JOURNAL OF BACTERIOLOGY, Nov. 1991, p. 7387-7390
Vol. 173, No. 22
0021-9193/91/227387-04$02.00/0 Copyright C 1991, American Society for Microbiology
Conversion of Bacillus subtilis 168 to a Subtilin Producer by Competence Transformation WEI LIU AND J. NORMAN HANSEN* Department of Chemistry and Biochemistry, University of Maryland, College Park, Maryland 20742 Received 19 June 1991/Accepted 4 September 1991
Subtilin is a ribosomally synthesized peptide antibiotic produced by Bacillus subtilis ATCC 6633. B. subtilis 168 was converted to a subtilin producer by competence transformation with chromosomal DNA from B. subtilis ATCC 6633. A chloramphenicol acetyltransferase gene was inserted next to the subtilin structural gene as a selectable marker. The genes that conferred subtilin production were derived from a 40-kb region of the B. subtilis ATCC 6633 chromosome that had flanking homologies to the B. subtilis 168 chromosome. The subtilin produced by the mutant was identical to natural subtilin in its biological activity, chromatographic behavior, amino acid composition, and N-terminal amino acid sequence.
mation was not determined. A random selection of transformants was tested for antimicrobial activity by using a halo assay (Fig. 3A) and for the presence of the subtilin gene by Southern blotting (Fig. 3B). Of several dozen transformants tested, more than 90% produced halos and contained the subtilin gene. That the transformants were derivatives of BR151 was confirmed by comparing restriction patterns of their chromosomal DNA with the pattern from BR151 (data not shown). One of the subtilin-producing transformants (LH45) was tested further. Successful transformation of a nonproducing strain to subtilin production indicates that all of the additional genes required for subtilin biosynthesis, over and above any endogenous genes, are contained within a single fragment of integrated DNA. Knowing the size of the integrated piece of DNA would establish an upper limit to the amount of DNA that encodes the set of new genes that confer subtilin production to strain BR151. The size was determined by carrying out transformations of strain BR151 by using LH45 DNA that had been restricted with a variety of rarely cutting restriction enzymes. Only those restriction fragments that contained homologies flanking both sides of the DNA containing the subtilin gene(s) should be capable of transformation. The strategy used and results obtained are presented in Fig. 2C and Fig. 3C. They establish that the subtilin gene(s) lie within a 60-kb MluI restriction fragment in LH45 and that removing 2 to 5 kb from either end of this fragment destroys its ability to transform. After taking account of the size of the
Subtilin is one of several ribosomally synthesized antimicrobial peptides that contains many unusual amino acids, including dehydroalanine, dehydrobutyrine, lanthionine, and 3-methyllanthionine (1, 2). Subtilin has a broad spectrum of action that encompasses many gram-positive species (10). The structure of subtilin is shown in Fig. 1. The unusual amino acids are introduced by posttranslational modifications of a precursor peptide that undergoes dehydrations of serines and threonines, formation of thioether cross-links, and cleavage of a leader peptide (1, 2). The natural producer of subtilin, Bacillus subtilis ATCC 6633, has not been genetically characterized. Further genetic studies of subtilin and the genes involved in its biosynthesis would be greatly facilitated if they could be transported to B. subtilis 168, for which much genetic information and a wide range of genetic methodologies are available (14). Moreover, since B. subtilis 168 neither contains the subtilin gene nor produces subtilin (1), its conversion to a subtilin producer by a technique such as competence transformation would provide evidence that all of the genes required for subtilin production are clustered together and can be transferred as a unit. This article reports the successful conversion of B. subtilis 168 to a producer of authentic subtilin by competence transformation by using DNA from subtilin-producing B. subtilis ATCC 6633. (This research was conducted by W. Liu in partial fulfillment of the requirements for a Ph.D. from the University of Maryland, College Park.) The chloramphenicol acetyltransferase (CAT) gene was used as a selectable marker for B. subtilis 168 mutants that produce subtilin. Accordingly, a CAT gene was inserted into the B. subtilis ATCC 6633 chromosome just downstream from the subtilin gene, as shown in Fig. 2A. Total chromosomal DNA from the insertion mutant was then used to transform B. subtilis BR151, which is a substrain of B. subtilis 168, to chloramphenicol resistance. Since the transforming chromosomal DNA was linear, the expected mechanism of acquiring chloramphenicol resistance was by integration of the CAT gene into the BR151 chromosome by a double crossover. Many chloramphenicol-resistant transformants were obtained, suggesting that appropriate flanking homologies are present, although the efficiency of transfor*
-Leu
SUBTiLiN
DI
S
I
9
S
4la-Phe-Leu-Gln-ABA
o -Val-ABA
Leu- Ala -ABA
o
Gly
S
sS
/\
Glu -Aia-Lys-Trp-NH
S
HOOC-Lys-DHA-lie-Lys-
%/C
&
D D
Asn-0 Gross et al. (6-8).
