Binding Specificity of the Lantibiotic-Binding Immunity Protein NukH

Abstract

NukH is a lantibiotic-binding immunity protein that shows strong binding activity against type A(II) lantibiotics. In this study, the binding specificity of NukH was analyzed by using derivatives of nukacin ISK-1, which is a type A(II) lantibiotic produced by Staphylococcus warneri ISK-1. Interactions between cells of Lactococcus lactis transformants expressing nukH and nukacin ISK-1 derivatives were analyzed by using a quantitative peptide-binding assay. Differences in the cell-binding rates of each nukacin ISK-1 derivative suggested that three lysine residues at positions 1 to 3 of nukacin ISK-1 contribute to the effective binding of nukacin ISK-1 to nukH-expressing cells. The binding levels of mutants with lanthionine and dehydrobutyrine substitutions (S11A nukacin_4-27 and T24A nukacin_4-27, respectively) to nukH-expressing cells were considerably lower than those of nukacin_4-27, suggesting that unusual amino acids play a decisive role in NukH recognition. Additionally, it was suggested that T9A nukacin_4-27, a mutant with a 3-methyllanthionine substitution, binds to NukH via an intermolecular disulfide bond after it is weakly recognized by NukH. We succeeded in the detection of specific type A(II) lantibiotics from the culture supernatants of various bacteriocin producers by using the binding specificity of nukH-expressing cells.

Bacteriocins are ribosomally synthesized antimicrobial peptides or proteins (31). Lantibiotics (class I bacteriocins) are produced by a variety of gram-positive bacteria. Lantibiotics are characterized by the presence of unusual amino acids such as dehydrated and lanthionine residues (5, 17, 19). These unusual amino acids confer high stability and various biological activities to peptides. Lantibiotic producers express immunity proteins on their cytoplasmic membrane for protection against self-produced lantibiotics. The self-immunity system consists of LanFEG and/or LanI. LanFEG is an ATP-binding cassette (ABC) transporter, which mediates the transport of lantibiotics from the cell membrane to the extracellular space (1, 4, 6, 10, 22, 23, 25, 28, 29). On the other hand, LanI is an immunity lipoprotein anchored to the membrane surface via a lipid-modified N-terminal cysteine residue (4, 11, 12, 16, 27-29, 32). One of the well-studied LanI proteins is NisI, which is involved in nisin immunity. The heterologous expression of NisI increases the amount of cell-bound nisin and the resistance level (28). These results suggest that NisI intercepts nisin before the latter attacks the cell membrane, thereby preventing pore formation. In addition, experiments using C-terminally truncated NisI mutants and a hybrid protein containing an immunity lipoprotein for subtilin (SpaI) have shown that the C terminus of NisI confers protection specifically against nisin (32). However, the details of the substrate-binding mechanism of NisI have not yet been explained.

Staphylococcus warneri ISK-1 produces a type A(II) lantibiotic, namely nukacin ISK-1 (3, 14, 15, 26). Type A(II) lantibiotics consist of two regions, a linear N-terminal region and a globular C-terminal portion containing dehydrated residues and lanthionine rings (5, 17, 19). Self-immunity for nukacin ISK-1 is conferred by a LanFEG-type protein, namely NukFEG, and a novel lantibiotic-binding immunity protein, namely NukH (3, 4). Previously, we demonstrated that NukFEG and NukH increase the immunity level of nukacin ISK-1-sensitive Lactococcus lactis; both proteins are required for full immunity (4). NukFEG transports cell-associated nukacin ISK-1, and NukH exhibits binding activity against nukacin ISK-1, as observed in the case of NisI (4). However, the features of NukH and NisI are quite different. NukH is a membrane protein with three transmembrane domains; the N terminus of the protein is on the cytoplasmic side of the cell membrane. Evaluation of the immunity levels and binding activities of various NukH mutants against nukacin ISK-1 has suggested that the entire structure of NukH, except for the N and C termini, is essential for its full immunity and that the third transmembrane helix is dispensable for its binding function (20). Recently, we demonstrated that a cooperative mechanism exists between NukFEG and NukH by using fluorescein-4-isothiocyanate (FITC)-labeled nukacin ISK-1 (FITC-nuk) (21). The results revealed that nukacin ISK-1 was captured by NukH and transported to the extracellular space by NukFEG in an energy-dependent manner. Additionally, it was suggested that the NukH protein recognizes the C-terminal ring region of nukacin ISK-1, since the interaction between FITC-nuk and nukH-expressing cells was inhibited by the truncation of the N-terminal region of nukacin ISK-1 but not by the addition of the synthesized N-terminal tail region.

