Vol. 131, No. 2 Printed in U.S.A.
JOURNAL OF BACTERIOLOGY, Aug. 1977, p. 699-701 Copyright C 1977 American Society for Microbiology
Isolation and Characterization of Four Types of Plasmids from Bacillus subtilis (natto) TERUO TANAKA* AND TAKAKO KOSHIKAWA Mitsubishi-Kasei Institute of Life Sciences, 11 Minamiooya, Machida-shi, Tokyo 194, Japan Received for publication 25 March 1977
Covalently closed circular deoxyribonucleic acids were found in 10 strains of Bacillus natto. The plasmids could be classified into four types on the basis ofthe molecular weights as well as the patterns in agarose gel electrophoresis after digestion with restriction endonucleases: (i) plasmids (seven were detected) with a molecular weight of 3.6 x 106; (ii) plasmids (two were detected) with a molecular weight of 4.0 x 106; (iii) plasmids (eight were detected) with a molecular weight of about 34 x 106; and (iv), a plasmid with an approximate molecular weight of 46 x 106. Out ofthe 10 plasmid-carrying strains, 6 (IF03009, IF03013, IF03335, IF013169, IAM1143, and IAM1207) harbored both type 1 and 3 plasmids; 2 (IAM1114 and IAM1168) harbored both type 2 and 3 plasmids, and IF03936 and IAM1163 carried type 1 and 4 plasmids, respectively. In attempts to ascertain the distribution of extrachromosomal deoxyribonucleic acid (DNA) elements in strains of Bacillus subtilis, two cryptic plasmids were found by Lovett and Bramucci (4). A similar screen for plasmids in our laboratory turned up at least three types of cryptic plasmids in four B. subtilis strains (8), and we have extended this effort to include an additional 15 strains of B. natto (now classified as B. subtilis [1]). In the present work we describe plasmids found in 10 of these strains. [3H]DNA from log-phase cells was banded by CsCl-ethidium bromide density gradient centrifugation to determine the percentage of total cellular DNA present in covalently closed circular (CCC) form. Based on radioactivity, the fraction of total banded DNA present in CCC form was 1.7, 2.5, 1.5, 0.65, 1.8, 1.8, 1.6, 1.7, 1.7, and 2. 0% for IF03009, IF03013, IF03335, IF03936, IF013169, IAM1143, IAM1207, IAM1114, IAM1168, and IAM1163, respectively (data not shown). The 3H-labeled CCC DNA was subjected to neutral sucrose gradient centrifugation to determine molecular weights as well as examine the heterogeneity of plasmid preparations (Table 1). Eight strains (IF03009, IF03013, IF03335, IF013169, IAM1143, IAM1207, IAM1114, and IAM1168) harbored two kinds of plasmids; whereas IF03936 and IAM1163 carried a single molecular species in each. A minimal number of copies of each plasmid per chromosome was calculated from the fraction of CCC counts banding in CsCl-ethidium bromide centrifugation and the ratio of each plasmid count to all the counts present in
a sucrose density gradient, assuming that a B. subtilis chromosome is 2 x 109 daltons (2). Results and plasmid designations are summarized in Table 1. The number of copies per chromosomal DNA of the large plasmids was below one, perhaps due to random nicking of supercoiled DNA during preparation. Thus, the number of copies should be taken as a lower limit of plasmid contents. To prepare plasmid DNA for digestion with restriction endonucleases, we used two methods; in method 1, cells were treated with lysozyme and Pronase followed by lysis with sodium dodecyl sulfate (8), and in method 2, cells were treated with lysozyme and ribonuclease A (RNase A) followed by lysis with Sarkosyl and were further incubated in the presence of Pronase. Chromosomal DNA precipitated by adding NaCl (method 1 [8]) or sodium dodecyl sulfate and NaCl (method 2, unpublished data) was removed by centrifugation. By method 1, small plasmids such as pLS15, pLS17, pLS19, pLS21, pLS22, pLS24, pLS26, pLS28, and pLS30 were preferentially obtained; whereas by method 2, both small and large plasmids were obtained. In an experiment with [3H]DNA, more than 90% of the large plasmids sedimented during centrifugation for removing chromosomal DNA in method 1, whereas approximately 60% of the small plasmids remained in the supernatant (data not shown). Plasmid preparations thus obtained were digested with three restriction endonucleases EcoRI (9), BamNI (6), and HindIII (7) and subjected to electrophoresis in agarose gels. The number of cleavage sites is
