INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

Lantibiotics are antibiotic peptides that contain the thioether amino acids lanthionine and/or methyllanthionine (37). The tricyclic peptide Pep5 (38) is produced by Staphylococcus epidermidis 5 and is classified along with nisin (19), subtilin (18), and epidermin (1) as a type A lantibiotic. This group of peptides comprises screw-shaped, positively charged, amphipathic molecules which exert their primary bactericidal action by the formation of pores in the cytoplasmic membrane of sensitive bacteria (36). Like all other known lantibiotics Pep5 is ribosomally synthesized. The biosynthetic gene cluster is located on the 20-kb plasmid pED503 (14, 22) and consists of the structural gene pepA as well as the genes encoding proteins required for posttranslational modification (pepB and pepC), proteolytic processing (pepP), transport (pepT), and immunity (pepI) (28). The Pep5 immunity gene pepI codes for a 69-amino acid peptide, which is characterized by a hydrophobic N-terminal segment and a strongly hydrophilic C-terminal part. PepI displays a high degree of similarity (74.2%) to EciI, the epicidin 280 immunity peptide (20), and confers cross-immunity to epicidin 280, suggesting a similar molecular self-protection mechanism for both lantibiotics. The nisin and subtilin immunity proteins NisI and SpaI consist of 245 and 165 amino acids, respectively, and have typical lipoprotein consensus sequences (23, 26). However, NisI and SpaI share no sequence homology and do not provide cross-immunity.

An important contribution to producer protection is apparently provided by dedicated ATP-binding cassette transporter systems. Such transporters have been identified in the nisin (43), subtilin (23), epidermin (29), and lacticin 481 (34) gene clusters and typically consist of two or three separate proteins, LanE, LanF, and/or LanG. NisFEG, SpaFG, EpiFEG, and LctFEG are thought to be located in the membrane and to act by exporting the lantibiotic out of the cytoplasmic membrane, thus keeping the concentration below a critical level (29). In contrast to these transporters, there are currently no theories that address how the lipoproteins NisI and SpaI or the small immunity peptides PepI and EciI could function at the molecular level to reduce producer strain sensitivity to the respective lantibiotics. However, PepI and EciI do not seem to be the only members of this unique class of immunity peptides. Recently, in the gene clusters of the structurally unrelated lantibiotic lactocin S (44) and the nonlantibiotic divergicin A (48) two genes were discovered which code for peptides of similar size, charge distribution, and significant sequence similarity (20). Particularly the presence of a related gene in a nonlantibiotic gene cluster suggests that such an immunity mechanism could be of general importance for bacteriocins of gram-positive bacteria.

Here, we report the molecular characterization of the Pep5 immunity system. One striking feature of the Pep5 system is the apparent coupling of the immunity phenotype with the production of Pep5, which led us to assume that pepA or its product, i.e., the unmodified prepeptide or as reported for nisin (25, 26) the mature lantibiotic, could be involved in the expression of this phenotype (32). We now demonstrate that pepI is sufficient for expression of Pep5 immunity. Coupling to Pep5 production is achieved at the transcriptional level through the stabilization of pepI-containing transcripts by means of an inverted repeat, which in the wild type, is located downstream of pepA. The presence of the terminator element rather than its position in the transcript was found to be important for mRNA stabilization, allowing the construction of hyperimmune strains and eventually hyperproducer strains for biotechnological purposes.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Bacterial strains and plasmids. The bacterial strains and plasmids used are listed in Table 1. All staphylococcal strains were maintained on blood agar or tryptone soy agar supplemented with the appropriate antibiotic. Staphylococcus carnosus TM300 (17) was used for heterologous expression, Escherichia coli BMH 71-18 (carrying mutS) was used as the host for recombinant DNA. The vector pALTER-1 (Altered Sites in vitro mutagenesis system; Promega, Madison, Wis.) was used for site-directed mutagenesis. The E. coli-Staphylococcus shuttle vector pCU1 and the staphylococcal expression vector pCX15 were kindly provided by R. Rosenstein and F. Götz, Tübingen, Germany.