FIG. 1. Structure of subtilin as determined by Abbreviations: ABA, aminobutyric acid; DHA, dehydroalanine; DHB, dehydrobutyrine (,B-methyldehydroalanine); Ala-S-Ala, lanthionine; ABA-S-Ala, P-methyllanthionine. Asterisks indicate residues that are in the D-stereoconfiguration.
Corresponding author. 7387
7388
NOTES
J. BACTERIOL.
integrated plasmid, we conclude that the DNA that confers subtilin production could amount to as much as 40 kb. However, this does not mean that genes for subtilin production occupy the entire 40 kb. The minimal size of the integrable DNA could reflect the location of flanking homologies rather than the coding function. Transformation of BR151 to a subtilin producer estab-
pTlZ9U plasmid
A
EcoRI
6633 chromosome
Campbell integration EcoRI
EcoRI
v
plZ19U
B 6633-CAT
LH4L
c
M: Mlu I K Kpn I S Sma I A: Apa I
IL
KAM
Ms
I1
S
subtilin gene region Kb
10
io
40
10
gO
60
1°0
g0
0
FIG. 2. (A) Strategy for placing a CAT selectable marker beside the subtilin precursor gene in B. subtilis ATCC 6633. The subtilin gene is the subtilin precursor peptide gene located upstream from an XbaI site. A promoter-containing CAT gene obtained from pC194 (4, 9) by using the polymerase chain reaction (kindly provided by Mark T. Steen) was ligated into the XbaI site of a 3.3-kb EcoRI-XbaI restriction fragment, in which the terminator of the subtilin precursor peptide gene lies about 300 bp upstream from the XbaI site (1). This construct was cloned into plasmid pTZ19U (which contains an origin of replication that functions in Escherichia coli but not in Bacillus spp.), creating pRXCM. After amplification in E. coli, the plasmid was introduced into B. subtilis ATCC 6633 by transformation by use of the polyethylene glycol method, and chloramphenicolresistant colonies were selected. Chloramphenicol resistance is conferred by a Campbell-type insertion of the plasmid into the chromosome. (B) Strategy for transformation of the subtilin operon into BR151. Total chromosomal DNA from the chloramphenicolresistant transformants of B. subtilis ATCC 6633 (panel A) was used to transform competent cells of BR151 to chloramphenicol resistance. Integration of the CAT gene into the linear chromosomal DNA is presumed to occur by a double recombination involving homologies in the flanking regions around the subtilin operon and corresponding homologies in the BR151 chromosome. (C) Restriction map of subtilin transforming region around the subtilin pre-
lishes that all the gene products required for the posttranslational modifications are present in the transformant. Although the halo assay indicates that the LH45 transformant is a subtilin producer, the complexity of precursor maturation requires verification that strain LH45 performs all of the posttranslational processing steps correctly. We employed the following tests of correct processing: biological activity, chromatographic behavior, amino acid composition, and N-terminal sequence analysis. Subtilin was isolated from strains ATCC 6633 and LH45 in the following way. Culture supernatants were heated for 2 to 3 min at 121°C to inactivate proteases and then applied to a carboxymethyl cellulose column in 50 mM NaPi (pH 5.8). The column was eluted with an NaCl gradient buffered with 50 mM NaPi. The major A254 peak was lyophilized, dissolved in 0.1% trifluoroacetic acid, and fractionated on a C-18 high-performance liquid chromatography column as described for nisin (13). The peak fractions were lyophilized immediately after separation, and the dried powder was stored at -80°C. These subtilin preparations were tested for their ability to inhibit spore