699
700
J. BACTERIOL.
NOTES
TABLE 1. Physical properties of plasmids from B. natto Mol wt (x 106)a
Strain
Plasmid
Plasmid type
gel Centrifuga_ Agarose electrophotion
No. of cop-
ies/chromosome
resis
IF03009 IFO3013
pLS15 pLS16
1 3 1 3 1 3
3.6
33
3.6
5
_C
0.5
No. of cleavage sitesb
EcoRI
BamNI
HindIII
1
2
4
16
5 2 3
-
3.9 pLS17 3.6 6 1 4 pLS18 33 0.9 15 IFO3335 pLS19 3.6 3.6 4 1 2 4 pLS20 34 0.4 17 4 1 3.6 IF03936 3.6 4 pLS21 1 2 4 1 IF013169 pLS22 3.4 3.6 4 1 2 4 pLS23 3 34 0.7 16 4 1 3.3 3.6 IAM1143 pLS24 3 1 2 4 pLS25 3 31 0.7 16 4 1 3.7 IAM1207 3.6 pLS26 4 1 2 4 pLS27 3 33 0.6 16 4 IAM1114 2 3.6 4.0 pLS28 5 2 1 5 pLS29 3 35 0.6 17 5 2 4.1 4.0 IAM1168 3 pLS30 2 1 5 pLS31 3 32 0.7 16 4 IAM1163 pLS32 4 46 0.9 20 9 a Molecular weights were determined by neutral sucrose gradient centrifugation and agarose gel electrophoresis. b Number of cleavage sites represents the number of DNA bands observable in agarose gels. DNA fragments smaller than 0.4 x 106 daltons could not be detected. c Not determined.
summarized in Table 1. Based on these results together with the digestion patterns in agarose gels and the molecular weights of the plasmids, we classified these plasmids into four types (Table 1). Although the plasmids belonging to type 3 displayed minor differences in the banding pattern and the number of cleavage sites, most of the DNA fragments were indistinguishable in electrophoretic mobilities (data not shown). These differences might be caused by insertion or deletion of DNA fragments into or from type 3 plasmid DNA. It is of interest to note that type 3 plasmid is present together with type 1 or 2 plasmid (Table 1). One possibility is that type 3 plasmid was transferred by some unknown mechanism into a cell where either type 1 or 2 plasmid had been present. It should be pointed out that B. natto is a bacillus isolated from "natto," a vegetable cheese prepared in Japan by fermentation of boiled soybeans. Thus, cells have the chance to contact each other during fermentation. Another possibility is that type 1 and 2 plasmids have a common ancestor and a cell carrying both the ancestor and type 3 plasmid has evolved into the present two kinds of cell; i.e., one containing type 1 and 3 plasmids and the other containing type 2 and 3 plasmids. Use of hybridization or restriction endonucleases, which give rise to shorter DNA fragments,
would provide means for comparing type 1 and 2 plasmids, which then would reveal whether they derived from a common ancestor or not. Type 3 and 4 plasmids were obtained successfully when lysates were prepared in the presence of RNase A (method 2). It has been shown that F plasmids are released from the folded chromosome when it is degraded by RNase treatment (3). Thus, type 3 or 4 plasmids seem to be embedded in the folded chromosome and were released when lysates were treated with RNase. Total cellular DNA from B. natto strains used in this study transformed arginine and leucine markers of B. subtilis 168 to the same extent as did homologous DNA (data not shown), indicating that the two organisms are closely related at least in terms of DNA sequences. This suggests the possibility that B. natto plasmids can reside in B. subtilis 168 too. In fact we constructed a recombinant DNA molecule in B. subtilis 168 consisting of one of the B. natto plasmids (pLS28) and a B.. subtilis leucine gene (manuscript in preparation). It should also be noted that a plasmid from B. pumilus has been inserted into and maintained by B. subtilis 168 (5). Neither bacteriocin production with B. subtilis 168 as a test organism nor resistance to common antibiotics, such as chloramphenicol,
NOTES
VOL. 131, 1977
tetracycline, streptomycin, and ampicillin, and heavy metal, HgCl2, was demonstrated in the plasmid-carrying strains (data not shown). So far, functions specified by the plasmids remain unknown. The plasmids reported in this communication are different from two plasmids reported by Lovett et al. (4) and four plasmids reported earlier by us (8), as judged by direct comparison of the restriction patterns in agarose gel electrophoresis. We are grateful to B. Weisblum for reading the manuscript. We are also grateful to K. Sakaguchi for interest and discussions and M. Kuroda-Tomizawa for expert technical assistance.
