NisI Maturation and Its Influence on Nisin Resistance in Lactococcus lactis

Nisin, a safe and natural antimicrobial peptide, has a long and impressive history as a food preservative and is also considered a novel candidate to alleviate the increasingly serious threat of antibiotic resistance. Nisin is produced by certain L. lactis strains. The nisin immunity protein NisI, a membrane-bound lipoprotein, is expressed by nisin producers to avoid suicidal action. Here, we report the roles of Lsp and Lgt in NisI maturation and nisin resistance of L. lactis F44. The results verified the importance of Lsp to NisI-conferred immunity and Lgt to localization. Our study revealed the process of NisI maturation in L. lactis and presented a potential strategy to enhance industrial nisin production.

ABSTRACT

NisI confers immunity against nisin, with high substrate specificity to prevent a suicidal effect in nisin-producing Lactococcus lactis strains. However, the NisI maturation process as well as its influence on nisin resistance has not been characterized. Here, we report the roles of lipoprotein signal peptidase II (Lsp) and prolipoprotein diacylglyceryl transferase (Lgt) in NisI maturation and nisin resistance of L. lactis F44. We found that the resistance of nisin of an Lsp-deficient mutant remarkably decreased, while no significant differences in growth were observed. We demonstrated that Lsp could cleave signal peptide of NisI precursor in vitro. Moreover, diacylglyceryl modification of NisI catalyzed by Lgt played a decisive role in attachment of NisI on the cell envelope, while it exhibited no effects on cleavage of the signal peptides of NisI precursor. The dissociation constant (K_D) for the interaction between nisin and NisI exhibited a 2.8-fold increase compared with that between nisin and pre-NisI with signal peptide by surface plasmon resonance (SPR) analysis, providing evidence that Lsp-catalyzed signal peptide cleavage was critical for the immune activity of NisI. Our study revealed the process of NisI maturation in L. lactis and presented a potential strategy to enhance industrial nisin production.

IMPORTANCE Nisin, a safe and natural antimicrobial peptide, has a long and impressive history as a food preservative and is also considered a novel candidate to alleviate the increasingly serious threat of antibiotic resistance. Nisin is produced by certain L. lactis strains. The nisin immunity protein NisI, a membrane-bound lipoprotein, is expressed by nisin producers to avoid suicidal action. Here, we report the roles of Lsp and Lgt in NisI maturation and nisin resistance of L. lactis F44. The results verified the importance of Lsp to NisI-conferred immunity and Lgt to localization. Our study revealed the process of NisI maturation in L. lactis and presented a potential strategy to enhance industrial nisin production.

INTRODUCTION

The increasingly serious threat of antibiotic resistance has prompted investigations into novel antimicrobial agents with multiple mechanisms of action. Lantibiotics, a class of lanthionine-containing antimicrobial peptides, are considered promising candidates due to their proven effectiveness and distinct modes of action (1–3). Lantibiotics are classified as class I bacteriocins, and their biosynthesis in bacteria generally involves precursor synthesis and posttranslational modifications, including dehydration and cyclization (4).

Nisin, classified as a type A lantibiotic, is produced by certain Gram-positive bacteria, including Lactococcus and Streptococcus species. Among the several reported natural variants of nisin, nisin A is the first identified and most extensively studied (5). Nisin Z, the closest variant of nisin A, exhibits a single amino acid change (6) and has a superior solubility, stability, and biological activity under neutral pH (7). The biosynthesis of nisin generally comprises four major steps. NisA/NisZ, a linear precursor peptide containing a leader peptide and a core peptide, is first generated from mRNA on the ribosome (8). Then a total of eight serines and threonines in the core region of NisA/NisZ are dehydrated by dehydratase NisB, forming dehydroalanine and dehydrobutyrine residues (9). Subsequently, catalyzed by the cyclase NisC, five lanthionine and methyllanthionine rings are generated through the nucleophilic addition of cysteinyl thiols to dehydroalanine and dehydrobutyrine (10). Finally, the leader peptide of nisin precursor is removed by lantibiotic protease NisP, leading to the formation of mature nisin (11).