outgrowth in a culture assay (13). Their specific activities were the same. They were subjected to amino acid composition and N-terminal sequence analyses. Because maturation of the subtilin precursor peptide requires dehydrations, cross-linking, and removal of the leader region, the amino acid composition would reflect these posttranslational modifications. If both subtilins were the same, their elution profiles from the amino acid analyzer would be the same, even though the presence of the unusual amino acids would give unidentifiable peaks. The elution profiles obtained were indistinguishable (data not shown). Moreover, the calculated amino acid compositions of the identifiable amino acids were those expected from the known structure of subtilin. The unusual amino acids also give characteristic patterns during N-terminal sequence analysis. The dehydro residues (dehydroalanine or dehydrobutyrine) irreversibly block subsequent rounds of cleavage (11, 12). Also, when the Ala of Ala-S-R is cleaved, it leaves a blank. Subtilin isolated from strains ATCC 6633 and LH45 performed identically during sequence analysis. Namely, there was a blank at residue 3, as expected if residue 3 is the first Ala-S-R of lanthionine and as expected for subtilin. Also, both showed Glu at residue 4, with a molar yield of 70% compared to the initial Trp. If there had been a dehydro residue at position 3, it would have been blocked, with Glu not appearing at all. Both preparations of subtilin, therefore, have a lanthionine at position 3. Sequencing further, both forms were blank at residue 5 and blank during subsequent cycles, showing that both forms cursor peptide gene. DNA from LH45 was digested with rarely cutting restriction enzymnes individually and in combination, and the restricted DNA was used to transform strain BR151. MluI fragments gave transformants, indicating that homologies on both sides were present. The size of the transforming MluI fragment is 60 kb, as shown in Fig. 3C. Treatment with SmaI, which removes 2 to 5 kb from the left side of the MluI fragment, or with Kpnl, which removes 2 to 5 kb from the right side of the MluI fragment, destroyed the ability of the MluI fragment to transform.
*5>.:t~68Kb
VOL. 173, 1991
A
~~BR151
,
~~~LH45
_
B
_
21.2 Kb
5.0 Kb
4.3 Kb
1 2 3 200 Kb
Kb_ 15i1 00 Kb_
_
_
4_60 Kb 'li1l
50Kb
FIG. 3. (A) Halo assay showing subtilin production in wild-type B. subtilis ATCC 6633 and several subtilin-producing transformants of BR151 (LH1, LH2, LH3, and LH45). Nontransformed BR151 is included to show the absence of subtilin production in the original recipient. LH45 was used for further studies. (B) Autoradiogram of a Southern blot of an agarose gel of EcoRI-digested total chromosomal DNA that had been hybridized with a 21-mer oligonucleotide probe that is complementary to the subtilin precursor peptide gene. The 6633 lane was DNA obtained from the CAT gene derivative of B. subtilis ATCC 6633 shown in Fig. 2A. LH45 and LH33 are transformants of BR151 that had been transformed with that 6633 CAT DNA and selected on chloramphenicol as described in Fig. 2B. The BR151 lane shows that the subtilin gene is absent from the original recipient strain. (C) Lanes: 1, ethidium-bromide-stained orthogonal-field-alternation (3) gel of LH45 DNA that had been digested with MluI; 2, concatomeric lambda DNA size standard with the sizes shown on the left; 3, autoradiogram of a Southern blot of lane 1 that had been hybridized with the nick-translated pRX fragment (cf. Fig. 2A) that contains the subtilin precursor peptide gene. The 60-kb size was estimated from the lambda standard.
