Previous Article | Next Article 
Applied and Environmental Microbiology, March 2007, p. 1792-1796, Vol. 73, No. 6
0099-2240/07/$08.00+0 doi:10.1128/AEM.02350-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
Production of Dehydroamino Acid-Containing Peptides by Lactococcus lactis
Rick Rink,1
Jenny Wierenga,1
Anneke Kuipers,1
Leon D. Kluskens,1
Arnold J. M. Driessen,2
Oscar P. Kuipers,3 and
Gert N. Moll1*
BiOMaDe Technology Foundation, Nijenborgh 4, 9747 AG Groningen, The Netherlands,1
Department of Microbiology,2
Department of Molecular Genetics, Biomolecular Sciences and Biotechnology Institute, University of Groningen, Kerklaan 30, 9751 NN Haren, Groningen, The Netherlands3
Received 5 October 2006/
Accepted 9 January 2007

ABSTRACT
Nisin is a pentacyclic peptide antibiotic produced by some
Lactococcus lactis strains. Nisin contains dehydroresidues and thioether
rings that are posttranslationally introduced by a membrane-associated
enzyme complex, composed of a serine and threonine dehydratase
NisB and the cyclase NisC. In addition, the transporter NisT
is necessary for export of the modified peptide. We studied
the potential of
L. lactis expressing NisB and NisT to produce
peptides whose serines and threonines are dehydrated.
L. lactis containing the
nisBT genes and a plasmid coding for a specific
leader peptide fusion construct efficiently produced peptides
with a series of non-naturally occurring multiple flanking dehydrobutyrines.
We demonstrated NisB-mediated dehydration of serines and threonines
in a C-terminal nisin(1-14) extension of nisin, which implies
that also residues more distant from the leader peptide than
those occurring in prenisin or any other lantibiotic can be
modified. Furthermore, the feasibility and efficiency of generating
a library of peptides containing dehydroresidues were demonstrated.
In view of the particular shape and reactivity of dehydroamino
acids, such a library provides a novel source for screening
for peptides with desired biological and physicochemical properties.

INTRODUCTION
Lantibiotics are (methyl)lanthionine-containing antibiotics
(
4). Lantibiotics contain (methyl)lanthionines, i.e., thioether-containing
amino acids, which are posttranslationally introduced by enzymes
coded for by genes of the lantibiotic's operon. These enzymes
dehydrate serine and threonine residues, thereby converting
them into dehydroalanine and dehydrobutyrine residues, respectively.
Subsequently, (methyl)lanthionines are formed by enzymatic coupling
of the dehydroresidues to cysteines. Functional reconstitution
of in vitro activity of one bifunctional enzyme, lacticin M
(
27), and of the cyclase NisC (
14) has been achieved. The structure
of NisC has also been determined (
14), providing detailed information
about its catalytic mechanism.
The pentacyclic lantibiotic nisin is produced by some Lactococcus lactis strains. Nisin is ubiquitously applied as a food preservative that inhibits the growth of mainly gram-positive spoilage and pathogenic bacteria. By binding to the pyrophosphate of lipid II with its N-terminal region, which encompasses the A and B rings, nisin exerts two different antimicrobial activities (1, 7). First, a transmembrane pore composed of both nisin and lipid II causes dissipation of the transmembrane ion gradients. Second, binding of nisin to lipid II causes cell killing by inhibition of cell wall synthesis. Nisin is composed of four methyllanthionines, one lanthionine, two dehydroalanines, one dehydrobutyrine, and twenty-six unmodified amino acids (6, 12). The dehydration reactions are carried out by the nisin dehydratase NisB. Replacement of the dehydroalanine at positions 5 of nisin (2) and subtilin (15) very much reduces the capacity to inhibit the outgrowth of spores. NisB can also dehydrate serine and threonine residues in peptides that are not at all related to lantibiotics such as, for instance, human peptide hormones, provided that these peptides are N terminally fused to the lantibiotic leader peptide (10). The presence of dehydroresidues in a nonlantibiotic peptide can influence the peptide's activity (16, 18, 21, 22, 25, 26), and they could be excellent starting points for chemical modification because of their reactivity. We have studied here the capacity of L. lactis cells containing the NisBT proteins to dehydrate and export model peptides, among which are semirandomized hexapeptides (Fig. 1). Our data indicate a high degree of versatility of NisB to dehydrate serine and threonine residues in peptides.