Nisin exhibits broad-spectrum antimicrobial activity against a majority of Gram-positive foodborne bacteria, such as Listeria and Clostridium (12, 13), and some Gram-negative pathogens, such as Escherichia coli and Salmonella spp., when combined with EDTA or physical treatments (14, 15). It has a long and impressive history as a food preservative (16, 17) and recently has shown promising laboratory and clinical results as a useful therapeutic agent (18). Nisin has an antimicrobial mechanism of action different from that of conventional antibiotics (19, 20). It can exert its antimicrobial activity in nanomolar concentrations by both pore formation on the surface of cells and inhibition of cell wall biosynthesis. Nisin interacts with the membrane-bound lipid II to form transient pores, which lead to dissipation of the membrane potential and/or the pH gradient and allow a rapid efflux of intracellular substances with low molecular weights, such as ATP and amino acids (21). Among the five specific lanthionine rings of nisin, the first two are responsible for lipid II binding (22), while the last two mainly act on the membrane for pore formation (23, 24). Since lipid II is an essential intermediate in peptidoglycan synthesis, nisin dramatically inhibits cell wall synthesis through binding to lipid II with high affinity (25). Additionally, the inhibition of bacterial spores’ outgrowth is the third distinct molecular mechanism of nisin depending on the presence of Dha residues in position 5 (26).

Indeed, the above-described action can also work on the membrane of Lactococcus lactis (27). To prevent such a suicidal mode of action, the nisin producers coordinately express a complex immunity system comprising a multiprotein ABC transporter complex, NisFEG, and a membrane-bound lipoprotein, NisI (28). NisFEG is reported to function as an immunity protein through expelling nisin from the cytoplasmic membrane before or during pore formation (29). NisI principally carries out its immune activity through preventing nisin from attacking the cell membrane. It can specifically interact with nisin to form an insoluble complex (30). Another immune mechanism for NisI is that the L. lactis cells start to cluster in the presence of both NisI and nisin. Thus, nisin is unable to reach the cell membrane (31). In addition to the NisI anchored to the cell membrane, there also exists NisI in a lipid-free form, which escapes lipid modification and is secreted to the extracellular environment. The lipid-free NisI (LF-NisI) possesses an additional mechanism of immunity (32).

For Gram-positive bacteria, the prolipoprotein diacylglyceryl transferase (Lgt) and the lipoprotein signal peptidase (Lsp) are the two key enzymes in posttranslational processing of lipoproteins. After translocation of the preprolipoprotein, Lgt transfers a diacylglyceryl moiety from phosphatidylglycerol to the sulfhydryl group of the prospective N-terminal cysteine residue of the prolipoprotein. Subsequently, the Lsp enzyme cleaves the signal peptide, resulting in a mature lipoprotein (33). Additionally, it was revealed that Lsp could cleave the signal peptide of lipoprotein in lgt-deficient Listeria monocytogenes, indicating that lipidation by Lgt is not a prerequisite for activity of Lsp (34). Mutation of either the lgt or the lsp gene in Gram-positive bacteria inactivates the maturation of lipoproteins, which might result in impaired growth and other phenotypes, such as virulence, adherence, antibiotic resistance, nutrient acquisition, and sensitivity to environmental stress (35, 36). For example, Bacillus anthracis deficient in Lgt-catalyzed lipidation was affected in spore germination, which markedly attenuated virulence. Reduced survival in human blood and increased sensitivity to oxidative stress in vitro were observed in an lsp mutant of Streptococcus pneumoniae (37). However, although both the lipoproteins PrtM and OppA are critical for growth of L. lactis in milk, an L. lactis lsp mutant was still viable in milk (38). Likewise, both Lgt and Lsp are not essential for the growth and viability of Bacillus subtilis, although PrsA lipoprotein is an essential protein in this organism (39, 40). The reason underlying the viability of lgt- or lsp-deficient Gram-positive strains is that some authentic lipoprotein precursors can be anchored in the cytoplasmic membrane and retain biological activity. Therefore, Lgt and Lsp exert different influences on membrane anchoring and physiological function of different lipoproteins.

The objective of this study was to elucidate the roles of Lgt and Lsp in NisI maturation and nisin resistance of L. lactis. Since the nisin immunity level of the strain is an important consideration for engineering efforts toward efficient production of nisin (41–45), this work provides a potential engineering strategy for industrial producer strains to improve nisin production through enhancing the immunity level.

