| Nucleic Acids Research | Pages |
Escherichia coli RNA polymerase translocation is accompanied by periodic bending of the DNA
Introduction
Materials And Methods
Preparation of the A1-promoter templates
Preparation of halted complexes for band-shift analysis
Band-shift analysis
RNA analysis
Rate of dead-end complex formation of halted complexes
Transcription competence of halted complexes after non-denaturing gel electrophoresis
Statistical correlation analysis
Results
Ternary complexes halted at registers 11-68 show a periodic change of their electrophoretic mobility
Changes of the electrophoretic mobility can be attributed to bending of the DNA
A periodic variation of the bending angle remains, even if the transcribed sequence is replaced
Origin of the changes of the DNA bending angles of DNA in ternary complexes
The ternary complex remains transcription competent after non-denaturing electrophoresis
Dead-end complex formation is accompanied by relaxation of the polymerase-induced DNA bending
Discussion
Structural variations of halted complexes: changes of the DNA bending angle
Sequence-dependent effects on the bending pattern
Origin of the changes of the bending angle
Acknowledgements
References
Escherichia coli RNA polymerase translocation is accompanied by periodic bending of the DNA
Received June 17, 1999; Revised and Accepted July 29, 1999
ABSTRACT RNA polymerase was halted in consecutive registers of RNA synthesis ranging from registers 11 to 68. Non-denaturing gel electrophoresis shows that the mobility of the complexes varies (up to 15%), indicating that halted complexes differ in their conformation. The electrophoretic mobility changes with an approximate 10-register periodicity. The change of the mobility can be attributed to relative changes of RNA polymerase-induced bending angle. We suggest that the periodicity of the bending angle reflects periodic changes of the conformation of the halted complexes that might have relevance for the translocation mechanism.
INTRODUCTION
Our view of how Escherichia coli RNA polymerase translocates along the DNA during RNA synthesis has changed dramatically over the last few years. The first translocation models assumed that RNA polymerase translocates continuously (1). This view was supported by exonuclease III footprinting data on halted complexes suggesting a continuous movement of RNA polymerase during elongation of RNA (2). This view was called into question by the observation that transcription complexes, when halted in consecutive registers of RNA synthesis, differ in their DNaseI and exonuclease III footprinting patterns (3-6). Differences in the footprints had earlier been observed on transcription complexes analyzed during transition from the abortive to the productive phase of synthesis. Based on this observation a stressed intermediate (7) and an `inchworm-like motion' of RNA polymerase were suggested (2). Discontinuities in the footprinting patterns observed in later stages of synthesis (3-6,8,9) led to the conclusion that translocation is a discontinuous process which can be described by an inchworm-like movement of RNA polymerase (4,10). However, the inchworm model was in turn questioned by findings that showed that the observed discontinuities in the footprints could be at least partially attributed to processes occurring after halt of RNA synthesis (11,12).
More recently Kashlev et al. (13) analyzed in detail the temporal dependence of conformational changes of complexes after halt of synthesis. They identified two conformations of the halted complex, an active and an inactive form. Both conformations are in a reversible equilibrium, permitting revival of the temporarily inactive complex. This behavior of the complex has been explained by an oscillation of RNA polymerase with respect to the position required for incorporation of the following nucleotide. A similar kind of oscillation of RNA polymerase was suggested previously for T7 RNA polymerase (14) and mammalian RNA polymerase (15). The concept that RNA polymerase oscillates after halt of synthesis also fits in well with the `spider' model (16) that has recently been suggested as a possible explanation of our OH-radical footprinting and single-strand accessibility studies (9).
The present study deals with the analysis of structural changes of the DNA, which we observed in halted transcription complexes. Our aim was to find out whether conformational changes in the DNA, such as bending, contribute to or participate in the translocation process.
We have analyzed the electrophoretic mobility of complexes halted in consecutive registers from 11 to 68. We show that a halted complex can remain transcription-competent even after several hours of electrophoresis under non-denaturing conditions if special precautions are taken.
Analysis of the electrophoretic mobility of halted complexes indicates that RNA polymerase induces bending of the DNA. The bending angle changes systematically with progress of RNA synthesis having a periodicity of approximately 10 registers.