MATERIALS AND METHODS
Bacterial strains and plasmids.
The strains and plasmids in this study are listed in Table
1.
Molecular cloning.
Nisin leader constructs (Fig.
1) were made by PCR-amplifying
leader peptide-encoding plasmid, pNZnisA-E3, using a phosphorylated
primer with a (nonannealing) peptide-encoding tail. DNA amplification
was carried out using Phusion DNA polymerase (Finnzymes, Finland).
Ligation was carried out with T4 DNA ligase (Roche). Electrotransformation
of
L. lactis was done as described previously (
8) using a Bio-Rad
gene pulser (Bio-Rad, Richmond, CA). Nucleotide sequence analysis
was performed by BaseClear (Leiden, The Netherlands).
Culturing.
L. lactis was grown at 30°C in M17 broth (24) supplemented with 0.5% glucose (GM17) or minimal medium (19) with or without chloramphenicol (5 µg/ml) and/or erythromycin (5 µg/ml). The concentration of the antibiotics was reduced to 4 µg/ml when both were present simultaneously. Prior to mass spectrometry, cells were cultured, and samples were prepared as follows. Overnight cultures of L. lactis NZ9000 grown in GM17 broth containing 4 µg of antibiotics/ml were diluted 1/100. After growth to an optical density at 600 nm of 0.4, cells were centrifuged, and the medium was replaced by minimal medium supplemented with 1/1,000 volume of filtered (0.45-µm pore size) overnight in L. lactis NZ9700 culture medium containing nisin for induction. Incubation was continued overnight.
Mass spectrometry.
Peptides were isolated from culture supernatants in a single step by applying the Ziptip procedure (C18 Ziptip; Millipore). Ziptips were wetted and equilibrated with 50% acetonitrile, followed by demineralized water. Subsequently, peptides from the medium were bound by subjecting 100 µl or more to Ziptip treatment, washed with 0.1% trifluoroacetic acid (TFA), eluted with a solution of 0.1% TFA with 50% acetonitrile, vacuum dried, and stored at 20°C until analysis. The dried Ziptip eluent was resuspended in 5 µl of 50% acetonitrile containing 0.1% (vol/vol) TFA, and 1 µl was applied to the target. Subsequently, 1 µl of matrix (5 mg of
-cyano-4-hydroxycinnamic acid/ml in 50% acetonitrile containing 0.1% [vol/vol] TFA) was added to the target and allowed to dry.
Mass spectra were recorded with a Voyager-DE PRO matrix-assisted laser desorption ionization-time of flight mass spectrometer (Applied Biosystems). In order to maintain high sensitivity, an external calibration was applied.
Hydrophobicity analysis of flanking residues of serines and threonines.
The hydrophobicity of individual amino acids was taken as the consensus values described by Eisenberg (5). The hydrohobicity of dehydroresidues was approached by taking the value of alanine. The hydrophobicity of lanthionine was approached by taking the average between cysteine and alanine. Analyses were based on the peptides listed in Table 3, on NisB-dehydrated peptides described previously (10, 19), and on the wild-type nisin molecule. The flanking residues of serines and threonines in the sequence ITSIS, which occurs in nisin and in constructs previously described (10), were counted only once, as were also the amino acids D and R as flanking residues of serines and threonines in designed peptides. Residues that flank N- or C-terminal serines or threonines were not taken into account.

RESULTS AND DISCUSSION
NisB dehydrates multiple flanking threonines.
We previously showed that NisB is equipped with a broad substrate
specificity (
10,
11,
19). To further evaluate the (lack of)
fidelity of this enzyme, we investigated whether NisB can dehydrate
multiple flanking threonines, which are not present in naturally
occurring lantibiotics. Therefore, we expressed a construct,
coding for the nisin leader extended with the amino acid sequence
ITTTTT (Fig.
1 and Fig.
2), together with pIL3BTC encoding the
nisin modifying and transporting enzymes. Mass spectrometric
analysis of the culture supernatant demonstrated successful
production of peptides. Partial N-terminal truncation of the
leader part was observed as described previously (
19). Up to
fivefold dehydration was observed for the peptide devoid of
the first five leader peptide residues, MSTKD (Fig.
