AEM
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Holtsmark, I.
Right arrow Articles by Brurberg, M. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Holtsmark, I.
Right arrow Articles by Brurberg, M. B.
Agricola
Right arrow Articles by Holtsmark, I.
Right arrow Articles by Brurberg, M. B.
Applied and Environmental Microbiology, September 2006, p. 5814-5821, Vol. 72, No. 9
0099-2240/06/$08.00+0     doi:10.1128/AEM.00639-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Purification, Characterization, and Gene Sequence of Michiganin A, an Actagardine-Like Lantibiotic Produced by the Tomato Pathogen Clavibacter michiganensis subsp. michiganensis

I. Holtsmark,1,3* D. Mantzilas,2 V. G. H. Eijsink,3 and M. B. Brurberg1

Norwegian Institute for Agricultural and Environmental Research, Plant Health and Protection Division, Ås, Norway,1 Department of Molecular Biosciences, University of Oslo, Oslo, Norway,2 Department of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, Ås, Norway3

Received 20 March 2006/ Accepted 19 June 2006


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Members of the actinomycete genus Clavibacter are known to produce antimicrobial compounds, but so far none of these compounds has been purified and characterized. We have isolated an antimicrobial peptide, michiganin A, from the tomato pathogen Clavibacter michiganensis subsp. michiganensis, using ammonium sulfate precipitation followed by cation-exchange and reversed-phase chromatography steps. Upon chemical derivatization of putative dehydrated amino acids and lanthionine bridges by alkaline ethanethiol, Edman degradation yielded sequence information that proved to be sufficient for cloning of the gene by a genome-walking strategy. The mature unmodified peptide consists of 21 amino acids, SSSGWLCTLTIECGTIICACR. All of the threonine residues undergo dehydration, and three of them interact with cysteines via thioether bonds to form methyllanthionine bridges. Michiganin A resembles actagardine, a type B lantibiotic with a known three-dimensional structure, produced by Actinoplanes liguriae, which is a filamentous actinomycete. The DNA sequence of the gene showed that the michiganin A precursor contains an unusual putative signal peptide with no similarity to well-known secretion signals and only very limited similarity to the (only two) available leader peptides of other type B lantibiotics. Michiganin A inhibits the growth of Clavibacter michiganensis subsp. sepedonicus, the causal agent of ring rot of potatoes, with MICs in the low nanomolar range. Thus, michiganin A may have some potential in biological control of potato ring rot.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Bacteriocins are antimicrobial peptides and proteins produced by bacteria and are active against closely related species. They are produced by almost all major lineages of bacteria (41), including plant pathogens (57). It has been proposed that the primary role of bacteriocins is to mediate population dynamics among closely related species (42). Bacteriocins inhibit bacterial growth by a variety of mechanisms, most notably by the formation of membrane pores (2, 9, 34). These compounds are receiving increasing attention because of their potential use as preservatives in the food industry, as antibiotics in medicine, or as biocontrol agents (10, 12, 20, 22, 29, 50).

Peptide bacteriocins are usually classified on the basis of structural and functional characteristics (16). Class I bacteriocins, called lantibiotics, carry characteristic posttranslational modifications, whereas class II bacteriocins lack such modifications (9, 16, 26, 31). The modifications in lantibiotics involve dehydration of threonine and serine residues, yielding dehydrobutyrine (Dhb) and dehydroalanine (Dha), respectively. These residues may react with sulfhydryl groups of cysteine residues to create ß-methyllanthionine or lanthionine ring structures, respectively (9, 31, 47). The lantibiotics consist of two subgroups. Group A contains cationic, elongated, flexible peptides, which have both pore-forming and cell wall synthesis-blocking activities. Group B contains noncationic globular molecules with head-to-tail cross-bridges conferring compactness, as well as relative robustness against proteolytic attack. Type B lantibiotics are further divided into the cinnamycin and mersacidin subtypes (9). Type B lantibiotics inhibit crucial enzyme functions through interaction with the respective substrates. For example, mersacidin forms a complex with lipid II, blocking transglycosylation and consequently inhibiting incorporation of lipid II in the cell walls of the target cells (8, 23).

Generally, relatively little is known about bacteriocin production in plant-pathogenic bacteria, despite the fact that these compounds in principle may play crucial roles in the ecological niches where infections take place and despite the fact that bacteriocins targeting plant pathogens could find important applications. The causal agent of bacterial wilt and canker on tomato, the actinomycete Clavibacter michiganensis subsp. michiganensis, has long been known to produce substances that inhibit the growth of close relatives (19, 35). Early studies revealed that at least some of these compounds have characteristics that are typical of small proteinaceous molecules (15, 19). One of the species inhibited by C. michiganensis subsp. michiganensis is Clavibacter michiganensis subsp. sepedonicus, which has status as an international quarantine organism and which is responsible for ring rot of potatoes, a devastating disease causing economic losses in North America, Asia, and Europe. In the European Union alone, ring rot outbreaks are currently estimated to cost {euro}15 million per year, due to eradication campaigns, increased research and development costs, and compensation to growers (54).

In the present study, we describe the purification and characterization of a peptide bacteriocin from C. michiganensis subsp. michiganensis with activity toward C. michiganensis subsp. sepedonicus. This peptide, called michiganin A, proved to be a type B lantibiotic, the structure of which could be modeled on the basis of the results of peptide sequencing and genetic analysis and by comparison with a related lantibiotic, actagardine, from the filamentous actinomycete Actinoplanes liguriae (63). Cloning and sequencing of the gene also revealed the sequence of the (atypical) prepeptide of michiganin A.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Bacterial strains and media.
Clavibacter michiganensis subsp. michiganensis strain NCPPB 1468 from the National Collection of Plant Pathogenic Bacteria, York, United Kingdom, was grown in LB broth for bacteriocin production. The indicator strain, C. michiganensis subsp. sepedonicus strain NCPPB 2136, also from the National Collection of Plant Pathogenic Bacteria, York, United Kingdom, was grown on nutrient broth yeast extract medium (containing [per liter] 8 g Difco nutrient broth, 2 g yeast extract [Difco], 5 g glucose, and 1 mM MgSO4). For plates, the media described above were solidified with 15 g/liter agar.

