2706–2714 Nucleic Acids Research, 2001, Vol. 29, No. 13
© 2001 Oxford University Press
Analysis of the open region of RNA polymerase II
transcription complexes in the early phase of
elongation
Ulrike Fiedler and H. Th. Marc Timmers*
Department of Physiological Chemistry, University Medical Center Utrecht, Stratenum Building, Universiteitsweg 100,
3584 CG Utrecht, The Netherlands
Received March 28, 2001; Revised and Accepted May 10, 2001
ABSTRACT
The RNA polymerase II (pol II) transcription complex
undergoes a structural transition around registers
20–25, as indicated by ExoIII footprinting analyses.
We have employed a highly purified system to
prepare pol II complexes stalled at very precise
positions during the initial stage of transcript elongation. Using potassium permanganate we analyzed
the open region (‘transcription bubble’) of complexes
stalled between registers 15 and 35. We found that
from register 15 up to 25 the transcription bubble
expands concomitantly with RNA synthesis. At registers 26 and 27 the bubble has a high tendency to
retract at the leading edge. Addition of transcription
elongation factor TFIIS re-extends the bubble to the
stall site, resulting in complexes competent for transcript elongation. These findings are discussed in
the light of the recently determined structures for
RNA polymerases.
INTRODUCTION
DNA transcription by RNA polymerases proceeds through
several steps before reaching productive elongation: preinitiation complex assembly, promoter opening, transcription initiation and promoter escape. These steps are remarkably similar
in general outline between eukaryotic nuclear RNA polymerases and the bacterial enzyme (for reviews see 1–4). Prior
to transcription initiation a promoter segment of 10–12 bp
predominantly upstream of the start site converts to the singlestranded state. This marks the closed to open complex transition. Transcription complexes synthesize RNA in abortive
mode until register 9–12. After this, RNA synthesis continues
in productive elongation mode. The mechanistic similarities
most likely originate from structural similarities between the
transcription complexes. Indeed, recent crystallographic
studies showed that a 10 subunit RNA polymerase II (pol II)
from yeast (5) is topologically very similar to the three subunit
RNA polymerase (RNAP) from the bacterium Thermus aquaticus
(6). However, important differences also exist between transcription systems.
In bacteria one protein, the σ-factor, suffices to guide the
enzyme through the above-described steps, but the pol II
system requires at least 15 proteins distributed over five basal
transcription factors (for reviews see 7–9). TBP (TATA boxbinding protein, also designated TFIID), TFIIB and TFIIF
allow promoter recognition by pol II. TFIIE and TFIIH are
required for opening of pol II promoters. In contrast to other
systems, transcription initiation by pol II depends on a β–γ
bond hydrolyzable form of ATP (ATP or dATP) as a cofactor
(10). The ATP cofactor requirement is coupled to the transcriptional requirement for TFIIH (11,12). TFIIH consists of nine
subunits and is the only basal factor with enzymatic activities,
which depend on ATP hydrolysis (for reviews see 13,14).
These comprise two presumed DNA helicases (XPB and XPD)
and a protein kinase (cdk7/cyclin H). It has been shown that the
XPB subunit of TFIIH and ATP hydrolysis are essential for
open complex formation (15–18). Recent crosslinking data
suggest that XPB induces opening by twisting promoter DNA
(19) and argues against a classical DNA helicase model
(discussed in 4). Additionally, TFIIH, by virtue of its XPB
subunit, can stimulate the transition from abortive to productive transcription under certain conditions (20–22). The cdk7
subunit can phosphorylate the C-terminal domain (CTD) of the
largest subunit of pol II. It has been proposed that this modification triggers the transition from initiation to elongation
mode. However, transcriptional dependence on CTD kinase is
promoter-specific (23–26) and it depends on the composition
of the pol II transcription system (27).
Many protein–protein interactions are involved in establishing a functional pol II preinitiation complex. Transcription
elongation complexes lack basal factors (with the exception of
TFIIF, which is also an elongation factor). Therefore, these
contacts need to be rearranged or disrupted during initiation,
which may mark structural transitions within the transcription
complex. An early study indicated that TFIIB and TFIIE dissociate during abortive transcription (28). Using the adenovirus
major late promoter (AdMLP) it was noted that transcription
becomes independent of TFIIH and of ATP hydrolysis after
register 4, which constitutes the second transition (29,30).
Nevertheless, TFIIH is still associated at this stage, as it was
*To whom correspondence should be addressed. Tel: +31 30 253 8981; Fax: +31 30 253 9035; Email: [email protected]
Present address:
Ulrike Fiedler, Institute for Molecular Oncology, Breisacherstrasse 117, D-79106 Freiburg, Germany
Nucleic Acids Research, 2001, Vol. 29, No. 13 2707
found to dissociate between +30 and +68 (28). Transition from
abortive to productive transcription occurs around register 10,
with concomitant collapse of the upstream (–9/–2) open region
(30). This coincides with attaining a mature RNA–DNA
hybrid, which spans 8–9 bp. Within bacterial RNAP a rudderlike structure has been proposed to split off the 5′-end of the RNA
(6) and direct it to an RNA-binding site on the enzyme. This
binding site extends 10–20 nt from the catalytic site (discussed in
5,6). This implies that RNA emanates from the polymerase when
it is >20 nt long. Interestingly, Samkurashvili and Luse proposed
that pol II complexes pass through a major structural transition
around register 25 (31). They found that stalled complexes
with 20–25 nt RNAs have an increased tendency to slide back
and become arrested. Similarly, bacterial RNAP complexes
move discontinuously between register 23 and 27 and are
increasingly sensitive to transcript cleavage induced by GreB
(32).
Studies from bacterial RNAP have indicated that a mature
transcription bubble is 14–22 nt in size (33). Previously we
reported that productive transcription complexes at register 11
contain a bubble of 9–12 nt (30). Clearly, this is not a mature
sized transcription bubble. In this paper we address the fate of
the open region during the early phase of productive elongation by pol II. Employing a highly purified pol II system we
stalled transcription complexes by incorporation of an RNA
chain terminator between registers 15 and 35. Single-stranded
regions in promoter DNA were detected by potassium permanganate modification of DNA. We observe that the transcription
bubble of pol II complexes stalled at +26, +27 and +31 is
highly unstable. Collapse of the bubble at the leading edge
indicates back-tracking by pol II. Interestingly, a functional
transcription bubble can be rescued by elongation factor TFIIS.