MATERIALS AND METHODS
RNA polymerase His-tagged at the C-terminus of the [beta][prime]-subunit was purified from the strain RL721 (provided by R. Landick) similar to that described by Kashlev et al. (17). The holoenzyme was separated from core enzyme by chromatography on a monoS resin in 50% glycerol.
Preparation of the A1-promoter templates
The promoter fragment containing A1 promoter of 220 bp length was prepared from the plasmid pDS-A1-220 (18) by BamHI hydrolysis, fractionation with polyethyleneglycol (19) and final purification on a monoQ column.
The templates A1-260, A1-340-1, A1-340-2, A1-340-3, A1-340-4 and A1-560 were prepared by PCR amplification using the plasmid pDS-A1-220 as template. The DNA fragments were purified by means of the QIAGEN PCR purification kit and concentrated by ethanol precipitation. The fragment A1-340-1M differs from A1-340-1 in the sequence downstream of position +46. The fragment is obtained by PCR using the plasmid pDS-A1-130 (19).
Preparation of halted complexes for band-shift analysis
A1 promoter fragment (1 µg, ~4.5 pmol) and RNA polymerase (4 µg, ~9 pmol) were incubated for 5 min at 37°C in 50 µl buffer BBh (10 mM HEPES, pH 7.5, 6 mM MgCl2, 50 mM NaCl and 1 mM mercaptoethanol) to form an open complex. RNA synthesis into register 11 or 20 was allowed by incubation with 20 µM ApUpC primer, 20 µM GTP (+CTP) and 5 µM [[alpha]-32P]ATP for 5 min at 37°C, 10 µl of the hard suspended Ni-NTA agarose was added. RNA polymerase was incubated for 10 min at room temperature to allow binding. The resin-bound complex was washed successively with BBh buffer supplemented with 1 M NaCl in order to destroy non-initiated complexes. The procedure was repeated three times using BBh containing 50 mM NaCl. RNA synthesis was propagated by addition of the proper NTP yielding a final concentration of 5 µM. Transcription was performed for 1 min at room temperature and the subsequent washing cycles were performed as described by Kashlev et al. (17).
Band-shift analysis
Samples of 10 µl of the Ni-agarose slurry containing ternary complexes were taken at each halted register, mixed with 3 µl of the stop/elution buffer containing 0.75 M imidazole, 25 mM EDTA, 20% glycerol and placed on ice. After finishing walking all samples were centrifuged simultaneously and 10 µl of the supernatant were applied on a non-denaturing gel. A 2 µl aliquot of the same supernatant was mixed with 4 µl formamide and analyzed on a denaturing gel.
Non-denaturing electrophoresis was performed in a 4% polyacrylamide gel (20 × 40 × 0.1 cm) containing TBE buffer (50 mM Tris-borate, pH 8.2 and 1 mM EDTA) at 5-10°C applying 5 W for 3-4 h. Under these conditions, the xylene cyanol dye migrated 40 cm. The gel was fixed in 10% acetic acid, washed with water, dried at 70°C and exposed to X-ray film. In the case of the complexes containing 11-20mer labeled RNA the fixation and drying steps were omitted.
RNA analysis
RNA was analyzed in a denaturing polyacrylamide gel (20 × 40 × 0.05 cm) containing TBE buffer and 8 M urea. A 20% gel was used for 11-20mer RNA, 15% for 20-44mers and 12% for RNAs longer than 44 nt.
Rate of dead-end complex formation of halted complexes
Complexes were formed as described above and incubated at 37°C. After different incubation time intervals elongation of RNA synthesis was permitted for 1 min at 37°C by addition of the proper substrate mix. The products were subsequently analyzed on a urea gel. The bands were quantified by means of a phosphoimager. Inactivation rate was calculated from the time dependence of the accumulation of non-elongated RNA normalized to the total amount of synthesized RNA. The values in Figure 4 represent the inactivation rate within 1 min. The values are corrected for the initially inactive fraction of the complexes (<5%).