Bacteriocin production.
A colony of C. michiganensis subsp. michiganensis 1468 was transferred from solid medium to 10 ml of LB medium and incubated at 20°C for 48 h at 100 rpm. The resulting culture was used to inoculate two 2-liter flasks, each containing 500 ml of LB broth. The cultures were incubated at 20°C for 4 days at 100 rpm.

Bacteriocin purification.
The peptide was precipitated from 1 liter of culture supernatant by ammonium sulfate precipitation (400 g added per liter). The precipitate was dissolved in 200 ml 20 mM sodium acetate buffer, pH 4 (buffer A), and applied to a 2.5- by 2.5-cm self-packed SP Sepharose Fast Flow cation-exchange column (GE Healthcare, Uppsala, Sweden) equilibrated with buffer A (flow rate, 3 ml/min). Ten milliliters of the flowthrough fraction was applied to a 3-ml Resource RPC column (GE Healthcare, Uppsala, Sweden) equilibrated with 0.1% (vol/vol) trifluoroacteic acid (TFA) and connected to an Äkta system (GE Healthcare, Uppsala, Sweden). After the column was eluted with a 20-column-volume (cv) gradient from 0 to 100% (vol/vol) 2-propanol in 0.1% (vol/vol) TFA (flow rate, 1 ml/min), the fractions with bacteriocin activity (containing about 45% 2-propanol) were pooled, diluted fivefold with 0.1% (vol/vol) TFA, and rechromatographed on a 1-ml pepRPC HR 5/5 column (GE Healthcare, Uppsala, Sweden). The pepRPC column was eluted by applying a 5-cv gradient from 0 to 25% 2-propanol, followed by a 20-cv gradient from 25 to 45% 2-propanol, in 0.1% (vol/vol) TFA (flow rate, 0.5 ml/min). The fractions with bacteriocin activity (containing about 32% 2-propanol) were pooled, diluted fourfold with 0.1% (vol/vol) TFA, and applied to a 0.35 ml (2.1- by 100-mm) µRPC C2/C18 column (GE Healthcare, Uppsala, Sweden) equilibrated with 10% 2-propanol in 0.1% (vol/vol) TFA and connected to a SMART system (GE Healthcare, Uppsala, Sweden). The column was eluted by applying a 6-cv gradient from 10 to 50% 2-propanol in 0.1% (vol/vol) TFA (flow rate, 50 µl/min). The eluted material was detected by measuring the UV absorbance at 214, 254, and 280 nm.

Activity assays.
Bacteriocin activities were approximated quantitatively by using a microtiter plate dilution assay, essentially as described by Geis et al. (18). Bacteriocin activity was expressed in units, with one bacteriocin unit being the amount of bacteriocin required to reduce the growth of the indicator strain by 50% under the conditions of the assay (200-µl culture volume). The MIC is defined as the peptide concentration producing 50% growth inhibition under the conditions of the assay. Peptide concentrations in purified samples were estimated on the basis of the results of amino acid composition analysis (see below).

Mass spectrometry.
For mass spectrometry, 1 µl of sample was mixed with 1 µl of a saturated solution of {alpha}-cyano-4-hydroxycinnamic acid in 0.1% (vol/vol) TFA on the sample plate and air dried. Analyses were performed on an Ultraflex matrix-assisted laser desorption ionization-time of flight instrument (Bruker Daltonik GmbH, Bremen, Germany) in positive electron mode, with 25-kV accelerating voltage. Spectra were calibrated using an external peptide calibration standard (Brukers Peptide calibration Standard II; Bruker Daltonics GmbH, Bremen, Germany).

Amino acid composition analysis.
Lyophilized peptide was vacuum hydrolyzed in 6 M HCl at 110°C for 24 h and subsequently analyzed using an Applied Biosystems Amino Acid Analyzer.

Amino acid sequencing.
Lyophilized peptide was resuspended and applied to a Polybrene-treated TFA filter for Edman degradation with an Applied Biosystems Protein Sequencer model 477A.

Ethanethiol derivatization of posttranslationally modified amino acids was performed essentially as described by Meyer et al. (33). Lyophilized peptide in a microcentrifuge tube (Treff AG, Degersheim, Switzerland) prefilled with argon was supplied with 15 µl of a reaction mixture (280 µl methanol, 200 µl H2O, 65 µl 5 M NaOH, 60 µl ethanethiol; all chemicals were GC grade, supplied by Sigma Aldrich). After incubation at 50°C for 1 hour, the suspension was acidified by the addition of 5 µl 70% (vol/vol) formic acid, applied to a Polybrene-treated TFA filter, and subjected to Edman degradation. Derivatives of modified amino acids were identified by comparison of the relative retention times with those presented and analyzed in previous studies (33, 48).

DNA manipulations.
General molecular biological techniques were performed essentially as described by Sambrook et al. (45). The DNA sequence of the michiganin A gene was obtained by using a genome-walking protocol, essentially as described by Vaughan et al. (56). Genomic DNA from C. michiganensis subsp. michiganensis 1468 and pUC18 DNA, both digested with the same restriction enzyme, were mixed and ligated, and the ligation mixture served as a template in PCRs with michiganin A-based degenerate primers in combination with vector primers. Sequences obtained through this approach were used to design new primers for PCR amplification of adjacent sequences. The forward and reverse vector primers were pUC1, 5'-GTTTTCCCAGTCACGAC-3', and M13 Reverse, 5'-CAGGAAACAGCTATGAC-3', respectively. In total, 16 degenerate primers were designed on the basis of the partial amino acid sequence of michiganin A, one of which, rev1d, 5'-GCRCADATDATNGTNCCACATTC-3', proved to be successful when combined with M13 Reverse. Primers Rdu 95, 5'-ACGCGGGCGAGGAC-3', and U2, 5'-CTGGCTCTGCACGCTCACC-3', were designed based on the DNA sequences of the PCR products obtained with the first primer pair. They were used in PCRs with vector primers pUC1L, 5'-CAAGGCGATTAAGTTGGGTAACG-3', and pUC2L, 5'-GAGCGGATAACAATTTCACACAGG-3', to amplify the C-terminal part of the michiganin A-encoding gene, which had not been covered by the degenerate primer.