MATERIALS AND METHODS
Materials
Nucleotides were obtained from Roche Molecular Biochemicals. Radioactive nucleotides were obtained from Amersham
or NEN/DuPont. Restriction enzymes, Klenow polymerase
and calf intestine phosphatase were purchased from Pharmacia. KMnO4 was from Jansen Chimica and piperidine from
Fluka. Oligonucleotides were from Pharmacia. The DNA
constructs were verified using the ABI Prism Big Dye Terminator Cycle Sequencing Ready Reaction Kit in a GeneAmp
PCR System 2400 and analyzed with an ABI Prism 310
Genetic Analyzer (all PE Applied Biosystems). The QiaQuick
gel extraction kit was purchased from Qiagen. Quantification
by phosphorimager analysis was performed on a Storm 820 gel
scanner from Molecular Dynamics using ImageQuaNT software v.4.2.
Purification of the transcription factors
Pol II was affinity purified from calf thymus using the 8WG16
antibody as described previously (34). Recombinant histidinetagged human TBP, recombinant human TFIIB and recombinant human TFIIE were expressed in Escherichia coli and
purified as described previously (12,16,34). Recombinant
δTFIIS (130–301) was a gift from Dr C. M. Kane (University
of California Berkeley). Recombinant TFIIF (human RAP74/
rat RAP30) was purified from baculovirus-infected insect cells
as described previously (16). TFIIH was purified from HeLa
cell extracts with the following modification of the protocol
described in Holstege et al. (16). The phenyl-Sepharose
column was replaced by an isopropyl-Sepharose column
(Pharmacia CL 4B, 100 × 16 mm). The column was developed
with a linear gradient from 1.2 to 0 M (NH4)2SO4 in T buffer
(20 mM Tris–HCl, pH 8.0, 20% glycerol, 1 mM EDTA, 1 mM
DTT) containing 50 mM KCl. TFIIH activity eluted between
800 and 700 mM (NH4)2SO4.
Preparations of TBP, TFIIB, TFIIE and TFIIF were homogeneous and TFIIH and pol II preparations were highly
purified as judged on silver stained protein gels. None of the
purified proteins contained nuclease, topoisomerase or other
basal factor activity.
Template DNA
The AdML promoter constructs were derived from pDNAdMLmut (16). The pDN-AdML+17G, +19G, +21G, +23G,
+25G, +26G, +27G, +28G, +29G, +31G, +33G and +35G
plasmids were constructed by inserting the appropriate PCR
fragment into EcoRI- and XbaI-digested plasmid pDNAdMLmut. PCR fragments were obtained using the
M13reverse primer (Stratagene), which anneals in the pDNAdMLmut plasmid upstream of the AdML promoter between
bases 108 and 128, and an oligonucleotide that anneals
between bases 30 and 50, introducing extra DNA sequence
directly upstream of the XbaI site. The XbaI site creates the
first G in the transcribed sequence. All constructs were verified
by DNA sequence analysis. For KMnO4 sensitivity assays
EcoRI- and HindIII-digested fragments were labeled at the
HindIII site using [α-32P]dATP and the Klenow fragment of
E.coli DNA polymerase I following standard procedures (35).
For transcription assays, promoter DNA fragments were
isolated from the appropriate pDN-AdMLmut derivative by
restriction enzyme digestion with PvuII. The fragments were
recovered using a QiaQuick gel extraction kit.
KMnO4 sensitivity assay
Radiolabeled probe (0.2–0.4 ng) was incubated at 30°C for
30 min with 25 ng TBP, 25 ng TFIIB, 80 ng TFIIF, 40 ng pol II,
30 ng TFIIE, saturating amounts of TFIIH and 12.5 ng δTFIIS
(where indicated) in a 20 µl reaction that contained 12 mM
HEPES–KOH, pH 7.9, 60 mM KCl, 5 mM MgCl2, 3 mM
(NH4)2SO4, 0.6 mM EDTA, 0.4 mM DTT and 90 µg/ml BSA.
Nucleotides were added for 10 min at the final concentrations
as indicated in the figure legends. The permanganate modification of the DNA fragments was restricted to 20 s but otherwise
performed as described previously (30).
Transcription assay
Template DNA fragments (5 ng PvuII fragments) were incubated at 30°C for 30 min with 25 ng TBP, 25 ng TFIIB, 80 ng
TFIIF, 40 ng pol II, 30 ng TFIIE, saturating amounts of TFIIH
and 12.5 ng δTFIIS (when indicated) in a 20 µl reaction that
contained 12 mM HEPES–KOH, pH 7.9, 60 mM KCl, 5 mM
MgCl2, 3 mM (NH4)2SO4, 0.6 mM EDTA, 0.7 mM DTT and
90 µg/ml BSA. After assembly, nucleotides were added to the
concentrations indicated in the figure legends and incubated at
30°C for the appropriate time as indicated. The reactions were
stopped and processed as described previously (36). Short
RNA products (up to 33 nt) were analyzed by electrophoresis
2708 Nucleic Acids Research, 2001, Vol. 29, No. 13
Figure 1. DNA sequence of the non-template strand of the pDN-AdML+nG
plasmids. The natural transcription start site (also depicted by the arrow) and
the first G residue in the RNA chain are indicated in bold. The templates are
identical outside the indicated region. Residues transcribed into RNA are
depicted as capitals. The asterisk identifies the position of the radiolabel in the
fragments used for footprinting.
on 21% polyacrylamide (acrylamide:bisacrylamide 19:2)–7 M
urea gels. Longer RNAs were analyzed on 7 or 10% polyacrylamide (acrylamide:bisacrylamide 19:1)–8.3 M urea gels.
RESULTS
Progression of the open region within pol II transcription
complexes (the ‘transcription bubble’) was analyzed in our
reconstituted basal transcription system. Previously we found
that in the early phase of initiation (RNA products up to 15 nt)
the leading edge of the bubble coincides with the site of NMP
addition (37). The trailing half of the bubble, the –9/–2 region
opened by TFIIH helicase, collapses when pol II reaches
register 11. The current study analyzes RNA product formation
and the transcription bubble of pol II complexes stalled
between registers 15 and 35. To this end, we constructed a set
of AdML promoter derivatives with their first G residue at +17
up to +35 (Fig. 1). As indicated in Figure 1, the DNA
sequences beyond the stall site are identical in all constructs.