Transcription competence of halted complexes after non-denaturing gel electrophoresis
Halted complexes containing the radioactive label in the RNA were subjected to non-denaturing gel electrophoresis. The RNA-containing bands were identified by autoradiography and cut out of the gel. The gel pieces were merged in a 100 µl transcription buffer (described above) containing the proper nucleotide mix, 20 µM (ATP + GTP + CTP) for complexes halted in the registers 11-18 and 48, and 20 µM (ATP + UTP + GTP) for complexes halted in the registers 20 and 50-58. RNA synthesis was permitted for 5 min at room temperature. The RNA was eluted from the gel pieces by shaking for 1 h at 37°C in the presence of 1% SDS and 10 µg/ml tRNA as carrier. The eluted and ethanol-precipitated RNA was analyzed by 20 or 12% PAGE in the presence of urea. Non-elongated complexes were treated as a control in the same way but without addition of NTP.
Statistical correlation analysis
Correlation between sequence and electrophoretic mobility pattern was performed by means of the image processing software SEMPER (20). For that purpose the sequence information was converted into mathematical functions, created for each single base and each combination of two bases. This was performed by setting the sequence function S(n) to 1, if a specific base or base combination was found in register n, otherwise S(n) was set to 0.
In order to enhance the correlation, the original mobility function (Fig. 1C and E) was modified. A parabolic function that fits to local mean values of the mobility function was subtracted yielding a mobility function Fm(n). Cross correlation of the sequence function S(n) and the mobility function Fm(n) was obtained by moving both functions register-by-register against each other and determining the normalized product functions.
Figure 1. Band-shift analysis of the ternary complexes halted at different registers. The fragments used in the following electrophoretic mobility experiments, A1-340-1 and A1-340-1M, are depicted in (A). Both contain the strong promoter A1 and differ with respect to the sequence downstream of position +46. Using these fragments gel electrophoretic mobility studies were performed with complexes halted in the register, as indicated. (B) Results obtained with A1-340-1 and (D) those obtained with A1-340-1M. The lower panels in (B) and (D) show the RNA products obtained in the different registers. (C) and (E) depict graphical representations of the mobility studies from (B) and (D). Arrows indicate the maxima of the mobility. The effects due to position-dependent changes of the mobility (calculated from the data in Fig. 2B) are indicated by the dotted lines.
RESULTS
Ternary complexes halted at registers 11-68 show a periodic change of their electrophoretic mobility
Bending of DNA, either intrinsic due to a specific sequence or induced by a protein, causes a change of the electrophoretic mobility of the DNA in a non-denaturing gel (21,22). This technique was previously used to visualize bending of the DNA in the binary and ternary complex (21,23-26). We applied this technique systematically to complexes halted in the registers 11-68. A DNA fragment (A1-340-1) carrying the promoter T7A1 was used in this study (Fig. 1A). The RNA polymerase was permitted to translocate from register 11, located in the middle of the fragment, towards register 68. RNA synthesis was halted in consecutive registers by using incomplete sets of substrate mixes, as described in Materials and Methods. Repeated application of this technique permitted RNA polymerase to `walk' along the DNA (17). The correct halt of synthesis was verified by analyzing the size of the synthesized RNA (Fig. 1B, lower panel). Figure 1B (upper panel) shows the electrophoretic mobility of the complexes in a non-denaturing gel. Mobility increases and decreases in the different registers, producing a wave-like pattern. Minima of the mobility are the registers 14-18, 25, 34-36, 45, 54 and 64 (Fig. 1C). Maxima are observed approximately every 10 registers, except around register 35 where an additional slight minimum is visible. The most pronounced changes are observed in the registers 40-65, where the maximum/minimum difference (amplitude) amounts to 15%.
A qualitative relationship between mobility and bending angle is well established: the lower the mobility the larger the bending angle (defined as deviation from the straight conformation) (27). However, a quantitative relationship between change of mobility and bending angle would require proper standards (28).
Changes of the electrophoretic mobility can be attributed to bending of the DNA
The observed mobility differences can be due to conformational changes of either DNA, RNA or protein. In order to study the role of DNA, ternary complexes were formed with DNA fragments having differently sized DNA regions flanking RNA polymerase.