In this study, the binding specificity of NukH was investigated by using nukacin ISK-1 derivatives. Nukacin_4-27 and nukacin_7-27—N-terminally truncated mutants of nukacin ISK-1—were bound to nukH-expressing cells after incubation with cells for 30 min; however, their binding rates were lower than that of nukacin ISK-1. Additionally, alanine substitutions of the unusual amino acids in nukacin ISK-1 decreased the binding activity of these amino acid substitution-containing mutants to nukH-expressing cells. We succeeded in detecting specific type A(II) lantibiotics from the culture supernatants of various bacteriocin producers by using the binding specificity of nukH-expressing cells.

MATERIALS AND METHODS

Bacterial strains and plasmids.

The bacterial strains used in this study are listed in Table 1. S. warneri ISK-1 and Pediococcus pentosaceus TISTR 536 were grown in MRS medium (Oxoid, Hampshire, United Kingdom) at 37°C. Lactobacillus sakei subsp. sakei JCM 1157^T, L. lactis subsp. lactis CNRZ 481, L. lactis subsp. lactis 61-14, and L. lactis subsp. lactis QU 5 were grown in MRS medium at 30°C. Staphylococcus hominis KQU-131 was grown in tryptic soy broth (TSB) medium (Difco, Sparks, MD) at 30°C. L. lactis NZ9000 was grown in M17 medium (Merck, Darmstadt, Germany) supplemented with 0.5% glucose at 30°C. By using a nisin-controlled expression system, nukH was expressed in L. lactis NZ9000 (7, 8). The methods for constructing pNZH, which is a pNZ8048 derivative containing nukH, have been described previously (20). For the selection of transformants carrying pNZH and its derivatives, we used chloramphenicol at a concentration of 10 μg/ml.

TABLE 1.

Bacterial strains used in this study

Strain	Description	Source or reference
Staphylococcus warneri ISK-1	Nukacin ISK-1 producer	26
Pediococcus pentosaceus TISTR 536	Pediocin PA-1 producer	TISTR, 30
Lactobacillus sakei subsp. sakei JCM 1157T	Indicator strain for bioassay	JCM
Lactococcus lactis subsp. lactis CNRZ 481	Lacticin 481 producer	24
L. lactis subsp. lactis 61-14	Nisin Q producer	35
L. lactis subsp. lactis QU 5	Lacticin Q producer	9
Staphylococcus hominis KQU-131	Nukacin KQU-131 producer	33
L. lactis NZ9000	MG1363 derivative; nisRK::pep	8
Escherichia coli BL21(DE3)	Expression host, F−ompT, hsdSB(rB− mB−) gal dcm (DE3)	Novagen

Preparation of nukacin ISK-1 derivatives.

