FEMS MicrobiologyLetters 93 (10021115-1211 ~ 1992 Federation of European Microbiological~cieties 11378-11~07/~2/$05.11~) Published by Elsevier
115
FEMSLE 114885
A pSClOl-par sequence-mediated study on the intracellular state of supercoiling of the pBR322 genome in Escherichia colt D N A topoisomerase I deletion mutant S a t o s h i lshii ~, T e t s u y a M u r a k a m i a n d K a z u o S h i s h i d o Departnu'nt ~f L~li"Science. Tokyo htxtinae of Tcchmdogv. Yokoham,.. lalntn
Received I11 FebruaD" It~t~2 Accepted 25 February 10o2 Key words: Supercoiling; D N A gyrase; Topoisomerase !: p S C l O l - p a r Sequence: pBR322 Genome: Escherichia colt
1. S U M M A R Y In Escherichia colt D N A topoisomerase i deletion mutant DM800, transcription of the tetracycline-resistance gene ( t e t ) in the pBR322 genome is thought to create and maintain two domains of positive supercoils ahead, and negative supercoils behind, the transcription complex. To assess the actual intracellular state of twin-supercoiled domains, p a r sequence (365 bp) of plasmid pSCI01, which shows a high affinity for D N A gyrase, was inserted into the E c o R l site upstream, or the A c a l site downstream, of the tet gene on the pBR322 genome. Analysis of the oxolinic acid-induced sites of cleavage by gyrase in DM890 revealed that the pBR322 derivatives are highly preferentially cleaved at the p a r sequence of the
Corrc~mdence to: K. Shishido. Department of Life Science. Tokyo Institute of Technology. Nagatsuta. Midori-ku. Yokohama 227. Japan. I Presellt address: Japan Tobacco Inc.. Life Ecience Research Laboratory. 6-2 Umegaoka. Midori-ku. Vokohama "27. Japan.
EcoR! site as well as the A t ' a l site and efficiently linearized whep. compared with pBR322. Assessment of the state of negative supercoiling of the pBR322 derivatives isolated suggested that the D N A (containing the A c a l site) ahead of the tet transcripts, is not so positively supercoiled and preferential interaction of gyrase with the E c o R i - p a r sequence does not result in removing negative superhclical turns so effectively as D N A topoisomcrase I does on pBR322 D N A in the isogenic wild-type cells.
2. I N T R O D U C T I O N Transcription was first postulated by Liu and Wang [I] to generate twin-supercoiled domains on the D N A template. As it proceeds, D N A in front of the transcription ensemble becomes positively supercoiled, and D N A behind the ensemble becomes negatively super~oiled [1-3]. Evidence for these domains has come from findings with Escherichia colt strains that are deficient in D N A topoisomerase 1 (top mutants), or that have
116 been treated with DNA gyrase inhibitors, pBR322 DNA isolated from E. coli strain DM800 carrying a deletion of the topoisomerase 1 gene (AtopA) with a compensatory mutation of the gyrase B gene (gyrB225) [4,5] is extremely heteiogeneous in linking number and highly negatively supercoiled [6,7]. In the absence of topoisomerase 1, which does remove the negative superhelical turns [8], the positively supercoiled domain may be efficiently relaxed by gyrase [9], resulting in a net accumulation of negative supercoils [2,3]. Highly positively supercoiled pBR322 DNA is isolated from an E. coli topA ~ gyr ÷ strain treated with gyrase inhibitors [2,10]. This is attributed to an exclusive elimination of negative supercoils by topoisomerase I. In pBR322, diffusion of positively and negativel~ supercoiled regions is considered to be prevented by both anchoring of the amino-terminal region of tetracycline-resistance (Tet) protein in the inner membrane, coupled to the plasmid DNA by transcription and transition of tet, and forming a barrier, possibly at replicalion origin [11,7], and by the divergent transcription of the bla and tet genes on the plasmid genome [2]. However, an intracellular state of twin-supercoiled domains in pBR322 genome has not been well-assessed. In order to address this question we have inserted the par sequence, the partition locus of plasmid pSCI01 which exhibits a high affinity for gyrase [12] at either the 3' or 5' end of the pBR322-tet transcription unit, and analyzed a mode of interaction of gyrase with the inscr',cd par sequence in E. coli DM800 and the extent of negative supercoiling of the pBR322 derivatives isolated from DM800. The results, taken together, suggest that the DNA ahead of the tet transcripts is not so positively supercoiled during a steady-state growth of E. coli DM800 cells.