This is important as recent structural studies indicate that 15 bp
of duplex DNA downstream of the active site are tightly
clamped by pol II (5). To prevent read-through we found that it
is essential to include the RNA chain terminator 3′-O-methylGTP (3′-OMeGTP) in the reactions. Simply omitting GTP
from the reactions results in a heterogeneous population of
RNAs extending well beyond the expected stall site (data not
shown). We believe that the longer RNAs are caused by pol II
read-through, as inclusion of elongation factor TFIIS reverses
these larger than expected RNAs (data not shown and see also
Fig. 7). This effect of TFIIS excludes the possibility that readthrough is caused by GTP contamination of the system. Previously, misincorporation had led to the false impression that
formation of the first phosphodiester bond results in a dramatic
expansion of the bubble (16,37). The templates were used in
reactions containing homogeneous preparations of recombinant TBP, TFIIB, TFIIE and TFIIF and highly purified HeLa
cell-derived TFIIH and calf thymus pol II. The DNA templates
were preincubated for 30 min with saturating amounts (except
for pol II) of these factors to allow preinitiation complex
assembly. Transcription was initiated by addition of
nucleotides and stopped after 10 min incubation. Figure 2
depicts a representative experiment, which shows that RNA
Figure 2. Analysis of the RNA products formed with the mutant AdMLP templates. Transcription initiation complexes were assembled on mutant AdML
templates as indicated at the top of the figure. Nucleotide triphosphates (60 µM
ATP, 10 µM UTP, 2 µM [α-32P]CTP and 120 µM 3′-OMeGTP) were added for
10 min. Subsequently, RNA products were processed and analyzed by denaturing PAGE as described in Materials and Methods.
synthesis in the presence of 3′-OMeGTP results in a homogeneous population of RNA products of the expected size.
Variability in the signals (Fig. 2, for example compare lanes 3
and 5) is due to variation in the efficiency of ethanol precipitation of such short RNAs and does not depend on the particular
template.
Open complex progression in elongating pol II complexes
Having established conditions that yield a homogeneous population of RNA products we analyzed the transcription bubble
of pol II complexes stalled at the different positions. Transcription complexes were assembled on DNA fragments with a
radiolabeled non-template strand under conditions for RNA
analysis. Transcription complexes stalled by 3′-OMeGTP were
briefly treated with potassium permanganate, which preferentially modifies single-stranded thymidines. Figure 3 shows the
results of this permanganate analysis using the different DNA
templates. The position of the first G residue is indicated to the
left of the figure. In accordance with our earlier study (30),
stalling pol II at register 15 by inclusion of 3′-OMeGTP in the
nascent mRNA leads to increased sensitivity to potassium
permanganate from position +3T to +13T (Fig. 3, lanes 1 and
2). Moving pol II to register 17 extends the sensitivity to +15T.
As expected the transcription bubble also expands further
downstream when transcription is stalled at registers 19, 21, 23
and 25. Surprisingly, the transcription bubble seems to be
retracted in complexes stalled at registers 26 and 27. Permanganate sensitivities can be observed from +5T to +19T and
+5T to +22T, respectively. At registers 28 and 29 this retraction is not observed and, additionally, sensitivity at positions
+3 and +5 is no longer apparent (Fig. 3, lanes 18 and 20; see
also Fig. 5A and B). Careful quantitation of multiple independent experiments indicates that the sensitivity of these positions is gradually lost when complexes are stalled beyond
register 28. Table 1 gives a summary of the analysis of the
open regions of stalled pol II complexes from between three
and five independent experiments. Please note that complexes
Nucleic Acids Research, 2001, Vol. 29, No. 13 2709
Figure 3. Permanganate sensitivity of pol II complexes stalled at different
positions. Transcription complexes were assembled with AdML+15G (lanes 1
and 2), AdML+17G (lanes 3 and 4), AdML+19G (lanes 5 and 6), AdML+21G
(lanes 7 and 8), AdML+23G (lanes 9 and 10), AdML+25G (lanes 11 and 12),
AdML+26G (lanes 13 and 14), AdML+27G (lanes 15 and 16), AdML+28G
(lanes 17 and 18) and AdML+29G (lanes 19 and 20). Reactions included
nucleotides (60 µM ATP, 10 µM UTP, 10 µM CTP and 120 µM 3′-OMe-GTP)
for 10 min at 30°C. After this, potassium permanganate was added for 20 s and
the reactions were processed as described in Materials and Methods. The G
residue creating the stall site is indicated by the arrow. Please note that the distance between stall site and labeling site is identical for all DNA template fragments (see also Fig. 1).
Table 1. Borders of the transcription bubble of stalled complexes
Template
Opening –TFIIS
Opening +TFIIS
15G
3–15
nd
17G
3–17
nd
19G
3–19
nd
21G
3–21
nd
23G
3–23
3–23
25G
3–25
8–25
26G
5–19
8–26
27G
5–20
11–27
28G
7–28
18–28
29G
7–29
18–29
31G
7/8–22
13–31
33G
14–33
22–33
35G
14–35
20–35
This table summarizes the data obtained in the permanganate sensitivity
assays using the mutated AdML templates. The numbers indicate the registers
which form the borders of the open region. For the most downstream position
we used the register of the stall site. Only positions with a permanganate sensitivity of twice the background were scored as sensitive. The assay was performed at least three times for each template. Complexes stalled at registers
25, 26, 27 and 28 were assayed more than five times. nd, not determined.
stalled at register 31 also show a retracted transcription bubble
(see also Fig. 5B).
