The antibiotic bicyclomycin affects the secondary RNA binding site of Escherichia coli transcription termination factor Rho.

The interaction of Rho and the antibiotic bicyclomycin was probed using in vitro transcription termination reactions, poly(C) binding assays, limited tryptic digestions, and the bicyclomycin inhibition kinetics of ATPase activity in the presence of poly(dC) and ribo(C)10. The approximate I50 value for the bicyclomycin inhibition of transcription termination at Rho-dependent sites within a modified trp operon template was 5 μM. At antibiotic concentrations near the I50 value, bicyclomycin inhibition of Rho-dependent transcripts was accompanied by the appearance of a new set of transcripts whose size was midway between the Rho-dependent transcripts and the readthrough transcripts. Bicyclomycin did not inhibit poly(C) binding to Rho. In the presence of poly(dC), bicyclomycin showed a reversible mixed inhibition of the ribo(C)10-stimulated ATPase activity. The extrapolated Ki for bicyclomycin was 2.8 μM without ribo(C)10 and increased to 26 μM in the presence of ribo(C)10. Correspondingly, the Km(app) for ribo(C)10 without bicyclomycin was 0.8 μM and with bicyclomycin was 5 μM at infinite inhibitor concentration. The data suggested that the antibiotic binds to Rho, influencing the secondary RNA binding (tracking) site on Rho and slows the tracking of Rho toward the bound RNA polymerase.

The interaction of Rho and the antibiotic bicyclomycin was probed using in vitro transcription termination reactions, poly(C) binding assays, limited tryptic digestions, and the bicyclomycin inhibition kinetics of ATPase activity in the presence of poly(dC) and ribo(C) 10 . The approximate I 50 value for the bicyclomycin inhibition of transcription termination at Rho-dependent sites within a modified trp operon template was 5 M. At antibiotic concentrations near the I 50 value, bicyclomycin inhibition of Rho-dependent transcripts was accompanied by the appearance of a new set of transcripts whose size was midway between the Rhodependent transcripts and the readthrough transcripts. Bicyclomycin did not inhibit poly(C) binding to Rho. In the presence of poly(dC), bicyclomycin showed a reversible mixed inhibition of the ribo(C) 10 -stimulated ATPase activity. The extrapolated K i for bicyclomycin was 2.8 M without ribo(C) 10 and increased to 26 M in the presence of ribo(C) 10 . Correspondingly, the K m(app) for ribo(C) 10 without bicyclomycin was 0.8 M and with bicyclomycin was 5 M at infinite inhibitor concentration. The data suggested that the antibiotic binds to Rho, influencing the secondary RNA binding (tracking) site on Rho and slows the tracking of Rho toward the bound RNA polymerase.
The structurally unique antibiotic bicyclomycin (Scheme 1) (1, 2) has been shown to target a broad spectrum of Gramnegative bacteria, such as Escherichia coli, Klebsiella, Salmonella, Shigella, and Citrobacter (1)(2)(3). The primary site of action for bicyclomycin was shown to be Rho transcription termination factor in E. coli (4,5). DNA coding for Rho protein from antibiotic-resistant mutants was able to confer drug resistance to otherwise sensitive cells (4). Bicyclomycin has been shown to inhibit Rho-poly(C)-dependent ATPase activity with simple noncompetitive kinetics with respect to ATP (6).
Rho protein is composed of six identical 46-kDa proteins of 419 amino acids (7) in a proposed planar, hexagonal, geometric D3 symmetry (8 -10). Rho transcription termination factor is required to stop transcription at several Rho-dependent termination sites, including the proximal region of the lacZ cistron (11), the tRI terminator (12), and the trp tЈ at the end of the tryptophan operon in E. coli (13).