3. MATERIALS AND METHODS 3.1. Construction of pBR322 derit'atit'es carrying the par sequence of pSCI01 Plasmid pMIKS-P2 (approx. 6.6 kbp), a generous gift from Dr. K. Yamaguchi, is an R6K [13]
~ar
EcoA ~ i l l l l l U H /,~co~ n , l ~ l ~ , ^ '., is
¢ Fig. 1. Structures of pBR322 and its derivativescarrying the pSClOl-par sequence constructed in this study. Location and direction of transcription of the tetracycline-resistance(tet) and ampicillin-rcsistance(bla) genes, location of the origin (ori) of replication, and restriction sites of EcoRl(4359), BamHl(375). At'al(1425), Pt'ull(2064),and Pst i(3607)on the pBR322 genome [26,27] are indicated. For the par fragment [12,14.15],A(At'al). Ha(Haell) and H(Hincll) are shown.
derivative with the 365-bp Al'al-Hincll par fragment of pSC101 [14,15] at its single E c o R l site. The par fragment was excised by digestion with EcoRI and inserted into the E c o R l site of pBR322 (Fig. 1). To insert the par fragment into the At'al site of pBR322, the AL'al-linearized pBR322 and the par fragment were converted to blunt ends by repair with the Kienow fragment of E. coli DNA polymerase 1 and ligated (Fig. 1). Based on the orientation of insertion of the par fragment, four pBR322 derivatives with a single par sequence at the E c o R l or the ht'al site were designated pBR-parEcoA, pBR-parEcoB, pBRparAt'aA, and pBR-parAvaB (Fig. 1). Restriction endonucleases and DNA modifying enzymes were all purchased from Takara Shuzo (Kyoto, Japan) and used according to the supplier's instructions. DNA samples of the pBR322 derivatives were used to transform E. coli JA221 [F- /euB6 A
117
trpE5 r e c A - h s d R - h s d M ÷ ] . Plasmid D N A w a s p r e p a r e d from the t r a n s f o r m a n t s selected by resistance to ampicillin and su~.jected to restriction analysis. Piasmids with expected s t r u c t u r e were used to t r a n s f o r m E. coli strains DM800 and DM4100.
3.2. Other methods E. coli W3110 derivatives DM800 (A(topAcysB)204 gyrB225 acrAl3) and DM4100 (cysB242(Am)) [4,51, g e n e r o u s gifts f r o m Drs. R. S t e r n g l a n z and ) Wang, respectively, w e r e used for analysis of linearization of plasmid D N A by oxolinie acid. A 9-ml culture o f E. coli cells
h a r b o r i n g plasmid w a s g r o w n at 37°C with aeration to the late exponential p h a s e ( A ~ = 0.7) in LB m e d i u m containing L-Cysteine (60 t . t g / m l ) and ampicillin (50 t z g / m l ) , a n d oxolinic acid (Sigma) was added to give a final c o n c e n t r a t i o n o f 50 p.g p e r ml o f culture. I n c u b a t i o n c o n t i n u e d for 30 min, at which point the cells were centrifuged at 4°C and s u s p e n d e d in 0.1 ml o f 50 m M T r i s - HCI ( p H 8.0) a n d 15% ( w / v ) sucrose. Total cellular D N A was p r e p a r e d from the cell s u s p e n s i o n according to the p r o c e d u r e described by O ' C o n n o r and Malamy [16]. Plasmid D N A used for analysis of negative superhelicity w a s p r e p a r e d from the cells grown to late-exponential p h a s e in LB m e d i u m according to a previously published pro-
(A) 1
2
3
4
(B) 5
6
7
)412345678 kb 23.13 9.42 6.56 4.36 3.25 2.32 2.03
1.11
0.40 0.31
Fig. 2. Highly preferential interaction of DNA gyrase with the pSClOI-par sequence inserted in the pBR322 genome in E, coil DM800. A. Linearization of pBR322 and its derivatives carrying the par sequence by oxolinic acid. 9 ml each of culture of E. coil DM800 containing plasmid was untreated (lanes I and 3) or treated with oxolinic acid (lane 2 and 4-7) and lysed. The lysate was then extracted with phenol and precipitated with ethanol. The pellet (total cellular DNA) was resuspended in 0.15 ml of TE buffer (10 mM Tris.HCI (pH 7.5), I mM EDTA). Samples (0-0.15 ml each) prepared from DMS00 harboring pBR322 (lanes I and 2), pBR-parEcoA (lanes 3 and 4), pBR-parEcoB (lane 5), pBR-parAl'aA (lane 6), and pBR-parAcaB (lane 7) were electrophoresed through a 1% agarose slab gel in Tris-acetate/EDTA buffer (pH 82) (41) mM Tris base, 20 mM sodium acetate, I mM EDTA). The capital letters N, L. and S are nicked circular, full-length linear, and super:oiled plasmid DNA, respectively. B Mapping oxolinic acid-induced cleavage sites on the par sequence-carrying pBR322 derivatives, pBR-parEcoA (lanes I and 5), pBR-parEcoB (lanes 2 and 6), pBR-parAraA (lanes 3 and 7), and pBR-parAl'aB (lanes 4 and 8) isolated from DM800. The section containing the unit-length linear plasmid molecules (see panel A) cut out and DNA was extracted from it. The resulting DNA samples (2 /.tg each) were digested with Psti (lanes 1-4) and Pruli (lanes 5-8) and electrophoresed through a 1.2:% agarose slab gel. Lane M contains the Hindill-digest of phage A DNA and Hincll- and Hapll-digests of pBR322 DNA to serve as molecular size markers. Sizes are indicateo to the left.