2A). The
mass peak of fourfold dehydration was most prominent, whereas
the one of fivefold dehydration was small. Hydrophilic residues
appear to disfavor dehydration (
19). C-terminal serine residues
are commonly not dehydrated in lantibiotics (
19). Likely due
to the hydrophilicity of the carboxy terminus, dehydration of
the C-terminal threonine is inefficient. The fourfold dehydration
could be clearly confirmed by four additions of ethanethiol
(Fig.
2B). These data clearly demonstrate that NisB can dehydrate
multiple flanking threonines, thus generating polydehydrobutyrine.
NisB is capable of dehydrating serine and threonine residues along the polypeptide chain.
Ser33 of nisin is the dehydratable residue in the NisA peptide,
which is most remote from the leader peptide, and has been reported
to escape dehydration in ca. 10% of the cases (
2,
20). The extent
of modification of Ser33 was reduced to 50% in [Trp30]nisin
A and [Lys27, Lys31]nisin A (
9). Plasmid-encoded overproduction
of NisB results in a higher extent of dehydration (
9). To determine
whether serine or threonine residues more distant from the leader
peptide than position 33 can be dehydrated, the leader peptide-nisin
construct was genetically C-terminally extended with nisin(1-14),
the lipid II-binding domain of nisin. This yielded plasmid pNZ-nisin-nisin(1-14).
In the supernatant of
L. lactis cells containing pILnisBT and
pNZ-nisin-nisin(1-14), a product with a mass peak of 6,979 Da
(M + H
+) (Fig.
3) was clearly observed under reducing conditions
in the presence of triscarboxyethyl phosphine (TCEP). This mass
peak corresponded to a 12-fold-dehydrated fusion product of
prenisin and nisin(1-14). These data indicate that in addition
to the known dehydration steps in nisin, at least four of the
five additional residues in the nisin(1-14) extensionThr36,
Ser37, Ser39, Thr42, and Thr47can be dehydrated. Hence,
NisB can also efficiently dehydrate residues that are more distant
from the leader peptide than Ser33 in nisin. These modifications
occur farther from the leader peptide than reported for any
other known lantibiotic or model peptide, showing a high versatility
of NisB.
A library of NisB-dehydrated peptides.
We subsequently investigated the possibility of generating a
library of peptides with dehydroresidues. Semirandomized primers
were used to produce constructs that encode the nisin leader
peptide fused to hexapeptides with four randomized positions.
These hexapeptides were composed of X(S/T)XXAX, in which X represents
any amino acid (Fig.
1 and Table
2). At least 75% of the obtained
transformants produced peptides that could be detected by direct
analysis of the supernatant by the matrix-assisted laser desorption
ionization-time of flight method. Sequence analysis of plasmids
of transformants that did not produce peptides with an expected
mass revealed that some of these contained a stop codon at the
start ofor withinthe hexapeptide. Indeed, corresponding
masses of the leader peptide without or with a few additional
amino acids could be detected. Hence, the remarkable export
of just the leader peptide itself, without propeptide fusion,
had also occurred. Interestingly, most produced hexapeptides
contained dehydrated residues, whereas some were completely
devoid of dehydration (Table
2). The latter peptides contained
mostly serine residues instead of threonine residues, a finding
consistent with our previous prediction that serine is a less
favorable substrate for NisB than threonine (
19).
In agreement with the previously obtained in silico and in vivo
data, hydrophobic flanking amino acid residues favor the dehydration
of serine and threonine residues (Table
2). Exceptions are the
hexapeptides VSPPAR, YTPPAL, and FSFFAF, which are not dehydrated.
These sequences suggest that the presence of proline residues
at positions 3 and 4 or flanked at both sides by a bulky phenylalanine
may be unfavorable for NisB-mediated dehydration. Other exceptions
to our previously proposed guidelines are the hexapeptides HTDLAD,
KSHYAM, and RTSHAA. These are dehydrated, whereas the modified
residues are flanked by two hydrophilic residues. Since neither
HTRRAE nor HSKQAG is dehydrated, it is currently unclear how
the presence of a histidine affects NisB-mediated dehydration.
