HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 31, 29104-29110, August 3, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/31/29104    most recent M103704200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Walter, W. Articles by Studitsky, V. M. Search for Related Content PubMed PubMed Citation Articles by Walter, W. Articles by Studitsky, V. M.

Facilitated Transcription through the Nucleosome at High Ionic Strength Occurs via a Histone Octamer Transfer Mechanism^*

Wendy Walter and Vasily M. Studitsky

From the Department of Biochemistry and Molecular Biology and Center for Molecular Medicine and Genetics, Wayne State University School of Medicine, Detroit, Michigan 48201

Received for publication, April 25, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The rate of transcription through the nucleosome, the fine structure of the nucleosomal barrier, and the fate of the nucleosome during transcription at different salt concentrations were analyzed using linear 227-base pair mononucleosomal templates containing a uniquely positioned nucleosome core. At lower ionic strength (30 mM NaCl), the nucleosome constitutes a strong barrier for SP6 RNA polymerase. At higher ionic strength (330 mM NaCl), the rates of transcription on nucleosomal and histone-free DNA templates are very similar. At both higher and lower ionic strengths, the complete histone octamer is transferred over the same distance by fundamentally similar mechanisms. The data indicate that even at the rate of transcription characteristic of histone-free DNA, the transfer intermediates can be formed quite efficiently. This suggests possible mechanisms that could facilitate transcription through the nucleosome at physiological ionic strength.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The most basic unit of DNA packaging in eukaryotes is the nucleosome. A nucleosome core particle consists of 146 bp¹ of DNA wrapped around a histone octamer. This octamer is composed of two units each of histones H2A, H2B, H3, and H4 (1). Where the DNA exits and enters the nucleosome core particle, one molecule of histone H1 binds to and stabilizes the nucleosome (2). Actively transcribed genes retain their nucleosomal packaging (see Refs. 3 and 4), but there is evidence to suggest that these genes are deficient in the levels of histone H1 (5, 6). Despite the obvious sterical problems of such compaction for RNA polymerases, transcription still occurs. The nature of this nucleosomal barrier to transcription has been the subject of intense investigation.

In vivo the rate of transcription of nucleosome-covered genes by RNA polymerase II is comparable with that of histone-free DNA in vitro (7, 8). The rate of transcription of nucleosomal templates in vitro, however, is much slower, suggesting that there must be mechanisms in vivo to allow the polymerase to overcome the nucleosomal barrier (9). It has been known for many years now that the rate of transcription by E. coli RNA polymerase as well as RNA polymerases I and II on chromatin templates in vitro increases at elevated ionic strength (10-13). Increasing the salt concentration to 0.3-0.5 M NaCl (0.15 M NaCl is considered to be physiological ionic strength) during transcription allows the rate of transcription of chromatin templates to reach those of naked DNA (11, 12), thus recapitulating the highly efficient transcription of chromatin templates in vivo. The mechanism of such efficient transcription through the nucleosome can be analyzed using this experimental approach. Chambon and co-workers (13) demonstrate that even at 0.5 M NaCl nucleosomes still remain associated with the DNA during transcription. There is no accumulation of nucleosomes in front of the polymerase, indicating that the polymerase does not just push the octamer forward (13). Furthermore, it has been shown that the nucleosomes seem to move relative to their initial positions on the DNA during transcription (14). Here we analyze the mechanism mediating this nucleosome mobility.

Previously, we have shown that transcription of mononucleosomal templates by SP6 RNA polymerase results in the transfer of the histone octamer from the promoter-distal end of the template to the promoter-proximal end by the "spooling" mechanism (15). Transfer occurs directly, without even transient release of the histone octamer into solution, via the formation of an intranucleosomal loop (15). This loop is formed and broken several times during transcription through the nucleosome, and the intermediates formed in the process constitute the barrier to transcription (16, 17). We have also established that this same mechanism is employed by the eukaryotic yeast RNA polymerase III (18), suggesting that the mechanistic studies of transcription through the nucleosome can be accomplished using the bacteriophage polymerase.