All PCRs were performed with a GenAmp PCR System 9700 (PE Applied Biosystems). PCR was performed with 2 µM of the degenerate primer and 0.2 µM of the specific primer in the presence of 1.5 mM MgCl2 in a total volume of 25 µl, with annealing at 58°C. Two subsequent reactions of 35 cycles, in which 1 µl of the first reaction product served as a template in the following reaction, were necessary to obtain a detectable product. PCR with nonambiguous primers was performed with primer concentrations of 0.2 µM and annealing at 66°C. The PCR products were analyzed on 1% agarose gels, from which fragments were isolated and sequenced. Cycle sequencing was performed using the BigDye Terminator Cycle Sequencing Ready Reaction Kit v1.1 (Applied Biosystems) according to the manufacturer's instructions. Analysis of sequence products was done with an ABI Prism 310 capillary sequenator (Applied Biosystems). Sequence analysis was performed using the Lasergene software package (DNAStar Inc.).

Nucleotide sequence accession number.
The GenBank accession number of the DNA sequence reported in this paper is DQ458780.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Purification of michiganin A.
The producer and indicator strain, as well as the growth conditions for optimal bacteriocin production, were selected after extensive testing (results not shown). After ammonium sulfate precipitation of the culture supernatant, several chromatographic procedures were tested. The bacteriocin showed no affinity to either positively or negatively charged ion-exchanging resins tested at pH 4.0, 5.0, or 6.0 or at pH 8.0, respectively. A cation-exhange step (SP Sepharose; pH 4.0) was nevertheless included in the purification protocol to remove some contaminants. The flowthrough fraction from the SP Sepharose run was used as the starting point for three consecutive reversed-phase chromatography steps, the results of which are depicted in Fig. 1. Fractions from every purification step were sampled and assayed for activity. Active fractions were selected as indicated in Fig. 1, pooled, diluted, and applied to the next column (see Materials and Methods). The apparent yields of the purification steps are given in Table 1.


Figure 1
View larger version (29K):
[in this window]
[in a new window]
  		FIG. 1. Purification of michiganin A. (A) Resource RPC 3-ml column. The fractions used in the next step are indicated by the box. (B) pepRPC HR 5/5 column. The active fractions indicated with an "M" were pooled and used in the next step. The other active fractions contained low-molecular-weight (<1,000-Da) compounds, possibly tunicamycin-like antibiotics (I. Holtsmark, D. Mantzilas, V. G. H. Eijsink, and M. B. Brurberg, unpublished data). (C) µRPC C2/C18 SC 2.1/10 column. Absorbance was measured at 214 nm (curves labeled with filled squares) and/or 280 nm (curves labeled with open squares); the gray bars in panels A to C indicate the relative antimicrobial activities in the fractions. Curves showing the 2-propanol concentration are labeled with a filled triangle. (D) Mass spectrometry analysis of an active fraction shown in panel C (all fractions, i.e., from both peaks, yielded similar results).

 

View this table:
[in this window]
[in a new window]
  		TABLE 1. Yield of antimicrobial activity for each purification step

 
In the final chromatography step, antimicrobial activity eluted evenly over a double peak. Mass spectrometry analyses of active fractions from independent purifications consistently showed a mass of 2,145.0 Da that was associated with antimicrobial activity (Fig. 1D). This activity is hereafter referred to as michiganin A.

Amino acid analysis and N-terminal sequencing.
Amino acid composition analysis of the purified bacteriocin revealed Glu, Ser, Gly, Ala, Arg, Ile, and Leu contents, as well as several unidentified peaks, which could reflect the presence of nonstandard amino acids (Table 2).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Amino acid composition of michiganin A

 
Edman degradation yielded the sequence S-X-S-G-X-L-(X)6- for several independently prepared samples, showing that the sequencing process was arrested somewhere beyond the leucine residue at position 6. It is known that Edman degradation is blocked by unsaturated amino acids, such as Dha and Dhb, found in lantibiotics (25). Dha and Dhb residues that have been further modified to become parts of lanthionine ring structures usually yield unrecognizable signals but do not block further Edman degradation (25). Of the common amino acids, Cys and Trp remain undetected during Edman degradation, but they do not block continuation of the process. Thus, the results suggest that the amino acids at positions 2 and 5 are Cys, Trp, or part of a lanthionine ring, whereas a free Dha or Dhb residue may occur somewhere beyond Leu6.

Modification of lantibiotics with ethanethiol under alkaline conditions results in conversion of Dha and Dhb to stable S-ethylcysteinyl (S-EC) and ß-methyl-S-ethylcysteinyl (ß-M-S-EC) derivatives, respectively (33). Thioether bridges are also derivatized, but it is not easy to predict a priori the preferred modification patterns upon ethanethiol treatment (33, 48). In lanthionine bridges, both the original Cys and Ser residues may end up as (detectable) S-EC or (undetectable) Cys. In methyllanthionine bridges, the original Cys residue may end up as undetectable Cys or as detectable S-EC, whereas the original Thr may end up as undetectable ß-methylcysteine or detectable ß-M-S-EC (33). Most importantly, none of these derivatives block Edman degradation.