Two sets of promoter-containing fragments were used (Fig. 2): (i) those with the promoter close to the center, where the length of the DNA regions flanking the promoter varied (Fig. 2A, upper panel); and (ii) those identical in size, where the promoter was located at different positions (Fig. 2B, upper panel).
Figure 2. Effect of the length of the DNA flanking RNA polymerase (A) and of the position of RNA polymerase (B) on the electrophoretic mobility. Upper panels in (A) and (B) show the fragments used. (A) These fragments carry the promoter (A1) close to the center of the fragment. The size of the fragments varies as indicated. The graph in the lower part of (A) depicts the register-dependent mobilities of the complexes halted using the fragments as indicated. The mobilities were normalized to that of the complex in register 11. (B) These fragments carry the promoter at different positions as indicated, but the size of the fragments remains the same. The gel pattern shows the mobility of the complexes using the different fragments.
Using set 1 of the fragments, complexes halted in the registers 11, 14, 16, 18 and 20 were formed and electrophoretically analyzed. Figure 2A (lower panel) shows that the mobilities of these complexes increase from register 11 to 18 and decrease again in register 20. The degree of increase and decrease depends on the size of the fragment, i.e. the size of the DNA region flanking the bound protein dictates the amplitude of mobility change. This result indicates that the DNA rather than RNA polymerase or RNA is primarily responsible for the observed wave-like changes in electrophoretic mobility.
Set 2 of the fragments is identical in size (340 bases) and carries the transcription start-site (+1 position) at different positions, as shown in Figure 2B (upper panel). Binary (binding) 11mer and 20mer complexes were formed and subjected to electrophoretic mobility studies. Figure 2B (higher panel) shows that the further RNA polymerase is positioned towards the end of the fragment, the lower the mobility is. This finding together with the finding described in the previous paragraph is in line with the view that the observed changes of the electrophoretic mobility at different registers are primarily due to changes in polymerase-induced DNA bending (22,27).
The mobilities in Figure 1C have not been corrected for the position-dependent effects, which are only slight compared to the register-dependent differences shown by this figure.
A periodic variation of the bending angle remains, even if the transcribed sequence is replaced
In order to analyze whether the periodicity of the bending angle depends on the sequence, a modified fragment (A1-340-1M, shown in Fig. 1A) was used. The new A1-340-1M fragment differs from the original A1-340 fragment by the sequence downstream of base position +46, which was replaced by an unrelated sequence. This replacement led to changes of the mobility pattern (Fig. 1C and E). DNA has the most extended conformation in the registers 11, 20, 32, 42, 47, 57 and 66 showing that the periodicity can increase up to 12 registers. The mobility in the range between register 47 and 57 changes only slightly, complicating conclusions about the periodicity in this register range.
Most interesting is the finding that the mobility pattern of the new construct (A1-340-1M) differs from that the mobility pattern of the original fragment (A1-340-1M) already in register 28 3, i.e. 18 3 registers in front of the replaced sequence. That means that RNA polymerase-induced DNA bending is sensitive towards sequence changes 18 base positions downstream of the polymerization site.
Origin of the changes of the DNA bending angles of DNA in ternary complexes
The bending angle of complexes halted in subsequent registers shows a regular increase and decrease of the electrophoretic mobility with fragment A1-340-1 as well as with fragment A1-340-1M (Fig. 1). The bending patterns show that the bending angle varies with a periodicity of 10 registers. But deviations are possible up to +2 and -5 registers from the regular 10-register periodicity. The amplitude changes are much more irregular, varying by as much as 15%.
We addressed the question of whether the observed periodicity depends on the sequence context by performing a correlation analysis of sequence and periodicity. The correlation between sequence and mobility pattern was determined analyzing all four single bases and all possible dinucleotide combinations. Correlation analysis with the peaks and the valleys of the patterns were performed and the results are shown in Figure 3. Since the replacement studies showed that the borders of the interaction regions of RNA polymerase and DNA play a role in bending, at least downstream, the relevant regions were subdivided into a core region (0 20) and border regions (+20 5 and -20 5). Figure 3 indicates that the correlation coefficients for the fragments A1-340-1 and A1-340-1M do not show any preference for a certain base or base combination, regardless of which criteria was applied. None of the single bases or base combinations yielded a higher correlation coefficient than 0.4, which is not too far from the average correlation coefficient. From this study we conclude that there is no evidence for a correlation between base sequence and mobility pattern within the statistical significance of the data.