In this study, we used SP6 RNA polymerase to transcribe mononucleosomal templates at lower and higher ionic strength in order to determine whether or not the spooling model still applies under conditions where the nucleosome becomes virtually invisible to the RNA polymerase. The relative rates of transcription were analyzed in reactions containing increasing concentrations of salt for both DNA and nucleosomal templates, and transcribed nucleosomal templates were analyzed for nucleosome displacement at both higher and lower ionic strength. We found that the amount of nucleosomal pausing decreases as the salt concentration is increased. As a result, there was an increased rate of transcription through the nucleosome with the increase in ionic strength. Furthermore, efficient histone octamer transfer was observed, even at higher ionic strength. The data indicate that the mechanism of transcription through the nucleosome is still the same under conditions where the nucleosome does not constitute as strong of a barrier to the polymerase.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Fragments and Plasmids-- pB22 was digested with SacI and NcoI, and the 227-bp fragment was gel-purified by electroelution from an agarose gel as described (15). This fragment was then end-labeled at the NcoI site using Klenow fragment. The plasmid DNA used as competitor was pBS1100 (15).

Nucleosomal Templates-- H2A/H2B dimers and H3/H4 tetramers of core chicken erythrocytes histones were purified as described (see Fig. 1B; Ref. 19). Nucleosomal cores were formed on end-labeled pB22-227-bp fragments using purified chicken erythrocyte core histones at a ratio of ~0.8:1 (weight to weight) of histones to DNA in a slightly modified version of the protocol used by the Bradbury laboratory (20, 21). Briefly, the histones and DNA (8 µg and 10 µg, respectively) were mixed in a final volume of 200 µl in buffer containing 10 mM Tris-HCl, pH 7.4, 0.2 mM EDTA, 0.1% IGEPAL, and 2 M NaCl. Dialysis was performed in 1-h steps against the same buffer but containing decreasing amounts of NaCl (2, 1.5, 1, 0.75, and 0.5 M). A final overnight dialysis step was done against the same buffer with only 10 mM NaCl and without IGEPAL. The reconstitute was digested with HaeII, and properly positioned nucleosomes were resolved by preparative native gel electrophoresis and purified as previously described (16).

In Vitro Transcription-- SP6 RNA polymerase was purified as described (22). The final enzyme preparation had a specific activity of 6 × 10⁶ units/mg of protein and was >95% pure as judged by protein electrophoresis (Fig. 1C). Stalled -C elongation complexes were formed by incubation of 5-20 ng of templates (cores or histone-free DNA isolated from the same native gel) in 10-60 µl of transcription buffer (45 mM HEPES, pH 8.0, 6 mM MgCl₂, and 2 mM spermidine; the ionic strength of the transcription buffer was experimentally determined to be equivalent to 30 mM NaCl based on conductivity) containing 0.5 mM ATP, GTP, UTP (in the case of experiments analyzing the labeled template) or 0.5 mM ATP and GTP and 5 µM UTP plus 10 µCi of [-³²P]UTP at 3000 Ci/mmol (PerkinElmer Life Sciences) (in the case of experiments analyzing the labeled transcripts), 1-5 units of RNase inhibitor (Roche Molecular Biochemicals), and 5-40 units of SP6 RNA polymerase for 10 min at room temperature. In control reactions, SP6 RNA polymerase was omitted or terminating nucleotide (3'-dATP; "mock" reactions, final concentration 0.5 mM) were present, as described under "Results." Then samples were transferred to an ice bath for at least 10 min. For multiple reactions, -C stalled complexes were formed in a single tube and then separated into aliquots (20 µl). After formation of the -C stalled complexes, ice-cold competitor DNA (final concentration 0.05 mg/ml) or NaCl (final concentration 300 mM, added dropwise from 1.5 M stock solution with immediate mixing) were added to some reactions, as described under "Results." To complete transcription of the templates, the reactions were supplemented with all four NTPs (final concentration 0.5 mM) and incubated on ice for 10 min or other time points indicated in the figures. In the case where labeled transcripts were analyzed, transcription was terminated by the addition of EDTA (to a final concentration of 10 mM). Samples were precipitated and analyzed by denaturing polyacrylamide gel electrophoresis as described (15). In experiments where the labeled template was analyzed, transcription was terminated by adding 3'-dATP to a final concentration of 0.5 mM. Reactions were diluted so that the higher ionic strength would not interfere with restriction enzyme digestion, and samples were incubated at 20 °C for 30 min in the presence or absence of HaeII restriction enzyme. Digestions were terminated with EDTA (final concentration 10 mM), and samples were analyzed in a native gel as described (15). Gels were dried and quantitated using Cyclone (Packard).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Kinetics of Transcription through the Nucleosome at Different Ionic Strengths-- Kinetic analysis of transcription through the nucleosome by SP6 RNA polymerase at lower (30 mM NaCl), moderate (130 mM NaCl), and higher (330 mM NaCl) ionic strengths was performed using the 227-bp template that was characterized previously (Fig. 1A; (15, 16)). This template contains an SP6 promoter and a C-less cassette corresponding to the first 14 nucleotides in the transcript. Transcription was initiated at 30 mM ionic strength at room temperature with an NTP mixture consisting of 0.5 mM ATP and GTP and 5 µM [-³²P]UTP. The polymerase stalls at the position where it needs to incorporate CTP (nucleotide 14), and transcription is synchronized on all of the templates. Transcription is resumed at 0 °C by the addition of all four unlabeled NTPs at a final concentration of 0.5 mM each in buffer containing the appropriate salt concentration. Thus, the RNA is labeled only in the first 14 nucleotides, and all of the transcripts in the reaction mix are comparable quantitatively, regardless of their length. Transcription was conducted at 0 °C because the rate of transcription at higher temperatures is too fast to allow reliable analysis of the time course (16).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Experimental system for analysis of the mechanism of transcription through the nucleosome at different salt concentrations. A, positions of nucleosomes on the 227-bp template before and after transcription at 30 mM ionic strength (15). The template was labeled at the NcoI end. The transcription start site is indicated with an arrow and is located 33 bp from the SacI end. There is a -CTP 14-mer tract immediately downstream of the transcription start site. The run-off transcript is 194 nt long. The position of the unique site for HaeII restriction enzyme is indicated. Panels B and C show isolated chicken erythrocytes core histones (H2A/H2B and H3/H4) and bacteriophage SP6 RNA polymerase, respectively, in SDS-polyacrylamide electrophoresis gels. +H1 Chromatin, chicken erythrocyte H1-containing chromatin (42). M, labeled RNA markers.