Nevertheless, it is generally observed among the peptides that
escape dehydration that the serine/threonine residues are flanked
by hydrophilic residues. Some peptides containing multiple serine
and/or threonine residues were only incompletely dehydrated.
Taken together, our analysis of random X(S/T)XXAX hexapeptides
indicates that hydrophobic flanking of serine or threonine favors
dehydration, whereas hydrophilic flankingespecially negatively
charged amino acidsdisfavors dehydration. Indeed, in
the case of dehydration, the average hydrophobicity of flanking
residues of serines and threonines is positive (0.40 and 0.13,
N and C side, respectively), and predominantly amino acids with
positive hydrophobicity are flanking the dehydroresidues (Table
3). In contrast, in the case of the absence of dehydration,
the average hydrophobicity of amino acids that flank unmodified
serines or threonines is negative (0.36 and 1.03,
N and C sides, respectively), and more flanking amino acids
with negative hydrophobicity are present than with positive
hydrophobicity (Table
3). Importantly, most of the cells containing
NisBT and the randomizing plasmid pTPhexa produced dehydroamino-acid-containing
peptide showing the high degree of promiscuity of NisB and NisT.
In the present study we demonstrate that nisin dehydratase and nisin transporter synthesizing L. lactis can produce a variety of dehydroresidue-containing peptides. Dehydration appears to be neither restricted by the presence of a series of flanking threonine residues nor restricted by the distance between the modifiable position and the nisin leader peptide. Previously, we showed that the threonine in a peptide composed of the nisin leader followed by the sequence DSRWARVALIDSQKAAVDKAITDIAEKL was the only modified residue (19). Apparently, if NisB processively scans the substrate, then it does not release it even when none of the first 21 propeptide residues can be dehydrated. Processive dehydration has been observed in the case of the bifunctional lantibiotic enzyme LctM (17). These results raise intriguing questions regarding how NisB interacts with the nisin leader and the substrate domain. NisB is active in the absence of NisT and NisC (11), and therefore recognition of the substrate should be the consequence of a direct interaction of the leader peptide with NisB. Possibly, the substrate region slides along the active center of NisB, but it is unclear whether the leader peptide remains associated with NisB during catalysis and how it is transferred to NisC. Our previous studies have shown that NisC is not needed for peptide release from NisB, since dehydrated peptides can be excreted in the medium even in the absence of NisC. In view of the relative high hydrophobicity of the flanking residue at the N side of the dehydratable residue, it is tempting to speculate on a directionality of NisB-mediated dehydration of the substrate region from proximal to distal to the leader peptide. Mechanistic information on lantibiotic-enzyme-mediated dehydration is only available for LctM, a bifunctional enzyme that performs both the dehydration and the cyclization reaction. LctM phosphorylates serine and threonine residues prior to dehydration (3). It should, however, be noted that LctM has no sequence homology with NisB. The data presented here involving the dehydration of multiple flanking threonines, the dehydration of distal serines and threonines, and the library of dehydroresidue-containing peptides consistently indicate the great versatility of NisB. (Poly-)dehydrobutyrine sequences can be very interesting starting points for the synthesis of other non-natural amino acid residues or serve as coupling sites in further organic synthesis. Moreover, since dehydroresidues can be important constituents of biologically active peptides due to their particular planar shape and reactivity (16, 18, 21, 22, 25, 26), screening of a dehydroresidue-containing peptide library for desired properties and activities is a realistic and attractive possibility.

ACKNOWLEDGMENTS
Leon D. Kluskens is supported by STW project 6927.

FOOTNOTES
* Corresponding author. Mailing address: BiOMaDe Technology Foundation, Nijenborgh 4, 9747 AG Groningen, The Netherlands. Phone: 31 50 3638070. Fax: 31 50 3634429. E-mail:
moll{at}biomade.nl.

Published ahead of print on 19 January 2007. 