DNA cloning and sequencing. Staphylococcal plasmid DNA was isolated by the method of Feliciello and Chinali (15). Lysis of staphylococci was achieved by adding 200 µg of lysostaphin per ml to solution I (50 mM glucose, 10 mM Tris-HCl [pH 8.0], 1 mM EDTA) and incubating the cells for 30 to 60 min at 37°C. Plasmid DNA of E. coli was purified with QIAprepspin columns (Qiagen, Hilden, Germany). E. coli and S. epidermidis 25 were transformed by electroporation (2), and S. carnosus protoplasts were transformed by the method of Götz and Schumacher (17). Double-stranded plasmid DNA was sequenced on an A.L.F. DNA sequencer (Pharmacia, Uppsala, Sweden) by the dideoxynucleotide chain termination method (40) with the AutoRead sequencing kit (Pharmacia). Restriction enzymes and T4 DNA ligase were supplied by Boehringer (Mannheim, Germany), and synthetic oligonucleotides were obtained from Eurogentec (Seraing, Belgium).

PCR amplification. For PCR amplification of pepI, pepA, and the terminator structure, we used pMR2 as a DNA template, Pwo DNA polymerase (Boehringer) or Goldstar Red DNA polymerase (Eurogentec), and the following primer pairs: pepI-SalI 5' [5'(CAATAATAAATGTCGACTTAGGCCATTTAATTTTTG)3'] and pepI-XbaI 3' [5'(CATTTTCTAGAATTAATTATTTAAACATACAAAG)3'], pepI-XbaI 5' [5'(ATTTTCTAGAAGGTATTAAAAAAATTTTAC)3'] and pepI-SalI 3' [5'(TATAATGTCGACACAATATAGAAAAAAAC)3'], pepA-EcoRI 5' [5'(GTTTAAAGAATTCATTATAAAAAATGTATTG)3'] and pepA-XbaI 3' [5'(GATTACTCTAGATTTTTTCTCCTGCATAC)3'], and Term-XbaI 5' [5'(CAGTGTCTAGAAAAGGAAAAAACGGATG)3'] and Term-EcoRI 3' [5'(TATTTGAATTCCATGCCCAGTGTAATCAC)3'] for amplifying the terminator structure (bold letters indicate deviations from the original sequence; the generated restriction sites are underlined). The correct introduction of the mutations was verified by sequencing the entire PCR product.

Site-directed mutagenesis. Site-directed mutagenesis of PepI and PepA was performed with a commercial phagemid system, as previously described for Pep5 mutant peptides (5). The following mutagenic oligonucleotides were employed (mismatches are underlined): 5'(CTTTATTTTTTGCTTTAAGAATTTTTATTGTAACTTAT)3' changes Ile17 of PepI into Arg, 5'(AAGAATAAATAGCAACTAAAAAGATAAACTTTAG)3' mutates Lys65 into a stop codon, and 5'(GTAATTTTAACTTCTTTAGATTTTGCTTTAAGAATTTT)3' was designed to change Phe13 into Asp. The oligonucleotide 5'(GAGGAGGTGGTTATATGGGAAAAATAACAAAAATT)3' exchanges the Met start codon of PepA for a Gly codon.

RNA isolation and Northern hybridization. Total RNA was prepared by using the RNeasy minikit (Qiagen) with the following modifications for isolation of RNA from staphylococci. The wild-type strain S. epidermidis 5 and mutants were grown in tryptic soy broth (Oxoid, Wesel, Germany) to an absorbance at 600 (A₆₀₀) nm of 1. Cultures (10 ml) were harvested by centrifugation (5,000 × g for 5 min at 4°C). The cells were resuspended in 600 µl of Tris-EDTA (pH 8.0) and, after addition of 0.4 mg of lysostaphin per ml and 32 U of RNAguard (Pharmacia), were incubated for 15 min at 37°C. After addition of 2.1 ml of lysis buffer RLT (Qiagen), the sample was mixed vigorously and centrifuged (2 min at 8,000 × g) to remove unlysed cells. The supernatant was mixed with 1.5 ml of ethanol (100%) and applied onto a spin column in several centrifugation steps. The following steps were performed according to the manufacturer's instructions.

Expression of pepI from the xylose-inducible pCX15 vector. For cloning in pCX15, pepI was amplified by PCR with pMR2 as a template and the following primers: pepI-BglII 5' [5'(CTATAAGATCTTCTAAATATATTTAAAAAGGG)3'] and pepI-EcoRI 3' [3'(CATTTTTTAGAATTCATTATTTAAACATACTAAAG)3'] (nucleotides in bold face are mutations introduced for the generation of restriction sites, which are underlined). After sequencing, the 0.284-kb BglII-EcoRI pepI fragment was cloned in pCX15 that had been restricted with BamHI and EcoRI.

