Portal de Periódicos da CAPES

Activation of the leu-500 Promoter by a Reversed Polarity tetA Gene

1998; Elsevier BV; Volume: 273; Issue: 1 Linguagem: Inglês

10.1074/jbc.273.1.653

1083-351X

Dongrong Chen, Sophie Bachellier‐Bassi, David M.J. Lilley,

Bacterial Genetics and Biotechnology

The leu-500 promoter is inactivated by a mutation in the −10 region but can be activated in topA Escherichia coli and Salmonella strains. We have found that the tetA gene plays a vital role in thetopA-dependent activation of a plasmid-borneleu-500 promoter. In previous studies, theleu-500 promoter and tetA gene have been arranged divergently. In this study we have reversed the polarity of the tetA gene, thus locating the leu-500promoter at the 3′ end of tetA. Despite being formally located in the downstream region of tetA, theleu-500 promoter is equally well activated in atopA strain in this environment, even though it is 1.6 kilobase pairs away from the promoter of the reversed tetAgene. Activation of the leu-500 promoter depends on transcription and translation of tetA but is largely insensitive to the function of other transcription units on the plasmid. These results require a change in viewpoint of the role oftetA, from local to global supercoiling. We conclude that transcription of the tetA gene is the main generator of transcription-induced supercoiling that activates theleu-500 promoter. Unbalanced relaxation of this supercoiling leads to a net increase in the negative linking difference of the plasmid globally, and there is a linear correlation between the change in global plasmid topology and the activation of theleu-500 promoter. Thus the leu-500 promoter appears to respond to the negative supercoiling of the plasmid overall. The leu-500 promoter is inactivated by a mutation in the −10 region but can be activated in topA Escherichia coli and Salmonella strains. We have found that the tetA gene plays a vital role in thetopA-dependent activation of a plasmid-borneleu-500 promoter. In previous studies, theleu-500 promoter and tetA gene have been arranged divergently. In this study we have reversed the polarity of the tetA gene, thus locating the leu-500promoter at the 3′ end of tetA. Despite being formally located in the downstream region of tetA, theleu-500 promoter is equally well activated in atopA strain in this environment, even though it is 1.6 kilobase pairs away from the promoter of the reversed tetAgene. Activation of the leu-500 promoter depends on transcription and translation of tetA but is largely insensitive to the function of other transcription units on the plasmid. These results require a change in viewpoint of the role oftetA, from local to global supercoiling. We conclude that transcription of the tetA gene is the main generator of transcription-induced supercoiling that activates theleu-500 promoter. Unbalanced relaxation of this supercoiling leads to a net increase in the negative linking difference of the plasmid globally, and there is a linear correlation between the change in global plasmid topology and the activation of theleu-500 promoter. Thus the leu-500 promoter appears to respond to the negative supercoiling of the plasmid overall. The activation of the leu-500 promoter provides a good illustration of the possible interrelationships between transcription and the topology of the DNA template in vivo.leu-500 is a leucine auxotroph of Salmonella typhimurium (1Mukai F.H. Margolin P. Proc. Natl. Acad. Sci. U. S. A. 1963; 50: 140-148Crossref PubMed Google Scholar) that results from an A to G transition in the −10 region of the promoter of the leu biosynthetic operon (2Gemmill R.M. Tripp M. Friedman S.B. Calvo J.M. J. Bacteriol. 1984; 158: 948-953Crossref PubMed Google Scholar). It was found that leucine prototrophy was restored inSalmonella bearing a supX mutation (3Dubnau E. Margolin P. Mol. Gen. Genet. 1972; 117: 91-112Crossref PubMed Scopus (39) Google Scholar). The later demonstration that supX was identical to topA(4Margolin P. Zumstein L. Sternglanz R. Wang J.C. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 5437-5441Crossref PubMed Scopus (40) Google Scholar), the gene encoding DNA topoisomerase I, provided a strong indication of a functional link between transcription and topology. Thus the increase in negative supercoiling that should arise from the loss of the supercoiling-relaxation activity from theSalmonella cell (5Richardson S.M.H. Higgins C.F. Lilley D.M.J. EMBO J. 1984; 3: 1745-1752Crossref PubMed Scopus (81) Google Scholar) might be expected to assist in the function of the leu-500 promoter, coupling the additional free energy of negative supercoiling to the opening of the more refractory −10 region of the mutant promoter (6Smith G.R. Cell. 1981; 24: 599-600Abstract Full Text PDF PubMed Scopus (110) Google Scholar, 7Pruss G. Drlica K. J. Bacteriol. 