RESULTS

Effects of Lsp, Lgt, and NisI on nisin resistance of L. lactis F44.

We deleted lgt and lspA of L. lactis F44 by double-crossover recombination. No significant difference in growth performance was detected among F44, F44 Δlgt, and F44 ΔlspA (Fig. 1). The growth performance at inhibition concentrations of nisin can represent the nisin resistance of the strains. According to our previous experiments, nisin concentrations of 4,000, 8,000, and 12,000 IU ml⁻¹ could all inhibit the growth of L. lactis F44 to different degrees. The inhibition effects at the same concentration are different again for liquid and solid media. Therefore, to demonstrate the effects of Lgt and Lsp on nisin resistance, we detected the growth performance of F44, F44 Δlgt, and F44 ΔlspA at nisin concentrations of 4,000, 8,000, and 12,000 IU ml⁻¹ in both liquid and solid media (Fig. 2). At the concentration of 4,000 IU ml⁻¹, no remarkable difference in cell densities between F44 and F44 Δlgt was observed in solid and liquid media. At the concentration of 8,000 IU ml⁻¹, the growth of F44 Δlgt was significantly retarded in the liquid medium. At the concentration of 12,000 IU ml⁻¹, the growth of both F44 and F44 Δlgt was remarkably inhibited in the liquid medium, while F44 Δlgt showed a 100-fold reduction of growth in the solid medium. Surprisingly, deletion of lspA was detrimental to the growth of L. lactis even at the lowest nisin concentration, 4,000 IU ml⁻¹.

FIG 1.

Comparison of growth curves among F44 and engineered strains. Average data from triplicate experiments are presented. Error bars represent standard deviations from three parallel replicates.

FIG 2.

Nisin resistance assay of F44 and engineered strains. Average data from triplicate experiments are presented. Error bars represent standard deviations from three parallel replicates.

To further confirm the effects of lgt and lspA, we conducted a growth performance experiment with overexpression strains. As shown in Fig. 1, overexpression of lgt, lspA, and/or nisI exhibited no or very modest effects on the growth performance of L. lactis F44. Overexpression of lgt, lspA, or nisI alone played no positive role in resistance to nisin and even decreased the cell density of L. lactis F44 at nisin concentrations of 8,000 and 12,000 IU ml⁻¹. However, coexpression of lspA-nisI and lgt-lspA-nisI both obviously increased the survival rates at the nisin concentration of 8,000 IU ml⁻¹ (Fig. 2).

Analysis of endopeptidase activity of Lsp on NisI.

To confirm the cleavage role of Lsp on NisI, SDS-PAGE was performed with pre-NisI as the substrate. To attain lipidated pre-NisI before Lsp-catalyzed cleavage of the signal peptides, purified Lgt was added to the reaction system with dioleoylphosphatidylglycerol (DOPG) as a lipid substrate. The sample containing only pre-NisI was used as a control. As shown in Fig. 3, the amount of pre-NisI was significantly reduced in the experimental sample (right lane) compared to the control sample (left lane) after 16 h. Lsp activity was evident by the appearance of bands corresponding to the cleaved NisI (26 kDa). After 20 h, only a small amount of pre-NisI existed in the experimental sample.

FIG 3.

Endopeptidase activity of Lsp with the substrate pre-NisI, generated in situ. To assay the proteolytic activity of Lsp, a 240-μl reaction mixture was set up containing 0.1 mg liter⁻¹ Lgt, 0.1 mg liter⁻¹ Lsp, 2 mg liter⁻¹ pre-NisI, and 150 mg liter⁻¹ DOPG in PBS and incubated at 30°C. The sample containing only pre-NisI was used as a control. The control sample (left lane) and experimental sample (right lane) were taken from the reaction mixture at 0 h, 16 h, and 20 h and then subjected to SDS-PAGE.

NisI localization in the F44 ΔlspA and F44 Δlgt strains.