The time courses of transcription on the DNA and nucleosomal template at different ionic strengths are shown in Fig. 2. As described previously (16), at lower (30 mM NaCl) ionic strength the nucleosome is a strong barrier to the transcribing polymerase. The rates of transcription on the nucleosomal template are much (about five times) slower than those of the naked DNA templates (Fig. 2, A and B; see (16)). At the 25-s time point, transcription on the DNA template is complete (with run-off transcripts 194 nucleotides in length), whereas the same time interval for the nucleosomal template yields relatively few run-off transcripts. Most of the polymerase is still paused within the nucleosome. The pausing by polymerase is so severe that many of the transcripts were not completed even at the longest point in the time course (10 min). Experiments with the detergent Sarcosyl, which removes the histones from the DNA but leaves elongation complexes intact and able to complete nascent transcripts, reveal that the polymerase is indeed paused and not in an arrested state or dissociated from the template. After removal of the histones during early time points, the polymerase is able to resume transcription, and the majority of the shorter transcripts are converted to the full-length run-off (Ref. 16 and data not shown).

View larger version (60K): [in this window] [in a new window]   Fig. 2.   Nucleosome-induced pausing of SP6 RNA polymerase is relieved at higher ionic strength. Analysis of pulse-labeled transcripts by denaturing polyacrylamide gel electrophoresis. Time courses of transcript elongation at different ionic strengths (30, 130, and 330 mM NaCl) on histone-free DNA and nucleosome cores are shown. Templates were transcribed for 4, 10, 25, 60, 180, or 600 s at 0 °C after the formation of the stalled complexes (-C) in the presence of labeled UTP. Mobility of labeled DNA used for normalization of loading is indicated. M, labeled RNA markers (sizes are indicated at the left). The position of the nucleosome is shown on the right, with the shading indicating the region of strong nucleosome-specific pausing and a horizontal line demonstrating the location of the dyad axis.

The rate of transcription of the free DNA slows down with the increasing salt concentration (Fig. 2C). This effect is minimal as the ionic strength is increased from 30 mM NaCl to 130 mM NaCl but becomes increasingly evident when the salt concentration is raised to 330 mM. This result is especially obvious in the 10-s time points. At 10 s in 30 mM NaCl, some of the transcripts are full-length, whereas at 130 mM NaCl, the longest transcripts are not quite full-length. At 330 mM NaCl, the transcripts are only about half their run-off length.