Nukacin ISK-1 was purified as described previously (4). Two N terminus-deletion derivatives lacking three lysine residues at positions 1 to 3 (nukacin_4-27) or lacking amino acids at positions 1 to 6 (nukacin_7-27) were obtained according to the methods described previously (2). We constructed three derivatives of nukacin_4-27 containing unusual amino acid substitutions (T9A nukacin_4-27, S11A nukacin_4-27, and T24A nukacin_4-27) by alanine substitutions of the respective residues that will be converted to the unusual amino acids. We used a nukA and nukM coexpression system to obtain the derivatives (18). Site-directed mutations were introduced into the appropriate region of nukA by using a QuikChange site-directed mutagenesis kit (Stratagene Cloning System, La Jolla, CA), according to the manufacturer's protocol (T9A and S11A mutations), or by using inverse PCR according to the method described previously (T24A mutation) (20). The plasmid pETnukAM (18) was used as a template, and the fragments containing mutated nukA were amplified by PCR using the following set of primers: 5′-GGAGTAATCCCAGCGGTGTCACACGATTGCCATA-3′ and 5′-ATCGTGTGACACCGCTGGGATTACTCCTGACTTT-3′ (T9A mutation), 5′-ATCCCAACTGTGGCGCACGATTGCCATATGAATT-3′ and 5′-ATGGCAATCGTGCGCCACAGTTGGGATTACTCCT-3′ (S11A mutation), and 5′-TAAAAAGGGGAATGTTGTAAAATCTGT-3′ and 5′-TGAACAACACGCAAATACAAATTGGAAAGA-3′ (T24A mutation). The mutated nukA was expressed as His₆-tagged peptides in Escherichia coli BL21(DE3); these peptides were subsequently purified by nickel affinity chromatography by using a HisTrap FF column (Amersham Pharmacia Biotech, Uppsala, Sweden) and reverse phase high-performance liquid chromatography (RP-HPLC) by using an Asahipak C8P-50 4E column (Shodex, Tokyo, Japan). The molecular mass of each peptide was analyzed by electrospray ionization-mass spectrometry (Accutof T100LC; JEOL, Tokyo, Japan). The existence of free cysteine residues in T9A nukacin_4-27 and S11A nukacin_4-27 was confirmed by 1-cyano-4-dimethylamino-pyridinium tetrafluoroborate (CDAP) treatment, followed by molecular-mass analysis (34). To remove the His₆-tag and leader peptide, purified peptides were cleaved by lysyl endopeptidase (Wako, Osaka, Japan), as described previously (2). Subsequently, T9A nukacin_4-27, S11A nukacin_4-27, and T24A nukacin_4-27 were purified by RP-HPLC using a Resource RPC column (Amersham Pharmacia Biotech).

Peptide-binding assay using nukacin ISK-1 and its derivatives.

The time-dependent binding of nukacin ISK-1 and its derivatives was analyzed using a previously described method with some modifications (20). Recombinant L. lactis cells harboring the plasmid pNZ8048 and its derivatives containing nukH (pNZH) were induced for 3 h. The cells were then harvested and washed twice with 50 mM sodium phosphate buffer (pH 7.0) containing 1% glucose (buffer A). The cell concentration was adjusted to an optical density at 600 nm (OD₆₀₀) of 10 using buffer A. The cell suspension (1 ml) was incubated with 5 μg/ml of nukacin ISK-1 or its derivatives with shaking (150 rpm) for 0, 5, 10, or 30 min at 30°C. After incubation, the cells were centrifuged at 20,000 × g for 5 min. The harvested cell pellets were gently mixed with 1 ml of 20% acetonitrile in water containing 0.1% trifluoroacetic acid and incubated with shaking (150 rpm) for 5 min at 30°C. The cells were removed by centrifugation at 20,000 × g for 5 min, and the supernatants were retained. Quantitative determination of the peptides in the supernatants was carried out by RP-HPLC using a Resource RPC column. The peptides were eluted with a linear gradient of 20 to 60% acetonitrile in water containing 0.1% trifluoroacetic acid for 30 min. The flow rate was 1 ml/min, and the eluates were monitored by measuring their absorbance at 220 nm.

Construction of the NukH mutant (NukHC43A).

A cysteine residue at position 43 of NukH was replaced by alanine using a QuikChange site-directed mutagenesis kit according to the manufacturer's protocol. The plasmid pNZH (20) was used as a template, and the fragments containing mutated nukH were amplified by PCR by using the following primer set: 5′-GTTAATATCATCGCGATACTTTTTACTATTGCTT-3′ and 5′-AGTAAAAAGTATCGCGATGATATTAACTATAGAG-3′. Transformation of L. lactis NZ9000 by using the purified DNA was performed according to the method developed by Holo and Nes (13).

Detection of type A(II) lantibiotic from culture supernatants of bacteriocin producers.