To evaluate the effects of lspA deletion and lgt deletion on the maturation process of NisI, Western blotting of proteins extracted from the membrane and cytoplasmic fractions was conducted at 6 and 12 h. The same amount of total protein from strains F44, F44 Δlgt, and F44 ΔlspA was loaded on SDS gels. The protein levels in Western blotting experiments were quantified by image analysis using ImageJ software. As shown in Fig. 4, the molecular weight of NisI in F44 ΔlspA was higher than that in F44, suggesting the presence of the uncleaved signal peptide. However, similar NisI contents were detected in the wild-type strain and F44 ΔlspA. This confirmed that Lsp played a decisive role in cleavage of the signal peptides of NisI precursor while exhibiting no effects on attachment of NisI on the cell envelope. Interestingly, the quantity of NisI protein from the membrane of F44 was higher than that of F44 ΔlspA both at 6 and at 12 h, so it seemed that pre-NisI with uncleaved signal peptide was more inclined to anchor on the cell membrane. Lgt is required for lipid modification of lipoprotein. Hence, we also assessed the effect of lgt mutation on the maturation of NisI. We found no effect of lgt deletion on size of NisI bands, indicating that diacylglyceryl modification of NisI is not a prerequisite for Lsp-catalyzed cleavage of the signal peptides. Notably, smaller amounts of NisI existed in F44 Δlgt both at 6 and at 12 h than in the wild-type strain, in both membrane and cytoplasmic fractions. It was speculated that NisI without lipid modification was secreted to the culture medium (46).

FIG 4.

Effects of ΔlspA mutation and Δlgt mutation on localization of NisI. Proteins were extracted from the membrane (M) and cytoplasmic (C) fractions of F44, F44 ΔlspA, and F44 Δlgt and probed by immunoblotting with anti-His tag antibody and anti-rabbit IgG antibody.

SPR analysis of interactions of pre-NisI and LF-NisI with nisin.

NisI can specifically interact with nisin to carry out its immune activity (30). Since lspA deletion significantly reduced the nisin resistance, we speculated that uncleaved NisI had a lower affinity for nisin than mature NisI. Therefore, we determined the strength of the interaction by surface plasmon resonance (SPR) analysis (Fig. 5). LF-NisI was obtained as previously described (32). The Reichert 4SPR system (Reichert Technologies, Depew, NY) was used to analyze the affinity of interaction of pre- and LF-NisI with nisin. Nisin was immobilized to the dextran SPR sensor chip surface to attain 490 micro-refractive index units (μRIU). The K_D (dissociation constant) for the interaction between nisin and pre-NisI was 4.61 μM, which was 2.8 times that between nisin and LF-NisI (1.65 μM), confirming that the cleavage of signal peptide could increase the affinity of NisI and nisin.

FIG 5.

Kinetic analysis of the interaction of pre-NisI and nisin (A) and LF-NisI and nisin (B) by SPR. Nisin (275 μg ml⁻¹) was bound to the sensor surface. Pre-NisI and LF-NisI at concentrations of 2.5, 1.25, 0.625, 0.3125, and 0.156 μM were injected. Dissociation of the nisin-bound pre-NisI and LF-NisI was determined for 5 min after injection.

DISCUSSION

Both the Lgt-catalyzed lipid modification and Lsp-catalyzed signal peptide cleavage have been shown to play essential roles in the survival of Gram-negative bacteria. In contrast, inactivation of lipoprotein maturation through mutations of lgt or lsp are not lethal in some Gram-positive bacteria even though some lipoproteins are essential for viability (34, 47–49). In this study, the results showed that the growth of L. lactis was not affected by complete knockout of either lgt or lspA (Fig. 1). Venema et al. demonstrated that Lsp is not required for the growth of L. lactis in rich medium (38), which is in accordance with our results. Their results also suggested that authentic lipoproteins, such as PrtM and OppA, retain function without signal peptide cleavage. Likewise, Bacillus subtilis mutants defective in lipid modification and signal peptide cleavage did not show a growth-deficient phenotype, although PrsA lipoprotein is an essential protein in this organism (39, 40).

It has been proven that the lipoprotein NisI is involved in the immunity to nisin (50). The nisI mutant exhibited a marked increase of sensitivity to nisin from the external environment (51). As shown in Fig. 2, knockout of either lgt or lspA significantly decreased the nisin resistance of L. lactis. Deletion of lspA even incapacitated the viability of L. lactis at a concentration of 12,000 IU ml⁻¹ on solid medium. Intriguingly, overexpression of lgt, lspA, or nisI alone did not increase the nisin tolerance and even resulted in a reduction in the concentration of 12,000 IU ml⁻¹ on solid medium. It is speculated that Lgt, Lsp, and NisI synergistically contribute to the maturation of NisI; overexpression of any alone cannot synthesize mature NisI and even increases metabolic burden. This hypothesis was verified by the resistance assay of F44::pLspA-NisI and F44::pLgt-LspA-NisI (Fig. 2). It also indicated that the cleavage of signal peptide is important for the immune ability of NisI.