118 cedure [17]. Agarose gel electrophoresis was carried out as described by Shishido et al. [17]. Chloroquine-agarose gel electrophoresis was carried out according to the method reported by Pruss [18]. DNA was recovered from gels by electroelution. Labeling of DNA by the random primer labeling method [19], DNA transfer to a nylon membrane Hybond-N (Amersham) [20] and DNA-DNA hybridization [21] were carried out according to the published procedure 122].
4. RESULTS AND DISCUSSION Exponentially growing E. coli DM800 cells harboring pBR322, or its derivatives with the pSC101-par sequence, were treated with 50 p.g/ml oxolinic acid for 30 min and then lysed by sequential treatment with lysozyme and sodium dodecyl sulfate (SDS). The total cellular DNAs were extracted from these cells (and from control cells not treated with oxolinic acid) and electrophoresed through a 1.0% (w/v) agarose slab gel (Fig. 2A). As reported by other investigators [16,23], around one-third of pBR322 molecules were linearized by treatment with oxolinic acid (laue 2). No linearization was apparent when oxolinic acid was omitted (lane 1). In the case of the par-carrying pBR322 derivatives, the majority of the DNAs were linearized by oxolinic acidtreatment (lanes 4-7). The separated DNA bands yielded Southern blot-hyoridization with a 32p. labeled pBR322 probe. The hybridization bands were cut out from the membranes and the radioactivity was measured. The degrees (approx. 80%) of plasmid linearizations of pBR-parAcaA (lane 6) and pBR-parAcaB) (lane 7) were only slightly higher than those (approx. 70%) of pBRparEcoA (lane 4) and pBR-parEcoB (lane 5). The proportion of linearized plasmid molecules was found to be invariant within the concentration range of l0 to 100 p.g/ml of oxolinic acid and it did not change over a time course of 10 to 60 rain of oxolinic acid treatment (data not shown). These results suggest that DNA gyrase preferentially interacts with the pSClOl-par sequence of the EcoR1 site (in the region that is expected to be
negatively supercoiled: negative domain) as well as the Acal site (positive domain) in the pBR322 genome. The following experiments confirm this. The linearized plasmid molecules were extracted from agarose gels and digested with restriction endonucleases Pstl and Pcull which have a single cleavage site on the plasmid DNAs (see Fig. 1). The resulting restriction fragments were analyzed by agarose gel electrophoresis (Fig. 2B). Both Pst- and Pcuil-digests of the linearized par-carrying pBR322 derivatives yielded two strong DNA bands. Analysis of the sizes of these DNA fragments revealed that gyrase-catalyzed double-strand cleavages occurred highly preferentially at the restricted site(s) within the pal" sequence. A number of faint DNA bands were also observed as if these were a background smear of DNA, because a large number of gyrase reaction sites are fairly evenly distributed on pBR322 DNA [16,23]. Southern blot-hybridization analysis revealed that around two-thirds of plasmid molecules of the par-carrying pBR322 derivatives were cleaved by gyrase at the par sequence and the rest of the molecules were cleaved at the many sites of pBR322 sequences. An orientation of insertion of the par sequence on pBR322 did net ~-;ffecteleavability of the DNA by gyrase. To study an intracellular state of the twin-supercoiled domains, DNAs of the par-carrying pBR322 derivatives prepared from DM800 were analyzed by a chloroquine-agarose gel electrophoresis, and compared with that of pBR322 from DMS00 (Fig. 3A). The derivatives pBRparAt'aA (lane 4) and pBR-parAcaB (lane 5) showed a supercoiling distribution containing a slightly larger amount of highly negatively supercoiled DNA species when compared with pBR322 (lane 1). These indicate that the DNA (containing the Acal site) ahead of the tet transcripts is positively supercoiled as expected, but its extent is probably low. If not so, one would expect gyrase to cause a significant increase of negative supercoiling. Oyrase is known to readily remove positive supercoils. In the case of the derivatives pBR-parEcoA (lane 2) and pBR-parEcoB (lane 3), on the other hand, they contained a slightly smaller amount of highly supercoiled DNA
119
(A)
12345
(S)
12
mid pMIK5-P2, E. cord strains DM800, DM4100, and JA221 respectively. We also grateful to Mr. H. A m a n o and Miss M. K o m a t s u for c o m p u t e r drawing of the figures and typing the m a n u s c r i p t , respectively. This work was s u p p o r t e d in part by a grant from the Ministry of E d u c a t i o n , Science and Culture of J a p a n .