REFERENCES
1 - Breukink, E., I. Wiedemann, C. van Kraaij, O. P. Kuipers, H.-G. Sahl, and B. de Kruijff. 1999. Use of the cell wall precursor lipid II by a pore-forming peptide antibiotic. Science 286:2245-2247.[Free Full Text]
2 - Chan, W. C., H. M. Dodd, N. Horn, K. Maclean, L.-Y. Lian, B. W. Bycroft, M. J. Gasson, and G. C. K. Roberts. 1996. Structure-activity relationships in the peptide antibiotic nisin: role of dehydroalanine 5. Appl. Environ. Microbiol. 62:2966-2969.[Abstract]
3 - Chatterjee, C., L. M. Miller, Y. L. Leung, L. Xie, M. Yi, N. L. Kelleher, and W. A. van der Donk. 2005. Lacticin 481 synthetase phosphorylates its substrate during lantibiotic production. J. Am. Chem. Soc. 127:15332-15333.[CrossRef][Medline]
4 - Chatterjee, C., M. Paul, L. Xie, and W. A. van der Donk. 2005. Biosynthesis and mode of action of lantibiotics. Chem. Rev. 105:633-684.[CrossRef][Medline]
5 - Eisenberg, D. 1984. Three-dimensional structure of membrane and surface proteins. Annu. Rev. Biochem. 53:595-623.[CrossRef][Medline]
6 - Gross, E., and J. L. Morrell. 1971. The structure of nisin. J. Am. Chem. Soc. 93:4634-4635.[CrossRef][Medline]
7 - Hasper, H. E., N. E. Kramer, J. L. Smith, J. D. Hillman, C. Zachariah, O. P. Kuipers, B. de Kruijff, and E. Breukink. 2006. An alternative bactericidal mechanism of action for lantibiotic peptides that target lipid II. Science 313:1636-1637.[Abstract/Free Full Text]
8 - Holo, H., and I. F. Nes. 1995. Transformation of Lactococcus by electroporation. Methods Mol. Biol. 47:195-199.[Medline]
9 - Karakas Sen, A., A. Narbad, N. Horn, H. M. Dodd, A. J. Parr, I. Colquhoun, and M. J. Gasson. 1999. Post-translational modification of nisin: the involvement of NisB in the dehydration process. Eur. J. Biochem. 261:524-532.[Medline]
10 - Kluskens, L. D., A. Kuipers, R. Rink, E. de Boef, S. Fekken, A. J. M. Driessen, O. P. Kuipers, and G. N. Moll. 2005. Post-translational modification of therapeutic peptides by NisB, the dehydratase of the lantibiotic nisin. Biochemistry 44:12827-12834.[CrossRef][Medline]
11 - Kuipers, A., E. de Boef, R. Rink, S. Fekken, L. D. Kluskens, A. J. M. Driessen, K. Leenhouts, O. P. Kuipers, and G. N. Moll. 2004. NisT, the transporter of the lantibiotic nisin, can transport fully modified, dehydrated and unmodified prenisin and fusions of the leader peptide with non-lantibiotic peptides. J. Biol. Chem. 279:22176-22182.[Abstract/Free Full Text]
12 - Kuipers, O. P., M. M. Beerthuyzen, R. J. Siezen, and W. M. de Vos. 1993. Characterization of the nisin gene cluster nisABTCIPR of Lactococcus lactis. Requirement of expression of the nisA and nisI genes for development of immunity. Eur. J. Biochem. 216:281-291.[Medline]
13 - Kuipers, O. P., P. G. De Ruyter, M. Kleerebezem, and W. M. de Vos. 1997. Controlled overproduction of proteins by lactic acid bacteria. Trends Biotechnol. 15:135-140.[CrossRef][Medline]
14 - Li, B., J. P. Yu, J. S. Brunzelle, G. N. Moll, W. A. van der Donk, and S. K. Nair. 2006. Structure and mechanism of the lantibiotic cyclase involved in nisin biosynthesis. Science 311:1464-1467.