ATPase and helicase activities associated with Rho are es-sential for transcription termination (14 -16). The ATPase activity is latent until RNA binds to Rho. RNA binding to both a primary and a secondary site is required to stimulate ATPase activity. Reports have proposed a tethered tracking mechanism in which, initially, Rho binds tightly to RNA at a rut site and tracks toward the stalled RNA polymerase using the secondary RNA-binding site (for reviews, see Refs. 15 and 16). According to this mechanism, RNA binds tightly to the primary site in Rho and remains bound during the Rho transcription termination process, but this binding is not sufficient to stimulate ATPase activity (17,18). RNA binding to the secondary or tracking site causes further change in the structure of Rho, which stimulates an ATPase activity (17,19). A 5Ј to 3Ј DNA: RNA helicase activity (20) is likely to be required for transcription termination. Primary RNA and ATP binding domains of Rho have been determined (21)(22)(23)(24). The N-terminal 151 amino acids bound trp tЈ and poly(C) (23) and the first 116 amino acids to ribo(C) 8 (25). Mutants that affect primary binding have been located within the first 116 amino acids (26,27), and mutagenesis of residues 62 and 64 decreased RNA binding activity (28). The ATP binding domain extends from residues 160 to 340 and contains sequence similarity to E. coli F1 ATPase ␣ and ␤ subunits and adenylate kinase (23). Mutations at residues 181, 184, and 265 affected ATP hydrolysis without changing RNA binding to the primary site (21). At present, little is known about the secondary RNA binding site. The Rho mutant suA1 has been identified as a secondary site mutant with a single amino acid change at residue 352. This finding suggests that this region of the sequence is involved in Rho tracking (29,30).
A working model of the mechanism of Rho transcription termination relies on the coupling of RNA binding, at both primary and secondary binding sites, to ATP hydrolysis, promoting RNA tracking, helicase activity and eventual transcription termination (14,31). Mutations that conferred bicyclomycin resistance (M218K, S266A, G337S) were located in the ATP binding domain (4); however, kinetic studies demonstrated that bicyclomycin was a reversible, simple, noncompetitive inhibitor of Rho with respect to ATP (6). To examine the effects of bicyclomycin on Rho activities further, in vitro transcription termination assay, poly(C) binding, and poly(dC)-ribo(C) 10stimulated ATPase assays were carried out. We report, herein, that bicyclomycin inhibits in vitro Rho transcription termination processes. Evidence is presented that the antibiotic affects the RNA binding to the secondary site in Rho and modifies the rate of RNA tracking.

EXPERIMENTAL PROCEDURES
Materials and Enzymes-Bicyclomycin was purified by three successive silica gel chromatographies using 20% methanol/chloroform as the eluant, as described (6). Dihydrobicyclomycin was prepared by catalytic hydrogenation of bicyclomycin (32). Rho protein was isolated from E. coli AR120 containing the overexpressing plasmid p39-AS (33) following previously published protocols (33). Rho purity was determined by SDS-PAGE, 1 and concentrations were determined according to the Lowry protein determination (34). T4 polynucleotide kinase was purchased from Promega Co. (Madison, WI). [␥-32 P]ATP and [␣-32 P]UTP (3000 Ci/mmol) were purchased from DuPont NEN; nucleotides and RNase inhibitor were from Ambion, Inc. (Austin, TX), polynucleotides were purchased from Pharmacia Biotech Inc., and ribo(C) 10 was purchased from Oligos Etc. (Wilsonville, OR). L-1-Tosylamido-2-phenylethyl chloromethyl ketone-treated trypsin was obtained from Sigma. HA-type nitrocellulose filters used for binding assays were purchased from Millipore Co. (Bedford, MA), and polyethyleneimine thin-layer chromatography plates used for ATPase assays were purchased from J. T. Baker, Inc. All other chemicals were reagent grade.
In Vitro Transcription Termination Assay-E. coli RNA polymerase was purified according to the method of Burgess and Jendrisak (35), with minor modification. The Bio-Gel A-5m column was replaced with a Sephacryl S400 column. The DNA substrate was a truncated form of the trp operon (13). The HaeIII-SalI fragment from a pWU5 plasmid was ligated into the EcoRV-SalI sites of pGEM5 resulting in pTRP5, and the isolated BamHI-SalI fragment from pTRP5 was used as the template for the assay.