Western blotting showed that deletion of lspA increased the amount of NisI from the membrane (Fig. 4), indicating an enhanced ability of NisI to anchor to the cytoplasmic membrane. However, deletion of lspA resulted in a 1,000-fold decrease of nisin tolerance at a concentration of 4,000 IU ml⁻¹ and even lost the viability at a concentration of 12,000 IU ml⁻¹ (Fig. 2). Since the kinetics of the interaction between nisin and LF-NisI has been examined (32), we speculated that the presence of signal peptide is not conducive to the binding of NisI and nisin. To confirm this speculation, SPR analysis was performed to compare the affinities of binding of pre-NisI and LF-NisI to nisin (Fig. 5). The results showed that the affinity for LF-NisI to nisin was 2.8 times higher than that of pre-NisI. This demonstrated that the cleavage of NisI signal peptide catalyzed by Lsp is crucial to prevent the binding of nisin to lipid II.

Many methods and technologies have been adopted to improve nisin yield and reduce production cost, such as fermentation optimization, random mutation and selection, and gene engineering (52). Among these endeavors, researchers have found that the nisin resistance of the strain is an important factor in enhancing the productive capacity of nisin. For example, introducing multicopy nisin immunity genes by certain engineered vectors into L. lactis could remarkably improve nisin production (41). A significant (1.4-fold) increase in nisin titer was achieved through improving nisin tolerance by genome shuffling (44). In this study, we demonstrated that coexpression of lspA and nisI could significantly increase the nisin resistance capacity of F44 (Fig. 2). The result may provide a promising approach to improve nisin production for industrial producer strains.

This study revealed the process of NisI maturation and determined its impact on nisin tolerance. The results clarified that Lsp is responsible for the immune function of NisI, while Lgt is responsible for localization. Further studies on why the precursor form of NisI has reduced affinity are recommended to generate new insights into the molecular mechanism of nisin-NisI interaction.

MATERIALS AND METHODS

Strains, media, and culture conditions.