NOTES
7389
had a dehydro residue at position 5. These conclusions were further supported by treating the subtilin preparations with N'-(methyl)mercaptoacetamide before sequencing. This mercaptan undergoes a Michael addition across the double bond of the dehydro residues, which stabilizes them against deamination during sequence analysis. When the mercaptantreated forms were sequenced, both gave Leu at residue 6, as expected for subtilin. It was concluded that both forms of subtilin have the same N-terminal sequence and have undergone the same posttranslational modifications in which the dehydro residues and thioether cross-linkages were formed and the leader sequence was removed. These results show that the LH45 transformant produces authentic subtilin. The integration of approximately 40 kb of ATCC 6633 DNA suggests that genes for functions other than subtilin production may have been incorporated as well. For example, we observed that strain LH45 grows well on medium A, which contains 10% sucrose, as was observed for ATCC 6633. Strain 168 normally grows poorly in sucrose-containing medium A, suggesting that LH45 has acquired gene(s) that affect sucrose metabolism. This is reminiscent of the observation that genes for nisin production in Lactococcus lactis are closely linked to genes for sucrose metabolism (5). The conversion of strain 168 to a subtilin producer by integration of a fragment of ATCC 6633 DNA does not conclusively prove that all of the genes required for subtilin production reside on the integrated DNA. The integrated DNA may merely complement endogenous functions that are already present in strain 168, or it may provide regulatory factors that switch on otherwise cryptic genes. The only thing that can be stated with certainty is that the DNA that has been integrated into strain 168 contributes all additional genes over and above those for endogenous host functions that are required for subtilin biosynthesis and that all of these genes are clustered together in a reasonably small region of the chromosome (
Vol. 173, No. 22
0021-9193/91/227387-04$02.00/0 Copyright C 1991, American Society for Microbiology
Conversion of Bacillus subtilis 168 to a Subtilin Producer by Competence Transformation WEI LIU AND J. NORMAN HANSEN* Department of Chemistry and Biochemistry, University of Maryland, College Park, Maryland 20742 Received 19 June 1991/Accepted 4 September 1991
Subtilin is a ribosomally synthesized peptide antibiotic produced by Bacillus subtilis ATCC 6633. B. subtilis 168 was converted to a subtilin producer by competence transformation with chromosomal DNA from B. subtilis ATCC 6633. A chloramphenicol acetyltransferase gene was inserted next to the subtilin structural gene as a selectable marker. The genes that conferred subtilin production were derived from a 40-kb region of the B. subtilis ATCC 6633 chromosome that had flanking homologies to the B. subtilis 168 chromosome. The subtilin produced by the mutant was identical to natural subtilin in its biological activity, chromatographic behavior, amino acid composition, and N-terminal amino acid sequence.