Edman degradation of ethanethiol-treated peptide yielded a 21-residue sequence, with 12 standard amino acids: Ser1-X2-Ser3-Gly4-X5-Leu6-X7-X8-Leu9-X10-Ile11-Glu12-X13-Gly14-X15-Ile16- Ile17-X18-Ala19-X20-Arg21. Using the results from Meyer et al. (33) and Smith et al. (48), the signal observed for residue 2 was interpreted as S-EC. Since Edman degradation of the unmodified peptide was not blocked at position 2 (excluding Dha), this result means that position 2 must contain a derivative of a Ser or a Cys involved in a lanthionine or ß-methyllanthionine bridge. Cycles 5, 13, 18, and 20 did not yield any signal, suggesting that the residues involved might be Trp, Cys, or residues involved in a lanthionine bridge. Cycle 7 yielded a peak close to leucine. Since leucine was not detected during Edman degradation of the unmodified peptide, residue 7 is likely to be a modified residue. Cycles 8, 10, and 15 yielded peaks close to the Phe peak (with slightly increased retention times), accompanied by relatively strong Leu signals. This pattern is characteristic of ß-M-S-EC (33, 48) (note that the amino acid composition analysis did not show any phenylalanine). This indicates that residues 8, 10, and 15 are Thr residues modified to either Dhb or a methyllanthionine moiety. The fact that Leu9 was not detected during Edman degradation of unmodified michiganin A shows that one of the two preceding residues must be a dehydrated residue not involved in a lanthionine bridge, blocking Edman degradation.

In conclusion, the amino acid sequence of michiganin A as deduced from the Edman degradation experiments was as follows: Ser1-Lan2-Ser3-Gly4-(Trp, Cys, Lan, MeLan)5-Leu6-X7-(Dhb/MeLan)8-Leu9-(Dhb/MeLan)10-Ile11-Glu12-(Trp, Cys, Lan, MeLan)13-Gly14-(Dhb/MeLan)15-Ile16-Ile17-(Trp, Cys, Lan, MeLan)18-Ala19-(Trp, Cys, Lan, MeLan)20-Arg21 (where Lan is lanthionine and MeLan is methyllanthionine). Searches of public sequence databases using PARALIGN software (43) revealed similarities with the type B lantibiotic actagardine, as depicted in Fig. 2 (see below for further discussion).


Figure 2
View larger version (8K):
[in this window]
[in a new window]
  		FIG. 2. Sequence alignment. Residues that differ between actagardine and michiganin A are printed in boldface. (1) Sequence of mature actagardine, as determined by nuclear magnetic resonance (NMR) (63). Lan, lanthionine; MeL, methyllanthionine. (2) The sequence of the structural gene encoding actagardine is not known. However, the stereochemical characteristics observed by NMR revealed the identity of the original residues involved in lanthionine bridge formation; these original residues are shown. (3) Sequence of michiganin A, as determined by Edman degradation of the ethanethiol-treated peptide. All residues with ambiguous assignments are represented by Xxx. (4) Identity of residues with ambiguous assignments and/or posttranslational modifications, as deduced from the sequence of the structural gene of michiganin A (see the text). (5) Proposed sequence of mature michiganin A (see the text).

 
Analysis of the michiganin A-encoding gene.
Only one of 16 degenerate primers that were designed on the basis of available sequence information for michiganin A and actagardine resulted in successful amplification of a part of the michiganin A-encoding gene. That primer was designed against the sequence Glu-Cys-Gly-Thr-Ile-Ile-Cys-Ala in reverse sense and yielded a 360-bp sequence, including the start of an open reading frame that ended in the primer binding site. This sequence was used to design primers used to amplify DNA fragments that included the C terminus of michiganin A. These experiments yielded a 375-bp DNA sequence with an open reading frame encoding the complete michiganin A precursor, which is depicted in Fig. 3.


Figure 3
View larger version (17K):
[in this window]
[in a new window]
  		FIG. 3. Nucleotide sequence and deduced amino acid sequence of the michiganin A gene. The 21 residues corresponding to the mature peptide are printed in boldface. An asterisk indicates the translation stop codon. The putative ribosome-binding site (rbs) is underlined. In-frame ATG codons upstream of the mature peptide are printed in boldface.

 
GC-rich organisms, such as C. michiganensis subsp. michiganensis (around 70% GC), are known to use GTG start codons more frequently and to have less conserved Shine-Dalgarno sequences than AT-rich organisms (38, 17). On the other hand, when the start codon is a typical ATG, a typical Shine-Dalgarno motif is often used, regardless of the genomic GC content (17). Since the DNA sequence upstream of the michiganin A coding sequence does not contain the alternative start codon GTG (or other known alternative start codons) in the correct reading frame, it is most likely that one of the four in-frame ATG codons upstream of the mature peptide functions as the initiation codon for the michiganin A precursor. Both the FrameD (46) and EasyGene (28) software packages predict the translational start at nucleotide 164, which would imply that the leader peptide of michiganin A contains 47 amino acids, with a net negative charge (–6 at pH 7). Analysis with SignalP (14) did not reveal any potential Sec-type leader peptide. Also, no typical promoter region could be found in the fragment.

Mass spectrometry.
The observed molecular mass of michiganin A (2,145.0 Da) is about 72 Da lower than the calculated mass of the 21-residue peptide deduced from the gene sequence (2,216.6 Da). Each dehydration of Ser or Thr would account for the loss of 18 Da, regardless of whether the dehydrated residue forms a lanthionine bridge. Combining the results of the Edman degradation experiments with the gene sequence (Fig. 2) shows that michiganin A contains four dehydrated amino acids (Ser2, Thr8, Thr10, and Thr15), which is in perfect agreement with the results from mass spectrometry. Since there are four Cys residues, all four dehydrated residues could be involved in lanthionine rings. As discussed above, the data from Edman degradation show that Ser2 is involved in a lanthionine ring. The fact that standard Edman degradation of the unmodified peptide was blocked between Leu6 and Leu9 strongly indicates that the dehydrated Thr8 is not involved in ring formation. The status of Thr10 and Thr15 cannot be deduced from the sequencing data, but the presence of three additional free Cys residues and the structure of actagardine strongly suggest the presence of lanthionine rings (see below for further discussion). The twin peak in the final chromatogram (Fig. 1C) suggests that there may be some heterogeneity, but this could not be detected at the mass level. Another possible explanation is the presence of stereoisomers (30, 52).