LITERATURE CITED 1. Gibson, T., and R. Gordon. 1974. Endospore-forming rods and cocci. Family I. Bacillacea, genus I. Bacillus Cohn, p. 529-550. In R. E. Buchanan and N. E. Gibsons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 2. Kavenoff, R. 1972. Characterization ofthe Bacillus sub-
3.
4. 5.
6.
7.
8. 9.
701
tilis w23 genome by sedimentation. J. Mol. Biol. 72:801-806. Kline, B. C., and J. R. Miller. 1975. Detection of nonintegrated plasmid deoxyribonucleic acid in the folded chromosome of Escherichia coli: physicochemical approach to studying the unit of segregation. J. Bacteriol. 121:165-172. Lovett, P. S., and M. G. Bramucci. 1975. Plasmid deoxyribonucleic acid in Bacillus subtilis and Bacillus pumilus. J. Bacteriol. 124:484-490. Lovett, P. S., E. J. Duvall, and K. M. Keggins. 1976. Bacillus pumilus plasmid pPL10: properties and insertion into Bacillus subtilis 168 by transformation. J. Bacteriol. 127:817-828. Shibata, T., and T. Ando. 1976. The restriction endonuclease in Bacillus amyloliquefaciens N strain. Substrate specificities. Biochim. Biophys. Acta 442:184196. Smith, H. O., and K. W. Wilcox. 1970. A restriction enzyme from Hemophilus influenzae. I. Purification and general properties. J. Mol. Biol. 51:379-391. Tanaka, T., M. Kuroda, and K. Sakaguchi. 1977. Isolation and characterization offour plasmids from Bacillus subtilis. J. Bacteriol. 129:1487-1494. Tanaka, T., and B. Weisblum. 1975. Construction of a colicin E1-R factor composite plasmid in vitro: means for amplification of deoxyribonucleic acid. J. Bacteriol. 121:354-362.
JOURNAL OF BACTERIOLOGY, Aug. 1977, p. 699-701 Copyright C 1977 American Society for Microbiology
Isolation and Characterization of Four Types of Plasmids from Bacillus subtilis (natto) TERUO TANAKA* AND TAKAKO KOSHIKAWA Mitsubishi-Kasei Institute of Life Sciences, 11 Minamiooya, Machida-shi, Tokyo 194, Japan Received for publication 25 March 1977
Covalently closed circular deoxyribonucleic acids were found in 10 strains of Bacillus natto. The plasmids could be classified into four types on the basis ofthe molecular weights as well as the patterns in agarose gel electrophoresis after digestion with restriction endonucleases: (i) plasmids (seven were detected) with a molecular weight of 3.6 x 106; (ii) plasmids (two were detected) with a molecular weight of 4.0 x 106; (iii) plasmids (eight were detected) with a molecular weight of about 34 x 106; and (iv), a plasmid with an approximate molecular weight of 46 x 106. Out ofthe 10 plasmid-carrying strains, 6 (IF03009, IF03013, IF03335, IF013169, IAM1143, and IAM1207) harbored both type 1 and 3 plasmids; 2 (IAM1114 and IAM1168) harbored both type 2 and 3 plasmids, and IF03936 and IAM1163 carried type 1 and 4 plasmids, respectively. In attempts to ascertain the distribution of extrachromosomal deoxyribonucleic acid (DNA) elements in strains of Bacillus subtilis, two cryptic plasmids were found by Lovett and Bramucci (4). A similar screen for plasmids in our laboratory turned up at least three types of cryptic plasmids in four B. subtilis strains (8), and we have extended this effort to include an additional 15 strains of B. natto (now classified as B. subtilis [1]). In the present work we describe plasmids found in 10 of these strains. [3H]DNA from log-phase cells was banded by CsCl-ethidium bromide density gradient centrifugation to determine the percentage of total cellular DNA present in covalently closed circular (CCC) form. Based on radioactivity, the fraction of total banded DNA present in CCC form was 1.7, 2.5, 1.5, 0.65, 1.8, 1.8, 1.6, 1.7, 1.7, and 2. 