mation was not determined. A random selection of transformants was tested for antimicrobial activity by using a halo assay (Fig. 3A) and for the presence of the subtilin gene by Southern blotting (Fig. 3B). Of several dozen transformants tested, more than 90% produced halos and contained the subtilin gene. That the transformants were derivatives of BR151 was confirmed by comparing restriction patterns of their chromosomal DNA with the pattern from BR151 (data not shown). One of the subtilin-producing transformants (LH45) was tested further. Successful transformation of a nonproducing strain to subtilin production indicates that all of the additional genes required for subtilin biosynthesis, over and above any endogenous genes, are contained within a single fragment of integrated DNA. Knowing the size of the integrated piece of DNA would establish an upper limit to the amount of DNA that encodes the set of new genes that confer subtilin production to strain BR151. The size was determined by carrying out transformations of strain BR151 by using LH45 DNA that had been restricted with a variety of rarely cutting restriction enzymes. Only those restriction fragments that contained homologies flanking both sides of the DNA containing the subtilin gene(s) should be capable of transformation. The strategy used and results obtained are presented in Fig. 2C and Fig. 3C. They establish that the subtilin gene(s) lie within a 60-kb MluI restriction fragment in LH45 and that removing 2 to 5 kb from either end of this fragment destroys its ability to transform. After taking account of the size of the
Subtilin is one of several ribosomally synthesized antimicrobial peptides that contains many unusual amino acids, including dehydroalanine, dehydrobutyrine, lanthionine, and 3-methyllanthionine (1, 2). Subtilin has a broad spectrum of action that encompasses many gram-positive species (10). The structure of subtilin is shown in Fig. 1. The unusual amino acids are introduced by posttranslational modifications of a precursor peptide that undergoes dehydrations of serines and threonines, formation of thioether cross-links, and cleavage of a leader peptide (1, 2). The natural producer of subtilin, Bacillus subtilis ATCC 6633, has not been genetically characterized. Further genetic studies of subtilin and the genes involved in its biosynthesis would be greatly facilitated if they could be transported to B. subtilis 168, for which much genetic information and a wide range of genetic methodologies are available (14). Moreover, since B. subtilis 168 neither contains the subtilin gene nor produces subtilin (1), its conversion to a subtilin producer by a technique such as competence transformation would provide evidence that all of the genes required for subtilin production are clustered together and can be transferred as a unit. This article reports the successful conversion of B. subtilis 168 to a producer of authentic subtilin by competence transformation by using DNA from subtilin-producing B. subtilis ATCC 6633. (This research was conducted by W. Liu in partial fulfillment of the requirements for a Ph.D. from the University of Maryland, College Park.) The chloramphenicol acetyltransferase (CAT) gene was used as a selectable marker for B. subtilis 168 mutants that produce subtilin. Accordingly, a CAT gene was inserted into the B. subtilis ATCC 6633 chromosome just downstream from the subtilin gene, as shown in Fig. 2A. Total chromosomal DNA from the insertion mutant was then used to transform B. subtilis BR151, which is a substrain of B. subtilis 168, to chloramphenicol resistance. Since the transforming chromosomal DNA was linear, the expected mechanism of acquiring chloramphenicol resistance was by integration of the CAT gene into the BR151 chromosome by a double crossover. Many chloramphenicol-resistant transformants were obtained, suggesting that appropriate flanking homologies are present, although the efficiency of transfor*
-Leu
SUBTiLiN
DI
S
I
9
S
4la-Phe-Leu-Gln-ABA
o -Val-ABA
Leu- Ala -ABA
o
Gly
S
sS
/\
Glu -Aia-Lys-Trp-NH
S
HOOC-Lys-DHA-lie-Lys-
%/C
&
D D
Asn-0 Gross et al. (6-8).
FIG. 1. Structure of subtilin as determined by Abbreviations: ABA, aminobutyric acid; DHA, dehydroalanine; DHB, dehydrobutyrine (,B-methyldehydroalanine); Ala-S-Ala, lanthionine; ABA-S-Ala, P-methyllanthionine. Asterisks indicate residues that are in the D-stereoconfiguration.