Despite the decrease in the rate of transcription on the DNA templates as the salt concentration increases, the rate of transcription on the nucleosomal templates increases with the increase in ionic strength. Although this effect is weak from 30 to 130 mM NaCl (Fig. 2, A and B), the amount of run-off transcript generated from the nucleosomal templates at 330 mM NaCl in 1 min (Fig. 2C) is much higher than that of the corresponding 30 and 130 mM NaCl points. The nucleosomal pausing at 330 mM NaCl is dramatically decreased, allowing the conversion into run-off transcript. In fact, at 330 mM NaCl, the rates of transcription on the nucleosomal and free DNA templates are very similar. However the residual nucleosome-specific pausing pattern at this ionic strength is very similar to the pausing pattern seen for transcription at lower concentrations of salt. This suggests that the mechanism for transcription through the nucleosome at all of the ionic strengths tested may essentially be the same.

We have shown previously that the polymerase pausing is due to difficulty in transcribing the DNA in the "pausing intermediates," and once the octamer has been transferred, there is no additional barrier for the polymerase (16, 17). The increase in ionic strength during transcription led to a decrease in the amount of pausing by the polymerase, suggesting that the intermediates were destabilized or the octamer did not remain associated with the template. We tested whether the nucleosome survived transcription at 330 mM NaCl and whether transfer of the octamer still occurred under these conditions that caused the polymerase to pause less in the nucleosome.

The Octamer Remains Bound to the Template at Higher Ionic Strength-- The addition of salt to the transcription reaction weakens the electrostatic interactions between the histones and the DNA. Thus, there are a few possible explanations for why the polymerase can more readily transcribe the nucleosomal template at higher ionic strength. First, the nucleosome could be destabilized so much that the passage of the polymerase could cause the octamer to dissociate from the template. Second, the increase in ionic strength could destabilize histone-histone as well as histone-DNA interactions in the nucleosome. The passage of the polymerase could result in formation of a sub-nucleosomal structure (for example, histone hexamer) giving polymerase greater access to the DNA. Alternatively, the histone-DNA electrostatic interactions in the paused intermediates (intermediate 2, see Fig. 5) could be weakened enough to allow polymerase to transcribe at much higher rate, but the DNA-histone interactions in the histone octamer transfer intermediate (intranucleosomal DNA loop; intermediate 3, Fig. 5) are still strong enough to allow efficient octamer transfer. Therefore, the nucleosomal barrier would be effectively removed, and the nucleosome-specific pausing by the polymerase would be strongly diminished.

To discriminate between these possibilities, the fate of the nucleosome during transcription at lower and higher ionic strength was analyzed. As shown in Fig. 3A, the nucleosome remains bound to the template even at higher ionic strength. The nucleosomal templates were labeled at the promoter distal end and analyzed in a nucleoprotein gel before or after transcription. All transcription reactions were initiated at 30 mM NaCl at room temperature with ATP, GTP, and UTP. The polymerase transcribes to the end of the C-less cassette and becomes stalled at nucleotide 14. Transcription was resumed at 0 °C at 30 or 330 mM NaCl in the presence or absence of saturating levels of competitor DNA with the addition of all NTPs. Transcription initiation cannot occur at 0 °C, so only those templates that have active elongation complexes formed on them will be transcribed. In this single-round assay, about 30-50% of the templates are transcribed (17). The controls without polymerase and with polymerase but no CTP as well as mock transcription in the presence of a terminating analog of ATP (3'-ATP) show the mobility of the nontranscribed nucleosomal template in a native gel. These nontranscribed controls also reveal the amount of free DNA present in the nucleosome preparation before transcription (25-30% free DNA). The mock transcription controls were performed in the presence of competitor DNA, demonstrating that incubation with competitor DNA does not deplete the nucleosomes from the template, even at the higher ionic strength.

View larger version (45K): [in this window] [in a new window]   Fig. 3.   Nucleosome cores survive transcription at higher ionic strength. Analysis of nucleosome cores in a native gel before and after transcription. A, labeled cores were transcribed at 30 or 330 mM NaCl in the presence or in the absence of a terminating analog of ATP (3'-dA; MO, MOCK reactions) and/or competitor DNA for 10 min at 0 °C (single-round transcription) after the formation of stalled complexes in a -C reaction. B, alternatively, cores were incubated at 30 mM NaCl at 20 °C (multiple-round transcription) in the presence or in the absence of RNA polymerase and then digested with HaeII restriction enzyme, as indicated. Mobilities of nucleosome cores and histone-free DNA are indicated. M, end-labeled MspI digest of pBR322. Note: the elongation complexes formed in the -C reaction are not stable enough to survive gel electrophoresis (17). M, labeled RNA markers.