Preparation of staphylococcal cell fractions. The preparation of membrane and soluble cytoplasmic fractions of S. epidermidis 5 and mutants was done by disruption of cells with glass beads and differential centrifugation, as previously described (32).

MIC determinations. The determination of MICs was performed in a microtiter plate assay with half-concentrated tryptone soy broth, as described by Bierbaum et al. (5).

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Transcriptional analysis of the Pep5 biosynthetic gene cluster. The production of Pep5 is encoded by the 20-kb plasmid pED503 of the producer strain S. epidermidis 5 (14). The Pep5 biosynthetic gene cluster covers approximately 7.9 kb and comprises pepTIAPBC. A putative terminator (56.9 kJ/mol) that may allow partial readthrough is located in the short noncoding segment between pepA and pepP (32). We identified five different mRNA transcripts within the biosynthetic gene cluster (Fig. 1) by Northern blot analysis. A single pepA mRNA of 0.3 kb was detected in large amounts. A pepT transcript of 0.9 kb and a transcript of 5.3 kb, covering pepA and the downstream genes for proteolytical processing and modification (pepP, pepB, and pepC), were identified in low concentrations. Two further transcripts hybridized with the pepI probe. One transcript (0.6 kb) contained the immunity gene pepI and the structural gene pepA, and the other transcript (1.0 kb) additionally covered a region upstream of pepI; a small transcript of approximately 0.3 kb covering only pepI was not detectable.

View larger version (7K): [in this window] [in a new window]   FIG. 1.   Organization of the Pep5 gene cluster (A) and transcription products (B). The arrows represent the direction of transcription relative to the structural gene pepA. The relative amounts of transcripts as judged from Northern blotting are indicated by the thicknesses of the lines.

Is PepA involved in the expression of immunity? Previous studies had shown that S. epidermidis 25(pMR9), containing only pepI of the Pep5 gene cluster, was sensitive to Pep5 to the same extent as its parent strain, S. epidermidis 25, which had had the Pep5 production plasmid pED503 removed (32). This was taken to indicate a role for pepA or its product in the expression of the immunity phenotype. In order to investigate whether both genes have to be transcribed in cis, pepI was cut out of pMR9 and cloned into the staphylococcal vector pT181mcs in S. carnosus TM300, generating pTMR9. We then transferred pTMR9 into S. epidermidis 25(pMR7), which harbors pepA in pCU1 and accumulates unmodified Pep5 prepeptide but is fully sensitive to Pep5. The resulting clone showed the same degree of sensitivity to Pep5 as the plasmidless strain S. epidermidis 25 (MIC, 0.6 µg of Pep5 per ml), demonstrating that immunity cannot be established by the complementation of pepA with pepI in trans.

View larger version (75K): [in this window] [in a new window]   FIG. 2.   Identification of Pep5 prepeptide by Western blot analysis. Total cell extracts of the respective strains were separated by SDS-polyacrylamide gel electrophoresis and immunoblotted with anti-Pep5 leader peptide antiserum. SDS-stable Pep5 prepeptide dimers and trimers (lanes 2 and 4) have been described previously (39). Lanes: 1, molecular size marker (numbers to the left of the gel indicate sizes in kilodaltons); 2, membrane fraction of S. epidermidis 25(pMR7) used as the standard; 3, S. epidermidis 25(pAG5/1); 4, S. epidermidis 25(pMR7, pTMR9).

View larger version (57K): [in this window] [in a new window]   FIG. 3.   Detection of PepI in cell extracts of induced S. carnosus and S. epidermidis cells by Western blotting. Lanes: 1, molecular size marker (numbers to the left of the gel indicate sizes in kilodaltons); 2, synthetic PepI (synthetic PepI was found to form SDS-stable multimers); 3, noninduced S. carnosus(pUP2); 4 and 5, induced S. carnosus(pUP2) (2 and 19 h after induction with 0.5% xylose, respectively); 6, noninduced S. epidermidis 25(pUP2); 7 and 8, induced S. epidermidis 25(pUP2) (2 and 19 h after induction with 0.5% xylose, respectively).