From these experiments it is clear that the transcription
bubble expands in a continuous fashion in the downstream
Figure 4. Analysis of the RNA products formed with the mutant AdMLP templates in the presence and absence of δTFIIS. Transcription initiation complexes were assembled in the presence and absence of δTFIIS on mutant
AdML templates as indicated at the top of the figure. The odd numbered reactions received no TFIIS while the even numbered reactions received δTFIIS
protein. Initiation complexes were assembled with AdML+15G (lanes 1 and
2), AdML+17G (lanes 3 and 4), AdML+19G (lanes 5 and 6), AdML+23G
(lanes 7 and 8), AdML+25G (lanes 9 and 10), AdML+26G (lanes 11 and 12),
AdML+27G (lanes 13 and 14), AdML+28G (lanes 15 and 16), AdML+29G
(lanes 17 and 18) and AdML+31G (lanes 19 and 20). The reactions included
nucleotide triphosphates (60 µM ATP, 10 µM UTP, 2 µM [α-32P]CTP and 120
µM 3′-OMeGTP) for 10 min. Subsequently, RNA products were processed
and analyzed on a 21% denaturing polyacrylamide gel as described in Materials and Methods.
direction. This process is concomittant with NMP addition to
the nascent RNA chain. In most cases the downstream border
of the transcription bubble corresponds to the length of the
RNA product. At certain positions the bubble retracts under the
conditions of the assay, although the synthesized mRNAs had
the expected size. The upstream part (+3 to +7) of the bubble
remains sensitive to permanganate. Around register 28 the
upstream part gradually starts to re-adopt the double-stranded
conformation and the bubble reached a size of 20–22 nt.
Effects of elongation factor TFIIS on stalled transcription
complexes
We decided to further investigate the retracted complexes
stalled at registers 26, 27 and 31. The observation that RNA
products from these templates had the expected sizes indicates
that after reaching their respective stall sites these transcription
complexes may have adopted a transcriptional arrest conformation. A hallmark of transcriptional arrest is that the 3′-end of
the nascent RNA becomes disengaged from the active site.
Elongation factor TFIIS can re-activate arrested complexes by
promoting reiterative cleavage and re-extension of nascent
transcripts (for reviews see 38–40). In this process cleavage
products ranging from 2 to 7 nt are released. In our system one
would expect that TFIIS would induce accumulation of these
short RNAs of complexes stalled at registers 26, 27 and 31.
To test this hypothesis we compared RNA products formed
in the presence and absence of human TFIIS from our set of
AdML templates. In these experiments we used a shortened
version of human TFIIS (residues 130–301; kindly provided
2710 Nucleic Acids Research, 2001, Vol. 29, No. 13
Figure 5. Template opening in the presence and absence of δTFIIS. Transcription initiation complexes were assembled in the presence and absence of δTFIIS on
radiolabeled mutant AdML templates as indicated at the top of the figure. (A) Initiation complexes were assembled with AdML+23G (lanes 1–3), AdML+25G
(lanes 4–6), AdML+26G (lanes 7–9), AdML+27G (lanes 10–12) and AdML+28G (lanes 13–15). The G residue creating the stall site is indicated by the arrow.
(B) Initiation complexes were assembled with AdML+29G (lanes 1–3), AdML+31G (lanes 4–6), AdML+33G (lanes 7–9), AdML+35G (lanes 10–12). The reactions included nucleotides (60 µM ATP, 10 µM UTP, 10 µM CTP and 120 µM 3′-OMeGTP) at 30°C for 10 min. After this, potassium permanganate was added
for 20 s and the reactions were processed as described in Materials and Methods. The last T residue before the stall site is indicated by the arrow (see also Fig. 1).
by Dr C. M. Kane, which is functionally indistinguishable from
the full-length protein (41). In the absence of TFIIS (Fig. 4, odd
numbered lanes) RNAs of the expected size are obtained. In
the presence of TFIIS (Fig. 4, even numbered lanes) two major
RNAs are obtained from each template. In addition to the
expected size, RNA molecules 1 nt shorter are observed. This
agrees with the observation that incorporation of 3′-OMeGTP
in nascent RNA is much slower compared to incorporation of
natural ribonucleotides (data not shown), but also suggests that
TFIIS induces continual removal of the terminal 3′-OMeGMP
residue. In the case of the 26G, 27G and 31G complexes high
levels of small RNAs ranging in size between 4 and 8 nt can be
observed in reactions that received TFIIS (Fig. 4, lanes 12, 14
and 20). In contrast, small RNAs do not accumulate to such
high levels with other complexes, although significant levels
can be observed with the 23G and 28G complexes. Upon
comparison of the yields of the full-length RNAs versus small
released RNAs we estimate that efficiency of cleavage is >5-fold
more efficient for the 26G and 27G complexes than for the
31G complex and >20-fold more efficient than for nonretracted complexes like the 23G complex. In combination
with the results from the permanganate sensitivity experiments
(Fig. 3), the RNA product analysis strongly suggests that the
transcription complexes stalled at registers 26, 27 and 31 have
a high tendency to back-track, resembling the transcriptional
arrest state.
Next, the effect of TFIIS on the transcription bubble of stalled
complexes was analyzed. We expect that TFIIS addition will
reverse bubble retraction. Figure 5 shows the results of the
permanganate analyses using different AdML mutant
templates (AdML 23G up to 35G) in the absence and presence
of TFIIS. The borders of the transcription bubble are not
affected by TFIIS when the transcription complexes were
stalled at register 23 (Fig. 5A, lanes 1–3). Interestingly, the
sensitivities of 18T, 19T and 21T were enhanced by TFIIS.
Such alterations are not observed in the +3/+14 region. This
result suggests that TFIIS alters the conformation of the transcription bubble but not its location. Alternatively, TFIIS may
alter the distribution of transcription complexes over distinct
functional states (see below). The effect of TFIIS on the
complex stalled at register 25 is quite similar to that on the 23G
complex (Fig. 5A, lanes 4–6). In contrast, dramatic effects of
TFIIS on the transcription bubble were observed when pol II
was stalled at registers 26, 27 and 31. TFIIS reverses retraction
of the transcription bubble of the complex stalled at register 26
(Fig. 5A, compare lanes 8 and 9) and the +7/+11 region
becomes less susceptible. These results are fully compatible
with the hypothesis that the 26G complex has become arrested.