15 - Liu, W., and J. N. Hansen. 1992. Enhancement of the chemical and antimicrobial properties of subtilin by site-directed mutagenesis. J. Biol. Chem. 267:25078-25085.[Abstract/Free Full Text]
16 - Lombardi, A., B. D'Agostino, F. Nastri, L. D. D'andrea, and A. Filippelli. 1998. A novel superpotent neurokinin A receptor antagonist containing dehydroalanine. Bioorg. Med. Chem. Lett. 8:1153-1156.[CrossRef][Medline]
17 - Miller, L. M., C. Chatterjee, W. A. van der Donk, and N. L. Kelleher. 2006. The dehydratase activity of lacticin 481 synthetase is highly processive. J. Am. Chem. Soc. 128:1420-1421.[CrossRef][Medline]
18 - Murkin, A. S., and M. E. Tanner. 2002. Dehydroalanine-based inhibition of a peptide epimerase from spider venom. J. Org. Chem. 67:8389-8394.[CrossRef][Medline]
19 - Rink, R., A. Kuipers, E. de Boef, K. J. Leenhouts, A. J. M. Driessen, G. N. Moll, and O. P. Kuipers. 2005. Lantibiotic structures as guidelines for the design of peptides that can be modified by lantibiotic enzymes. Biochemistry 44:8873-8882.[CrossRef][Medline]
20 - Rollema, H. S., P. Both, and R. J. Siezen. 1991. NMR and activity studies of nisin degradation products, p. 123-130. In G. Jung and H.-G. Sahl (ed.), Nisin and novel lantibiotics. ESCOM, Leiden, The Netherlands.
21 - Schuster, B., and J. Rétey. 1995. The mechanism of action of phenylalanine ammonia-lyase: the role of prosthetic dehydroalanine. Proc. Natl. Acad. Sci. USA 92:8433-8437.[Abstract/Free Full Text]
22 - Siddiqui, M. I., S. Kataria, V. Ahuja, and G. S. Rao. 2001. A peptide inhibitor of HIV-1 protease using alpha, beta-dehydro residues: a structure based computer model. Indian J. Biochem. Biophys. 38:90-95.[Medline]
23 - Simon, D., and A. Chopin. 1988. Construction of a vector plasmid family and its use for molecular cloning in Streptococcus lactis. Biochimie 70:559-566.[Medline]
24 - Terzaghi, B. E., and W. E. Sandine. 1975. Improved medium for lactic streptococci and their bacteriophages. Appl. Microbiol. 29:807-813.[Medline]
25 - Tomkinson, B., L. Grehn, B. Fransson, and Ö. Zetterqvist. 1994. Use of dehydroalanine-containing peptide as an efficient inhibitor of tripeptidyl peptidase II. Arch. Biochem. Biophys. 314:276-279.[CrossRef][Medline]
26 - Vollbrecht, L., H. Steinmetz, G. Hofle, L. Oberer, G. Rihs, G. Bovermann, and P. Matt. 2002. Argyrins, immunosuppressive cyclic peptides from myxobacteria. II. Structure elucidation and stereochemistry. J. Antibiot. 55:715-721.[Medline]
27 - Xie, L., L. M. Miller, C. Chatterjee, O. Averin, N. L. Kelleher, and W. A. van der Donk. 2004. Lacticin 481: in vitro reconstitution of lantibiotic synthetase activity. Science 303:679-681.[Abstract/Free Full Text]
Applied and Environmental Microbiology, March 2007, p. 1792-1796, Vol. 73, No. 6
0099-2240/07/$08.00+0 doi:10.1128/AEM.02350-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
This article has been cited by other articles:
-
Kuipers, A., Meijer-Wierenga, J., Rink, R., Kluskens, L. D., Moll, G. N.
(2008). Mechanistic Dissection of the Enzyme Complexes Involved in Biosynthesis of Lacticin 3147 and Nisin. Appl. Environ. Microbiol.
74: 6591-6597
[Abstract]
[Full Text]
-
van den Berg van Saparoea, H. B., Bakkes, P. J., Moll, G. N., Driessen, A. J. M.
(2008). Distinct Contributions of the Nisin Biosynthesis Enzymes NisB and NisC and Transporter NisT to Prenisin Production by Lactococcus lactis. Appl. Environ. Microbiol.
74: 5541-5548
[Abstract]
[Full Text]
-
Lubelski, J., Overkamp, W., Kluskens, L. D., Moll, G. N., Kuipers, O. P.
(2008). Influence of Shifting Positions of Ser, Thr, and Cys Residues in Prenisin on the Efficiency of Modification Reactions and on the Antimicrobial Activities of the Modified Prepeptides. Appl. Environ. Microbiol.
74: 4680-4685
[Abstract]
[Full Text]