Role of the inverted repeat downstream of pepA. To identify the element that is essential for the self-protection mechanism, in addition to the immunity gene pepI, we constructed various gene arrangements (Fig. 4). First, the polarity of the structural gene was reversed, in order to test whether pepI has to be transcribed together with pepA as one transcript for the production of a functional mRNA. Both genes were amplified by PCR, introducing SalI-XbaI and EcoRI-XbaI restriction sites, respectively, and subsequently both PCR products were subcloned into pCU1 digested with SalI-EcoRI. The resulting recombinant plasmid pUP4 contained pepI and, on the complementary strand in a head-to-tail arrangement, pepA (Fig. 4). This variant strain was only slightly less insensitive to Pep5 than the wild-type strain.

View larger version (18K): [in this window] [in a new window]   FIG. 4.   pepI-containing S. epidermidis 25 variants and their sensitivities to Pep5. The arrangement of pepI and pepA and the locations of the pepA terminator element in different S. epidermidis 25 strains are correlated with their respective MICs.

View larger version (10K): [in this window] [in a new window]   FIG. 5.   Terminator element downstream of pepA. The terminator downstream of pepA has a calculated free energy of 56.9 kJ/mol.

View larger version (59K): [in this window] [in a new window]   FIG. 6.   Northern blot analysis of pepI transcription. (A) Lanes: 1, size standard (numbers to the left of the gels indicate sizes in kilobases); 2, S. epidermidis 25(pUP4); 3, S. epidermidis 25(pUP6); 4, S. epidermidis 25(pUP7); 5, S. epidermidis 25(pAG5/1); 6, S. epidermidis 25(pGB8). A pepI-specific probe was used for detection. In S. epidermidis 25(pUP7) the identical pattern was obtained with a pepA-specific probe; additionally, a single pepA transcript of 0.3 kb was detectable (data not shown). (B) Lanes: 1, size standard; 2, S. epidermidis 25; 3, S. epidermidis 5; 4, S. epidermidis 25(pCU1); 5, S. epidermidis 25(pMR2); 6, S. epidermidis 25(pMR9). A pepI-specific probe was used for detection. (C) Northern blot analysis of the 1.1 kb pepI transcript, with a probe that hybridizes with a region upstream of pepI. Lanes: 1, size standard; 2, S. epidermidis 5; 3, S. epidermidis 25(pGB8).

Site-directed mutagenesis of PepI. PepI is characterized by a striking charge distribution. Whereas the N-terminal segment contains a 20-amino acid stretch of apolar residues, the C-terminal region is very hydrophilic, with a net positive charge. These features are shared by the recently discovered epidicin 280 immunity peptide, EciI, which confers cross-immunity to Pep5 (20). Interestingly, the degree of similarity between PepI and EciI is even higher in the N-terminal region (80%), which was proposed to mediate interactions with the cytoplasmic membrane (32). The exchange of Ile17, which is conserved in both peptides, for Arg (pAG1/1) caused a significant reduction in the level of immunity (Table 2). Nevertheless, the mutant strain was less sensitive to Pep5 than the cloning host, S. epidermidis 25. Western blot analysis of membrane and soluble cell fractions (Fig. 7) did not indicate that the decreased level of immunity was due to reduced PepI production or to a reduced association of PepI with the membrane. With both native PepI and I17R PepI, about one-third of the peptide was detected in the soluble cytoplasmic fraction and two-thirds was detected in the membrane fraction. An additional mutation, introduced in the N-terminal region by exchanging the conserved Phe13 for Asp (pAG4/1), resulted in partially improved immunity compared to that caused by the single I17R mutation. By Western blot analysis the shortened F13D-I17R PepI was detected, indicating a proteolytic truncation of the peptide. We also introduced a stop codon at position 65 of PepI, thereby deleting the four terminal, charged amino acids, three of which are not present in EciI. Again, the strain displayed a significantly decreased level of immunity. The amount of PepI 1-65 was reduced to a level below the detection limit in Western blots, although S. epidermidis 25(pAG2/1) was less sensitive than S. epidermidis 25. 