Transcription bubble retraction is also reversed by TFIIS on
the 27G and 31G templates (Fig. 5A, compare lanes 11 and 12;
Fig. 5B, compare lanes 5 and 6). In the presence of TFIIS, transcription complexes are no longer retracted and the leading
edge of the transcription bubble corresponds to the expected
position. In addition, we observed that in most cases the position of the upstream border of the open transcription
complexes changes. The upstream border seems to shift 2–3 nt
Nucleic Acids Research, 2001, Vol. 29, No. 13 2711
Figure 6. Arrest at register 27 is enhanced by 3′-OMeGTP. Initiation complexes were assembled on the AdML+27G template in the presence and
absence of δTFIIS as indicated at the top of the figure. After 30 min preincubation at 30°C the reactions received no nucleotides (lane 1), 60 µM ATP, 10
µM CTP and 10 µM UTP for 10 min (lanes 2 and 3) or 60 µM ATP, 10 µM
CTP, 10 µM UTP and 120 µM 3′-OMeGTP for 10 min (lanes 4 and 5). After
this time potassium permanganate was added for 20 s and the reactions were
processed as described in Materials and Methods. The G residue creating the
stall site is indicated by the arrow.
further downstream (for a summary of the permanganate analysis see Table 1). Inclusion of TFIIS during preincubation was
not needed to observe these effects. Similar results were
obtained when TFIIS was added just before the NTPs (data not
shown). From these results we conclude: (i) TFIIS reverses
arrest of transcription complexes stalled at registers 26, 27 or
31; (ii) TFIIS affects the permanganate sensitivities of singlestranded thymidine residues in the transcription bubble such
that the leading half of the bubble becomes more susceptible;
(iii) TFIIS seems to restrict the size of the transcription bubble.
Transcription bubble collapse is influenced by 3′OMeGTP and correlates with inefficient chase of RNA
products
We investigated whether registers 26 and 27 also represent
sites with a higher probability of transcriptional arrest or
pausing during the normal elongation cycle. This was examined in standard run-off experiments by lowering the concentration of UTP, CTP or ATP beyond the apparent Km value.
Also, we shortened transcription elongation times and lowered
the reaction temperature. None of these conditions led to an
accumulation of products, which would be indicative of
stalling or arrest at the expected registers (data not shown).
Apparently, even under these limiting conditions transcript
elongation is favored over transcriptional stalling or arrest.
Next, we examined whether transcription bubble retraction
is influenced by incorporation of 3′-OMeGTP into the
nascent mRNA. The Hawley group showed that unusual
nucleotides like ITP slow down the kinetics of mRNA
synthesis and can induce pausing (42). Even at saturating
concentrations 3′-OMeGTP is incorporated at a slower rate in
nascent mRNA than the standard nucleotides (Fig. 4 and data
not shown). In the experiment shown in Figure 6 we examined
Figure 7. Only part of the RNA products formed by the 26G complex can be
chased into longer products. Initiation complexes were assembled on
AdML+25G and AdML+26G templates in the presence and absence of δTFIIS
as indicated at the top of the figure. Reactions included nucleotides (60 µM
ATP, 10 µM UTP and 2 µM [α-32P]CTP). Lanes 4, 5, 7, 8, 11, 12, 15 and 16
included 120 µM 3′-OMeGTP in addition. After 10 min incubation at 30°C,
GTP was added at 600 µM to the reactions of the even numbered lanes and
incubation was continued for another 10 min. Subsequently, RNA products
were processed and analyzed on a 21% denaturing polyacrylamide gel as
described in Materials and Methods. To the left are indicated the lengths of the
RNA products.
the effect of 3′-OMeGTP on transcriptional arrest in the
permanganate sensitivity assay. When transcription was
performed on the 27G template with ATP, CTP and UTP but
without 3′-OMeGTP most of the complexes will stall at
register 26 and some of them will read through (∼20–40%; see
Fig. 7 and data not shown). In the permanganate sensitivity
experiments we observed template opening spanning +5T to
+25T (Fig. 6, lane 2). Addition of TFIIS in the absence of
3′-OMeGTP results in a more efficient modification of singlestranded thymidine residues in the leading half of the transcription bubble (Fig. 6, lane 3). When reactions were performed in
the presence of 3′-OMeGTP a more pronounced retraction of
the transcription bubble was observed (Fig. 6, compare lanes 2
and 4), which can be reversed by TFIIS (Fig. 6, lane 5). The
notion that 3′-OMeGTP enhances bubble retraction is also
supported by comparison of 27G complexes formed in the
absence of 3′-OMeGTP (Fig. 6, lane 2) with 26G complexes in
the presence of 3′-OMeGTP (Fig. 3, lane 14; Fig. 5A, lane 8).
The majority of both complexes are stalled at register 26, but
bubble retraction is more pronounced in 3′-OMeGTP-stalled
26G complexes. It seems likely that this is related to the
reduced stability of 3′-OMeGMP-capped RNAs in the
catalytic site of pol II. However, the difference could also be
caused by a portion of the transcription complexes reading
through the stall site in the absence of 3′-OMeGTP (Fig. 7 and
data not shown). However, in that case one would expect to
observe sensitivity at +29T, which is not detected (Fig. 6, lane
2).
Together our analyses indicate that pol II complexes stalled
by 3′-OMeGTP at registers 26, 27 and 31 resemble arrested
transcription complexes. Such complexes do not resume transcription upon addition of ‘missing’ nucleotides, which is in
contrast to stalled complexes. Our stalling strategy precludes a
direct discrimination between stalling and arrest. While it yields
homogeneous populations of RNAs, the 3′-OMeGMP cap
precludes further extension into longer products. We therefore
stalled complexes in the absence or presence of 3′-OMeGTP
2712 Nucleic Acids Research, 2001, Vol. 29, No. 13
and tested whether addition of TFIIS was required to allow
extension into longer products in the following way. Subsequent to the 10 min incubation with nucleotides, excess GTP
was added to stalled complexes in the absence or presence of
TFIIS for 10 min and RNA products were analyzed. Figure 7
shows the results for the 25G and 26G transcription
complexes. Whereas all RNA products of the 25G complex
can be chased by GTP, a significant portion of 26G products
fails to become extended (Fig. 7, compare lanes 1 and 2 with
lanes 9 and 10). Importantly, addition of TFIIS to the 26G
complex allows extension of these products (Fig. 7, lanes 13
and 14). This TFIIS dependence indicates that this portion of
the 26G products requires re-alignment with the catalytic
center of pol II. As expected, inclusion of 3′-OMeGTP
prevents the GTP chase in the absence of TFIIS but not in its
presence. Products formed by the 27G complex behaved similarly to 26G products (data not shown).