Figure 3. Distribution of cross correlation coefficients of sequence and bending patterns. The histogram shows the extrema of the cross correlation coefficients obtained by correlating the bending pattern with the sequence pattern of single bases or dinucleotide combinations. It was searched for correlation (maxima) and anti-correlation (minima) as well. The x-axis of the histogram shows ranges of the correlation coefficients in steps of 0.05. Each cross represents a correlation extremum (maximum or minimum) obtained by relative shifts of the bending and sequence patterns, the core region ranging from position -20 to +20 (B) and the border regions ranging from position -15 to -25 (A) and position +15 to +25 (C) respectively. The top panels show the correlation coefficients of fragment A1-340-1 and the bottom panels of the fragment A1-340-1M.
Figure 4. Halted complexes at the `decision point'. Schematic representation of possible reaction pathways of halted transcription complexes. (A) The halted complex, which can enter the elongation status (B). Alternatively, the complex (A) can adopt the status (A[prime]) that represents a transiently inactive status as suggested by Komissarova and Kashlev (13). This complex is in equilibrium with the transcription competent complex (A), but can also enter the dead-end status, complex (C). If the RNA is cleaved, the dead-end complex (C) can become transcription competent, leading to complex (D) that enters the elongation status [complex (B)].
The 10-register periodicity of the electrophoretic mobility is somewhat puzzling because the DNA helix shows the same periodicity. This finding prompted us to analyze whether a simple explanation could account for the observed 10-register periodicity in the mobility pattern, namely a superposition of a potential static bend in the DNA and a movable bend induced by RNA polymerase. Such a double bend with a monotonous change of its orientation would lead, according to Zinkel and Crothers (29), to a 10-register periodicity. Results obtained with fragments identical in size but differing with respect to their promoter positions shown in Figure 2B (upper panel) tend to rule out this possibility. The mobilities of these fragments are the same (data not shown) as expected for fragments containing no static bend. Another argument against this hypothesis is the finding that the mobility pattern deviates from the strict 10-register periodicity.
By excluding that the bending pattern is a consequence of the periodicity or the sequence of the transcribed DNA the view is strengthened that the observed periodic bending is an intrinsic feature of the translocation process.
The ternary complex remains transcription competent after non-denaturing electrophoresis
One criticism of the results obtained by electrophoretic mobility studies of halted complexes is that a possible conformational change of the complex can occur during electrophoresis. This criticism is not unreasonable, since it is established that halted complexes can undergo a number of side reactions. Figure 4 shows known intermediates and possible reaction pathways of a halted complex. RNA polymerase can slide back and form a dead-end-complex that has lost transcription competence (Fig. 4C) (26,30). Cleavage of the nascent transcript by RNA polymerase can reactivate the complex (Fig. 4D) (31-33). The complex can enter an inactive (silent) state after removal of nucleotides and can be reactivated by addition of nucleotides (Fig. 4A[prime]) (11).
The rate of dead-end complex formation in solution was analyzed for complexes halted in the registers 11-58. Figure 5 shows that halted complexes are essentially stable, with the exception of registers 16/18 and 26/27. The latter are prone to dead-end complex formation, a finding that is in line with published results (11,12,26). But dead-end complex formation can be largely suppressed even in these registers by lowering the temperature and by omitting Mg++ (26; and data not shown).
Figure 5. Rate of dead-end complex formation of halted complexes. Complexes were halted at the registers as indicated. The inactivation rate per minute was determined by time-dependent activity studies as described in Materials and Methods.
In order to prove conclusively that no inactivation occurred during our 4 h electrophoresis, the transcription competence of the complexes halted in the registers 11-20 and 48-58 was analyzed in the gel after electrophoresis, as described in Materials and Methods. Figure 6A and B shows that all complexes resumed synthesis after electrophoresis. This finding confirms that under the applied conditions transcription competence of complexes is not affected, even if electrophoresis lasts several hours.