View larger version (41K): [in this window] [in a new window]   FIG. 7.   Western blot analysis of PepI. Immunoblot analyses of wild-type PepI (lanes 3 and 4), PepI 1-65 (lanes 5 and 6), F13D-I17R PepI (lanes 7 and 8), and I17R PepI (lanes 9 and 10) in soluble and membrane cytoplasmic fractions of cell extracts, respectively, are shown. Standard peptides (sizes are indicated to the left of the gel in kilodaltons) and synthetic PepI are shown in lanes 1 and 2, respectively.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

An efficient self-protection system is vital for a bacteriocin-producing bacterial strain. Channel-forming colicins (e.g., colicins A, E1, and B) produced by E. coli are known to be specifically antagonized by stoichiometric complex formation with immunity proteins residing in the cytoplasmic membrane (16, 42, 45). The structural genes are located upstream of the respective immunity genes, which are transcribed, however, in the opposite direction (9, 31). In the lantibiotic gene clusters, specific immunity peptides and/or ATP-binding cassette transporter systems are encoded (23, 26, 29, 32, 34, 43). Regulation of immunity to nisin, subtilin, and epidermin is mediated through the two-component regulators NisR-SpaR and NisK-SpaK and the transcriptional activator EpiQ, respectively (13, 23, 29). An intact nisin structural gene was found to be necessary for the expression of full immunity (26), which was later shown to be due to the fact that mature nisin activates transcription of the biosynthetic gene cluster via NisR and NisK (11, 25). Also, disruption of the spaA structural gene, which is the last gene in the spaBTCA operon, resulted in subtilin-sensitive cells (23). Coordinate expression of the structural and immunity genes had been established for Pep5 as well (32) and is shown here to be achieved through a specific stabilization of the mRNA transcripts. In the presence of an inverted repeat, which in wild-type S. epidermidis 5 is localized downstream of the structural gene pepA, pepI transcripts are stable and can be translated into the immunity peptide.

mRNA stability plays an important role in regulation of bacterial gene expression. While the majority of ribonucleases are involved in the maturation of rRNA and tRNA, a few ribonucleases, acting either as exonucleases or endonucleases, have been shown to be implicated in mRNA decay. Processive degradation by the 3' exonucleases RNaseII and polynucleotide phosphorylase is impaired by RNA secondary structures (4, 30). Similarly, stable pepI transcripts could not be detected when the terminator was missing, while placing the terminator downstream of the immunity gene (pUP6 [Fig. 6]) led to the production of a stable pepI mRNA. Another example was observed by Klug and Cohen (24), who found two hairpins at the 3' terminus of the puf operon of Rhodobacter capsulatus that function in both transcription termination and stabilization of upstream puf mRNA sequences. The transcriptional terminator of the crystal protein gene (cry) from Bacillus thuringiensis versus Kurstaki HD-1 was shown to function as a positive retroregulator in controlling gene expression in both Bacillus subtilis and E. coli (47). The replacement of the penP (penicillinase gene) terminator with the cry terminator resulted in enhanced mRNA stability. The stabilizing function of the retroregulator was observed to be independent of its orientation of insertion with respect to the target gene. This is also true for the pepA terminator, since it was able to stabilize the pepI mRNA in spite of a reversed orientation, as in the case of the mutant pUP4. However, the pepA inverted repeat lost its function as a transcription termination element. The terminator function is strongly dependent on a stretch of at least four U at the 3' end (35), which after the change of orientation of the pepA repeat, is placed at the 5' end.

Cloning of the pepA terminator upstream of the immunity gene (pUP7 [Fig. 6]) resulted in the appearance of a stable pepAI mRNA and a small (0.3-kb) pepA transcript (data not shown). Several structural elements, such as 5' untranslated regions (7, 12) or polypurine sequences (21), also function as mRNA stabilizers. These probably act by blocking access of 5' binding endonucleases like RNase E to the transcript, presumably by stalling ribosomes at the 5' end. In pUP7 an equivalent structural element, which could explain the high stability of the transcript comprising pepA and pepI in this order, is not present at the 5' end of pepA. The pepA terminator was placed upstream of pepI, but since it was also localized upstream of the pepI promoter sequences, it could not be included in a potential pepI transcript; consequently a small transcript containing only pepI was not detectable. The presence of a pepAI mRNA in this construct also suggests that the progressive degradation in the 5'-to-3' direction does not play a crucial role in the decay of the pepI transcripts. These transcripts rather seem to be degraded to a high extent by the endonucleolytic attack of ribonucleases, which do not necessarily bind to the 5' end of a transcript but are blocked by intercistronic secondary structures. Additionally, 3'-to-5' exonucleases may be involved in degradation. Intercistronic stem-loop structures in the puf operon were also shown to be involved in stabilizing mRNA (10); however, these loops influenced only upstream sequences and not, as in the case of pepI, downstream sequences.