In conclusion, the GTP chase analysis of Figure 7 also
supports the conclusion that transcription complexes stalled
around register 26 have a high probability of entering a transcriptional arrest state.
DISCUSSION
Preinitiation complex assembly involves a multitude of interactions between basal factors, pol II and DNA. Many of these
interactions need to be broken during the early phase of transcription. When transcription converts from abortive to
productive mode between registers 9–11 (37), several basal
factors have already dissociated (28). However, at this stage
the transcription bubble has yet to achieve its mature size.
In this study we have followed the structure of the DNA
template during the early phase of pol II transcription. Using
potassium permanganate we have probed for the singlestranded region of transcription complexes stalled up to
register 35. Complexes were assembled in a reconstituted
sytem with purified pol II and the minimal set of basal transcription factors. First we discuss our findings from experiments lacking TFIIS. We found that between registers 11 and
25 the bubble expands up to 22 nt in size. This is consistent
with the reported size of the bacterial RNAP bubble of 14–22 nt
(33). The registers at which pol II attains sliding clamp mode
appear to be between 25 and 28 (Table 1). It is tempting to
speculate that the bubble retraction we observed with the 26G
and 27G complexes (Fig. 3) is related to conversion to the
sliding clamp mode of transcription elongation. However, the
observation that the bubble of the 31G complex (Fig. 5) is also
retracted argues against a strong causal relationship. Bubble
retraction is not solely dependent on the distance from the transcription start site but is also influenced by the sequence of the
DNA template. Alteration of template bases at registers 21 and
22 affects bubble retraction of 3′-OMeGTP-stalled 27G
complexes (data not shown). In addition, if registers 25–28
mark a soft spot in early initiation, one would expect that part
of the complexes would pause or arrest at these sites. We failed
to observe accumulation of RNA products of 25–28 nt during
normal elongation (see also Fig. 2), even under conditions of
low NTP concentrations, low temperature or restricted elongation times (data not shown). Neverthless, our results are in
accordance with observations by Samkurashvili and Luse, who
proposed that pol II complexes pass through a structural transition
around register 25 (31). They observed that complexes with
20, 23 and 25 nt RNAs have ExoIII footprint borders very
close to the site of NMP addition, reminiscent of arrested
complexes. In their paper transcription complexes were assembled in crude HeLa cell extracts, stalled by NTP omission,
rinsed with sarkosyl and purified by gel filtration. A direct
comparison with our results is complicated by differences in
the stalling protocols (NTP depriviation versus 3′-OMeGTP
chain termination), uncertainty in transcription complex
composition and detection method (ExoIII footprint versus
permanganate sensitivity). Despite these differences, similarities with our 26G and 27G complexes assembled in a minimal
pol II system and in the absence of TFIIS are quite remarkable.
An important issue in the potassium permanganate sensitivity assay concerns the homogeneity in the stalled complexes
(for a discussion see 43). In the interpretation of the data it is
assumed that all individual transcription complexes have the
same DNA–protein structure. Alternatively, the data are the
average of the most prominent configurations. This would
complicate the interpretation considerably. For several reasons
we do not think that heterogeneity is a complicating factor in
our study. First of all, the permanganate incubation was minimized to 20 s and the length of NTP incubation was optimized
for the highest yield of RNA product. Secondly, with the inclusion of 3′-OMeGTP as a RNA chain terminator <10% of the
RNA products have lengths other than those expected (Fig. 2).
Thirdly, we never observed permanganate sensitivity of the
thymidine residue 3 bp beyond the stall site (see for example
Figs 5A and 7), which indicates that no significant portion of
complexes transcribed beyond the stall site. Finally, only a
very small portion of 3′-OMeGTP-stalled complexes can be
chased by an excess of GTP in the absence of TFIIS (Fig. 7 and
data not shown). This means that almost all of the RNA products formed are capped at their 3′-terminus by 3′-OMeGMP.
All of these stalled RNAs can be chased to run-off products
when TFIIS is added together with excess GTP (Fig. 7 and data
not shown), implying that TFIIS efficiently removes the
3′-OMeGMP cap to allow chain elongation. Finally, the
assumption that our complexes are homogeneous is compatible
(see below) with the dimensions of the ‘grooves’ on pol II
proposed to harbor nascent RNA, the RNA–DNA hybrid and
the remaining lagging half of the bubble (5).
Addition of TFIIS to the 26G, 27G and 31G complexes
reverses retraction of the transcription bubble. This is concomitant with an accumulation of short RNAs (Fig. 4), which are
most likely cleavage products. These observations are fully
explained by assuming that these complexes have become
arrested. This is also supported by the chase experiment of
Figure 6. Arrested complexes can be reactivated by TFIISinduced cleavage releasing short oligonucleotides at the 3′-end
of the nascent RNA, which becomes disengaged from the catalytic center of the polymerase upon arrest (reviewed in 38,39).
The trimmed RNA becomes re-aligned with respect to the
catalytic center. The nascent RNA can then be re-extended,
which induces permanganate sensitivity of non-template residues up to the stall site. Our observations confirm the model of
TFIIS-induced re-alignment with the active center. TFIIS
reverses 26G and 27G retraction, suggesting that TFIIS can
assist transcription complexes to attain the sliding clamp mode
of elongation. TFIIS not only reverses back-tracking of the
26G, 27G and 31G complexes, but also seems to limit the size
Nucleic Acids Research, 2001, Vol. 29, No. 13 2713
of the transcription bubble at its lagging edge (Figs 5 and 7).