1985; 164: 947-949Crossref PubMed Google Scholar). More recent work in this laboratory has identified an additional level of complexity in this process. The demonstration of a direct requirement for a null topA background (8Richardson S.M.H. Higgins C.F. Lilley D.M.J. EMBO J. 1988; 7: 1863-1869Crossref PubMed Scopus (49) Google Scholar) led to the suggestion that the leu-500 promoter might be activated by variations in template supercoiling arising from transcriptional-induced supercoiling due to the transcription of a nearby gene (9Lilley D.M.J. Higgins C.F. Mol. Microbiol. 1991; 5: 779-783Crossref PubMed Scopus (53) Google Scholar, 10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar). According to the twin-supercoiled domain model of Liu and Wang (11Liu L.F. Wang J.C. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7024-7027Crossref PubMed Scopus (1502) Google Scholar), a rotationally hindered RNA polymerase in the elongation phase of transcription will tend to generate positive supercoiling ahead of its passage and negative supercoiling in its wake. These domains will be relaxed by DNA gyrase and topoisomerase I, respectively, in eubacteria, but unbalanced relaxation by topoisomerase activity due to either inhibition or mutation will lead to alteration in the local level of DNA supercoiling (12Lockshon D. Morris D.R. Nucleic Acids Res. 1983; 11: 2999-3017Crossref PubMed Scopus (113) Google Scholar, 13Pruss G.J. Drlica K. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 8952-8956Crossref PubMed Scopus (154) Google Scholar). Thus theleu-500 promoter might be activated by negative supercoiling arising from the transcription of the putative nearby gene, which would be less efficiently relaxed in topA cells. Although this model could explain the activation of the chromosomalleu-500 promoter in topA Salmonella, a further complication came to light when we sought to reproduce the effect on a plasmid. We found that we could only obtain significant activity of theleu-500 promoter when the plasmid also bore the gene encoding resistance to tetracycline, tetA. Using such plasmids we could achieve topA-dependent activation of the promoter in either Salmonella (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar) orEscherichia coli (14Chen D. Bowater R. Lilley D.M.J. J. Bacteriol. 1994; 176: 3757-3764Crossref PubMed Google Scholar). This implied a key role for thetetA gene, and a number of studies have indicated that the coupled transcription, translation, and membrane insertion of thetetA gene product are essential for efficient oversupercoiling of plasmids in topA eubacterial cells (13Pruss G.J. Drlica K. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 8952-8956Crossref PubMed Scopus (154) Google Scholar,15Lodge J.K. Kazik T. Berg D.E. J. Bacteriol. 1989; 171: 2181-2187Crossref PubMed Google Scholar, 16Lynch A.S. Wang J.C. J. Bacteriol. 1993; 175: 1645-1655Crossref PubMed Scopus (130) Google Scholar, 17Ma D. Cook D.N. Pon N.G. Hearst J.E. J. Biol. Chem. 1994; 269: 15362-15370Abstract Full Text PDF PubMed Google Scholar) due to the anchorage of the transcribing RNA polymerase to the membrane. We showed that activation of the leu-500 promoter on a plasmid did indeed require transcription and translation of thetetA gene and insertion of the TetA polypeptide into the membrane (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar, 18Chen D. Bowater R. Lilley D.M.J. Biochemistry. 1993; 32: 13162-13170Crossref PubMed Scopus (24) Google Scholar). We can conceive of two roles for the tetA gene in the activation of the leu-500 promoter on a plasmid. First, transcription of the tetA gene could be the primary generator of supercoiling; tethering RNA polymerase to the membrane would be a particularly effective way in which to hinder its rotation about the DNA template, and thus efficient induction of supercoiling might be expected. The second role could be more passive: to provide a barrier to the diffusion of supercoiling. If negative and positive domains of supercoiling were generated by transcription elsewhere on the plasmid, these could diffuse around the circle and cancel each other by rotation about the duplex axis, providing a highly efficient nonenzymatic relaxation mechanism. However a point of anchorage (such as the insertion of the nascent TetA polypeptide into the membrane) should provide a barrier to the diffusion of supercoiling around the plasmid and might thus increase local levels of DNA supercoiling. In the plasmid pLEU500Tc, with which we first achieved thetopA-dependent activation of theleu-500 promoter (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar), the tetA gene was oriented divergently to the leu-500 promoter, with a short distance between the promoters. This places the leu-500 promoter immediately upstream of the tetA promoter, which would be consistent with a very local effect whereby