In Vitro transcription was carried out in a 10-l volume containing 20 mM Tris acetate, pH 7. . The samples were incubated at 37°C (20 min), diluted with 100 l of 0.3 M sodium acetate, 1 mM EDTA, and carrier tRNA at 0.8 g/l, extracted with phenol, precipitated with ethanol, and dissolved in loading buffer (80% formamide, 0.1% xylene cyanol, 0.1% bromphenol blue, and 2 mM EDTA). The RNA products were separated on a 5% polyacrylamide, 8 M urea gel and visualized using autoradiography. The relative amounts of radioactive incorporation of each band were determined with densitometry.
RNA Binding Assays-Polycytidylate (poly(C)) was labeled at the 5Ј end using [␥-32 P]ATP in reactions containing 50 pmol of poly(C), 70 mM Tris⅐HCl, pH 7.6, 10 mM MgCl 2 , 5 mM DTT, 30 Ci of [␥-32 P]ATP (3000 Ci/mmol), and 10 units of T4 polynucleotide kinase. Reaction mixtures were incubated at 37°C (30 min) and then 70°C (10 min) and loaded on polyacrylamide gels. The desired length of oligomer (approximately 130 nt) was cut out, eluted, and recovered by ethanol precipitation. Binding of poly(C) to Rho was done in binding buffer (40 mM Tris⅐HCl, pH 8.0, 25 mM KCl, 10 mM MgCl 2 , 0.1 mM DTT, 0.1 mM EDTA) and 0 -113 nM Rho in a total volume of 100 l. Two procedures were employed. First, bicyclomycin (400 M) or dihydrobicyclomycin (1.5 mM) was added to the Rho solution, and the solution was preincubated at room temperature (5 min) prior to adding approximately 25 nM 5Ј end 32 P-labeled poly(C). In the second method, Rho, bicyclomycin (400 M) or dihydrobicyclomycin (1.5 mM), and approximately 25 nM 5Ј end 32 P-labeled poly(C) were sequentially added without preincubation. Each reaction mixture was then incubated at 37°C (1 min) followed by filtration of 50 l of the reaction solution using 25-mm HA nitrocellulose filters presoaked with 0.1 mg/ml RNA. The filters were washed twice with 0.5 ml of buffer, dried briefly, and measured in a scintillation counter.
Tryptic Digestion and Gel Electrophoresis-One microliter of freshly prepared aqueous trypsin solution (0.1 mg/ml) was added to a 10-l solution containing 4 g of Rho in digestion buffer (40 mM Tris⅐HCl, pH 8.0, 50 mM KCl, 5 mM MgCl 2 , 0.1 mM DTT). The mixture was incubated at 37°C for various time periods. The reactions were stopped by the addition of 10 l of stop buffer (6 M urea, 5% 2-mercaptoethanol, 2% SDS, 10% glycerol, 25 mM Tris⅐HCl, pH 6.8) and heated at 90°C (5 min). The mixture was electrophoresed in 15% urea-SDS-PAGE at 120 V for approximately 8 h. The proteins were visualized using Coomassie Blue R-250. One microliter of 1 M poly(C), 1 l of 3.2 M poly(dC), 1 l of 0.2 mM ribo(C) 10 , 1 l of 10 mM ATP, and 1 l of 4 mM bicyclomycin were added to the appropriate samples in a total volume of 10 l.
Poly(dC)-Ribo(C) 10  Reactions were preincubated at 32°C for 2 min prior to the addition of ATP. Aliquots (2 l) were removed at various times (15,30,45,60, 75, and 90 s) during the reaction and spotted onto polyethyleneimine TLC sheets that had been prerun in H 2 O and dried. [␥-32 P]ATP and 32 P i were separated by chromatography on the polyethyleneimine sheets using 0.75 M KH 2 PO 4 , pH 3.5, as the mobile phase and then located by autoradiography. The radioactive spots were either cut out and counted by liquid scintillation, according to published methods, or the developed TLC plates were used to expose PhosphorImager plates (15-30 min) and scanned using a Fuji BAS 1000 Bio Imaging Analyzer and analyzed using the Macintosh BAS analysis program. The initial rates of reactions were determined by plotting the amount of ATP hydrolyzed versus time.