MIC.
With the correct amino acid sequence available, the peptide concentrations in samples of purified michiganin A could be approximated on the basis of the results of amino acid composition analysis. The concentration of michiganin A required to inhibit the growth of C. michiganensis subsp. sepedonicus 2136 by 50% was approximately 30 pmol/ml (30 nM).


   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Despite the potential of peptide bacteriocins in biological control, very little is known about the production of these compounds by plant pathogens. Some antibacterial substances from gram-negative plant pathogens have been investigated (39, 53), whereas ipomicin, produced by Streptomyces ipomoeae (61), is to our knowledge the only bacteriocin from a gram-positive plant pathogen that has been purified and characterized so far. The present study provides the first example of a lantibiotic from a plant (tomato) pathogen with proven activity toward another plant (potato) pathogen. Interestingly, recent studies indicate that the production of virulence factors and of antibiotics or bacteriocins may be coregulated (32, 36, 37, 58). Understanding the metabolism of bacteriocins may thus be a powerful portal to combating bacterial diseases in crops.

Figure 4 shows the known structures of actagardine (62, 63) and mersacidin (4, 40) and a hypothetical model of michiganin A based on the experimental data and on the structure of actagardine. The assignment of ring structures in michiganin A is based on analogy with actagardine. The presence of more than the one ring structure (involving residue 2) in michiganin A is most probable, since it is highly unlikely that the peptide contains three free cysteine residues. There are several differences between actagardine and michiganin A. Actagardine has a methyllanthionine ring connecting Thr7 and Cys12, while the analogous Thr8 in michiganin A is only dehydrated without engaging in a methyllanthionine ring. This means that at least one Cys residue in michiganin A, probably Cys13 (Fig. 4), has a free sulfhydryl group. Other differences involve two conservative internal mutations (two Val residues in actagardine are replaced by Leu and Ile in michiganin A), as well as one-residue N- and C-terminal extensions. The addition of a C-terminal charged arginine residue is potentially important in a peptide thought to interact with (negatively charged) membranes. In fact, many bacteriocins contain positively charged residues in their N- and/or C-terminal ends, and site-directed mutagenesis studies have shown that such residues are important for functionality of both lantibiotics and nonlantibiotics (3, 55). Actagardine and mersacidin interfere with the incorporation of lipid II into peptidoglycan (6, 7, 23, 49). Considering the high sequence similarity between michiganin A and actagardine, and considering the conservation of a glutamate residue thought to be important for lipid II interaction (at positions 11 and 12 in actagardine and michiganin A, respectively (2, 51), it is likely that michiganin A and actagardine exert their antimicrobial actions through similar mechanisms.


Figure 4
View larger version (17K):
[in this window]
[in a new window]
  		FIG. 4. Primary structures of type B lantibiotics. (A) Actagardine (62, 63). (B) Mersacidin (4, 40). (C) Michiganin A (model inferred from data presented here combined with the structure of actagardine; see the text). Abu-S is the threonine-derived moiety of a methyllanthionine ring. Ala-S is either the serine-based half of a lanthionine ring or the cysteine-derived moiety of either type of ring. The mass of the predicted mature michiganin A molecule is 2,144.6 Da, whereas the predicted mass of the unmodified gene product without the leader peptide is 2,216.6 Da. The 72-Da difference reflects four dehydration reactions that reduce the mass by 18 Da each; see the text for further details.

 
The DNA sequence of the structural gene for michiganin A is only the third for a type B lantibiotic; the other two known genes are those encoding mersacidin (1) and cinnamycin (24, 59), while no DNA sequence related to actagardine is known. Both programs used for finding the translational start site predicted a (rather long) 47-residue leader sequence with a net negative charge (the latter property is common for lantibiotics (4). So far, three different types of leader peptides have been observed in lantibiotics. Leader peptides of about 25 residues containing a conserved FNLD sequence are typical of lantibiotics that are modified by LanB and LanC enzymes and processed by a dedicated serine protease, LanP, such as nisin and subtilin (4, 44). Other type A lantibiotics contain a so-called "double-glycine" leader peptide of around 20 residues (21), which is recognized and processed by a chimeric transporter, LanT, with a protease domain. In terms of length, the putative leader peptide of michiganin A resembles those of the type B lantibiotics mersacidin and cinnamycin, whose leader peptides contain as many as 48 and 59 residues, respectively. Because an internal fragment of the mersacidin leader peptide shows some resemblance to a double-glycine leader and because the mersacidin gene cluster encodes a LanT protein, it has been suggested that processing of the mersacidin precursor involves the processing of a "double-glycine"-type leader, followed by further N-terminal proteolytic processing, possibly by a less specific protease (1). Such double processing has been found in a few other cases (5, 13). On the other hand, cinnamycin does not contain anything like a "double-glycine"-type cleavage site, nor does the gene cluster contain an open reading frame with similarity to LanT. In this case, processing by the Sec machinery has been suggested (59).

The putative leader sequence of michiganin A, which is only the third leader sequence of a type B lantibiotic, does not make the situation any clearer. First, it is not possible to make a sensible alignment of the three leader sequences. Secondly, the michiganin A leader sequence does not show any detectable known leader peptide characteristics. The only convincing similarity we could find was between short sequences in the C-terminal part of the michiganin A leader and the N-terminal part of the cinnamycin precursor (Fig. 5). In addition to interacting with the transport machinery, the leader peptides of lantibiotics presumably serve several functions related to self-protection (i.e., keeping the peptide inactive until it is secreted) and/or posttranslational modification (e.g., by directing the prepeptide to specific enzymes or by stabilizing structural properties of the precursor that are essential for correct posttranslational modification) (27, 60). It is possible that some of these functions involve conserved sequence elements, such as the one depicted in Fig. 5; shuffling of the positions of these elements in different leader peptides is not necessarily inconsistent with function. The general picture that emerges from this study is that the leader peptides of type B lantibiotics show large variation among themselves and that they do not resemble the better-known leader peptides of type A lantibiotics.