Figure 6. Transcription competence of halted complexes after non-denaturing gel electrophoresis. Complexes halted at the registers as indicated were subjected to gel electrophoresis for 3 h as described in Materials and Methods. The complexes containing radioactively labeled RNA were cut out and the gel slices were merged into buffer containing MgCl2 and NTP, permitting limited synthesis. The RNA was eluted from the slices, precipitated and applied on a PAGE in urea [a 20% polyacrylamide agarose gel (A) and a 12% polyacrylamide agarose gel (B)].
Dead-end complex formation is accompanied by relaxation of the polymerase-induced DNA bending
A comparison of the electrophoretic mobility of dead-end complexes and transcription competent complexes in a non-denaturing gel is shown in Figure 7. Complexes in the registers 26 and 27 were chosen, since they are especially prone to dead-end complex formation (Fig. 5). Figure 7A (upper panel) shows that the 27mer complex is converted within 15 min into a faster-moving complex. `Self cleavage' by RNA polymerase can be excluded as an explanation of this mobility change, since the RNA remains intact, as gel electrophoresis of the complex under denaturing conditions revealed (Fig. 7A, lower panel). Figure 7B shows that the 26mer complex as well as the 27mer complex isomerizes after 15 min into a second, faster moving form. The RNA extension assay (Fig. 7B, lower panel) shows that only the upper band contains a transcription-competent complex that facilitates elongation of RNA. We conclude that the lower band containing the fast moving species represents the dead-end complex. The increase in the mobility indicates that the DNA conformation is converted into a more extended conformation for the reasons pointed out above. We conclude that the DNA in the dead-end complex, at least in the 26mer/27mer complex, is in a more extended conformation than in the transcription-competent complex.
Figure 7. Electrophoretic mobility change upon dead-end complex formation. The 27mer complex obtained by elongating the 25mer complex was incubated for 0, 2, 5 or 15 min at 37°C and subsequently subjected to non-denaturing gel electrophoresis (A), upper panel. For comparison purposes, the mobilities of the complexes halted in the registers 20, 23 and 25 applied immediately after halt of synthesis are shown. The lower panel in (A) shows the corresponding RNA products. (B) The upper panel shows the electrophoretic pattern of complexes halted in the registers 26 and 27. They were applied either immediately (0) or after 15 min incubation at 37°C (15). The lower panel in (B) shows the RNA products obtained without addition of nucleotides (-) or with nucleotides (+) added to gel slices after the end of the electrophoresis as described in Materials and Methods. The lower and the upper bands of the complexes halted in registers 26 and 27 were analyzed. For comparison purposes, the complexes in the registers 20, 23, 25 and 34 are also shown.
DISCUSSION
Our study shows that RNA polymerase induces changes of the DNA bending angle that varies periodically with progress of RNA synthesis. This finding was obtained from gel retardation studies on halted complexes.
Complexes in the registers 16, 18 and 26, 27 show the most pronounced tendency of dead-end-complex formation with a rate of 15 and 20%/min respectively (Fig. 5) at 37°C. We could reduce this reaction to <3%/h, probably due to the lower temperature and the absence of Mg++ during electrophoresis, which enhances the stability of the complexes. The halted complexes remain transcription-competent during this procedure, as verified by transcription activity studies in the gel after electrophoresis.
We conclude from these studies that non-denaturing gel electrophoresis can be used to obtain information about the conformation of halted complexes without the danger of impairment by dead-end complex formation.
Structural variations of halted complexes: changes of the DNA bending angle
The electrophoretic mobility differences in halted complexes suggest structural variations of the complexes. The mobility differences can be attributed to RNA polymerase-induced changes of the DNA bending angle, as mobility studies on complexes with different DNA fragments have revealed.
That RNA polymerase can induce bending of the DNA in binary as well as ternary complexes was demonstrated previously by applying different techniques, such as gel electrophoresis (24,25,29,34), neutron solution scattering (35) electro-optical birefringence (36) and atomic force microscopy (37).