Inhibition of Rho Transcription Termination by Bicyclomycin
and Dihydrobicyclomycin-The ability of Rho to terminate in vitro transcription processes using a modified trp operon was first reported by Platt and co-workers (13). As we have diagrammed, Platt found that the read-through (run-off) transcript from the 800-base pair BamHI-SalI fragment starting from the trp p promoter consists of 545 nt (Band C) (see Fig. 1). Two terminators are present on this template DNA, a Rhoindependent terminator in which approximately 25% of the message terminates at 245 nt (trp t, Band A) and a Rho-dependent terminator at 470 nt (trp tЈ, Band B).
E. coli polymerase transcriptional processes were monitored by the incorporation of [␣-32 P]UTP, as described previously (13), and visualized using autoradiography (Fig. 1). Addition of Rho in the reaction led to the Rho-terminated transcript at trp tЈ (Fig. 1, lane 2, Band B) whereas the 545-nt read-through transcript was observed when Rho was not present (Fig. 1, lane 1 The abbreviations used are: PAGE, polyacrylamide gel electrophoresis; K i , dissociation constant for the enzyme-inhibitor complex; DTT, dithiothreitol; I 50 , concentration of inhibitor at 50% inhibition; K m , Michaelis constant; K m(app) , apparent Michaelis constant; nt, nucleotides. 1, Band C). Inclusion of bicyclomycin with Rho-containing solutions inhibited Rho-dependent termination processes (Fig. 1,  lanes 3-11). The approximate I 50 value for bicyclomycin was 5 M, where the I 50 value corresponded to the amount of bicyclomycin that gave 50% of the transcript terminated at the trp tЈ site.
There were two interesting findings for bicyclomycin inhibition of Rho transcription termination. First, the amount of bicyclomycin used to inhibit transcription termination reactions (I 50 ϳ5 M) was more than 10 times less than the amount needed to inhibit Rho-poly(C)-dependent ATPase activity (I 50 ϭ 60 M) (4). We have attributed the difference in the I 50 values in these two assays, in part, to the different sensitivities of the Rho-RNA complexes toward bicyclomycin. We found that the I 50 value for bicyclomycin in the poly(U)-dependent ATPase assay (I 50 ϭ 10 M) 2 was lower than that observed in the poly(C)-dependent ATPase assay (I 50 ϭ 60 M). Poly(U) is a weaker activator of ATPase activity than poly(C) (17). Also, oligoribonucleotides, ribo(U n C n ) 8 , varying in U and C composition, affect the K m of binding at the secondary site and the V max of Rho ATPase activity (19). The K m increased and the V max decreased when an increasing number of ribo(U)s was placed at the 5Ј site of the octooligoribonucleotide. These results suggest that poly(U) binds weaker to the secondary site of Rho than poly(C) and that decreased levels of bicyclomycin are required to inhibit Rho-mediated processes that employ poorer ATPaseactivating RNA substrates. Consistent with this notion, the trp tЈ is a weaker activator of Rho ATPase activity than poly(C) (21).
Second, we found appreciable amounts of a new set of intermediate-size transcripts when bicyclomycin concentrations close to the observed I 50 value were used (Fig. 1, lanes 5-8,  Band D). These transcripts were not observed at low or high bicyclomycin concentrations. The detection of these transcripts was consistent with the notion that bicyclomycin concentrations near the I 50 value sufficiently slowed down but did not totally abolish the Rho tracking process. Significantly, in order for transcription to effectively terminate at the trp tЈ site, the rate at which Rho transverses 5Ј to 3Ј toward the stalled polymerase must be faster than the rate at which the polymerase proceeds through the stall site on the DNA. If either the rate of the polymerase is increased or the rate of Rho translocation is decreased, termination would not happen. These results are consistent with the notion that bicyclomycin may affect the Rho secondary RNA binding site and slow the rate of tracking toward the polymerase. This phenomenon leads to the production of a new set of transcripts that is somewhat larger than 470 nt in length.