REFERENCES
Fig. 3. Chloroquine-agarose gel electrophoretie analysis of the degree of plasmid DNA supercoiling found in pBR322 and its derivatives canting the single tmr sequence. (A) 2 p,g each of pBR322 (lane I), pBR-parEcoA (lane 2), pBR-parEcoB (lane 3), pBR-parAvaA (lane 4), and pBR-lntrAvaB (lane 5) isolated from DM800; and (B) 2 ~g each of pBR322 isolated from DM4100 (lane I) and DM800, (lane 2) were elcctrophoresed through a 1~ agarose gel containing chloroquine 112 .ug/ml) under the conditions described by Pruss [18]. The lowest pBR322 band (marked with the arrow) is the most supercoiled, the intermediate bands are less supercoiled, and the top band (marked with an arrow) is nicked circular or completely relaxed DNA. The distribution of topoisomers shifts upwards with less supercoiling. species w h e n c o m p a r e d with pBR322; the level of negative supercoiling o f these derivatives, how. ever, was clearly m u c h h i g h e r than that of pBR322 isolated from the isogenic wild-type strain DM4100 (Fig. 3B). T h e s e results s h o w that the D N A (containing the E c o R l site) b e h i n d of the tet t r a n s c r i p t s (and also of the bla (/3-1actamase (ampicillin-resistance)) transcripts) is negatively supercoiled. G y r a s e actually r e m o v e s intracellular negative turns. However, a highly preferential interaction of gyrase with the EcoRl-par seq u e n c e d o e s not result in removing negative t u r n s so effectively as t o p o i s o m e r a s e ! does on pBR322 D N A in DM4100.
ACKNOWLEDGEMENTS W e t h a n k Drs. K. Y a m a g u c h i , R. Sternglanz, J. W a n g and N. G u n g e for kindly providing plas-
[I] Liu, LF. and Wang. J.C. 119871 Prtm. Natl. Acad. ,~i. USA 84, 7024-7027. [2] Wu, H.-Y., Shyy. S.. Wang. J.C. and Liu. L.F. (19881 Cell. 53, 433-440. [3] Tsao, Y.-P., Wu, H.-Y. and Liu, L.F. (1989) Cell 56. II I-liB. [4] Sternglanz, R., DiNardo, S.. Vt~elkeL K.A.. Nisnimura. Y., Hirota. Y., Becherer, K., Zumstein, L. and Wang, J.C. (Iq81) Proc. Natl. Acad. Sci. USA 78, 2747-2751. [5] DiNardo, S.. Voelkel. K.A.. Sternglanl. R.. Reynolds, A.E. and Wright. A. 119821 Cell. 31, 43-51. [6] Pruss. G.J. and Drlica, K. (1986) Proc. Natl. Acad. Sci. USA 83. 8952-8956. [7] Shishido, K.. IshiL S and Komiyama. N. 119891 Nucleic Acids Reg. 17, 9749-9759. [8] Wang, J.C. (19711J. Mol. Biol. 55, 523-533. [9] Gellcrt, M.. Mizuuchi, K., O'Dea, M.H. and Nash, H.A. (1976) Proc. Natl. Acad. Sci. USA 73, 3872-3876. [l()] Lockshon, D. and Morris. D.R. (1`483) Nucleic Acids Res. I 1, 2990-31117. [I 1] Lodge, J.K.. Kazic, T. and Berg, D.E. (1989) J. Bacteriol. 171, 2181-2187. [12] Wahle. E. and Kornberg, A. (19881 EMBO J. 7. 188`41895. [13] Filutowicz, M.. McEachern. M., Greener. A., Mukhopadhyay. P.. Uhlenhopp. E., Dudand. R. and Helinski, D.R. (1985} In: Plasmids in bacteria, (Helinski. D.R, Cohen. S.N.. Clewell, D.B.. Jackson. D.A. and Hollaender, A., Eds.), pp. 125-2611. Plenum Press, New York. [14] Meacock, P.A. and Cohen, S.N. (1980) Cell 20. 529-542. [15] Armstrong, K.A., Acosta, R., Ledner, E.. Machida. Y., Pancotto, M., McCormick, M., Ohtsul~'~. H. and Ohtsubo, E. (1`4841J. Mol. Biol. 175, 331-347. [16] O'Connor. M.B. and Malamy, M.H. (1985)J. Mol. Biol. 18 I. 545 -5511. [17] Shishido. K., Komiyama. N. and Ikawa, S. (Iq87)J. Mol. Biol. 195, 215-218. [18] Pruss, G.J. 11985) J. Mol. Biol. 185, 51-63. [1'4] Feinberg. A.P. and Vogelstein. B.A. (1~'83) Anal. Bit*chera. 132. 6-13. [211] Reed, K.C. and Mann, D.A, (1985) Nucleic Acids Reg. 13, 72112-7221. [21] Southern. E. 11'4751J. Mol. Biol. 79, 23 -248. [22] Amano, H., lvcs, C.L., Bott. K.F. and SiAshido. K. 11991)
1211 Biochim. Biophys. Acta 11188,251-258. [23] la)ckshon. D, and Morris. D.R. (1985)J, Mol. Biol. 181, 63-74. [24] Higgins. N.P.. Peebles. C.I.. Sugino, A. and Cozzarelli, N.R. (1978) Proc. Natl. Acad. Sci. USA, 75. 1773-1777,
[25] Gellert, M., Fisher, L.M. and O'Dea, M.H. (1979) Proc. Natl. Acad. Sci. USA 83. 7152-7156. [26] Sutcliffe, J.G. (1979) Cold Spring Harbor Syrup. Quant. Biol. 43, 77-90. [27] Watson, N. (1988) Gene 7(I. 399-403.