All bacterial strains used in this study are listed in Table 1. The parent strain was L. lactis F44, a nisin-producing strain which was constructed through genome shuffling of L. lactis YF11 (China General Microbiological Culture Collection Center accession number CGMCC7.52) in our previous study (44). Escherichia coli TG1, used as the host for plasmid construction, was cultured in Luria-Bertani (LB) medium. Micrococcus flavus ATCC 10240, used as an indicator strain for the bioassay of nisin, was grown in LB medium. Its agar diffusion bioassay medium (pH 7.0) contained (per liter) 8 g tryptone, 5 g glucose, 5 g yeast extract, 5 g NaCl, 2 g Na₂HPO₄, and 15 g agar. L. lactis F44 and the engineered strains were cultured in 100 ml seed medium (pH 7.2) containing (per liter) 15 g glucose, 15 g peptone, 15 g yeast extract, 20 g KH₂PO₄, 1.5 g NaCl, and 0.15 g MgSO₄·7H₂O. The fermentation medium (pH 7.2) for L. lactis strains contained (per liter) 15 g glucose, 15 g peptone, 15 g yeast extract, 20 g KH₂PO₄, 1.5 g NaCl, 0.15 g MgSO₄·7H₂O, 3 g corn steep liquor, and 0.26 g cysteine. E. coli BL21(DE3) was used for protein purification. To provide selective pressure or maintain the plasmid stability of recombinant L. lactis strains, the medium was supplemented with 5 μg ml⁻¹ erythromycin or 5 μg ml⁻¹ chloramphenicol. For E. coli, 150 μg ml⁻¹ erythromycin, 50 μg ml⁻¹ chloramphenicol, 50 μg ml⁻¹ kanamycin, or 100 μg ml⁻¹ ampicillin sodium was added to the medium when required. The shake flask fermentation was conducted statically in 250-ml flasks containing 100 ml fermentation medium at 30°C. The shake flask fermentation experiments were conducted in triplicate.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Description	Reference or source
Strains
L. lactis
F44	Derived from L. lactis YF11	44
F44::pLgt	F44 harboring pLE1	This study
F44::pLspA	F44 harboring pLE2	This study
F44::pNisI	F44 harboring pLE3	This study
F44::pLgt-NisI	F44 harboring pLE13	This study
F44::pLspA-NisI	F44 harboring pLE23	This study
F44::pLgt-LspA	F44 harboring pLE12	This study
F44::pLgt-LspA-NisI	F44 harboring pLE123	This study
F44 Δlgt	Derivative of F44 containing a lox72 replacement of lgt	This study
F44 ΔlspA	Derivative of F44 containing a lox72 replacement of lspA	This study
FC	F44 harboring pLE301	This study
ΔC1	F44 Δlgt harboring pLE301	This study
ΔC2	F44 ΔlspA harboring pLE301	This study
E. coli
TG1	Cloning host	Laboratory stock
BL21(DE3)	Protein purification	Laboratory stock
Micrococcus flavus	Indicator strain for bioassay of nisin	ATCC 10240
Plasmids
pLEB124	Expression vector; Emr	54
pLE1	pLEB124 carrying the lgt gene from F44; Emr	This study
pLE2	pLEB124 carrying the lspA gene from F44; Emr	This study
pLE3	pLEB124 carrying the nisI gene from F44; Emr	This study
pLE13	pLEB124 carrying the lgt and nisI genes from F44; Emr	This study
pLE23	pLEB124 carrying the lspA and nisI genes from F44; Emr	This study
pLE12	pLEB124 carrying the lgt and lspA genes from F44; Emr	This study
pLE123	pLEB124 carrying the lgt, lspA, and nisI genes from F44; Emr	This study
pLE301	pLEB124 carrying the nisI gene from F44 and 6×His; Emr	This study
pNZ5319	For gene deletion; Emr Cmr	55
pNZ11	pNZ5319 containing homologous region up- and downstream of F44 lgt; Emr Cmr	This study
pNZ22	pNZ5319 containing homologous region up- and downstream of F44 lspA; Emr Cmr	This study
pNZTS-Cre	Carries the cre gene; Emr	56
pET22b	Expression vector; Ampr	Novagen
pET22b-Pre	pET22b carrying the nisI gene from F44; Ampr	This study
pET22b-LF	pET22b carrying the nisI gene from F44 not containing signal peptide; Ampr	This study
pET28a	Expression vector; Kanr	Novagen
pET28a-1	pET28a carrying the lgt gene from F44; Kanr	This study
pET28a-2	pET28a carrying the lspA gene from F44; Kanr	This study
pET28a-3	pET28a carrying the nisI gene from F44; Kanr	This study

Construction of plasmids and strains.

All the plasmids constructed in this study are summarized in Table 1. All the primers used for cloning were synthesized by GENEWIZ Inc. (Beijing, China) and are listed in Table 2. Genomic DNAs were isolated from L. lactis F44 using the TIANamp bacterial DNA kit, purchased from Tiangen Biotech (Beijing) Co., Ltd. All of the genes used in this study were amplified from genomic DNAs of L. lactis F44.

TABLE 2.