The identification of the stem-loop structure as a stabilizing element and the apparent correlation between RNA stability, quantity of PepI produced, and resulting levels of immunity have several interesting implications for biotechnology and may provide clues to how PepI and related immunity peptides work. The construction of hyperimmune strains is a prerequisite for increasing the production rate of bacteriocins. In this respect, placing bacteriocin structural gene expression under the control of efficient promoters and increasing mRNA half-life, by placing the hairpin in appropriate positions, are valuable tools for increasing the amount of transcript for these genes. Moreover, such stem loops, which are found in many gene clusters of unmodified and modified bacteriocins, could be generally useful for enhancing protein yields, e.g., in heterologous expression hosts such as the food-grade S. carnosus bacteria used here.

The obvious correlation between the level of immunity and the amount of PepI may indicate a direct interaction between Pep5 and PepI and stoichiometric antagonization similar to the mechanism found for the channel-forming colicins (16, 42, 45). In this case, the colicins bind specifically to an immunity protein which is anchored within the cytoplasmic membrane by means of four membrane-spanning helices. The channel-forming domain interacts with an external loop of the immunity protein which prevents this domain from folding into a functional pore. It is tempting to assume by analogy that PepI, as well as the related peptides encoded by eciI, orf57, and dviA, could stoichiometrically antagonize Pep5 or the respective peptide bacteriocins. The N-terminal stretch of hydrophobic amino acids could provide the membrane anchor, while the C-terminal segment of PepI could mediate the interaction with Pep5. However, our mutagenesis results do not necessarily favor such a model. Although the introduction of an Arg residue in the hydrophobic segment strongly reduced the level of immunity, the distribution of the mutant peptide between soluble fractions and membrane-associated fractions was not changed; even the peptide containing a second charge in this segment was still mostly associated with the membrane.

Generally, the interpretation of the mutagenesis experiments was hampered by the lack of a molecular model of the activity of PepI and related peptides, which in turn seems difficult to establish because of our limited knowledge of the mode of action of the bacteriocins. While it is widely accepted that the peptide antibiotics destabilize the cytoplasmic membrane, it remains largely unclear which molecular interactions take place at or within a bacterial membrane. Recently, we were able to demonstrate that, in vivo, nisin and epidermin use the undecaprenol-bound cell wall precursors, the so-called lipid I and lipid II, as docking molecules for subsequent pore formation (8). The high-affinity binding of the lantibiotics to lipids I and II seems to facilitate energetically the formation of pores and may explain the high sensitivity of some bacterial species, e.g., Micrococcus luteus, which has MICs in the nanomolar range. Pep5 does not make use of the cell wall precursors (8) and is not very active against Micrococcus. However, Staphylococcus simulans and S. carnosus are sensitive in nanomolar concentrations to this lantibiotic. As such low MICs are generally not observed with amphipathic peptides and may only be achievable through high-affinity binding to particular components of bacterial membranes, we are currently trying to identify a docking molecule for Pep5. Such a concept for the molecular activity of at least some lantibiotics could not only help to explain the enormous variation in sensitivities of gram-positive bacteria to different bacteriocins but may also provide a new role for an immunity peptide, in that it may function by preventing the molecular target from binding the lantibiotic.

We currently do not have a satisfactory explanation for the observation that in the double mutant the immunity was partially restored in spite of an apparent truncation of the twofold mutated peptide; also, the instability of PepI after shortening of the C terminus remains to be explained. PepI was localized outside the cells (32), and it is conceivable that the mutations interfere with the export of the peptides, resulting in prolonged exposure to intracellular proteases, as observed with some altered Pep5 prepeptides (6). However, the processes that lead to the translocation of PepI across the membrane are also not understood. Finally, it could also be of functional importance that the PepI-type immunity peptides seem to be associated with lantibiotics which are processed inside the cell (28, 44). It is obvious that many questions about these peptides remain unanswered and that, in general, producer self-protection of gram-positive bacteria at the molecular level is the least understood topic in bacteriocin research. Certainly, these phenomena are of interest for fundamental research and applied disciplines and deserve more attention in the future.