Sequence-dependent effects on the bending pattern
While RNA polymerase-induced bending of DNA is well established, the mechanism by which the bending angle is changed is unclear. That the sequence affects the bending behavior, is obvious from the comparison of the bending patterns of fragment A1-340-1 and fragment A1-340-1M. They differ by the sequence downstream of position +46. The bending pattern changes with respect to the phasing and especially with respect to the amplitude of the bending angle. It is interesting to note that the bending pattern differs already 18 3 registers before the polymerization active site has reached the changed sequence. Interestingly, the region 18 3 base positions downstream of the 3[prime] position of the nascent RNA coincides roughly with the downstream edge of the footprint (9,16). This coincidence suggests that the sequences flanking RNA polymerase, at least at the downstream border of the polymerase-DNA interaction region, affect the bending behavior.
Origin of the changes of the bending angle
The registers 11, 20, 30, 40, 50 and 60 of the wild-type sequence and 11, 20, 32, 42, 47 and 54 of the modified sequence are the registers where the DNA has the most extended conformation. Between these registers, DNA adopts a bent conformation. Thus, the bending angle changes its value periodically if synthesis proceeds.
It is unlikely that these changes of the bending angle can be attributed solely to the sequence context in which an RNA polymerase is located. Correlation analysis of sequence and bending pattern suggests that there is no interdependence between both patterns. However, this does not exclude the possibility that the sequence modulates the bending pattern, especially with respect to the amplitude.
It is widely acknowledged that halted complexes can be subject to conformational changes (11-13). This could also apply to the bending. Is the change of the bending angle an intrinsic feature of the translocation process occurring immediately after nucleoside incorporation or is it a process occurring after halt of synthesis? The gel electrophoresis studies do not answer this question, since a run requires several hours. Despite this uncertainty, we can conclude that the observed bending pattern reflects differences in the conformation of the ternary complexes.
We can only speculate about the origin of the variation in the bending angle. One possibility could be a register-dependent variation in the transcription bubble that affects the bending angle. Modeling studies suggest that the DNA is bent by at least 20° in order to accommodate all three nucleic acid components (16) of the classical transcription bubble, template strand, non-template strand and about nine bases of the RNA hybridized to the template strand (1). Further opening of the DNA, as observed in the registers 11-18 (9) may be partly compensated by an increase of the bending angle, as suggested by molecular modeling studies on the transcription bubble without RNA polymerase (16). Support for the view that the transcription bubble plays a functional role in the translocation process is the finding that the transcription bubble is enlarged between registers 11 and 18. The bubble collapses in the upstream region if synthesis proceeds from register 18 to 20, adopting the original size of 12 bp (9). Thus, the process of opening and closing of the transcription bubble runs parallel with increase and decrease of the bending angle (this study) at least in the registers 11-20. These results show a direct relationship between bending and size of the transcription bubble during progress of RNA synthesis. We suggest that the size and geometry of the transcription bubble depends not only on the progress of RNA synthesis but is also modulated by the sequence. This might explain why the two different templates show differences in amplitude and phasing.
The hypothesis that the periodic change of the bending angle is connected with the transcription mechanism is further supported by the finding that the transition of the halted complex from a transcription-competent complex to a dead-end complex is connected with a change in the bending angle. The DNA enters a more stretched configuration upon dead-end complex formation, perhaps by breaking contacts of RNA polymerase to one of two previously suggested DNA binding sites (16).
ACKNOWLEDGEMENTS
We thank Dr R. Landick for providing the His-RNA polymerase strain. This work was supported by DFG grant He 1285/7-3 and INTAS-RFFI grant (#95-1150).
REFERENCES
*To whom correspondence should be addressed. Tel: +49 89 8578 2216; Fax: +49 89 8578 2822; Email: heumann{at}biochem.mpg.de
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  N. R. Forde, D. Izhaky, G. R. Woodcock, G. J. L. Wuite, and C. Bustamante Using mechanical force to probe the mechanism of pausing and arrest during continuous elongation by Escherichia coli RNA polymerase PNAS, September 3, 2002; 99(18): 11682 - 11687. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||