Figure 5
View larger version (5K):
[in this window]
[in a new window]
  		FIG. 5. Sequence similarity between the putative leader sequences of michiganin A and cinnamycin. Identical residues are shaded black; gray shading indicates strong similarity. The sequences are numbered from –1 (the residue immediately preceding the putative cleavage site).

 
In-depth molecular work is required to determine how the primary gene products of type B lantibiotics, such as michiganin A, are secreted and converted into mature, active lantibiotics. Such studies will not only provide further insight into the mechanism of bacteriocin production, but are also essential for further exploration of the potential use of these compounds in plant disease management.


   ACKNOWLEDGMENTS
 
This work was supported by a grant from the Norwegian Research Council and The Norwegian Crop Research Institute.

We thank Sissel Kjæraas, Knut Sletten, and Ola R. Blingsmo at the Biotechnology Centre, University of Oslo, for help with sequencing and amino acid analysis, and we are grateful to Morten Skaugen, Norwegian University of Life Sciences, for help with mass spectrometry. Thanks are due to Arild Sletten for the kind provision of C. michiganensis subsp. michiganensis and C. michiganensis subsp. sepedonicus strains.


   FOOTNOTES
 
* Corresponding author. Mailing address: Norwegian Institute for Agricultural and Environmental Research, Plant Health and Protection Division, Ås, Norway. Phone: 47 64949218. Fax: 47 6494922. E-mail: Ingrid.Holtsmark{at}Bioforsk.no. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Altena, K., A. Guder, C. Cramer, and G. Bierbaum. 2000. Biosynthesis of the lantibiotic mersacidin: organization of a type B lantibiotic gene cluster. Appl. Environ. Microbiol. 66:2565-2571.[Abstract/Free Full Text]
  2. Bauer, R., and L. M. T. Dicks. 2005. Mode of action of lipid II-targeting lantibiotics. Int. J. Food Microbiol. 101:201-216.[CrossRef][Medline]
  3. Bhugaloo-Vial, P., J. P. Douliez, D. Molle, X. Dousset, P. Boyaval, and D. Marion. 1999. Delineation of key amino acid side chains and peptide domains for antimicrobial properties of divercin V41, a pediocin-like bacteriocin secreted by Carnobacterium divergens V41. Appl. Environ. Microbiol. 65:2895-2900.[Abstract/Free Full Text]
  4. Bierbaum, G., H. Brötz, K. P. Koller, and H. G. Sahl. 1995. Cloning, sequencing and production of the lantibiotic mersacidin. FEMS Microbiol. Lett. 127:121-126.[CrossRef][Medline]
  5. Booth, M. C., C. P. Bogie, H. G. Sahl, R. J. Siezen, K. L. Hatter, and M. S. Gilmore. 1996. Structural analysis and proteolytic activation of Enterococcus faecalis cytolysin, a novel lantibiotic. Mol. Microbiol. 21:1175-1184.[CrossRef][Medline]
  6. Brötz, H., G. Bierbaum, A. Markus, E. Molitor, and H. G. Sahl. 1995. Mode of action of the lantibiotic mersacidin: inhibition of peptidoglycan biosynthesis via a novel mechanism? Antimicrob. Agents Chemother. 39:714-719.[Abstract]
  7. Brötz, H., G. Bierbaum, P. E. Reynolds, and H. G. Sahl. 1997. The lantibiotic mersacidin inhibits peptidoglycan biosynthesis at the level of transglycosylation. Eur. J. Biochem. 246:193-199.[Medline]
  8. Brötz, H., M. Josten, I. Wiedemann, U. Schneider, F. Gotz, G. Bierbaum, and H. G. Sahl. 1998. Role of lipid-bound peptidoglycan precursors in the formation of pores by nisin, epidermin and other lantibiotics. Mol. Microbiol. 30:317-327.[CrossRef][Medline]
  9. 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]
  10. Cotter, P. D., C. Hill, and R. P. Ross. 2005. Bacteriocins: developing innate immunity for food. Nat. Rev. Microbiol. 3:777-788.[CrossRef][Medline]
  11. Creighton, T. E. 1993. Proteins: structures and molecular properties. W. H. Freeman and Company, New York, N.Y.
  12. Delves-Broughton, J., P. Blackburn, R. J. Evans, and J. Hugenholtz. 1996. Applications of the bacteriocin, nisin. Antonie Leeuwenhoek 69:193-202.[CrossRef][Medline]
  13. Diep, D. B., L. S. Håvarstein, and I. F. Nes. 1996. Characterization of the locus responsible for the bacteriocin production in Lactobacillus plantarum C11. J. Bacteriol. 178:4472-4483.[Abstract/Free Full Text]
  14. Dyrlov Bendtsen, J., H. Nielsen, G. von Heijne, and S. Brunak. 2004. Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 340:783-795.[CrossRef][Medline]
  15. Echandi, E. 1976. Bacteriocin production by Corynebacterium michiganense. Phytopathology 66:430-432.
  16. Eijsink, V. G. H., L. Axelsson, D. B. Diep, L. S. Håvarstein, H. Holo, and I. F. Nes. 2002. Production of class II bacteriocins by lactic acid bacteria; an example of biological warfare and communication. Antonie Leeuwenhoek 81:639-654.[CrossRef][Medline]
  17. Fuglsang, A., and J. Engberg. 2003. Non-randomness in Shine-Dalgarno regions: links to gene characteristics. Biochem. Biophys. Res. Commun. 302:296-301.[CrossRef][Medline]
  18. Geis, A., J. Singh, and M. Teuber. 1983. Potential of lactic streptococci to produce bacteriocin. Appl. Environ. Microbiol. 45:205-211.[Abstract/Free Full Text]