A similar set of experiments was conducted with dihydrobicyclomycin (Scheme 1). Dihydrobicyclomycin is a bicyclomycin analog in which the C(5)-C(5a) exomethylene group has been reduced. We observed that dihydrobicyclomycin inhibited Rho transcription termination and that the approximate I 50 value was 20 M (data not shown). The increase in the I 50 value for dihydrobicyclomycin compared with bicyclomycin in the transcription termination assay was similar to the observed differ-ences between these two compounds in the poly(C)-stimulated ATPase assay (bicyclomycin, I 50 ϭ 60 M (4), dihydrobicyclomycin, I 50 ϭ 120 M (5)). Once again, we observed intermediate-size transcripts at dihydrobicyclomycin concentrations near the I 50 value.
Effect of Bicyclomycin and Dihydrobicyclomycin on the Poly(C) Binding to Rho-The effect of bicyclomycin and dihydrobicyclomycin on Rho-poly(C) binding was studied using a nitrocellulose filter binding assay (17). Rho was incubated (room temperature, 5 min) with high levels of either bicyclomycin (400 M) or dihydrobicyclomycin (1.5 mM) and labeled poly(C) both without and with 5 mM ATP, and the reaction solutions were filtered. These antibiotic levels are sufficient to inhibit 95% of the poly(C)-dependent ATPase activity. Fig. 2 shows the percent of poly(C) bound to Rho as a function of Rho concentration without bicyclomycin, with bicyclomycin, with ATP, and with bicyclomycin and ATP. We observed no inhibition of poly(C) binding in the presence of bicyclomycin or dihydrobicyclomycin (data not shown). Finally, no detectable loss of poly(C) binding was observed even when bicyclomycin (400 M) was not preincubated with Rho (data not shown). These results indicated that the antibiotic did not affect RNA binding to the primary RNA binding site in Rho.
Effect of Bicyclomycin on Limited Tryptic Digestion of Rho-Conformational changes in Rho due to ligand binding can be probed by limited tryptic digestion (24,36). The tryptic digestion patterns of Rho change upon addition of poly(C) and/or ATP. Accordingly, we conducted a study of the effects of bicyclomycin binding on the trypsin digestion of Rho protein. Additional ligands included in our study were poly(C), ATP, poly(dC), and ribo(C) 10 . Without bicyclomycin our results were in agreement with previous studies (24,36). Inclusion of bicyclomycin alone or with another ligand(s) provided no significant changes in the digestion patterns (Fig. 3). We conclude that tryptic digestion does not permit the detection of any bicyclomycin-mediated Rho conformational changes.
Bicyclomycin and Dihydrobicyclomycin Inhibition of Rho-Poly(dC)-Ribo(C) 10 -stimulated ATPase Activity-The effects of bicyclomycin on Rho-dependent transcription termination suggest that bicyclomycin may affect the secondary RNA binding (tracking) site. The secondary binding site can be distinguished from the primary binding site (rut binding site) with kinetics using poly(dC) and ribo(C) (18). Poly(dC) binds to Rho at the primary binding site without stimulating ATP hydrolysis. Upon further addition of ribo(C) 7-10 the ATPase activity is stimulated. Fig. 4A shows the effects of bicyclomycin concentration on the poly(dC)-ribo(C) 10 -stimulated ATPase activity. We observed that increasing concentrations of bicyclomycin decreased the rate of ATP hydrolysis. The kinetics showed mixed inhibition when 1/V was plotted against 1/[ribo(C) 10 ]. The binding of ribo(C) 10 is dependent on the amount of bicyclomycin present. The K m(app) for ribo(C) 10 without bicyclomycin was 0.8 M. This value was lower than that previously observed (30,37,38). The extrapolated K m(app) for ribo(C) 10 was 5 M at infinite bicyclomycin concentration. This increase in the K m(app) for ribo(C) 10 indicated that bicyclomycin influenced the ribo(C) 10 binding. A K i value of 2.8 M, the dissociation equilibrium constant for the interaction of bicyclomycin with the free enzyme (Rho plus poly(dC) without ribo(C) 10 ), was determined from the plot of the slopes versus bicyclomycin concentration (Fig. 4B) derived from Fig. 4A. The corresponding K i value in the presence of ribo(C) 10 was 26 M, calculated from a plot of the intercepts (1/V max ) versus bicyclomycin concentration (Fig. 4B). This value was comparable with the K i value of 20 M previously determined for ATP hydrolysis in the presence of poly(C) (6).