Primers used in this study

Primer	Sequence (5′–3′)
lgt-F	CCCAAGCTTGACCGAAAATTCTGTCAGTAAAT
lgt-R	CGGGATCCTTAATTTTTTTTCATTCTGCGA
LspA-F	CCCAAGCTTAAACTGTTCTAGCGAGCTATC
LspA-R	CGCGGATCCATGTTAAATAAAACTTTCTGTCAG
NisI-F	CGCGGATCCCTTATTGGAGACAAGCACTGTTA
NisI-R	CATGCCATGGCTAGTTTCCTACCTTCGTTGC
O-lgt-F	CCCAAGCTTGACCGAAAATTCTGTCAGTAAAT
O-lgt-R	CTAGAACAGTTTTTAATTTTTTTTCATTCTGCGA
O-lspA-F	AAAAAAAATTAAAAACTGTTCTAGCGAGCTATCA
O-lspA-R	CGGGATCCTTATTTTGTTTCTTTGTCTGTCAA
L-nisI-F	CGGGATCCCTTATTGGAGACAAGCACTGTTA
L-nisI-R	CATGCCATGGCTAGTTTCCTACCTTCGTTGC
lgt-up-F	CCGCTCGAGATAGGGGAGACTTACAATGG
lgt-up-R	AGCTTTGTTTAAACATTTTTTATCAGTTTTCTAAATTT
lgt-down-F	ATCGATCGTTGTTCTCTTTATTTATCGCAG
lgt-down-R	GAAGATCTAACGACAGCATCTTTAGTATCAT
F44-LspA-Up-F	CCGCTCGAGTTCTTTGAAATGTGACTGCG
F44-LspA-Up-R	AGCTTTGTTTAAACGCTTTAATTATAGCATATTTCGC
F44-LspA-Down-F	TCCCCCGGGATTGATTAAAAAATCACTGACAGAA
F44-LspA-Down-R	ATCGATCGAACTCACAATGGCTGTCATTAA
NisI-22-F	GGAATTCCATATGAGAAAATATTTAATACTTATTGTG
NisI-22-R	CCGCTCGAGGTTTCCTACCTTCGTTGCA
LF-NisI-22-F	GGAATTCCATATGTATCAAACAAGTCAAAAAAAGG
LF-NisI-22-R	CCGCTCGAGGTTTCCTACCTTCGTTGCA
NisI-ch-F	GGAATTCCATATGATGAGAAAATATTTAATACTTATTGTG
NisI-ch-R	CGGGATCCCTAGTTTCCTACCTTCGTTGC
LspA-ch-F	GGAATTCCATATGATGAAAAAACTACTGTCACTTGTTA
LspA-ch-R	CCGCTCGAGTTATTTTGTTTCTTTGTCTGTCAA
Lgt-ch-F	GGAATTCCATATGATGAATAATCTATTTCCCTTTTTAG
Lgt-ch-F	CCGCTCGAGTTAATTTTTTTTCATTCTGCG
NisI-wb-F	CGCGGATCCCTTATTGGAGACAAGCACTGTTA
NisI-wb-R	CATGCCATGGCTAGTGGTGGTGGTGGTGGTGGTTTCCTACCTTCGTTGC

The lgt and lspA genes were separately cloned into the expression vector pLEB124 using restriction sites HindIII and BamHI, resulting in pLE1 and pLE2. The lgt and lspA genes were linked by overlap PCR, resulting in lgt-lspA, and subsequently cloned into pLEB124 using restriction sites HindIII and BamHI, yielding pLE12. The nisI gene was cloned into pLEB124, pLE1, pLE2, and pLE12 using restriction sites BamHI and NcoI, resulting in pLE3, pLE13, pLE23, and pLE123 (see Fig. S1 in the supplemental material).

For construction of the knockout vectors of target gene, a 1,000-bp fragment of the upstream sequence and a 1,000-bp fragment of the downstream sequence of target gene were amplified by PCR. Subsequently, the fragments of the upstream and downstream sequences were separately ligated into the XhoI-PmeI and SacI-BglII restriction sites of pNZ5319. The resulting plasmid was transformed into electroporation-competent L. lactis F44 to obtain the mutant containing a lox66-P₃₂-cat-lox71 replacement of the target gene. Plasmid pNZTS-Cre was transformed into the mutant, yielding the desired strains F44 Δlgt and F44 ΔlspA (see Fig. S2 in the supplemental material).

Nisin resistance assay.

Overnight stationary-phase cultures of L. lactis were harvested by centrifugation at 6,580 × g for 5 min and washed twice with 0.9% NaCl. Then L. lactis cells were resuspended in an equal volume of 0.9% NaCl, serially diluted 10⁷-fold in 10-fold increments, and then plated on seed medium agar plates containing 4,000, 8,000, and 12,000 IU ml⁻¹ nisin. Plates were incubated at 30°C for 48 h. The high-purity nisin (>99.5%) was kindly provided by Chihon Biotechnology Co., Ltd. (Luoyang, China).

Purification of proteins.