115
FEMSLE 114885
A pSClOl-par sequence-mediated study on the intracellular state of supercoiling of the pBR322 genome in Escherichia colt D N A topoisomerase I deletion mutant S a t o s h i lshii ~, T e t s u y a M u r a k a m i a n d K a z u o S h i s h i d o Departnu'nt ~f L~li"Science. Tokyo htxtinae of Tcchmdogv. Yokoham,.. lalntn
Received I11 FebruaD" It~t~2 Accepted 25 February 10o2 Key words: Supercoiling; D N A gyrase; Topoisomerase !: p S C l O l - p a r Sequence: pBR322 Genome: Escherichia colt
1. S U M M A R Y In Escherichia colt D N A topoisomerase i deletion mutant DM800, transcription of the tetracycline-resistance gene ( t e t ) in the pBR322 genome is thought to create and maintain two domains of positive supercoils ahead, and negative supercoils behind, the transcription complex. To assess the actual intracellular state of twin-supercoiled domains, p a r sequence (365 bp) of plasmid pSCI01, which shows a high affinity for D N A gyrase, was inserted into the E c o R l site upstream, or the A c a l site downstream, of the tet gene on the pBR322 genome. Analysis of the oxolinic acid-induced sites of cleavage by gyrase in DM890 revealed that the pBR322 derivatives are highly preferentially cleaved at the p a r sequence of the
Corrc~mdence to: K. Shishido. Department of Life Science. Tokyo Institute of Technology. Nagatsuta. Midori-ku. Yokohama 227. Japan. I Presellt address: Japan Tobacco Inc.. Life Ecience Research Laboratory. 6-2 Umegaoka. Midori-ku. Vokohama "27. Japan.
EcoR! site as well as the A t ' a l site and efficiently linearized whep. compared with pBR322. Assessment of the state of negative supercoiling of the pBR322 derivatives isolated suggested that the D N A (containing the A c a l site) ahead of the tet transcripts, is not so positively supercoiled and preferential interaction of gyrase with the E c o R i - p a r sequence does not result in removing negative superhclical turns so effectively as D N A topoisomcrase I does on pBR322 D N A in the isogenic wild-type cells.
2. I N T R O D U C T I O N Transcription was first postulated by Liu and Wang [I] to generate twin-supercoiled domains on the D N A template. As it proceeds, D N A in front of the transcription ensemble becomes positively supercoiled, and D N A behind the ensemble becomes negatively super~oiled [1-3]. Evidence for these domains has come from findings with Escherichia colt strains that are deficient in D N A topoisomerase 1 (top mutants), or that have
116 been treated with DNA gyrase inhibitors, pBR322 DNA isolated from E. coli strain DM800 carrying a deletion of the topoisomerase 1 gene (AtopA) with a compensatory mutation of the gyrase B gene (gyrB225) [4,5] is extremely heteiogeneous in linking number and highly negatively supercoiled [6,7]. In the absence of topoisomerase 1, which does remove the negative superhelical turns [8], the positively supercoiled domain may be efficiently relaxed by gyrase [9], resulting in a net accumulation of negative supercoils [2,3]. Highly positively supercoiled pBR322 DNA is isolated from an E. coli topA ~ gyr ÷ strain treated with gyrase inhibitors [2,10]. This is attributed to an exclusive elimination of negative supercoils by topoisomerase I. In pBR322, diffusion of positively and negativel~ supercoiled regions is considered to be prevented by both anchoring of the amino-terminal region of tetracycline-resistance (Tet) protein in the inner membrane, coupled to the plasmid DNA by transcription and transition of tet, and forming a barrier, possibly at replicalion origin [11,7], and by the divergent transcription of the bla and tet genes on the plasmid genome [2]. However, an intracellular state of twin-supercoiled domains in pBR322 genome has not been well-assessed. In order to address this question we have inserted the par sequence, the partition locus of plasmid pSCI01 which exhibits a high affinity for gyrase [12] at either the 3' or 5' end of the pBR322-tet transcription unit, and analyzed a mode of interaction of gyrase with the inscr',cd par sequence in E. coli DM800 and the extent of negative supercoiling of the pBR322 derivatives isolated from DM800. The results, taken together, suggest that the DNA ahead of the tet transcripts is not so positively supercoiled during a steady-state growth of E. coli DM800 cells.