When the nucleosomal templates were transcribed at both 30 and 330 mM salt in the absence of competitor DNA, the nucleosomes remained bound to the template, and the mobilities of the nucleosomes in the native gel before and after transcription were very similar. This indicates that the nucleosome has either not moved or that the octamer has been transferred in cis to the promoter distal end, a symmetrical position on the template. The octamer itself has not been disrupted by transcription at the increased salt concentration because a histone hexamer or tetramer would have a different mobility in the native gel (16). There is a slight increase in the amount of free DNA present after transcription at 330 mM salt, indicating that a very small amount of dissociation has occurred. However, this slight increase in free DNA is also seen in the mock control, suggesting that this is probably not a transcription-dependent phenomenon but rather the effect of incubation in the presence of higher salt. The small amount of histone dissociation might occur due to a temporary increase of ionic strength during mixing of the high salt stock solution into the sample. Transcription in the presence of excess unlabeled competitor DNA gives rise to more free DNA at both lower and higher ionic strengths. This increase in the amount of histone-free DNA is entirely transcription-dependent, indicating that the histone octamer has been transferred to unlabeled competitor DNA (15). The release of labeled free DNA indicates that nucleosomal templates were efficiently transcribed. Moreover, the amount of labeled DNA released as a result of transcription also allows for an approximate estimation of the fraction of templates that were transcribed. In the presence of excess competitor DNA, ~70% of the octamers are transferred to the competitor DNA (15). In the experiments shown in Fig. 3, 30 and 40% of the templates were utilized in the single-round transcription assay at 330 and 30 mM ionic strengths, respectively. Slightly lower nucleosomal template utilization at 330 mM is probably related with partial transcription-independent DNA dissociation from the histone octamer occurring at the higher salt concentration. In summary, the data indicate that nucleosomes almost quantitatively survive transcription at the higher ionic strength.

The Octamer Is Transferred at Higher Ionic Strength-- The position of a nucleosome on a template can be determined based on its mobility in the native gel (20). However, this method cannot distinguish between two symmetrically equivalent nucleosome positions, such as those on the promoter-proximal end versus the promoter-distal end of the template. The way to distinguish between the two symmetrical positions is to employ the restriction enzyme sensitivity assay (15). If a nucleosome is positioned over a restriction enzyme site, the enzyme cannot cleave the DNA. We have shown previously that when this 227-bp template is transcribed by SP6 polymerase or RNA polymerase III at physiological ionic strength or lower, the octamer is transferred from the promoter-distal end of the template to the promoter-proximal end over a distance of about 80 bp (15-18). Our 227-bp template is labeled at the promoter-distal end and has a uniquely positioned nucleosome also located at this end. The nucleosome covers a HaeII site (Fig. 1A). The mobility of the nucleosomal template in a native gel is the same after transcription in 30 mM NaCl as it is without transcription (Fig. 3B). Before transcription, the template is resistant to cleavage by HaeII (as seen by the retention of label in the nucleosomal band rather than the HaeII fragment band on the native gel). After transcription, the octamer is transferred to the promoter end of the template, making the HaeII site accessible (as seen by the appearance of the label in the HaeII fragment band and the decrease in the amount of label in the nucleosomal band). Note that in Fig. 3B, transcription and restriction enzyme digestion are performed at 20 °C. Therefore, initiation can occur at any time, leading to a higher efficiency of template utilization as compared with reactions performed at 0 °C. A major fraction of the templates have been transcribed in this experiment. This level of template utilization cannot be achieved during transcription at 330 mM NaCl because the polymerase cannot initiate at this concentration of salt (data not shown).