Corresponding author. 7387
7388
NOTES
J. BACTERIOL.
integrated plasmid, we conclude that the DNA that confers subtilin production could amount to as much as 40 kb. However, this does not mean that genes for subtilin production occupy the entire 40 kb. The minimal size of the integrable DNA could reflect the location of flanking homologies rather than the coding function. Transformation of BR151 to a subtilin producer estab-
pTlZ9U plasmid
A
EcoRI
6633 chromosome
Campbell integration EcoRI
EcoRI
v
plZ19U
B 6633-CAT
LH4L
c
M: Mlu I K Kpn I S Sma I A: Apa I
IL
KAM
Ms
I1
S
subtilin gene region Kb
10
io
40
10
gO
60
1°0
g0
0
FIG. 2. (A) Strategy for placing a CAT selectable marker beside the subtilin precursor gene in B. subtilis ATCC 6633. The subtilin gene is the subtilin precursor peptide gene located upstream from an XbaI site. A promoter-containing CAT gene obtained from pC194 (4, 9) by using the polymerase chain reaction (kindly provided by Mark T. Steen) was ligated into the XbaI site of a 3.3-kb EcoRI-XbaI restriction fragment, in which the terminator of the subtilin precursor peptide gene lies about 300 bp upstream from the XbaI site (1). This construct was cloned into plasmid pTZ19U (which contains an origin of replication that functions in Escherichia coli but not in Bacillus spp.), creating pRXCM. After amplification in E. coli, the plasmid was introduced into B. subtilis ATCC 6633 by transformation by use of the polyethylene glycol method, and chloramphenicolresistant colonies were selected. Chloramphenicol resistance is conferred by a Campbell-type insertion of the plasmid into the chromosome. (B) Strategy for transformation of the subtilin operon into BR151. Total chromosomal DNA from the chloramphenicolresistant transformants of B. subtilis ATCC 6633 (panel A) was used to transform competent cells of BR151 to chloramphenicol resistance. Integration of the CAT gene into the linear chromosomal DNA is presumed to occur by a double recombination involving homologies in the flanking regions around the subtilin operon and corresponding homologies in the BR151 chromosome. (C) Restriction map of subtilin transforming region around the subtilin pre-
lishes that all the gene products required for the posttranslational modifications are present in the transformant. Although the halo assay indicates that the LH45 transformant is a subtilin producer, the complexity of precursor maturation requires verification that strain LH45 performs all of the posttranslational processing steps correctly. We employed the following tests of correct processing: biological activity, chromatographic behavior, amino acid composition, and N-terminal sequence analysis. Subtilin was isolated from strains ATCC 6633 and LH45 in the following way. Culture supernatants were heated for 2 to 3 min at 121°C to inactivate proteases and then applied to a carboxymethyl cellulose column in 50 mM NaPi (pH 5.8). The column was eluted with an NaCl gradient buffered with 50 mM NaPi. The major A254 peak was lyophilized, dissolved in 0.1% trifluoroacetic acid, and fractionated on a C-18 high-performance liquid chromatography column as described for nisin (13). The peak fractions were lyophilized immediately after separation, and the dried powder was stored at -80°C. These subtilin preparations were tested for their ability to inhibit spore outgrowth in a culture assay (13). Their specific activities were the same. They were subjected to amino acid composition and N-terminal sequence analyses. Because maturation of the subtilin precursor peptide requires dehydrations, cross-linking, and removal of the leader region, the amino acid composition would reflect these posttranslational modifications. If both subtilins were the same, their elution profiles from the amino acid analyzer would be the same, even though the presence of the unusual amino acids would give unidentifiable peaks. The elution profiles obtained were indistinguishable (data not shown). Moreover, the calculated amino acid compositions of the identifiable amino acids were those expected from the known structure of subtilin. The unusual amino acids also give characteristic patterns during N-terminal sequence analysis. The dehydro residues (dehydroalanine or dehydrobutyrine) irreversibly block subsequent rounds of cleavage (11, 12). Also, when the Ala of Ala-S-R is cleaved, it leaves a blank. Subtilin isolated from strains ATCC 6633 and LH45 performed identically during sequence analysis. Namely, there was a blank at residue 3, as expected if residue 3 is the first Ala-S-R of lanthionine and as expected for subtilin. Also, both showed Glu at residue 4, with a molar yield of 70% compared to the initial Trp. If there had been a dehydro residue at position 3, it would have been blocked, with Glu not appearing at all. Both preparations of subtilin, therefore, have a lanthionine at position 3. Sequencing further, both forms were blank at residue 5 and blank during subsequent cycles, showing that both forms cursor peptide gene. DNA from LH45 was digested with rarely cutting restriction enzymnes individually and in combination, and the restricted DNA was used to transform strain BR151. MluI fragments gave transformants, indicating that homologies on both sides were present. The size of the transforming MluI fragment is 60 kb, as shown in Fig. 3C. Treatment with SmaI, which removes 2 to 5 kb from the left side of the MluI fragment, or with Kpnl, which removes 2 to 5 kb from the right side of the MluI fragment, destroyed the ability of the MluI fragment to transform.