3. MATERIALS AND METHODS 3.1. Construction of pBR322 derit'atit'es carrying the par sequence of pSCI01 Plasmid pMIKS-P2 (approx. 6.6 kbp), a generous gift from Dr. K. Yamaguchi, is an R6K [13]
~ar
EcoA ~ i l l l l l U H /,~co~ n , l ~ l ~ , ^ '., is
¢ Fig. 1. Structures of pBR322 and its derivativescarrying the pSClOl-par sequence constructed in this study. Location and direction of transcription of the tetracycline-resistance(tet) and ampicillin-rcsistance(bla) genes, location of the origin (ori) of replication, and restriction sites of EcoRl(4359), BamHl(375). At'al(1425), Pt'ull(2064),and Pst i(3607)on the pBR322 genome [26,27] are indicated. For the par fragment [12,14.15],A(At'al). Ha(Haell) and H(Hincll) are shown.
derivative with the 365-bp Al'al-Hincll par fragment of pSC101 [14,15] at its single E c o R l site. The par fragment was excised by digestion with EcoRI and inserted into the E c o R l site of pBR322 (Fig. 1). To insert the par fragment into the At'al site of pBR322, the AL'al-linearized pBR322 and the par fragment were converted to blunt ends by repair with the Kienow fragment of E. coli DNA polymerase 1 and ligated (Fig. 1). Based on the orientation of insertion of the par fragment, four pBR322 derivatives with a single par sequence at the E c o R l or the ht'al site were designated pBR-parEcoA, pBR-parEcoB, pBRparAt'aA, and pBR-parAvaB (Fig. 1). Restriction endonucleases and DNA modifying enzymes were all purchased from Takara Shuzo (Kyoto, Japan) and used according to the supplier's instructions. DNA samples of the pBR322 derivatives were used to transform E. coli JA221 [F- /euB6 A
117
trpE5 r e c A - h s d R - h s d M ÷ ] . Plasmid D N A w a s p r e p a r e d from the t r a n s f o r m a n t s selected by resistance to ampicillin and su~.jected to restriction analysis. Piasmids with expected s t r u c t u r e were used to t r a n s f o r m E. coli strains DM800 and DM4100.
3.2. Other methods E. coli W3110 derivatives DM800 (A(topAcysB)204 gyrB225 acrAl3) and DM4100 (cysB242(Am)) [4,51, g e n e r o u s gifts f r o m Drs. R. S t e r n g l a n z and ) Wang, respectively, w e r e used for analysis of linearization of plasmid D N A by oxolinie acid. A 9-ml culture o f E. coli cells
h a r b o r i n g plasmid w a s g r o w n at 37°C with aeration to the late exponential p h a s e ( A ~ = 0.7) in LB m e d i u m containing L-Cysteine (60 t . t g / m l ) and ampicillin (50 t z g / m l ) , a n d oxolinic acid (Sigma) was added to give a final c o n c e n t r a t i o n o f 50 p.g p e r ml o f culture. I n c u b a t i o n c o n t i n u e d for 30 min, at which point the cells were centrifuged at 4°C and s u s p e n d e d in 0.1 ml o f 50 m M T r i s - HCI ( p H 8.0) a n d 15% ( w / v ) sucrose. Total cellular D N A was p r e p a r e d from the cell s u s p e n s i o n according to the p r o c e d u r e described by O ' C o n n o r and Malamy [16]. Plasmid D N A used for analysis of negative superhelicity w a s p r e p a r e d from the cells grown to late-exponential p h a s e in LB m e d i u m according to a previously published pro-
(A) 1
2
3
4
(B) 5
6
7
)412345678 kb 23.13 9.42 6.56 4.36 3.25 2.32 2.03
1.11
0.40 0.31
Fig. 2. Highly preferential interaction of DNA gyrase with the pSClOI-par sequence inserted in the pBR322 genome in E, coil DM800. A. Linearization of pBR322 and its derivatives carrying the par sequence by oxolinic acid. 9 ml each of culture of E. coil DM800 containing plasmid was untreated (lanes I and 3) or treated with oxolinic acid (lane 2 and 4-7) and lysed. The lysate was then extracted with phenol and precipitated with ethanol. The pellet (total cellular DNA) was resuspended in 0.15 ml of TE buffer (10 mM Tris.HCI (pH 7.5), I mM EDTA). Samples (0-0.15 ml each) prepared from DMS00 harboring pBR322 (lanes I and 2), pBR-parEcoA (lanes 3 and 4), pBR-parEcoB (lane 5), pBR-parAl'aA (lane 6), and pBR-parAcaB (lane 7) were electrophoresed through a 1% agarose slab gel in Tris-acetate/EDTA buffer (pH 82) (41) mM Tris base, 20 mM sodium acetate, I mM EDTA). The capital letters N, L. and S are nicked circular, full-length linear, and super:oiled plasmid DNA, respectively. B Mapping oxolinic acid-induced cleavage sites on the par sequence-carrying pBR322 derivatives, pBR-parEcoA (lanes I and 5), pBR-parEcoB (lanes 2 and 6), pBR-parAraA (lanes 3 and 7), and pBR-parAl'aB (lanes 4 and 8) isolated from DM800. The section containing the unit-length linear plasmid molecules (see panel A) cut out and DNA was extracted from it. The resulting DNA samples (2 /.tg each) were digested with Psti (lanes 1-4) and Pruli (lanes 5-8) and electrophoresed through a 1.2:% agarose slab gel. Lane M contains the Hindill-digest of phage A DNA and Hincll- and Hapll-digests of pBR322 DNA to serve as molecular size markers. Sizes are indicateo to the left.