the leu-500promoter responds to a domain of negative supercoiling directly upstream of tetA. We therefore wondered if theleu-500 promoter would still be activated in topAcells if the orientation of the tetA gene were reversed. We find that the leu-500 promoter is activated to the same level under these circumstances and that the activity remains fully dependent on the function of the tetA gene. We conclude that transcription of the tetA gene is the major source of negative supercoiling that activates the leu-500 promoter, but that this is mediated through the global topology of the plasmid. E. coli strains HB101 (F−,hsdS20 (r−B, m−B),recA13, ara-14, proA2,lacYI, galK2, rpsL20(Smr), xyl-5, mtl-1,supE44, λ−), and DM800(Δ(topA-cysB)204 acrA13 gyrB225) (25Pruss G.J. Manes S.H. Drlica K. Cell. 1982; 31: 35-42Abstract Full Text PDF PubMed Scopus (239) Google Scholar, 26DiNardo A. Voelkel K.A. Sternglanz R. Reynolds A.E. Wright A. Cell. 1982; 31: 43-51Abstract Full Text PDF PubMed Scopus (289) Google Scholar) have been used in the experiments reported here. Bacteria were cultured at 37 °C with aeration in LB medium or grown on 1.2% LB agar plates. Media were supplemented with antibiotics as required; ampicillin and kanamycin were both used at 50 μg/ml and tetracycline was used at 10 μg/ml (except for strains related to E. coli DM800, where this was reduced to 2 μg/ml tetracycline). E. coli strains were transformed with plasmids using the calcium chloride procedure (27Cohen S. Chang A. Hsu L. Proc. Natl. Acad. Sci. U. S. A. 1972; 69: 2110-2114Crossref PubMed Scopus (1839) Google Scholar). The plasmids used in this work are summarized in TableITable IList of plasmids used in this workpLEU500TcOriginal plasmid containingleu-500 promoter and clockwise tetApL500TRFully functional anticlockwise tetApL500TR.ΔPtet.revpL500TR with deletion of anticlockwise tetA promoterpL500TR.Tet48pL500TR with translation terminator in NheI site of anticlockwisetetApL500TR.Tet96pL500TR with translation terminator in BamHI site of anticlockwise tetApL500TR.Tet188pL500TR with translation terminator inSalI site of anticlockwise tetApL500TR.Tet296pL500TR with translation terminator inNruI site of anticlockwise tetApL500TR.ΔblapL500TR with 30% N-terminal deletion of blapL500TR.ΔPtetpL500TR with deletion of clockwise tetA promoterpL500TR.ΔblaΔPtetpL500TR with deletions of bla and clockwise tetA promoterpL500TR.ΔPtetΔPtet.revpL500TR with deletion of clockwise and anticlockwise tetA promoterspL500TR.Bla12pL500TR with translation terminator in Eco57 site of blapL500TR.Bla80pL500TR with translation terminator inScaI site of bla Open table in a new tab The plasmid pLEU500Tc (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar) was cleaved withNheI and BalI, and pAT153 (28Twigg A.J. Sherratt D. Nature. 1980; 283: 216-218Crossref PubMed Scopus (650) Google Scholar) was digested withEcoRI and BalI. The NheI and EcoRI termini were rendered flush by incubation with 2.5 units of VentR DNA polymerase (NEB) at 72 °C for 30 min. The smallerEcoRI-BalI fragment of pAT153, containing the entire tetA gene, and the largerNheI-BalI fragment of plasmid pLEU500Tc were isolated by preparative gel electrophoresis. The two blunt-ended fragments were then ligated together with T4 DNA ligase, and the resulting plasmid was transformed into E. coli HB101. The plasmid containing the complete tetA gene oriented anticlockwise (see Fig. 1) was identified by restriction digestion of isolated plasmid DNA. pL500TR was cleaved withClaI, and the resulting linear DNA was digested with mung bean nuclease (35 units/μl for 25 min at 37 °C). The blunt-ended DNA was religated to generate a plasmid that contained the modified anticlockwise tetA promoter. To remove the −10 region of the original (clockwise) tetA promoter, pL500TR was partially cleaved with HindIII followed by digestion with mung bean nuclease (35 units/μl for 25 min at 37 °C). The blunt-ended DNA was religated and transformed into E. coliHB101. Since the HindIII cleavage could occur at either of the target sites, restriction digests and DNA sequencing were used to identify the deletion of clockwise-oriented tetApromoter. This plasmid contains deletions of both tet promoters of pL500TR. It contains a 4-bp 1The abbreviation used is: bp, base pair(s). deletion at the HindIII site overlapping the clockwise-orientedtetA promoter and a 4-bp deletion at the ClaI site of the anticlockwise-oriented tetA promoter. This plasmid contains a deletion between theSspI site and the ScaI site of the blagene. pL500TR was cleaved at the SspI and ScaI sites, the largest blunt-ended fragment was isolated by preparative electrophoresis in an agarose gel, and circularized with T4 DNA ligase. This plasmid combines the deletion between the SspI site and the ScaI site of the bla gene and the 4-bp deletion at theHindIII site of the clockwise-oriented tetApromoter in pL500TR. Pairs of complementary oligonucleotides (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar) were ligated into the plasmid pL500TR linearized by the appropriate restriction enzyme; NheI (partial digestion was required since there are two NheI sites in pL500TR),BamHI, SalI, and NruI, generating pL500TR.Tet48, pL500TR.Tet96, pL500TR.Tet188, and pL500TR.Tet296, respectively. These plasmids contain translation termination codons inserted into either theEco57 or ScaI sites within the blagene of pL500TR. Self-complementary oligonucleotides encoding a universal translation terminator (18Chen D. Bowater R. Lilley D.M.J. Biochemistry. 