A cell suspension of L. lactis NZ9000 harboring the plasmid pNZ8048 or pNZH was prepared as mentioned above and adjusted to an OD₆₀₀ of 20 by using buffer A containing 1 M NaCl. Subsequently, 400 microliters of the cell suspension was mixed and incubated with the same volume of bacteriocin-containing culture supernatants of S. warneri ISK-1 (15, 26), L. lactis subsp. lactis CNRZ 481 (24), S. hominis KQU-131 (33), L. lactis subsp. lactis 61-14 (35), P. pentosaceus TISTR 536 (30), and L. lactis subsp. lactis QU5 (9) or TSB medium containing 50 μg/ml of purified mersacidin (as a substitution for the culture supernatant) for 30 min at 30°C. After incubation, the cells were harvested by centrifugation at 20,000 × g for 5 min and washed with buffer A. The harvested cell pellets were mixed with 400 μl of 20% acetonitrile in water containing 0.1% trifluoroacetic acid and incubated with shaking (150 rpm) for 5 min at 30°C. The cells were removed by centrifugation at 20,000 × g for 5 min, and the supernatants were retained. To detect bacteriocin activity in the supernatants, 10 μl of each supernatant was spotted on soft agar (Lactobacilli Agar AOAC; Difco Laboratories, Detroit, MI) that had been seeded with an overnight culture of L. sakei subsp. sakei JCM 1157^T (100-fold dilution). After incubation at 30°C for 18 h, a clear zone was observed in the culture plates.

RESULTS

Construction of nukacin ISK-1 derivatives.

To analyze the binding specificity of NukH, we constructed various nukacin ISK-1 derivatives. The structures of nukacin ISK-1 and its derivatives used in this study are shown in Fig. 1. Previously, we showed that nukacin_4-27 and nukacin_7-27, which are N-terminally truncated derivatives, exhibit suppressed and no antimicrobial activity, respectively (2). In this study, we constructed nukacin ISK-1 derivatives by using an E. coli nukA and nukM coexpression system; these derivatives contained mutations at unusual amino acids located in the ring region of nukacin ISK-1 (18). Using this coexpression system, all derivatives were expressed with leader peptides; however, the leader peptides were removed together with three lysine residues at the N terminus by lysyl endopeptidase treatment that cleaves the C-terminal region of lysine residues. To construct T9A nukacin_4-27 and S11A nukacin_4-27, threonine at position 9 and serine at position 11, which are dehydrated and cyclized, respectively, were replaced by alanine. The CDAP treatment followed by molecular-mass analysis indicated that both derivatives contained a newly introduced alanine and a free cysteine residue (data not shown). Similarly, when threonine at position 24 was substituted with alanine, T24A nukacin_4-27 was obtained. These three derivatives with unusual amino acid substitutions did not show any antimicrobial activities against L. lactis transformants used in this study (data not shown).

FIG. 1.

Putative structures of nukacin ISK-1 and its derivatives. The dark-shaded residues indicate modified amino acids. Ala-S-Ala, lanthionine; Abu-S-Ala, 3-methyllanthionine; and Dhb, dehydrobutyrine.

Three lysine residues in the N-terminal region of nukacin ISK-1 contribute to the effective binding of nukacin ISK-1 to nukH-expressing cells.

To investigate the function of the tail region of nukacin ISK-1 on its binding to recombinant L. lactis cells, the binding rates of the N-terminally truncated derivatives of nukacin ISK-1 were analyzed using a peptide-binding assay. It was found that nukacin ISK-1 bound to control cells (harboring pNZ8048) to a greater extent than did nukacin_4-27 and nukacin_7-27 (Fig. 2A). Notably, the binding rate of nukacin ISK-1 was very high; i.e., ca. 40% of the added nukacin ISK-1 was collected from the cell fraction immediately after the initiation of incubation. This result corresponds to that of our previous study, indicating that the three lysine residues in the N-terminal region of nukacin ISK-1 contribute to the interaction between nukacin ISK-1 and the cytoplasmic membrane (2). We presume that there is no degradation of peptides during this experiment, because remaining peptides were recovered (>90%) from the supernatants after 30 min of incubation (data not shown). On the other hand, the amounts of all three peptides bound to nukH-expressing cells were increased during incubation; consequently, almost all of the added peptides were bound to the cells after 30 min (Fig. 2B). However, the binding rates of the three peptides were different (nukacin ISK-1 > nukacin_7-27 > nukacin_4-27). According to a previous study, nukacin ISK-1 inhibits the binding of FITC-nuk, which interacts specifically with nukH-expressing cells, to a greater extent than nukacin_7-27 and nukacin_4-27 (21). Taken together, we conclude that the lysine residues at positions 1 to 3 of nukacin ISK-1 play an important role in electrostatic interaction between nukacin ISK-1 and the cytoplasmic membrane and are consequently involved in the interaction with NukH located at the cytoplasmic membrane.