118 cedure [17]. Agarose gel electrophoresis was carried out as described by Shishido et al. [17]. Chloroquine-agarose gel electrophoresis was carried out according to the method reported by Pruss [18]. DNA was recovered from gels by electroelution. Labeling of DNA by the random primer labeling method [19], DNA transfer to a nylon membrane Hybond-N (Amersham) [20] and DNA-DNA hybridization [21] were carried out according to the published procedure 122].
4. RESULTS AND DISCUSSION Exponentially growing E. coli DM800 cells harboring pBR322, or its derivatives with the pSC101-par sequence, were treated with 50 p.g/ml oxolinic acid for 30 min and then lysed by sequential treatment with lysozyme and sodium dodecyl sulfate (SDS). The total cellular DNAs were extracted from these cells (and from control cells not treated with oxolinic acid) and electrophoresed through a 1.0% (w/v) agarose slab gel (Fig. 2A). As reported by other investigators [16,23], around one-third of pBR322 molecules were linearized by treatment with oxolinic acid (laue 2). No linearization was apparent when oxolinic acid was omitted (lane 1). In the case of the par-carrying pBR322 derivatives, the majority of the DNAs were linearized by oxolinic acidtreatment (lanes 4-7). The separated DNA bands yielded Southern blot-hyoridization with a 32p. labeled pBR322 probe. The hybridization bands were cut out from the membranes and the radioactivity was measured. The degrees (approx. 80%) of plasmid linearizations of pBR-parAcaA (lane 6) and pBR-parAcaB) (lane 7) were only slightly higher than those (approx. 70%) of pBRparEcoA (lane 4) and pBR-parEcoB (lane 5). The proportion of linearized plasmid molecules was found to be invariant within the concentration range of l0 to 100 p.g/ml of oxolinic acid and it did not change over a time course of 10 to 60 rain of oxolinic acid treatment (data not shown). These results suggest that DNA gyrase preferentially interacts with the pSClOl-par sequence of the EcoR1 site (in the region that is expected to be
negatively supercoiled: negative domain) as well as the Acal site (positive domain) in the pBR322 genome. The following experiments confirm this. The linearized plasmid molecules were extracted from agarose gels and digested with restriction endonucleases Pstl and Pcull which have a single cleavage site on the plasmid DNAs (see Fig. 1). The resulting restriction fragments were analyzed by agarose gel electrophoresis (Fig. 2B). Both Pst- and Pcuil-digests of the linearized par-carrying pBR322 derivatives yielded two strong DNA bands. Analysis of the sizes of these DNA fragments revealed that gyrase-catalyzed double-strand cleavages occurred highly preferentially at the restricted site(s) within the pal" sequence. A number of faint DNA bands were also observed as if these were a background smear of DNA, because a large number of gyrase reaction sites are fairly evenly distributed on pBR322 DNA [16,23]. Southern blot-hybridization analysis revealed that around two-thirds of plasmid molecules of the par-carrying pBR322 derivatives were cleaved by gyrase at the par sequence and the rest of the molecules were cleaved at the many sites of pBR322 sequences. An orientation of insertion of the par sequence on pBR322 did net ~-;ffecteleavability of the DNA by gyrase. To study an intracellular state of the twin-supercoiled domains, DNAs of the par-carrying pBR322 derivatives prepared from DM800 were analyzed by a chloroquine-agarose gel electrophoresis, and compared with that of pBR322 from DMS00 (Fig. 3A). The derivatives pBRparAt'aA (lane 4) and pBR-parAcaB (lane 5) showed a supercoiling distribution containing a slightly larger amount of highly negatively supercoiled DNA species when compared with pBR322 (lane 1). These indicate that the DNA (containing the Acal site) ahead of the tet transcripts is positively supercoiled as expected, but its extent is probably low. If not so, one would expect gyrase to cause a significant increase of negative supercoiling. Oyrase is known to readily remove positive supercoils. In the case of the derivatives pBR-parEcoA (lane 2) and pBR-parEcoB (lane 3), on the other hand, they contained a slightly smaller amount of highly supercoiled DNA
119
(A)
12345
(S)
12
mid pMIK5-P2, E. cord strains DM800, DM4100, and JA221 respectively. We also grateful to Mr. H. A m a n o and Miss M. K o m a t s u for c o m p u t e r drawing of the figures and typing the m a n u s c r i p t , respectively. This work was s u p p o r t e d in part by a grant from the Ministry of E d u c a t i o n , Science and Culture of J a p a n .