1993; 32: 13162-13170Crossref PubMed Scopus (24) Google Scholar) were inserted into theEco57 or the ScaI sites within the blagene of pL500TR, generating pL500TR.Bla12 and pL500TR.Bla80, respectively. RNA was isolated using essentially the method described previously (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar). RNA was prepared from 200-μl cultures in the mid-exponential growth phase by the addition of an equal volume of 20 mm sodium acetate (pH 5.2), 2% SDS, 0.3 msucrose and transferring to a boiling water bath for 1 min. The sample was then extracted twice with phenol/chloroform, and the nucleic acids were precipitated with ethanol. After the addition of 0.2 pmol of the appropriate radioactively [5′-32P]-labeled DNA primer, the sample was heated to 90 °C in 4.5 μl of 50 mmTris-HCl (pH 8.0), 50 mm KCl, and rapidly cooled. 25 units of RNasin (0.5 μl) were added, and the solution was incubated at 43 °C for 20 min before the addition to 12 μl of 70 mmTris-HCl (pH 8.0), 70 mm KCl, 15 mmMgCl2, 15 mm dithiothreitol, 1.3 mmdeoxynucleoside triphosphate mixture containing 50 units of moloney murine leukemia virus reverse transcriptase (Superscript Plus; Life Technologies, Inc.) and incubated at 42 °C for 2 h. cDNA transcripts were electrophoresed in 6% polyacrylamide gels in 90 mm Tris borate (pH 8.3), 10 mm EDTA (TBE) containing 7 m urea, next to sequence markers generated by dideoxy sequence reactions (29Sanger F. Nicklen S. Coulson A.R. Proc. Natl. Acad. Sci. U. S. A. 1977; 74: 5463-5467Crossref PubMed Scopus (52581) Google Scholar) using the same primer. After drying the gels, radioactive fragments were visualized by autoradiography at −70 °C with intensifier screens or with storage phosphor screens and a 400 S PhosphorImager (Molecular Dynamics). Quantitation of radioactivity was performed directly upon the phosphorimage using ImageQuant (Molecular Dynamics). E. coli cells were grown in 30 ml of LB plus appropriate antibiotics to mid-exponential growth phase, and the plasmid DNA was extracted using the Wizard Plus DNA extraction system (Promega). The purified DNA was electrophoresed in 1% agarose gels in TBE containing 2 μg/ml chloroquine. After electrophoresis, the gels were subjected to extensive washing in water followed by staining in 1 μg/ml ethidium bromide and further washing in water. The stained gels were photographed under UV illumination with red and green filters to remove background fluorescence. The photographic negatives were scanned electronically, and a negative image was presented. In previous studies we showed that the activation of the leu-500 promoter on the plasmid pLEU500Tc in topA S. typhimurium was dependent on the function of the adjacent tetracycline resistance genetetA (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar). The orientation of the tetA gene in pLEU500Tc is opposite to that of the leu-500 promoter,i.e. the leu-500 promoter is located immediately upstream of the tetA gene. Thus transcription oftetA might be the major generator of negative supercoiling in this local region, by the mechanism of Liu and Wang (11Liu L.F. Wang J.C. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7024-7027Crossref PubMed Scopus (1502) Google Scholar). Activation of the leu-500 promoter in pLEU500Tc required the coupled transcription and translation of tetA and the membrane insertion of its product (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar, 18Chen D. Bowater R. Lilley D.M.J. Biochemistry. 