FIG. 2.

Time-dependent peptide-binding assay using nukacin ISK-1 and its derivatives. Cell suspensions of L. lactis harboring the plasmid pNZ8048 (A) or pNZH (B) were incubated with nukacin ISK-1 (•), nukacin_4-27 (▵), and nukacin_7-27 (▴) at concentrations of 5 μg/ml with shaking for 0, 5, 10, and 30 min at 30°C. After incubation, the cells were harvested by centrifugation at 20,000 × g for 5 min. The cell pellets were gently mixed with 1 ml of 20% acetonitrile in water containing 0.1% trifluoroacetic acid and incubated with shaking for 5 min at 30°C. The cells were removed by centrifugation at 20,000 × g for 5 min, and the supernatants were retained. The concentration of nukacin ISK-1 in the supernatants was determined by RP-HPLC. Relative values calculated from the total amount of peptides added (5 μg) have been represented. This assay was performed in duplicate, and the average values are plotted with error bars.

Importance of unusual amino acids of nukacin ISK-1 for interaction with NukH.

Next, we investigated the importance of the unusual amino acids in the ring region of nukacin ISK-1 with regard to the interaction between nukacin ISK-1 and NukH by using derivatives with unusual amino acid substitutions, as shown in Fig. 1. In the peptide-binding assay, all peptides did not bind to a significant degree to the control cells (<20%), and the added peptides were collected at a high recovery rate (>90%) (data not shown). The amounts of the S11A nukacin_4-27 and T24A nukacin_4-27 peptides bound to nukH-expressing cells were considerably lower than those of the nukacin_4-27 peptides (Fig. 3A). It seems that small amounts of the S11A nukacin_4-27 and T24A nukacin_4-27 peptides were bound to the membranes of nukH-expressing cells nonspecifically, since there were not much differences in the amounts of the cell-associated peptides between control and nukH-expressing cells (data not shown). These results indicate that lanthionine and dehydrobutyrine residues in the ring region of nukacin ISK-1 are important for the recognition of nukacin ISK-1 by NukH. It should be noted that the rate of recovery of the added peptides (nukacin_4-27, S11A nukacin_4-27, and T24A nukacin_4-27) from nukH-expressing cells was high (>90%); however, the recovery rate of T9A nukacin_4-27 was very low (ca. 50%) (Fig. 3A). From these results, it was speculated that T9A nukacin_4-27 that bound to the cells was not fully recovered even after washing with 20% acetonitrile because the binding mechanism of T9A nukacin_4-27 is different from those of other peptides, perhaps involving disulfide bond formation between cysteine residues.

FIG. 3.

Quantitative peptide-binding assay using nukH-expressing cells and nukacin ISK-1 derivatives. After the induction of recombinant L. lactis cells harboring pNZH (A) or pNZHC43A (B), the cell suspensions (OD₆₀₀ = 10) were incubated with nukacin ISK-1 derivatives at a concentration of 5 μg/ml with shaking for 30 min at 30°C. The amounts of nukacin ISK-1 derivatives remaining in the supernatants (white bars) and the amounts of cell-bound nukacin ISK-1 derivatives (black bars) were determined by RP-HPLC. This assay was performed in duplicate, and the average values are represented with error bars.

Disulfide bond formation between two cysteine residues, one each in T9A nukacin_4-27 and NukH.

NukH contains a cysteine residue at position 43, located in the middle of the second transmembrane helix (20). Similarly, T9A nukacin_4-27 also contains a free cysteine residue at position 14 (Fig. 1). Based on these facts, we hypothesized that the low recovery rate of T9A nukacin_4-27 might be due to the formation of an intermolecular disulfide bond between these cysteine residues. As a first attempt, we added 10 to 50 mM dithiothreitol into the cell-washing solution to disrupt the complex through chemical reduction of the disulfide bond; however, we could not recover a large enough amount of the peptide from nukH-expressing cells (data not shown). We speculated that this is due to the low effect of dithiothreitol through the membrane. Therefore, we replaced the cysteine residue in NukH with alanine, and the obtained mutant (NukHC43A) was expressed in L. lactis NZ9000. In the peptide-binding assay performed by using nukHC43A-expressing cells, the binding profiles and recovery rates of the nukacin_4-27, S11A nukacin_4-27, and T24A nukacin_4-27 peptides were observed to be almost similar to those obtained in the experiments using nukH-expressing cells (Fig. 3B). This indicates that the cysteine residue at position 43 is not important for the binding function of NukH. On the other hand, the amount of the T9A nukacin_4-27 peptide recovered from the supernatant fraction was increased drastically, resulting in a recovery rate of 90%. Thus, T9A nukacin_4-27, after being recognized by NukH, binds to it via a disulfide bond.