To determine whether the octamer is transferred at higher ionic strength and to be able to directly compare the amount of transfer that occurs (if any) with that at lower ionic strength, transcription was limited to a single round of initiation (Fig. 4). Transcription of NcoI end-labeled nucleosomal templates was initiated at room temperature in lower salt in reactions lacking CTP and then resumed at 0 °C in the presence of 30 or 330 mM NaCl with the addition of all NTPs. Initiation cannot occur at 0 °C (data not shown), so the number of transcribed templates should be approximately equal at both high and low salt concentrations. Transcription was terminated with the addition 3'-dATP so that further initiation could not occur in the low salt reaction during incubation with restriction enzyme, and the samples were incubated in the presence or absence of HaeII at 20 °C for 30 min at room temperature. Controls without polymerase and without extension after the formation of early elongation complexes (-C) at 30 mM NaCl show how much free DNA is in the original nucleosome preparation. Mock transcription controls were performed at both 30 and 330 mM NaCl by incubation in the presence of 3'-dATP after the formation of early elongation complexes to prevent further transcription. In these controls, there is 25-30% free DNA. The slight increase in the amount of free DNA at the higher salt concentration is not transcription-dependent. The HaeII digest of the sample before transcription reveals that the amount of label in the HaeII fragment band is equal to the amount of label in the free DNA band, demonstrating that the nucleosome is at the NcoI end of the template and resistant to cleavage by HaeII.

View larger version (35K): [in this window] [in a new window]   Fig. 4.   Nucleosome cores are transferred during transcription at higher ionic strength. Schematic of the experimental design is shown at the top (see "Results" for detail). Labeled cores were transcribed at 30 or 330 mM NaCl in the presence or absence of a terminating analog of ATP (3'dA; MO, MOCK reactions) for 10 min at 0 °C (single-round transcription) in the presence or in the absence of excess of competitor DNA after the formation of stalled complexes in a -C reaction. Then transcription was terminated by the addition of excess of 3'-dA, and cores were digested with HaeII restriction enzyme, as indicated. Analysis of nucleosome cores in a native gel before and after single-round transcription using the restriction enzyme protection assay is shown. M, labeled RNA markers; Pol, the absence of polymerase.

As shown in Fig. 4, transcription at both 30 and 330 mM NaCl yields nucleosomes that have the same mobility after transcription as before transcription. This indicates that the nucleosomes are located on the end of the template. The histone octamer is not disrupted during transcription at higher ionic strength because sub-nucleosomal structures would have a different mobility in the native gel (15). There is very little or no free DNA released as a result of transcription, so the histone octamers remain bound to the template during transcription. Digestion with HaeII reveals that some of the octamers have been transferred to the promoter end of the template. The amount of label in the HaeII band after transcription is greater than the amount of label in the free DNA band after transcription, indicating that the nucleosomes have been translocated away from the HaeII site, making this site accessible to the enzyme. The amount of transfer that occurs at 330 mM NaCl and 30 mM NaCl is similar, indicating that transfer can still occur very efficiently when the polymerase transcribes at a more rapid pace through the nucleosome.

Since octamer transfer still occurs during transcription at higher ionic strength but polymerase pausing through the nucleosome is diminished, it is likely that the pausing intermediates are destabilized during transcription at higher ionic strength. The mechanism of octamer transfer is just as efficient when the intermediates that cause polymerase to pause during transcription are destabilized and the nucleosome is almost invisible to the polymerase.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In summary, we have shown that an increase in ionic strength progressively weakens the nucleosomal barrier to transcription (Fig. 2). Although the pausing through the nucleosome is greatly diminished at 330 mM NaCl, the residual pausing pattern at higher ionic strength is very similar to the pattern seen at physiological or lower ionic strength. Thus, the fine structure of the barrier (nucleosome-specific patterns of pausing) remains principally unchanged, even when the barrier is almost undetectable. Since the polymerase pausing pattern reflects the pathway of nucleosome transfer during transcription (16, 17), this suggests that the overall mechanism of nucleosome transfer remains unchanged. Moreover, direct analysis of the fate of the nucleosome during transcription (Figs. 3 and 4) indicates that at lower (30 mM NaCl) and higher (330 mM NaCl) ionic strengths the intact histone octamer is transferred over very similar distances to the promoter-proximal end of the DNA with equal efficiency. Therefore, an increase in ionic strength results in very efficient transcription through the nucleosome accompanied by transfer of the histone octamer.

