variouslaboratory conditions, which increased in stressful environments(12). It may be possible that plasmid insertion into tra981Iin this study resulted in altered transposition events, leadingto increased fitness of the cell. It is also worth noting thatymcF is also followed by a transposase, tra905. However, asIS905 has not previously been reported to undergo transposition(12), its positioning is unlikely to be of relevance to thephenotype of this mutant. The role of the YmcF protein willneed to be established in order to determine the precise mechanismvia which it contributes to nisin resistance. Based on its impairedability to utilize maltose and gluconate, the enhanced resistancedue to mutation of yjjC is most likely to be a result of analteration in the metabolism of the strain. A role for bioenergeticsin determining nisin resistance has already been established(6). The mleS gene encodes the inducible malolactic enzyme inL. lactis. It was confirmed, by reverse transcription-PCR (primersmleSpORI and mlePR [Table 1]) using cells grown to log phasein GM17 broth, that mleS was transcribed in the wild-type strainunder the conditions used (data not shown), and antibiotic diskassays on GM17 agar demonstrated that the pORI19::mleS mutantdisplayed enhanced neomycin resistance (data not shown). A relationshipbetween lacticin 3147 resistance, neomycin resistance, and acidsensitivity has been previously documented (17), while a relationshipbetween altered acid sensitivity and lantibiotic resistancein L. monocytogenes mutants has also been reported (5, 8, 9,23, 29). As lacticin 3147 has increased activity against energizedcells and the presence of a proton motive force promotes theinteraction of the bacteriocin with the cell membrane (21),a reduction in the proton motive force of the mutant strainmay explain its enhanced lacticin 3147 resistance.

In conclusion, this study has identified novel genes with arole in resistance to lantibiotics. Notably, the majority ofthe mutants isolated (i.e., all bar the more resistant tra981Imutant) exhibited resistance only in the presence of relativelylow levels of lacticin 3147, a trend which is in accordancewith earlier investigations of spontaneously arising lacticin3147-resistant L. lactis IL1403 derivatives (17). It is particularlyinteresting that two of the mutants that show a small increasein resistance to lacticin 3147 display a more dramatic increasedresistance to nisin. This correlates with an observed greaterfrequency and level of spontaneous resistance to nisin thanto lacticin 3147 (data not shown). As mentioned above, the abilityto create stable lantibiotic-resistant starter cultures hasrelevance in the food industry. This could potentially givecontrol over the biota in food fermentations with the additionof lantibiotics for food safety purposes. This principle waspreviously investigated using a spontaneously arising lacticin3147-resistant variant of Lactobacillus paracasei in conjunctionwith a lacticin 3147-producing starter and resulted in increasedcontrol of the nonstarter lactic acid bacteria in cheese manufacture(27). However, as the stability of a spontaneously generatedmutant can be variable, the creation of stable lantibiotic-resistantstarters and adjuncts could be greatly beneficial. In this study,two mutants are created by nonpolar deletions which result instable food-grade strains with increased resistance to bothlacticin 3147 and nisin. This is potentially a strategy thatcould be employed by the food industry to increase food safety,through the addition of lantibiotics, without impacting adverselyon food quality.

ACKNOWLEDGMENTS

We acknowledge the funding received from the Irish Governmentunder the National Development Plan 2000-2006 and through fundingof the Alimentary Pharmabiotic Centre by the Science Foundationof Ireland Centres for Science Engineering and Technology (CSET)scheme.

FOOTNOTES

* Corresponding author. Mailing address: Department of Microbiology, University College Cork, Cork, Ireland. Phone: 353-21-4901373. Fax: 353-21-4903101. E-mail: c.hill{at}ucc.ie

Published ahead of print on 25 May 2007.

Present address: Laboratory for Bacterial Evolution and Pathogenesis,Centre for Infectious Diseases, Chancellor's Building, Universityof Edinburgh, Edinburgh EH16 4SB, United Kingdom.

REFERENCES

Top