  19. Gross, D. C., and A. K. Vidaver. 1979. Bacteriocins of phytopathogenic Corynebacterium species. Can. J. Microbiol. 25:367-374.[Medline]
  20. Guinane, C. M., P. D. Cotter, C. Hill, and R. P. Ross. 2005. Microbial solutions to microbial problems; lactococcal bacteriocins for the control of undesirable biota in food. J. Appl. Microbiol. 98:1316-1325.[CrossRef][Medline]
  21. Håvarstein, L. S., D. B. Diep, and I. F. Nes. 1995. A family of bacteriocin ABC transporters carry out proteolytic processing of their substrates concomitant with export. Mol. Microbiol. 16:229-240.[Medline]
  22. Hert, A. P., P. D. Roberts, M. T. Momol, G. V. Minsavage, S. M. Tudor-Nelson, and J. B. Jones. 2005. Relative importance of bacteriocin-like genes in antagonism of Xanthomonas perforans tomato race 3 to Xanthomonas euvesicatoria tomato race 1 strains. Appl. Environ. Microbiol. 71:3581-3588.[Abstract/Free Full Text]
  23. Hsu, S. T., E. Breukink, G. Bierbaum, H. G. Sahl, B. de Kruijff, R. Kaptein, N. A. J. van Nuland, and A. M. J. J. Bonvin. 2003. NMR study of mersacidin and lipid II interaction in dodecylphosphocholine micelles. Conformational changes are a key to antimicrobial activity. J. Biol. Chem. 278:13110-13117.[Abstract/Free Full Text]
  24. Kaletta, C., K. D. Entian, and G. Jung. 1991. Prepeptide sequence of cinnamycin (Ro 09-0198): the first structural gene of a duramycin-type lantibiotic. Eur. J. Biochem. 199:411-415.[Medline]
  25. Kellner, R., G. Jung, T. Horner, H. Zahner, N. Schnell, K. D. Entian, and F. Gotz. 1988. Gallidermin: a new lanthionine-containing polypeptide antibiotic. Eur. J. Biochem. 177:53-59.[Medline]
  26. Klaenhammer, T. R. 1993. Genetics of bacteriocins produced by lactic-acid bacteria. FEMS Microbiol. Rev. 12:39-86.[Medline]
  27. 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]
  28. Larsen, T., and A. Krogh. 2003. EasyGene—a prokaryotic gene finder that ranks ORFs by statistical significance. BMC Bioinformatics 4:21.[Medline]
  29. Lavermicocca, P., S. L. Lonigro, F. Valerio, A. Evidente, and A. Visconti. 2002. Reduction of olive knot disease by a bacteriocin from Pseudomonas syringae pv. ciccaronei. Appl. Environ. Microbiol. 68:1403-1407.[Abstract/Free Full Text]
  30. Liu, S., E. Cheung, M. Rajopadhye, N. E. Williams, K. L. Overoye, and D. S. Edwards. 2001. Isomerism and solution dynamics of Y-90-labeled DTPA-biomolecule conjugates. Bioconjug. Chem. 12:84-91.[CrossRef][Medline]
  31. McAuliffe, O., R. P. Ross, and C. Hill. 2001. Lantibiotics: structure, biosynthesis and mode of action. FEMS Microbiol. Rev. 25:285-308.[CrossRef][Medline]
  32. McGowan, S. J., A. M. L. Barnard, G. Bosgelmez, M. Sebaihia, N. J. L. Simpson, N. R. Thomson, D. E. Todd, M. Welch, N. A. Whitehead, and G. P. C. Salmond. 2005. Carbapenem antibiotic biosynthesis in Erwinia carotovora is regulated by physiological and genetic factors modulating the quorum sensing-dependent control pathway. Mol. Microbiol. 55:526-545.[CrossRef][Medline]
  33. Meyer, H. E., M. Heber, B. Eisermann, H. Korte, J. W. Metzger, and G. Jung. 1994. Sequence analysis of lantibiotics: chemical derivatization procedures allow a fast access to complete Edman degradation. Anal. Biochem. 223:185-190.[CrossRef][Medline]
  34. Moll, G. N., W. N. Konings, and A. J. M. Driessen. 1999. Bacteriocins: mechanism of membrane insertion and pore formation. Antonie Leeuwenhoek 76:185-198.[CrossRef][Medline]
  35. Nelson, G. A., and G. Semeniuk. 1964. Antagonistic variant of Corynebacterium insidiosum + some properties of inhibitor. Phytopathology 54:330.
  36. Nguyen, A. H., T. Tomita, M. Hirota, T. Sato, and Y. Kamio. 1999. A simple purification method and component analyses for carotovoricin Er, a phage-tail-like bacteriocin from the plant pathogen Erwinia carotovora Er. Biosci. Biotechnol. Biochem. 63:1360-1369.[CrossRef][Medline]
  37. Nguyen, H. A., J. Kaneko, and Y. Kamio. 2002. Temperature-dependent production of carotovoricin Er and pectin lyase in phytopathogenic Erwinia carotovora subsp carotovora Er. Biosci. Biotechnol. Biochem. 66:444-447.[CrossRef][Medline]
  38. Nielsen, P., and A. Krogh. 2005. Large-scale prokaryotic gene prediction and comparison to genome annotation. Bioinformatics 21:4322-4329.[Abstract/Free Full Text]
  39. Parret, A. H. A., G. Schoofs, P. Proost, and R. De Mot. 2003. Plant lectin-like bacteriocin from a rhizosphere-colonizing Pseudomonas isolate. J. Bacteriol. 185:897-908.[Abstract/Free Full Text]
  40. Prasch, T., T. Naumann, R. L. M. Markert, M. Sattler, W. Schubert, S. Schaal, M. Bauch, H. Kogler, and C. Griesinger. 1997. Constitution and solution conformation of the antibiotic mersacidin determined by NMR and molecular dynamics. Eur. J. Biochem. 244:501-512.[Medline]
  41. Riley, M. A., C. M. Goldstone, J. E. Wertz, and D. Gordon. 2003. A phylogenetic approach to assessing the targets of microbial warfare. J. Evol. Biol. 16:690-697.[CrossRef][Medline]
  42. Riley, M. A., and D. M. Gordon. 1999. The ecological role of bacteriocins in bacterial competition. Trends Microbiol. 7:129-133.[CrossRef][Medline]