1993; 32: 13162-13170Crossref PubMed Scopus (24) Google Scholar). This suggested that membrane insertion of the TetA protein was essential to provide an anchorage point, which might act as a topological barrier against the diffusion of DNA supercoiling. These two related yet distinct roles for thetetA gene might be dissected if its polarity were reversed in the plasmid, and we therefore constructed a new plasmid pL500TR that contains a tetA gene oriented anticlockwise in the conventional depiction of pBR322-based plasmids. The reversedtetA gene is fully functional, and transformed cells have normal levels of resistance to tetracycline. pL500TR still contains the original clockwise tetA promoter, but the gene (tetA′) is truncated at the 48th codon. It also contains the anticlockwise antitet promoter. The plasmid map of pL500TR is shown in Fig. 1. In our earlier study, we demonstratedtopA-dependent activation of theleu-500 promoter carried on plasmid pLEU500Tc containing a clockwise tetA gene. To investigate the effect of a reversed polarity tetA gene on the activity of the leu-500promoter, RNA was isolated from pL500TR-carrying topA ortop + E. coli cells in mid-exponential growth, and transcripts initiated at the leu-500 promoter were sought. This was achieved by means of run-off reverse transcription using a primer that hybridizes to the vector sequence upstream of the S. typhimurium DNA (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar). A cDNA corresponding to RNA initiated at the leu-500 promoter should be 191 nuceotides in length. Since the antitetpromoter (the tetR promoter transcribing the same strand as the leu-500 promoter) is retained on pL500TR, cDNA corresponding to initiation at this promoter would be 281 nucleotides in length and provides a useful reference for quantitation. The results of the reverse transcription analysis are shown in Fig.2 A. There is a clear band of cDNA corresponding to initiation at the leu-500 promoter in DM800 (ΔtopA) cells, but the intensity of this species is very much lower for the RNA extracted from SD108 (top +). The cDNA band corresponding to initiation at the antitet promoter is of similar intensity in both top + and ΔtopA experiments. Thus the leu-500 promoter was activated by the reversed polarity tetA gene, and this activation was dependent on theΔtopA background. The activity of the leu-500 promoter as a function of the polarity of the tetA gene is directly compared in Fig.2 B using cells carrying either pLEU500Tc or pL500TR. RNA was extracted from E. coli DM800 (ΔtopA) in exponential growth and subjected to reverse transcription analysis as before. The level of initiation at the leu-500 promoter is closely similar in both plasmids. Thus thetopA-dependent activation of theleu-500 promoter does not depend on the orientation of thetetA gene. In its original orientation in pLEU500Tc, thetetA gene must be transcribed to activate theleu-500 promoter (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar). We therefore investigated whether this was also required when tetA was reversed in pL500TR. Plasmid pL500TR.ΔPtet.rev was constructed containing a 4-bp deletion in the ClaI site upstream of the reversedtetA promoter (anticlockwise). Cells containing the modified plasmid are sensitive to tetracycline, demonstrating the inactivation of the tetA gene. Cellular RNA was extracted from DM800 (ΔtopA) harboring pL500TR or pL500TR.ΔPtet.rev and analyzed by reverse transcription as before. The results (Fig. 3) show that initiation of transcription at the leu-500 promoter in pL500TR was significantly reduced by the promoter deletion in the reversed tetA promoter. Thus thetopA-dependent activation of theleu-500 promoter in pL500TR requires transcription of the reversed tetA gene. By analogy with the role of the clockwisetetA gene of pLEU500Tc, it seemed probable that translation would be required in the reversed gene of pL500TR. This was examined by provoking premature termination of translation of the reversedtetA gene by introducing translation terminators at various positions in the coding sequence. This was achieved by introducing complementary oligonucleotides into the NheI,BamHI, SalI, and NruI restriction sites along the tetA gene, thereby generating truncated TetA polypeptides of 48, 96, 188, and 296 amino acids, respectively. These can be compared with the full-length TetA that is 394 amino acids in length. These plasmids are called pL500TR.Tet48, pL500TR.Tet96, pL500TR.Tet188, and pL500TR.Tet296, respectively. These plasmids were transformed into E. coli DM800 (ΔtopA), RNA was prepared from cells in exponential growth, and the initiation of transcription from the leu-500promoter was analyzed by reverse transcription as before. Electrophoretic analysis of the cDNA (Fig.4 A) showed that the level of activity of the leu-500 promoter became lower as the length of the translated reversed TetA polypeptide was reduced. The data were quantified by phosphorimaging and are presented graphically in Fig.4 B. Evidently the function of the leu-500promoter is dependent on translation of the reversed tetAgene, and the level of the activation of the leu-500promoter is approximately linearly dependent on the size of the TetA polypeptide synthesized. Thus the topA-dependent activation of the leu-500 promoter depends both on transcription and translation of the reversed tetA gene. This closely parallels the situation where the tetA gene was oriented clockwise in the original construct pLEU500Tc, suggesting that a similar mechanism of activation of the leu-500 promoter is involved in both cases. When plasmids carrying a functioningtetA gene are isolated from topA E. coli orS. typhimurium in exponential growth and their linking number distribution examined by electrophoresis in agarose gels containing the intercalator chloroquine, it is generally observed that there is a bimodal distribution of topoisomers, one fraction of which is very highly negatively supercoiled. We have previously shown this to be the case for pLEU500Tc and demonstrated a correlation between the degree of activation of the leu-500 promoter and the extent of this hypersupercoiled fraction (14Chen D. Bowater R. Lilley D.M.J. J. Bacteriol. 