Since the overall mechanism of transcription through the nucleosome is preserved at higher ionic strength, similar intermediates must be formed there. It has been shown previously that the nucleosomal barrier to transcription is formed by the intermediates that are produced when the intranucleosomal loop breaks behind the polymerase before transfer of the octamer is completed (Fig. 5 and see Ref.). They are formed when the DNA, which was partially displaced, is rewrapped along with the polymerase onto the surface of the octamer. The loop collapses in such a way that there are a limited number histone-DNA interactions surrounding the polymerase (40-60 bp in front of and behind the polymerase; Fig. 5, intermediate 2) and presumably interfering with polymerase movement along the DNA. One of these interactions has to be broken to allow further progression of the polymerase. Before histone octamer transfer is completed, the interactions behind the polymerase are preferentially broken (17). The amount of DNA-histone contacts behind the polymerase in the intermediates (~40 bp; intermediate 2) is much less than in the complete nucleosome core (~150 bp DNA; Ref. 1). This may explain why a moderate increase in the salt concentration preferentially destabilizes these intermediates but not the complete nucleosome core, which remains intact even at the highest ionic strength used in this work (23). After DNA-histone interactions behind the polymerase are broken, the polymerase escapes from the pause, and transfer intermediates (intermediate 3) are formed again. The diminished pausing of polymerase through the nucleosome at 330 mM NaCl suggests that pausing intermediates which interfere with polymerase progression (intermediate 2) are destabilized to allow for more efficient transcription. These intermediates may be weakened enough to allow the polymerase to more easily escape, or the intranucleosomal loop itself might not collapse as readily under these conditions. Surprisingly, the higher ionic strength does not seem to affect the stability of the transfer intermediates, despite the fact that they consist of a relatively small number of DNA-histone interactions (as compared with the complete nucleosome core). Our results demonstrate that the transfer intermediates are stable enough at higher ionic strength to support efficient translocation of the nucleosome. This suggests that the number of DNA-histone contacts involved in intranucleosomal loop formation could be greater than the number of contacts in the pausing intermediates, thus allowing the transfer intermediate to be more stable at higher ionic strength than the pausing intermediates. Alternatively, the small number of DNA-histone contacts remaining in the transfer intermediate at higher ionic strength could be sufficient for efficient nucleosome translocation. After polymerase transcribes ~60 bp into the nucleosome, DNA-histone interactions in front of the enzyme are preferentially broken, completing nucleosome translocation (Fig. 5, intermediate 4).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Mechanism facilitating transcription through the nucleosome at higher ionic strength. After RNA polymerase (RP) initiates transcription at the promoter (P) on the 227-bp template (1), the polymerase enters the nucleosome core and induces formation of the paused intermediates (2). In the paused intermediates, RNA polymerase is surrounded by a relatively small number of DNA-histone interactions extending over 40-60 bp of DNA (17). In contrast, DNA-histone interactions in the complete nucleosome core occupy ~150 bp of DNA (1). Therefore, a moderate increase in ionic strength that has almost no effect on the stability of nucleosome cores can dramatically destabilize the paused intermediates and facilitate transcription through the nucleosome. An intranucleosomal DNA loop is then formed (3), and the octamer is transferred (4) as described previously (17).

In the cell, RNA polymerases must remain processive in the context of chromatin. RNA polymerase II can transcribe chromatin in the cell at a rate of 25 nucleotides/s (7, 8), but the nucleosome and even the H3/H4 tetramer are absolute blocks to polymerase II transcription in highly purified in vitro systems (9, 24). Only naked DNA can be transcribed at a physiological elongation rate in vitro, and this only occurs if the reaction is supplemented with elongation factors (reviewed in Refs. 25-27). The same elongation factors have no effect on the rate of transcription through nucleosomes, indicating that there are other mechanisms in vivo that help the polymerases transcribe through chromatin (reviewed in Refs. 25-27). By increasing the salt concentration of the transcription reactions, we have demonstrated that mononucleosomes can be transcribed at about the same rate as the histone-free DNA templates. Thus, the nucleosome is no longer a strong barrier to the transcribing polymerase under these conditions. Even at a higher rate of transcription, the entire octamer is transferred, suggesting that our model for the mechanism of transcription through the nucleosome is still likely to be accurate at physiological elongation rates. This raises a possibility that the in vivo mechanism for transcription through a nucleosome involves factors that can destabilize the electrostatic interactions between the histones and the DNA, allowing polymerase to overcome the barrier.