  43. Sæbø, P. E., S. M. Andersen, J. Myrseth, J. K. Laerdahl, and T. Rognes. 2005. PARALIGN: rapid and sensitive sequence similarity searches powered by parallel computing technology. Nucleic Acids Res. 33:W535-W539.[Abstract/Free Full Text]
  44. Sahl, H. G., and G. Bierbaum. 1998. Lantibiotics: biosynthesis and biological activities of uniquely modified peptides from gram-positive bacteria. Annu. Rev. Microbiol. 52:41-79.[CrossRef][Medline]
  45. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
  46. Schiex, T., J. Gouzy, A. Moisan, and Y. de Oliveira. 2003. FrameD: a flexible program for quality check and gene prediction in prokaryotic genomes and noisy matured eukaryotic sequences. Nucleic Acids Res. 31:3738-3741.[Abstract/Free Full Text]
  47. Schnell, N., K. D. Entian, U. Schneider, F. Gotz, H. Zahner, R. Kellner, and G. Jung. 1988. Prepeptide sequence of epidermin, a ribosomally synthesized antibiotic with 4 sulfide-rings. Nature 333:276-278.[CrossRef][Medline]
  48. Smith, L., J. Novak, J. Rocca, S. McClung, J. D. Hillman, and A. S. Edison. 2000. Covalent structure of mutacin 1140 and a novel method for the rapid identification of lantibiotics. Eur. J. Biochem. 267:6810-6816.[Medline]
  49. Somma, S., W. Merati, and F. Parenti. 1977. Gardimycin, a new antibiotic inhibiting peptidoglycan synthesis. Antimicrob. Agents Chemother. 11:396-401.[Abstract/Free Full Text]
  50. Stern, N. J., E. A. Svetoch, B. V. Eruslanov, Y. N. Kovalev, L. I. Volodina, V. V. Perelygin, E. V. Mitsevich, I. P. Mitsevich, and V. P. Levchuk. 2005. Paenibacillus polymyxa purified bacteriocin to control Campylobacter jejuni in chickens. J. Food Prot. 68:1450-1453.[Medline]
  51. Szekat, C., R. W. Jack, D. Skutlarek, H. Farber, and G. Bierbaum. 2003. Construction of an expression system for site-directed mutagenesis of the lantibiotic mersacidin. Appl. Environ. Microbiol. 69:3777-3783.[Abstract/Free Full Text]
  52. Tamaki, M., S. Akabori, and I. Muramatsu. 1994. Studies of gramicidin-S analogs having various ring sizes by reversed-phase high-performance liquid-chromatography. J. Chromatogr. 685:237-241.[CrossRef][Medline]
  53. Tudor-Nelson, S. M., G. V. Minsavage, R. E. Stall, and J. B. Jones. 2003. Bacteriocin-like substances from tomato race 3 strains of Xanthomonas campestris pv. vesicatoria. Phytopathology 93:1415-1421.
  54. van der Wolf, J. M., J. G. Elphinstone, D. E. Stead, M. Metzler, P. Müller, A. Hukkanen, and R. Karjalainen. 2005. Epidemology of Clavibacter michiganensis subsp. sepedonicus in relation to control of bacterial ring rot. Report 95. Plant Research International B.V., Wageningen, The Netherlands.
  55. van Kraaij, C., E. Breukink, H. S. Rollema, R. J. Siezen, R. A. Demel, B. DeKruijff, and O. P. Kuipers. 1997. Influence of charge differences in the C-terminal part of nisin on antimicrobial activity and signaling capacity. Eur. J. Biochem. 247:114-120.[Medline]
  56. Vaughan, A., V. G. H. Eijsink, and D. van Sinderen. 2003. Functional characterization of a composite bacteriocin locus from malt isolate Lactobacillus sakei 5. Appl. Environ. Microbiol. 69:7194-7203.[Abstract/Free Full Text]
  57. Vidaver, A. K. 1983. Bacteriocins—the lure and the reality. Plant Dis. 67:471-475.
  58. von Bodman, S. B., W. D. Bauer, and D. L. Coplin. 2003. Quorum sensing in plant-pathogenic bacteria. Annu. Rev. Phytopathol. 41:455-482.[CrossRef][Medline]
  59. Widdick, D. A., H. M. Dodd, P. Barraille, J. White, T. H. Stein, K. F. Chater, M. J. Gasson, and M. J. Bibb. 2003. Cloning and engineering of the cinnamycin biosynthetic gene cluster from Streptomyces cinnamoneus cinnamoneus DSM 40005. Proc. Natl. Acad. Sci. USA 100:4316-4321.[Abstract/Free Full Text]
  60. 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]
  61. Zhang, X., C. A. Clark, and G. S. Pettis. 2003. Interstrain inhibition in the sweet potato pathogen Streptomyces ipomoeae: purification and characterization of a highly specific bacteriocin and cloning of its structural gene. Appl. Environ. Microbiol. 69:2201-2208.[Abstract/Free Full Text]
  62. Zimmermann, N., and G. Jung. 1997. The three-dimensional solution structure of the lantibiotic murein-biosynthesis-inhibitor actagardine determined by NMR. Eur. J. Biochem. 246:809-819.[Medline]
  63. Zimmermann, N., J. W. Metzger, and G. Jung. 1995. The tetracyclic lantibiotic actagardine 1H-NMR and 13C-NMR assignments and revised primary structure. Eur. J. Biochem. 228:786-797.[Medline]

Applied and Environmental Microbiology, September 2006, p. 5814-5821, Vol. 72, No. 9
0099-2240/06/$08.00+0     doi:10.1128/AEM.00639-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Holtsmark, I.
Right arrow Articles by Brurberg, M. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Holtsmark, I.
Right arrow Articles by Brurberg, M. B.
Agricola
Right arrow Articles by Holtsmark, I.
Right arrow Articles by Brurberg, M. B.

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
J. Bacteriol. Microbiol. Mol. Biol. Rev. Eukaryot. Cell All ASM Journals