1994; 176: 3757-3764Crossref PubMed Google Scholar). We therefore examined the plasmids carrying the reversed tetA gene to see if these were similarly subject to hypersupercoiling. Plasmid DNA was isolated from E. coli DM800 (ΔtopA) in exponential growth and analyzed by electrophoresis in 1% agarose in TBE buffer containing 2 μg/ml chloroquine (Fig. 5). The distribution of pL500TR topoisomers was clearly bimodal, with a significant fraction of hypersupercoiled DNA. Reversing the polarity of the tetAgene has not changed its effect on the overall topology of the plasmid. Interference with the function of the tetA gene reduces the extent of this fraction of highly supercoiled plasmid. The proportion was severely reduced for pL500TR.ΔPtet.rev (the plasmid containing a 4-bp deletion in the promoter of the reversedtetA gene), demonstrating the role of transcription of the reversed tetA in generating hypersupercoiled DNA. Translation of the tetA gene is also important for the hypersupercoiling, since the fraction of hypersupercoiled DNA was reduced in the plasmids where the tetA gene was interrupted by translation terminators; the shorter the translated TetA polypeptide, the smaller the fraction of hypersupercoiled DNA. Overall, there was a reasonable correlation between the fraction of hypersupercoiled DNA and the activity of the leu-500promoter for the different plasmid constructs containing a reversedtetA gene (see “Discussion”). Analysis of the topA-dependent activation of the leu-500 promoter in pL500TR clearly highlights the importance of the reversed tetA gene. We discussed two conceivable roles for this gene: as a direct generator of negative supercoiling by transcription with hindered rotation of RNA polymerase and as a topological barrier against the diffusion of negative supercoiling. Since the promoter of the tetA gene is a significant distance from the leu-500 promoter in pL500TR, it is possible that the primary role is the latter function and that other more local promoters are important in the generation of supercoiling. We therefore turned our attention to other gene expression occurring within the vicinity of the leu-500promoter. This arises primarily from the bla gene and the original tetA gene of which the promoter is retained in pL500TR. To determine the effect of local gene expression on the activity of theleu-500 promoter, a number of new plasmids were constructed. pL500TR.Δbla contains a 30% deletion in thebla gene, generated by removing the fragment between theSspI site and the ScaI site of the blagene in pL500TR. The bla promoter is not directly affected by this deletion. pL500TR.ΔPtet contains a 4-bp deletion in the HindIII site at the clockwise tetApromoter in pL500TR. This deletion is known to inactivate the promoter of the tetA gene (10Chen D. Bowater R. Dorman C. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8784-8788Crossref PubMed Scopus (64) Google Scholar). pL500TR.ΔblaΔPtet combines the deletions in the bla gene and in the clockwise-oriented tetApromoter. pL500TR.ΔPtet.rev contains a 4-bp deletion at the ClaI site at the anticlockwise-oriented tetApromoter; this plasmid has been discussed above. pL500TR.ΔPtetΔPtet.rev contains both the deletion at the HindIII site at the clockwise-orientedtetA promoter and the deletion at the ClaI site at the anticlockwise-oriented tetA promoter. These plasmids were transformed into E. coli DM800 (ΔtopA), cellular RNA was isolated from cells in mid-exponential growth, and the initiation of transcription at theleu-500 promoter was analyzed by primer extension as before (Fig. 6 A). The activity of theleu-500 promoter in pL500TR.Δbla (lane 4) was not significantly less than that in pL500TR, indicating that bla was not important in thetopA-dependent activation of theleu-500 promoter. Deletion of the clockwise-orientedtetA promoter that remains in pL500TR (pL500TR.ΔPtet; lane 5) also had very little effect on the activity of the leu-500 promoter. Even the combination of both bla and clockwise tetApromoter deletions (pL500TR.