Rapid screening of type A(II) lantibiotic producers.

According to the previous result (4), nukH-expressing cells show binding activity against nukacin ISK-1 and lacticin 481, which are type A(II) lantibiotics, but not against nisin, which is a type A(I) lantibiotic. Therefore, by using the binding specificity of nukH-expressing cells against type A(II) lantibiotics, we attempted to establish a screening system for type A(II) lantibiotics from the culture supernatants of bacteriocin producers. The following bacteriocins were obtained from the bacteriocin-producing strains: type A(II) lantibiotic (nukacin ISK-1, lacticin 481, and nukacin KQU-131), type A(I) lantibiotic (nisin Q), and class IIa bacteriocin (pediocin PA-1 and lacticin Q). The culture supernatants and TSB medium containing type B lantibiotics (mersacidin) exhibited antimicrobial activities against L. sakei subsp. sakei JCM 1157^T (Fig. 4A). Active supernatants were incubated with control cells (L. lactis harboring pNZ8048) or nukH-expressing cells; subsequently, the antimicrobial activities of the cell-bound peptides were analyzed. In this assay, NaCl was added to a final concentration of 0.5 M to reduce nonspecific binding of the bacteriocins to the cells (4, 22, 28, 29). It was found that the activities of nukacin ISK-1, lacticin 481, and nukacin KQU-131 from nukH-expressing cells were remarkably higher than those from control cells (Fig. 4B). On the other hand, no clear difference was observed between control cells and nukH-expressing cells treated with nisin Q and lacticin Q (Fig. 4B). This result indicates that these two bacteriocins bound to the cells regardless of nukH expression. The samples extracted from the cells treated with both mersacidin and pediocin PA-I showed considerably lower binding activity (Fig. 4B). Taken together, we concluded that nukH-expressing cells show high binding specificity against type A(II) lantibiotic, and this can be applied to the screening of type A(II) lantibiotics from culture supernatants of various bacteriocin producers.

FIG. 4.

Rapid screening of type A(II) lantibiotics from culture supernatants of bacteriocin producers. (A) Antimicrobial activities of the supernatants of bacteriocin producers against L. sakei subsp. sakei JCM 1157^T. 1, S. warneri ISK-1 (nukacin ISK-1); 2, L. lactis subsp. lactis CNRZ 481 (lacticin 481); 3, S. hominis KQU-131 (nukacin KQU-131); 4, L. lactis subsp. lactis 61-14 (nisin Q); 5, TSB medium containing 50 μg/ml of purified mersacidin; 6, L. lactis subsp. lactis QU 5 (lacticin Q); and 7, P. pentosaceus TISTR 536 (pediocin PA-1). (B) Antimicrobial activities of cell-bound bacteriocins against L. sakei subsp. sakei JCM 1157^T. IA(II), IA(I), IB, and II indicate type A(II) lantibiotic, type A(I) lantibiotic, type B lantibiotic, and class II bacteriocin, respectively.