0% for IF03009, IF03013, IF03335, IF03936, IF013169, IAM1143, IAM1207, IAM1114, IAM1168, and IAM1163, respectively (data not shown). The 3H-labeled CCC DNA was subjected to neutral sucrose gradient centrifugation to determine molecular weights as well as examine the heterogeneity of plasmid preparations (Table 1). Eight strains (IF03009, IF03013, IF03335, IF013169, IAM1143, IAM1207, IAM1114, and IAM1168) harbored two kinds of plasmids; whereas IF03936 and IAM1163 carried a single molecular species in each. A minimal number of copies of each plasmid per chromosome was calculated from the fraction of CCC counts banding in CsCl-ethidium bromide centrifugation and the ratio of each plasmid count to all the counts present in
a sucrose density gradient, assuming that a B. subtilis chromosome is 2 x 109 daltons (2). Results and plasmid designations are summarized in Table 1. The number of copies per chromosomal DNA of the large plasmids was below one, perhaps due to random nicking of supercoiled DNA during preparation. Thus, the number of copies should be taken as a lower limit of plasmid contents. To prepare plasmid DNA for digestion with restriction endonucleases, we used two methods; in method 1, cells were treated with lysozyme and Pronase followed by lysis with sodium dodecyl sulfate (8), and in method 2, cells were treated with lysozyme and ribonuclease A (RNase A) followed by lysis with Sarkosyl and were further incubated in the presence of Pronase. Chromosomal DNA precipitated by adding NaCl (method 1 [8]) or sodium dodecyl sulfate and NaCl (method 2, unpublished data) was removed by centrifugation. By method 1, small plasmids such as pLS15, pLS17, pLS19, pLS21, pLS22, pLS24, pLS26, pLS28, and pLS30 were preferentially obtained; whereas by method 2, both small and large plasmids were obtained. In an experiment with [3H]DNA, more than 90% of the large plasmids sedimented during centrifugation for removing chromosomal DNA in method 1, whereas approximately 60% of the small plasmids remained in the supernatant (data not shown). Plasmid preparations thus obtained were digested with three restriction endonucleases EcoRI (9), BamNI (6), and HindIII (7) and subjected to electrophoresis in agarose gels. The number of cleavage sites is
699
700
J. BACTERIOL.
NOTES
TABLE 1. Physical properties of plasmids from B. natto Mol wt (x 106)a
Strain
Plasmid
Plasmid type
gel Centrifuga_ Agarose electrophotion
No. of cop-
ies/chromosome
resis
IF03009 IFO3013
pLS15 pLS16
1 3 1 3 1 3
3.6
33
3.6
5
_C
0.5
No. of cleavage sitesb
EcoRI
BamNI
HindIII
1
2
4
16
5 2 3
-
3.9 pLS17 3.6 6 1 4 pLS18 33 0.9 15 IFO3335 pLS19 3.6 3.6 4 1 2 4 pLS20 34 0.4 17 4 1 3.6 IF03936 3.6 4 pLS21 1 2 4 1 IF013169 pLS22 3.4 3.6 4 1 2 4 pLS23 3 34 0.7 16 4 1 3.3 3.6 IAM1143 pLS24 3 1 2 4 pLS25 3 31 0.7 16 4 1 3.7 IAM1207 3.6 pLS26 4 1 2 4 pLS27 3 33 0.6 16 4 IAM1114 2 3.6 4.0 pLS28 5 2 1 5 pLS29 3 35 0.6 17 5 2 4.1 4.0 IAM1168 3 pLS30 2 1 5 pLS31 3 32 0.7 16 4 IAM1163 pLS32 4 46 0.9 20 9 a Molecular weights were determined by neutral sucrose gradient centrifugation and agarose gel electrophoresis. b Number of cleavage sites represents the number of DNA bands observable in agarose gels. DNA fragments smaller than 0.4 x 106 daltons could not be detected. c Not determined.