Dihydrobicyclomycin was also used as an inhibitor of Rhopoly(dC)-ribo(C) 10 -stimulated ATPase activity and gave a similar kinetic profile. The double-reciprocal plot is provided in Fig. 5A. The K m(app) for ribo(C) 10 without dihydrobicyclomycin was 0.9 M while the extrapolated K m app for ribo(C) 10 was 3.3 M at infinite dihydrobicyclomycin concentration. Values for K i with and without ribo(C) 10 were 45 and 10 M, respectively (Fig. 5B). These results indicated that both bicyclomycin and dihydrobicyclomycin displayed comparable inhibitory properties in the ATPase assay when poly(dC)-ribo(C) 10 was used in place of poly(C). The K i for dihydrobicyclomycin without ribo(C) 10 was about 4 times higher than the corresponding value for bicyclomycin. This difference in activities for these two compounds was similar to the differences previously observed in the I 50 values for Rho-dependent transcription termination and the K i values for inhibition of poly(C)-dependent ATPase activity.

DISCUSSION
In vitro Rho-dependent transcription termination reactions were inhibited by bicyclomycin with an I 50 value of ϳ5 M (Fig.  l). At these concentrations, intermediate-size transcripts midway in length between the Rho-terminated and the readthrough transcripts were seen. This observation suggested bicyclomycin interfered with the tracking rate at which Rho binds to RNA at the secondary site and translocates down the mRNA from 5Ј to 3Ј toward the stalled RNA polymerase. A slower tracking rate would permit the polymerase to move through the stall site before transcription termination. The continued synthesis of RNA would lead to transcripts of intermediate length provided Rho could catch the polymerase before the end of the template. This kinetic coupling mechanism has been used to explain mutations in the ␤ subunit of RNA po- lymerase that complement Rho mutations (39). Kinetic complementation was suggested over a direct contact between Rho and the RNA polymerase. The mutation in the ␤ subunit of RNA polymerase slowed down the polymerase rate so that a slowly tracking mutant of Rho could catch up and effectively terminate transcription (39).
The hypothesis that bicyclomycin affects the secondary RNA binding (tracking) site is supported by several additional observations. We found that the percent of poly(C) bound to Rho was not influenced by bicyclomycin. Bicyclomycin (400 M) sufficient to inhibit 95% of the ATPase activity did not detectably decrease the amount of poly(C) bound to Rho compared with a similar experiment done without bicyclomycin (Fig. 2). A similar result was observed when ATP was included in the reaction. Comparable findings were found for dihydrobicyclomycin. These results suggest that bicyclomycin and dihydrobicyclomycin did not affect RNA binding to the primary RNA binding site. Ligand binding (poly(C), ATP) to Rho has been shown to change the degradation pattern of Rho affecting the rates and appearance of partial digestion bands (24,36). The addition of either poly(C) or poly(dC) to the trypsin digestions of Rho led to conformational changes that rendered the protein more susceptible to trypsin, whereas the addition of ATP provided partial protection. We did not observe pronounced changes in the digestion pattern when bicyclomycin was added (Fig. 3) with or without other ligands, signifying that its addition did not affect the primary RNA and ATP binding sites in Rho.