The lgt, lspA, and nisI genes used in this study were amplified from genomic DNAs of L. lactis F44 and cloned into the expression vector pET28a using restriction sites NdeI and XhoI, respectively. The pET28a-1/2/3 recombinant plasmid was transformed into E. coli BL21(DE3) by electroporation transformation. Cells containing pET28a-1/2 were incubated at 37°C to an optical density at 600 nm (OD₆₀₀) of 0.4 to 0.6 and induced with 0.1 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) at 16°C and 180 rpm for 18 h. The pre-NisI was produced in E. coli BL21(DE3) incubated at 37°C to an OD₆₀₀ of 0.6 and following induction with 0.1 mM IPTG at 28°C and 180 rpm for 2 h to avoid lipidation and cleavage by E. coli Lgt and Lsp.

Cells were harvested by centrifugation at 6,580 × g for 10 min at 4°C and washed twice with binding buffer (50 mM Tris-HCl [pH 8.0], 500 mM NaCl, 10 mM imidazole). Then cells were resuspended in binding buffer supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF) and sonicated on ice (5 s on, 5 s off, 20% power) until visual assessment indicated that the lysis was complete. Supernatants were separated into soluble and membrane fractions by centrifugation at 14,800 × g for 20 min at 4°C. The supernatant was used for protein purification by a GE Healthcare ӒKTA purifier system (GE Healthcare, USA). The sample was filtered through a 0.22-μm-pore-size filter prior to injection into a HisTrap column (GE Healthcare), prepacked with precharged Ni Sepharose 6 Fast Flow, by washing with washing buffer (50 mM Tris-HCl [pH 8.0], 500 mM NaCl, 20 mM imidazole) and eluting with elution buffer (50 mM Tris-HCl [pH 8.0], 500 mM NaCl, 500 mM imidazole). All the collected protein fractions were analyzed by SDS-PAGE, and protein was quantified using a bicinchoninic acid protein assay kit purchased from Solarbio (Beijing, China).

SDS-PAGE analysis.

Twenty microliters of each sample was mixed with 5× SDS-PAGE sample buffer (Beyotime, Beijing, China) and heated for 10 min at 99°C. Then 50 μg of protein was separated by electrophoresis on an SDS-containing 12% polyacrylamide gel at 100 V for 90 min. The gel was subsequently stained with Coomassie blue R-250 and decolorized with decoloring solution (ethanol-acetate-distilled water at 1:1:8).

Endopeptidase activity assay of Lsp.

To assay the proteolytic activity of Lsp, the purified proteins were desalted by washing with phosphate-buffered saline (PBS; pH 7.2). Then a 240-μl reaction mixture was set up containing 0.1 mg liter⁻¹ Lgt, 0.1 mg liter⁻¹ Lsp, 2 mg liter⁻¹ pre-NisI, and 150 mg liter⁻¹ DOPG in PBS and incubated at 30°C. Samples were taken from the reaction mixture at 0 h, 16 h, and 20 h and then subjected to SDS-PAGE.

Immunoblotting.

Proteins were separated by SDS-PAGE as described above and then transferred to a methanol-presoaked polyvinylidene difluoride (PVDF) membrane using a Bio-Rad semidry transfer cell. Immunoblotting was carried out with anti-His tag antibody and anti-rabbit IgG antibody and detected with the ECL system (Amersham Biosciences). Quantitation of protein on gels was done by image analysis using ImageJ (53).

SPR analysis.

The experiments were performed using a Reichert 4SPR refractometer system (Reichert Technologies, Depew, NY) equipped with a dextran SPR sensor chip (Reichert; carboxymethyl dextran P/N 13206066). The surface was stabilized with PBS-Tween (PBST) buffer (pH 7.4) at a flow rate of 25 μl min⁻¹ (25°C) until a constant refractive index was obtained. Then 275 μg ml⁻¹ nisin in 10 mM sodium acetate (pH 4.5) was immobilized over the sensor chip surface. LF-NisI or pre-NisI solution at different concentrations (0.156, 0.3125, 0.625, 1.25, and 2.5 μM) prepared in PBST was injected into the cell over both channels at a flow rate of 25 μl min⁻¹ with 1 min of association and 2 min of dissociation. Data analysis was performed with TraceDrawer software.

Statistical analysis.

Statistical analysis was conducted using SPSS 18.0 software (SPSS, Chicago, IL). All experiments were repeated at least three times. One-way analysis of variance (ANOVA) was performed to estimate the differences for OD₆₀₀ and cell counts between control and experimental groups. A P value less than 0.05 was considered statistically significant.