Several explanations are possible. First, a small portion of
stalled transcription complexes may have a tendency to backtrack. TFIIS reverses this process, resulting in a homogeneous
population of complexes. In this scenario one would expect a
TFIIS-induced accumulation of small RNAs in all cases, which
is not observed (see Fig. 4, lanes 2, 4, 6 and 18). Secondly,
TFIIS could bind directly to the non-template strand, thus
protecting it from permanganate modification. It has been
suggested that TFIIS can bind to single-stranded nucleic acids
through its zinc ribbon domain, as discussed (44). Finally, a
mere association of TFIIS to the transcription complex could
alter the pol II–DNA contacts within the DNA exit channel,
resulting in stabilization of double-stranded DNA at the
lagging part of the transcription bubble. In this respect it is
noteworthy that TFIIS has been shown to functionally interact
with Rbp6 (45). This subunit lies at the base of an inverted
funnel-shaped cavity of pol II, which is on the opposite side of
the DNA to the entry and exit channels (5). This funnel was
proposed to harbor the 3′-end of arrested RNAs and also to
provide access to TFIIS (5). These structural considerations
make it unlikely that TFIIS contacts the DNA directly. Rather,
it seems that TFIIS association induces altered pol II–DNA
contacts in the DNA exit channel, as suggested (39).
From biophysical and biochemical studies it has become
clear that bacterial and eukaryotic RNA polymerases are quite
similar. As mentioned, the size and position of the open region
prior to transcription initiation are almost identical (see 4).
Also, transcription shifts from abortive to productive mode at
very similar registers (37,46). In addition, our present results
strongly contribute to the view that pol II transcription
complexes undergo an important structural transition around
registers 25–27 (31), as was also noted previously for bacterial
RNAP (32).
What could be the reason for this transition? In this respect it
is interesting to review the structural models for T.aquaticus
RNAP (6) and yeast pol II (5) and the proposed paths of the
DNA template and RNA product. The DNA entry channel can
harbor ∼9 bp (RNAP) or ∼20 bp (pol II) of double-stranded
DNA, which it has been proposed is propelled towards the
active center by a helical screw rotation (5). In pol II this
downstream DNA would be retained by a flexible ‘clamp’
within the enzyme comprised of the N-terminal region of Rpb1
and the C-terminal region of Rpb2 (5). Compatible with the
observed continuous template opening (this paper and 37),
downstream DNA melts just before the active center. In both
RNA polymerases it has been proposed that the incoming
DNA duplex and the RNA–DNA hybrid make an angle of
almost 90° at the active center. The RNA–DNA hybrid is
placed at the upstream edge of the flexible clamp of pol II. An
important feature of the RNA–DNA hybrid exit channel of
T.aquaticus RNAP is a ‘rudder’-like structure formed by
region C of β′ at a distance from the active center compatible
with a RNA–DNA hybrid of 8–9 bp. A corresponding rudderlike structure has been detected in improved electron density
maps of pol II (P.Cramer and R.Kornberg, personal communication). It was proposed that in RNAP this rudder redirects
protruding RNA away from the DNA template under a flexible
flap structure formed by βF,G,H (6). In the case of pol II the
rudder probably directs the RNA towards groove 1 as it
extends from the edge of the clamp structure (P.Cramer and
R.Kornberg, personal communication). Groove 1 in pol II can
accomodate nascent RNA 10–20 nt from the active center.
This curved groove is located at the base of the flexible clamp
and leads back to downstream DNA (5). Filling of this groove
could tighten the grip of the flexible clamp on downstream
DNA. Conceivably, the groove is filled to completion when
the RNA is 26–27 nt long. This would obviously mark a structural transition in the transcription complex and could explain
why complexes stalled at these positions are liable to backtracking and transcriptional arrest.
The path of the non-template strand could also be involved
in the instability of the 26/27 complexes. This strand has to be
outside the RNA–DNA hybrid exit channel, because this
cannot accommodate both the RNA–DNA hybrid and the nontemplate strand. What happens with the template strand when
the 5′-end of a 9–10 nt RNA encounters the rudder? The
template DNA strand may be threaded through groove 2 (5).
The profound curvature of this groove (and of groove 1) is only
compatible with binding single-stranded nucleic acids. The
non-template strand could re-anneal only after the template
strand has passed through the narrow part of groove 2. The
narrow part of groove 2 (<20 Å in diameter) extends for 60 Å
from the end of the proposed RNA–DNA hybrid before
widening (P.Cramer and R.Kornberg, personal communication). This model has two attractive features. Localizing the
non-template strand away from polymerase structures explains
its sensitivity to modifying agents like permanganate (16).
Secondly, combining the length of the RNA–DNA hybrid with
the dimensions of the narrow part of groove 2 indicates a
minimal length for the transcription bubble of ∼20 nt (with no
base stacking interactions occuring within the bubble). Base
stacking interactions would allow a maximal bubble size of
27 nt. These considerations are compatible with a model in
which filling of grooves 1 and 2 is complete when the RNA is
25–27 nt long. This possibly constitutes the structural change
observed around registers 25–27.
Thus, it is very possible that clamp movement induced by
RNA binding in groove 1 and filling of groove 2 by the DNA
template strand are coordinated and result in a stable configuration of the transcription complex. It is also possible that these
structural changes are accompanied by TFIIH dissociation
from the complex. However, it is unlikely that TFIIH action is
a prerequisite for inducing the structural change required for
promoter escape, because pol II can productively transcribe
negatively supercoiled or premelted DNA templates in the
absence of TFIIH (11,12,16,47). Resolving the details of the
structural transition around registers 25–27 further requires a
combination of biochemical experiments, such as precise
DNA–protein crosslinking studies and three-dimensional
structure determination of pol II transcription complexes at
these stages of the transcription cycle.
ACKNOWLEDGEMENTS
We are grateful to Nell Shimazaki and Caroline Kane (University of California Berkeley) for supplying recombinant human
TFIIS protein. We would like to thank Roger Kornberg,
Patrick Cramer, Richard Ebright, Donal Luse, Peter van der
Vliet and members of our laboratory for stimulating discussions. In addition, we thank Patrick Cramer, Robert Sijbrandi
and Thomas Albert for critical review of this manuscript.
2714 Nucleic Acids Research, 2001, Vol. 29, No. 13
U.F. was supported by a long-term fellowship from EMBO. In
addition, U.F. and M.T. were supported by a Pionier grant
from the Netherlands Organization for Scientific Research.
REFERENCES
1. Eick,D., Wedel,A. and Heumann,H. (1994) From initiation to elongation:
comparison of transcription by prokaryotic and eukaryotic RNA
polymerases. Trends Genet., 10, 292–296.