Further information concerning the effect of bicyclomycin on Rho function was obtained from kinetic studies. Poly(dC) has been shown to bind only at the primary RNA binding site in Rho, and this binding is not sufficient to stimulate Rho-dependent ATPase activity (17,18). The ATPase activity can be restored in Rho-poly(dC) mixtures by the addition of ribo(C) (e.g. 10-mers). These findings indicated that the stimulated ATPase activity resulted from oligoribonucleotide binding to the secondary RNA binding site. The activation of ATPase activity by small oligoribonucleotides provided a direct measure of RNA binding to the tracking site in Rho. Since bicyclomycin or dihydrobicyclomycin inhibition of Rho-dependent transcription termination intimated that the antibiotic affected the tracking rate of Rho toward the polymerase, we determined whether bicyclomycin or dihydrobicyclomycin inhibited ATP hydrolysis in the presence of poly(dC) and ribo(C) 10 . The kinetics were measured with near saturating amounts of ATP (0.2 mM). We have reported that bicyclomycin or dihydrobicyclomycin inhibits ATP hydrolysis by a simple noncompetitive inhibition pathway with respect to ATP in the presence of poly(C) (6). This finding allows us to assume that differing ATP concentrations did not affect bicyclomycin binding. We observed that bicyclomycin changed both the V max and K m(app) for ribo(C) 10 , and a mixed inhibition model with respect to ribo(C) 10 was used to fit the data. The K m(app) for ribo(C) 10 changed from 0.8 to 5 M as the bicyclomycin concentration was extrapolated from 0 to infinite concentrations. Analysis of the slopes and the intercepts from the double-reciprocal plot (Fig. 4A) indicated that the extrapolated K i value, the dissociation equilibrium constant for the interaction of bicyclomycin with the free enzyme (Rho plus poly(dC) without ribo(C) 10 ), was 2.8 M, and the ␣K i value in the presence of ribo(C) 10 was 26 M.
Since bicyclomycin inhibition of ribo(C) 10 -stimulated ATPase activity followed a mixed inhibition model with respect to ribo(C) 10 , the affinity of ribo(C) 10 binding to the Rho-poly(dC) complex decreased in the presence of bicyclomycin, and the affinity of bicyclomycin to the Rho-poly(dC) complex in turn decreased in the presence of ribo(C) 10 . This result leads to the conclusion that bicyclomycin affects the RNA tracking rate at the secondary RNA binding site by binding to a probable nonoverlapping site, which nonetheless is capable of reciprocal interactions depending on substrate binding. This is consistent with the results of the transcription termination and poly(C) binding experiments.
Our findings showed that dihydrobicyclomycin displayed properties similar to bicyclomycin in the transcription termination, Rho-poly(C) binding, and poly(dC)-ribo(C) 10 -stimulated ATPase assays. These results indicated that bicyclomycin expressed its inhibitory activity in in vitro Rho-mediated processes principally by a noncovalent binding process. This observation is important since bicyclomycin readily reacts with nucleophiles at the C(5)-C(5a) exomethylene group to give covalent adducts (40,41). Reduction of this structural unit in bicyclomycin to give dihydrobicyclomycin eliminates this pathway. Significantly, our findings do not exclude the possibility that covalent modification plays a role in the in vivo action of bicyclomycin.
Recently, several laboratories have demonstrated that Rhomediated processes were sensitive to bicyclomycin. Transcription termination reactions catalyzed by a Rho isolated from a Gram-positive bacterium Micrococcus luteus were sensitive to bicyclomycin (42). Also, bicyclomycin-resistant mutants were localized to rho and rpoB (4,43). These findings are consistent with the notion that bicyclomycin slows the tracking rate of Rho from the rut (binding) site toward the stalled RNA polymerase because a reduction in the transcription rates would allow a more slowly moving Rho to catch up to the polymerase and terminate transcription. Our results, however, do not rule out that additional factors (e.g. NusG) (43)(44)(45)(46) that control the rate of Rho transcription termination could also influence bicyclomycin resistance in E. coli.