DISCUSSION

In this study, we investigated the lantibiotic-binding mechanism of NukH by using nukacin ISK-1 and its derivatives. A time-dependent peptide-binding assay performed using N-terminally truncated nukacin ISK-1 derivatives and nukH-expressing cells clearly indicated that NukH recognizes the C-terminal ring region of nukacin ISK-1. After incubation for 30 min, most nukacin ISK-1 variants bound to nukH-expressing cells, but their binding rates were different (nukacin ISK-1 > nukacin_7-27 > nukacin_4-27) (Fig. 2B). It should be noted that after 30 min of incubation, large amounts of nukacin ISK-1 (ca. 40% of the added peptide) rapidly bound to the control cells, while nukacin_7-27 and nukacin_4-27 hardly bound to the control cells (Fig. 2A). This finding raises a question about the binding target of nukacin ISK-1 in nukH-expressing cells, i.e., whether nukacin ISK-1 interacts with the cell membrane or NukH in nukH-expressing cells. Previously, we investigated the comparative effect of the three peptides on the binding of FITC-nuk, which specifically interacts with nukH-expressing cells (21). We measured cell fluorescence after incubating FITC-nuk and nukH-expressing cells in the presence of each of the three peptides at various concentrations for 60 min. It was found that, at any concentration, nukacin ISK-1 showed the highest inhibitory effect, followed by nukacin_7-27 and nukacin_4-27 (21). This finding suggests that nukacin ISK-1 interacts with NukH rather than the cell membrane of nukH-expressing cells. Taken together, lysine residues at positions 1 to 3 of nukacin ISK-1 appear to be involved in rapid interaction with the cytoplasmic membrane, thereby contributing to the effective binding of NukH to nukacin ISK-1.

Next, the relevance of the structure of the ring region in nukacin ISK-1 to its recognition by NukH was investigated by using unusual amino acid-substituted nukacin ISK-1 derivatives. The level of binding to nukH-expressing cells was decreased by S11A and T24A mutation (Fig. 3A), indicating that lanthionine and dehydrobutyrine at positions 11 and 24, respectively, are important for the recognition of nukacin ISK-1 by NukH. Since the S11A mutation is expected to have an effect on the structure of the peptide, it can also be said that the conformation of the ring region stabilized by the lanthionine ring might be important for the recognition. T9A mutation also resulted in a decrease in the amount of cell-bound peptide; however, the recovery rate was lower than those of the other three derivatives. It was speculated that this phenomenon was due to the formation of an intermolecular disulfide bond between two cysteine residues, one each in T9A nukacin_4-27 and NukH. As expected, mutation in a single cysteine residue in NukH resulted in an increase in the recovery rate (Fig. 3B). Based on these results, we hypothesized the following mechanism underlying the binding of T9A nukacin_4-27 by NukH: NukH recognizes T9A nukacin_4-27, and subsequently, a disulfide bond is formed between the two cysteine residues, one each in NukH and T9A nukacin_4-27. Without disulfide bond formation, this interaction appears to be very weak even if T9A nukacin_4-27 is recognized by NukH, because we have detected most portions of T9A nukacin_4-27 in the supernatant fraction in this experiment by using nukHC43A-expressing cells (Fig. 3B). S11A nukacin_4-27 also contains a free cysteine residue; however, this free cysteine residue is not bound to nukH-expressing cells. Compared to the T9A mutation, the S11A mutation might cause drastic structural changes in the ring region of nukacin ISK-1, preventing the recognition of S11A nukacin_4-27 by NukH. The difference between the positions of cysteine residues (C14 and C25) of T9A and S11A mutants might also be a reason for this phenomenon. Based on the T9A nukacin_4-27-NukH binding model, we can speculate that the binding mechanism of nukacin ISK-1 to NukH, i.e., the ring region of nukacin ISK-1, might be inserted into the membrane, which consequently may interact with the second transmembrane helix of NukH.

We established a novel screening method for type A(II) lantibiotics from culture supernatants of bacteriocin producers by using nukH-expressing cells. By using this method, we successfully identified lacticin 481, which contains the same unusual amino acids at the same positions as nukacin ISK-1 and four amino acids that are different from those present in the ring region of nukacin ISK-1 (Fig. 4B). Similarly, nukacin KQU-131, in which lanthionine and methionine are replaced by 3-methyllanthionine and valine, respectively, in the ring region of nukacin ISK-1 could also be identified. These results suggest that NukH recognizes the characteristic ring pattern of type A(II) lantibiotics and that this system can be applied to the screening of various type A(II) lantibiotics. To date, many bacteriocins have been isolated and characterized. However, structural determination of bacteriocins (e.g., purification, DNA sequencing of the structural gene, and amino acid sequencing) is time and energy consuming. The system proposed by us might enable rapid screening of type A(II) lantibiotics from various bacteriocin-containing culture supernatants.