Δbla ΔPtet;lane 6) resulted in a relatively minor reduction inleu-500 promoter activity. By contrast, as we have seen above, deletion of the promoter of the reversed tetA gene (lane 7) results in a marked reduction in activity of the leu-500 promoter, and combination of the deletions of both tetA promoters results in a very similar low level of leu-500 promoter activity (lane 8). We conclude that the dominant effector of thetopA-dependent activation of theleu-500 promoter is the reversed tetA gene. Previous experiments showed that in the original construct with a clockwisetetA gene (pLEU500Tc), initiation of transcription at theleu-500 promoter was influenced by translation of thebla gene under some circumstances (18Chen D. Bowater R. Lilley D.M.J. Biochemistry. 1993; 32: 13162-13170Crossref PubMed Scopus (24) Google Scholar). We therefore examined the effect of modulating the function of the blagene on the activation of the leu-500 promoter in the presence of the reversed tetA gene. Two new plasmids were constructed to examine the influence of bla translation. pL500TR.Bla12 and pL500TR.Bla80 contain translation termination codons inserted into the bla coding sequences at theEco57 and the ScaI sites, respectively, generating β-lactamase polypeptides shortened from 263 amino acids to 12 or 80 amino acids, respectively. The plasmids were transformed into DM800 (ΔtopA), cellular RNA was isolated from cells in mid-exponential growth, and the initiation of transcription at the leu-500 promoter was analyzed by reverse transcription as before (Fig. 6 B). The activity of the leu-500 promoter was not significantly reduced in either of these plasmids, indicating that translation of thebla gene is relatively unimportant in the activation of theleu-500 promoter in pL500TR. Our results clearly demonstrate that the leu-500promoter can be activated on a plasmid in topA E. coli by the presence of a tetA gene in either orientation. Activation requires the full function of the tetA gene, but the leu-500 promoter can be located in a position that can be regarded either as primarily upstream or one that is downstream of this gene. Moreover the role of the tetA gene is paramount; although other promoters present in pL500TR are of relatively minor consequence, inactivation of tetA function reduces the activity of the leu-500 promoter to background levels. In summary, the tetA gene is essential for thetopA-dependent activation of theleu-500 promoter, but its orientation is unimportant. It might be regarded as surprising that this effect is independent oftetA orientation; that the activation of theleu-500 promoter is equally efficient when it is placed in what is formally the domain of positive supercoiling (downstream oftetA) (11Liu L.F. Wang J.C. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 7024-7027Crossref PubMed Scopus (1502) Google Scholar), as when it is located in the upstream domain of negative supercoiling. We therefore change our perspective from a local view of variation in superhelix density to a more global view. The local view supposes that the leu-500 promoter must be located directly within the domain of negative supercoiling to be activated. In the global view, unbalanced relaxation of transcriptional-induced supercoiling from the tetA gene results in a net reduction in the linking difference of the plasmid. If the tetA gene is the primary generator of supercoiling (because of its membrane anchoring effect), then it will create local domains of negative and positive supercoiling. If only the latter can be relaxed in a topA cell, the overall effect will be to lower the linking number of the plasmid. If the leu-500promoter is responding to this global change in topology, then it will do so independent of relative orientation or separation. We arrive at the same conclusion following a second line of argument. As we discussed in the introduction, an alternative role of membrane anchorage by coupled transcription, translation, and insertion of TetA could be to provide a topological barrier so that the domains of positive and negative supercoiling generated by transcription (in theory from any promoter) cannot diffuse around the circular plasmid and undergo self-cancellation by a simple rotation of the helix. If this were true, it would require the existence of a second barrier on the opposite side of the circular plasmid, and it has been suggested that the replication origin might function in this way (19Gartenberg M.R. Wang J.C. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 10514-10518Crossref PubMed Scopus (26) Google Scholar). The combined effect of two such barriers would effectively isolate the lower half of the plasmid in topological terms. However, in pL500TR, the promoter of the reversed tetA gene would be located in this domain, isolated topologically from the leu-500promoter. Yet we have shown that the single most important factor on the plasmid for the topA-dependent activation of the leu-500 promoter is the tetA promoter. We therefore conclude that it cannot be located in a separate domain and that the barrier model does not hold. We are left with the primary role of membrane anchorage as the provision of rotational hindrance to RNA polymerase transcribing the tetA gene. Since thetetA and leu-500 promoters are separated by more than 1.6 kbp, this must be considered as an essentially global phenomenon in the plasmid. The global view of the activation is consistent with measurement of the linking difference of isolated plasmids (e.g. Fig. 5), which is a measure of the global topology by definition. This shows that the fraction of hypersupercoiled plasmid DNA is generated whenever thetetA gene is present in cis, whatever its orientation. Indeed, we obtain a linear correlation between the level of activation of the leu-500 promoter in topA E. coli with the fraction of hypersupercoiled plasmid DNA isolated from the cells (Fig. 7). In situ probing of the formation of cruciform structures by alternating adenine-thymine ((AT) n ) sequences can be used as a means of testing local negative superhelix density in cellular DNA (20McClellan J.A. Boublikova P. Palecek E. Lilley D.M.J. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 8373-8377Crossref PubMed Scopus (150) Google Scholar), and we have shown that reporter (AT) n sequences introduced in the region corresponding to that upstream oftetA in pLEU500Tc detect unconstrained oversupercoiling intopA strains (21Bowater R.P. Chen D. Lilley D.M.J. Biochemistry. 1994; 33: 9266-9275Crossref PubMed Scopus (34) Google Scholar). However, contrary to initial expectations, we also detected elevated negative supercoiling at (AT) n sequences placed downstream of the tetAgene, 2R. P. Bowater, D. Chen, and D. M. J. Lilley, unpublished data. i.e. in the region that might be expected to experience transcriptional induction of positive supercoiling. Once again this result is more consistent with a global view of the induction of negative plasmid supercoiling in topA cells. The topA-dependent activation of theleu-500 promoter in pL500TR does differ in some respects from that in the original pLEU500Tc containing the clockwisetetA gene. One is the effect of bla expression; we observed that bla deletion lowered the level ofleu-500 promoter activation in pLEU500Tc (18Chen D. Bowater R. Lilley D.M.J. Biochemistry. 1993; 32: 13162-13170Crossref PubMed Scopus (24) Google Scholar), whereas there is little influence of bla in the presence of the anticlockwise tetA gene of pL500TR. However, we found that the effect of bla deletion on the leu-500promoter in pLEU500Tc could be removed when a tac promoter was introduced into this plasmid, suggesting that subtle effects may occur in this region. Another difference is the effect of spacing. When we introduced random DNA fragments between the leu-500 and tetA promoters of pLEU500Tc, this reduced the level of initiation of transcription at the former, whereas in pL500TR, the crucial Ptet.rev is almost diametrically opposite to theleu-500 promoter. At present we are unable to account for this difference. There have been reports of activation of the leu-500promoter in topA cells using plasmids that do not include the tetA gene (22Tan J. Shu L. Wu H.-Y. J. Bacteriol. 1994; 176: 1077-1086Crossref PubMed Google Scholar, 23Spirito F. Bossi L. J. Bacteriol. 1996; 178: 7129-7137Crossref PubMed Google Scholar). We find these observations perplexing, because in our experiments the role of the tetAgene is paramount. It is conceivable that other factors play a role in these constructs, but it is possible that the overall level of activation of transcriptional initiation was lower in those investigations. It is beyond question that in the plasmids based upon pLEU500Tc, the role of the tetA gene is essential for the observed level of activation and cannot be replaced by any other gene that we have explored. Moreover, correlation with the physical level of hypersupercoiling in our plasmids has been independently confirmed by the experiments of Mojica and Higgins (24Mojica F.J.M. Higgins C.F. Mol. Microbiol. 1996; 22: 919-928Crossref PubMed Scopus (16) Google Scholar), who measured the level of unconstrained plasmid supercoiling using an intercalation assay. In summary, the leu-500 promoter is activated highly efficiently in topA cells when it is borne on a plasmid carrying the tetA gene in cis, irrespective of orientation. The most probable explanation is that it is activated by negative supercoiling arising from the transcription of thetetA gene and that this process is most effective when RNA polymerase is effectively tethered due to coordinate transcription, translation, and membrane insertion. The coupling between the promoters can be fully explained by topological effects operating within the plasmid globally. We thank Dr. Richard Bowater for discussions and the Medical Research Council and Cancer Research Campaign for financial support.