*5>.:t~68Kb
VOL. 173, 1991
A
~~BR151
,
~~~LH45
_
B
_
21.2 Kb
5.0 Kb
4.3 Kb
1 2 3 200 Kb
Kb_ 15i1 00 Kb_
_
_
4_60 Kb 'li1l
50Kb
FIG. 3. (A) Halo assay showing subtilin production in wild-type B. subtilis ATCC 6633 and several subtilin-producing transformants of BR151 (LH1, LH2, LH3, and LH45). Nontransformed BR151 is included to show the absence of subtilin production in the original recipient. LH45 was used for further studies. (B) Autoradiogram of a Southern blot of an agarose gel of EcoRI-digested total chromosomal DNA that had been hybridized with a 21-mer oligonucleotide probe that is complementary to the subtilin precursor peptide gene. The 6633 lane was DNA obtained from the CAT gene derivative of B. subtilis ATCC 6633 shown in Fig. 2A. LH45 and LH33 are transformants of BR151 that had been transformed with that 6633 CAT DNA and selected on chloramphenicol as described in Fig. 2B. The BR151 lane shows that the subtilin gene is absent from the original recipient strain. (C) Lanes: 1, ethidium-bromide-stained orthogonal-field-alternation (3) gel of LH45 DNA that had been digested with MluI; 2, concatomeric lambda DNA size standard with the sizes shown on the left; 3, autoradiogram of a Southern blot of lane 1 that had been hybridized with the nick-translated pRX fragment (cf. Fig. 2A) that contains the subtilin precursor peptide gene. The 60-kb size was estimated from the lambda standard.
NOTES
7389
had a dehydro residue at position 5. These conclusions were further supported by treating the subtilin preparations with N'-(methyl)mercaptoacetamide before sequencing. This mercaptan undergoes a Michael addition across the double bond of the dehydro residues, which stabilizes them against deamination during sequence analysis. When the mercaptantreated forms were sequenced, both gave Leu at residue 6, as expected for subtilin. It was concluded that both forms of subtilin have the same N-terminal sequence and have undergone the same posttranslational modifications in which the dehydro residues and thioether cross-linkages were formed and the leader sequence was removed. These results show that the LH45 transformant produces authentic subtilin. The integration of approximately 40 kb of ATCC 6633 DNA suggests that genes for functions other than subtilin production may have been incorporated as well. For example, we observed that strain LH45 grows well on medium A, which contains 10% sucrose, as was observed for ATCC 6633. Strain 168 normally grows poorly in sucrose-containing medium A, suggesting that LH45 has acquired gene(s) that affect sucrose metabolism. This is reminiscent of the observation that genes for nisin production in Lactococcus lactis are closely linked to genes for sucrose metabolism (5). The conversion of strain 168 to a subtilin producer by integration of a fragment of ATCC 6633 DNA does not conclusively prove that all of the genes required for subtilin production reside on the integrated DNA. The integrated DNA may merely complement endogenous functions that are already present in strain 168, or it may provide regulatory factors that switch on otherwise cryptic genes. The only thing that can be stated with certainty is that the DNA that has been integrated into strain 168 contributes all additional genes over and above those for endogenous host functions that are required for subtilin biosynthesis and that all of these genes are clustered together in a reasonably small region of the chromosome (