Destabilization of the arrested intermediates in vivo can be accomplished through several possible means. One mechanism is through the neutralization of the positively charged N-terminal tails of the histones via acetylation, a mechanism that is probably equivalent to the addition of salt to transcription reactions in vitro. The histone-DNA interactions are weakened by such modifications and should presumably have a positive effect on the rate of elongation through the nucleosome. Histone acetyltransferases are known to be associated with elongating RNA polymerases (28, 29), indicating that acetylation can play an important role in elongation. Recently, Widom and co-workers (30) have shown that acetylation of the histone tails increases the rate of transcription on a mononucleosomal template with T7 RNA polymerase. Although this group did not directly determine whether the acetylated octamer was transferred, our results strongly suggest that transfer would occur under these conditions. The facilitation of polymerase progression on these templates is likely due to disruption of the intermediates during transcription through the nucleosome, similar to our results with increased ionic strength.

Another possible mechanism for overcoming the nucleosomal barrier in vivo is through protein factors that destabilize the nucleosome itself or the intermediates formed during transcription through the nucleosome. Recently, an elongation factor called FACT has been shown to stimulate transcription elongation by RNA polymerase II on chromatin templates (31, 32). The chromatin remodeling complex, SWI/SNF, is also able to facilitate elongation through nucleosomes (33). SWI/SNF promotes nucleosome mobility through the disruption of DNA-histone contacts (34, 35), and this unwrapping of the DNA from the octamer could help RNA polymerases overcome the nucleosomal barrier. RSC, a related remodeling factor, has been shown to mediate octamer transfer (36) and could also facilitate progression of RNA polymerase through nucleosomes. Similarly, the HMG14 protein has been shown to stimulate transcription through chromatin (37). Spt4, Spt5, and Spt6 proteins may also play a role in the destabilization of transcription intermediates in chromatin. Both genetic studies and biochemical analysis have shown that Spt4, Spt5, and Spt6 effect elongation (38, 39). Recent localization of these factors in association with elongation-competent, phosphorylated RNA polymerase II and throughout the entire genes in transcriptionally active sites on Drosophila polytene chromosomes also supports the idea that Spt4, Spt5, and Spt6 play roles in elongation through chromatin (40, 41).

In summary, we have shown that higher ionic strength destabilizes the intermediates that cause RNA polymerases to pause and preserves the pathway that leads to octamer transfer. The polymerase is able to transcribe the nucleosome and free DNA templates at nearly equivalent rates, thus recapitulating the highly efficient transcription of chromatin in vivo. Modifications that moderately destabilize the histone-DNA interactions can dramatically stimulate transcription through the nucleosome and allow for efficient octamer transfer. Such destabilization of the intermediates can be accomplished in vivo by a number of possible candidates.

	ACKNOWLEDGEMENTS

We thank Drs. D. Clark, M. Kashlev, M. Kireeva, and L. Lutter for valuable discussions and comments on the manuscript.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant GM58650 (to V. M. S.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology and Center for Molecular Medicine and Genetics, Wayne State University School of Medicine, 540 E. Canfield Ave., Rm. 5123, Detroit, MI 48201. Tel.: 313-993-7818; Fax: 313-577-2765; E-mail: vstudit@wayne.med.edu.

Published, JBC Papers in Press, June 4, 2001, DOI 10.1074/jbc.M103704200

	ABBREVIATIONS

The abbreviation used is: bp, base pair(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

This article has been cited by other articles:

				R. J. Sims III, R. Belotserkovskaya, and D. Reinberg Elongation by RNA polymerase II: the short and long of it Genes & Dev., October 15, 2004; 18(20): 2437 - 2468. [Abstract] [Full Text] [PDF]

				Y. Xie and J. N. Reeve Transcription by an Archaeal RNA Polymerase Is Slowed but Not Blocked by an Archaeal Nucleosome J. Bacteriol., June 1, 2004; 186(11): 3492 - 3498. [Abstract] [Full Text] [PDF]

				W. Walter, M. L. Kireeva, V. M. Studitsky, and M. Kashlev Bacterial Polymerase and Yeast Polymerase II Use Similar Mechanisms for Transcription through Nucleosomes J. Biol. Chem., September 19, 2003; 278(38): 36148 - 36156. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/31/29104    most recent M103704200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Walter, W. Articles by Studitsky, V. M. Search for Related Content PubMed PubMed Citation Articles by Walter, W. Articles by Studitsky, V. M.