summarized in Table 1. Based on these results together with the digestion patterns in agarose gels and the molecular weights of the plasmids, we classified these plasmids into four types (Table 1). Although the plasmids belonging to type 3 displayed minor differences in the banding pattern and the number of cleavage sites, most of the DNA fragments were indistinguishable in electrophoretic mobilities (data not shown). These differences might be caused by insertion or deletion of DNA fragments into or from type 3 plasmid DNA. It is of interest to note that type 3 plasmid is present together with type 1 or 2 plasmid (Table 1). One possibility is that type 3 plasmid was transferred by some unknown mechanism into a cell where either type 1 or 2 plasmid had been present. It should be pointed out that B. natto is a bacillus isolated from "natto," a vegetable cheese prepared in Japan by fermentation of boiled soybeans. Thus, cells have the chance to contact each other during fermentation. Another possibility is that type 1 and 2 plasmids have a common ancestor and a cell carrying both the ancestor and type 3 plasmid has evolved into the present two kinds of cell; i.e., one containing type 1 and 3 plasmids and the other containing type 2 and 3 plasmids. Use of hybridization or restriction endonucleases, which give rise to shorter DNA fragments,
would provide means for comparing type 1 and 2 plasmids, which then would reveal whether they derived from a common ancestor or not. Type 3 and 4 plasmids were obtained successfully when lysates were prepared in the presence of RNase A (method 2). It has been shown that F plasmids are released from the folded chromosome when it is degraded by RNase treatment (3). Thus, type 3 or 4 plasmids seem to be embedded in the folded chromosome and were released when lysates were treated with RNase. Total cellular DNA from B. natto strains used in this study transformed arginine and leucine markers of B. subtilis 168 to the same extent as did homologous DNA (data not shown), indicating that the two organisms are closely related at least in terms of DNA sequences. This suggests the possibility that B. natto plasmids can reside in B. subtilis 168 too. In fact we constructed a recombinant DNA molecule in B. subtilis 168 consisting of one of the B. natto plasmids (pLS28) and a B.. subtilis leucine gene (manuscript in preparation). It should also be noted that a plasmid from B. pumilus has been inserted into and maintained by B. subtilis 168 (5). Neither bacteriocin production with B. subtilis 168 as a test organism nor resistance to common antibiotics, such as chloramphenicol,
NOTES
VOL. 131, 1977
tetracycline, streptomycin, and ampicillin, and heavy metal, HgCl2, was demonstrated in the plasmid-carrying strains (data not shown). So far, functions specified by the plasmids remain unknown. The plasmids reported in this communication are different from two plasmids reported by Lovett et al. (4) and four plasmids reported earlier by us (8), as judged by direct comparison of the restriction patterns in agarose gel electrophoresis. We are grateful to B. Weisblum for reading the manuscript. We are also grateful to K. Sakaguchi for interest and discussions and M. Kuroda-Tomizawa for expert technical assistance.
LITERATURE CITED 1. Gibson, T., and R. Gordon. 1974. Endospore-forming rods and cocci. Family I. Bacillacea, genus I. Bacillus Cohn, p. 529-550. In R. E. Buchanan and N. E. Gibsons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 2. Kavenoff, R. 1972. Characterization ofthe Bacillus sub-
3.
4. 5.
6.
7.
8. 9.
701
tilis w23 genome by sedimentation. J. Mol. Biol. 72:801-806. Kline, B. C., and J. R. Miller. 1975. Detection of nonintegrated plasmid deoxyribonucleic acid in the folded chromosome of Escherichia coli: physicochemical approach to studying the unit of segregation. J. Bacteriol. 121:165-172. Lovett, P. S., and M. G. Bramucci. 1975. Plasmid deoxyribonucleic acid in Bacillus subtilis and Bacillus pumilus. J. Bacteriol. 124:484-490. Lovett, P. S., E. J. Duvall, and K. M. Keggins. 1976. Bacillus pumilus plasmid pPL10: properties and insertion into Bacillus subtilis 168 by transformation. J. Bacteriol. 127:817-828. Shibata, T., and T. Ando. 1976. The restriction endonuclease in Bacillus amyloliquefaciens N strain. Substrate specificities. Biochim. Biophys. Acta 442:184196. Smith, H. O., and K. W. Wilcox. 1970. A restriction enzyme from Hemophilus influenzae. I. Purification and general properties. J. Mol. Biol. 51:379-391. Tanaka, T., M. Kuroda, and K. Sakaguchi. 1977. Isolation and characterization offour plasmids from Bacillus subtilis. J. Bacteriol. 129:1487-1494. Tanaka, T., and B. Weisblum. 1975. Construction of a colicin E1-R factor composite plasmid in vitro: means for amplification of deoxyribonucleic acid. J. Bacteriol. 121:354-362.