REFERENCES
Fig. 3. Chloroquine-agarose gel electrophoretie analysis of the degree of plasmid DNA supercoiling found in pBR322 and its derivatives canting the single tmr sequence. (A) 2 p,g each of pBR322 (lane I), pBR-parEcoA (lane 2), pBR-parEcoB (lane 3), pBR-parAvaA (lane 4), and pBR-lntrAvaB (lane 5) isolated from DM800; and (B) 2 ~g each of pBR322 isolated from DM4100 (lane I) and DM800, (lane 2) were elcctrophoresed through a 1~ agarose gel containing chloroquine 112 .ug/ml) under the conditions described by Pruss [18]. The lowest pBR322 band (marked with the arrow) is the most supercoiled, the intermediate bands are less supercoiled, and the top band (marked with an arrow) is nicked circular or completely relaxed DNA. The distribution of topoisomers shifts upwards with less supercoiling. species w h e n c o m p a r e d with pBR322; the level of negative supercoiling o f these derivatives, how. ever, was clearly m u c h h i g h e r than that of pBR322 isolated from the isogenic wild-type strain DM4100 (Fig. 3B). T h e s e results s h o w that the D N A (containing the E c o R l site) b e h i n d of the tet t r a n s c r i p t s (and also of the bla (/3-1actamase (ampicillin-resistance)) transcripts) is negatively supercoiled. G y r a s e actually r e m o v e s intracellular negative turns. However, a highly preferential interaction of gyrase with the EcoRl-par seq u e n c e d o e s not result in removing negative t u r n s so effectively as t o p o i s o m e r a s e ! does on pBR322 D N A in DM4100.
ACKNOWLEDGEMENTS W e t h a n k Drs. K. Y a m a g u c h i , R. Sternglanz, J. W a n g and N. G u n g e for kindly providing plas-
[I] Liu, LF. and Wang. J.C. 119871 Prtm. Natl. Acad. ,~i. USA 84, 7024-7027. [2] Wu, H.-Y., Shyy. S.. Wang. J.C. and Liu. L.F. (19881 Cell. 53, 433-440. [3] Tsao, Y.-P., Wu, H.-Y. and Liu, L.F. (1989) Cell 56. II I-liB. [4] Sternglanz, R., DiNardo, S.. Vt~elkeL K.A.. Nisnimura. Y., Hirota. Y., Becherer, K., Zumstein, L. and Wang, J.C. (Iq81) Proc. Natl. Acad. Sci. USA 78, 2747-2751. [5] DiNardo, S.. Voelkel. K.A.. Sternglanl. R.. Reynolds, A.E. and Wright. A. 119821 Cell. 31, 43-51. [6] Pruss. G.J. and Drlica, K. (1986) Proc. Natl. Acad. Sci. USA 83. 8952-8956. [7] Shishido, K.. IshiL S and Komiyama. N. 119891 Nucleic Acids Reg. 17, 9749-9759. [8] Wang, J.C. (19711J. Mol. Biol. 55, 523-533. [9] Gellcrt, M.. Mizuuchi, K., O'Dea, M.H. and Nash, H.A. (1976) Proc. Natl. Acad. Sci. USA 73, 3872-3876. [l()] Lockshon, D. and Morris. D.R. (1`483) Nucleic Acids Res. I 1, 2990-31117. [I 1] Lodge, J.K.. Kazic, T. and Berg, D.E. (1989) J. Bacteriol. 171, 2181-2187. [12] Wahle. E. and Kornberg, A. (19881 EMBO J. 7. 188`41895. [13] Filutowicz, M.. McEachern. M., Greener. A., Mukhopadhyay. P.. Uhlenhopp. E., Dudand. R. and Helinski, D.R. (1985} In: Plasmids in bacteria, (Helinski. D.R, Cohen. S.N.. Clewell, D.B.. Jackson. D.A. and Hollaender, A., Eds.), pp. 125-2611. Plenum Press, New York. [14] Meacock, P.A. and Cohen, S.N. (1980) Cell 20. 529-542. [15] Armstrong, K.A., Acosta, R., Ledner, E.. Machida. Y., Pancotto, M., McCormick, M., Ohtsul~'~. H. and Ohtsubo, E. (1`4841J. Mol. Biol. 175, 331-347. [16] O'Connor. M.B. and Malamy, M.H. (1985)J. Mol. Biol. 18 I. 545 -5511. [17] Shishido. K., Komiyama. N. and Ikawa, S. (Iq87)J. Mol. Biol. 195, 215-218. [18] Pruss, G.J. 11985) J. Mol. Biol. 185, 51-63. [1'4] Feinberg. A.P. and Vogelstein. B.A. (1~'83) Anal. Bit*chera. 132. 6-13. [211] Reed, K.C. and Mann, D.A, (1985) Nucleic Acids Reg. 13, 72112-7221. [21] Southern. E. 11'4751J. Mol. Biol. 79, 23 -248. [22] Amano, H., lvcs, C.L., Bott. K.F. and SiAshido. K. 11991)
1211 Biochim. Biophys. Acta 11188,251-258. [23] la)ckshon. D, and Morris. D.R. (1985)J, Mol. Biol. 181, 63-74. [24] Higgins. N.P.. Peebles. C.I.. Sugino, A. and Cozzarelli, N.R. (1978) Proc. Natl. Acad. Sci. USA, 75. 1773-1777,
[25] Gellert, M., Fisher, L.M. and O'Dea, M.H. (1979) Proc. Natl. Acad. Sci. USA 83. 7152-7156. [26] Sutcliffe, J.G. (1979) Cold Spring Harbor Syrup. Quant. Biol. 43, 77-90. [27] Watson, N. (1988) Gene 7(I. 399-403.