2. Geiduschek,E.P. and Kassavetis,G.A. (1995) Comparing transcriptional
initiation by RNA polymerases I and III. Curr. Opin. Cell Biol., 7, 344–351.
3. Zawel,L. and Reinberg,D. (1995) Common themes in assembly and
function of eukaryotic transcription complexes. Annu. Rev. Biochem., 64,
533–561.
4. Fiedler,U. and Timmers,H.T.M. (2000) Peeling by binding or twisting by
cranking: models for promoter opening and transcription initiation by
RNA polymerase II. Bioessays, 22, 316–326.
5. Cramer,P., Bushnell,D.A., Fu,J., Gnatt,A.L., Maier-Davis,B.,
Thompson,N.E., Burgess,R.R., Edwards,A.M., David,P.R. and
Kornberg,R.D. (2000) Architecture of RNA polymerase II and
implications for the transcription mechanism. Science, 288, 640–649.
6. Zhang,G., Campbell,E.A., Minakhin,L., Richter,C., Severinov,S. and
Darst,S.A. (1999) Crystal structure of Thermus aquaticus core RNA
polymerase at 3.3 angstrom resolution. Cell, 98, 811–824.
7. Conaway,R.C. and Conaway,J.W. (1993) General initiation factors for
RNA polymerase II. Annu. Rev. Biochem., 62, 161–190.
8. Orphanides,G., Lagrange,T. and Reinberg,D. (1996) The general
transcription factors of RNA polymerase II. Genes Dev., 10, 2657–2683.
9. Roeder,R.G. (1996) The role of general initiation factors in transcription
by RNA polymerase II. Trends Biochem. Sci., 21, 327–335.
10. Bunick,D., Zandomeni,R., Ackerman,S. and Weinmann,R. (1982)
Mechanism of RNA polymerase II-specific initiation of transcription in vitro:
ATP requirement and uncapped runoff transcripts. Cell, 29, 877–886.
11. Goodrich,J.A. and Tjian,R. (1994) Transcription factors IIE and IIH and
ATP hydrolysis direct promoter clearance by RNA polymerase II. Cell,
77, 145–156.
12. Timmers,H.T.M. (1994) Transcription initiation by RNA polymerase II
does not require hydrolysis of the β–γ phosphoanhydride bond of ATP.
EMBO J., 13, 391–399.
13. Coin,F. and Egly,J.-M. (1998) Ten years of TFIIH. Cold Spring Harbor
Symp. Quant. Biol., 63, 105–110.
14. Svejstrup,J.Q., Vichi,P. and Egly,J.-M. (1996) The multiple roles of
transcription/repair factor TFIIH. Trends Biochem. Sci., 21, 346–350.
15. Wang,W., Carey,M. and Gralla,J.D. (1992) Polymerase II promoter
activation: closed complex formation and ATP-driven start site opening.
Science, 255, 450–453.
16. Holstege,F.C.P., van der Vliet,P.C. and Timmers,H.T.M. (1996) Opening of
an RNA polymerase II promoter occurs in two distinct steps and requires the
basal transcription factors IIE and IIH. EMBO J., 15, 1666–1677.
17. Tirode,F., Busso,D., Coin,F. and Egly,J.-M. (1999) Reconstitution of the
transcription factor TFIIH: assignment of function for the three enzymatic
subunits, XPB, XPD and cdk7. Mol. Cell, 3, 87–95.
18. Winkler,G.S., Araujo,S.J., Fiedler,U., Vermeulen,W., Coin,F., Egly,J.-M.,
Hoeijmakers,J.H.J., Wood,R.D., Timmers,H.T.M. and Weeda,G. (2000)
TFIIH with inactive XPD helicase functions in transcription initiation but
is defective in DNA repair. J. Biol. Chem., 275, 4258–4266.
19. Kim,T.K., Ebright,R.H. and Reinberg,D. (2000) Mechanism of ATPdependent promoter melting by transcription factor IIH. Science, 288,
1418–1422.
20. Dvir,A., Conaway,R.C. and Conaway,J.W. (1997) A role for TFIIH in
controlling the activity of early RNA polymerase II elongation complexes.
Proc. Natl Acad. Sci. USA, 94, 9006–9010.
21. Kumar,K.P., Akoulitchev,S. and Reinberg,D. (1998) Promoter proximal
stalling results from the inability to recruit transcription factor IIH to the
transcription complex and is a regulated event. Proc. Natl Acad. Sci. USA,
95, 9767–9772.
22. Bradsher,J., Coin,F. and Egly,J.-M. (2000) Distinct roles for helicases of
TFIIH in transcript intiation and promoter escape. J. Biol. Chem., 275,
2532–2538.
23. Akoulitchev,S., Makela,T.P., Weinberg,R.A. and Reinberg,D. (1995)
Requirement for TFIIH kinase activity in transcription by RNA
polymerase II. Nature, 377, 557–560.
24. Mäkelä,T.P., Parvin,J.D., Kim,J., Huber,L.J., Sharp,P.A. and
Weinberg,R.A. (1995) A kinase-deficient transcription factor TFIIH is
functional in basal and activated transcription. Proc. Natl Acad. Sci. USA,
92, 5174–5178.
25. Serizawa,H., Makela,T.P., Conaway,J.W., Conaway,R.C., Weinberg,R.A.
and Young,R.A. (1995) Association of Cdk-activating kinase subunits
with transcription factor TFIIH. Nature, 374, 280–282.
26. Jiang,Y., Yan,M. and Gralla,J.D. (1996) A three-step pathway of
transcription initiation leading to promoter clearance at an activated RNA
polymerase II promoter. Mol. Cell. Biol., 16, 1614–1621.
27. Li,Y. and Kornberg,R.D. (1994) Interplay of positive and negative
effectors in function of the C-terminal repeat domain of RNA polymerase
II. Proc. Natl Acad. Sci. USA, 91, 2362–2366.
28. Zawel,L., Kumar,P. and Reinberg,D. (1995) Recycling of the general
transcription factors during RNA polymerase II transcription. Genes Dev.,
9, 1479–1490.
29. Stelzer,G., Goppelt,A., Lottspeich,F. and Meisterernst,M. (1994)
Repression of basal transcription by HMG2 is counterparted by TFIIHassociated factors in an ATP-dependent process. Mol. Cell. Biol., 14,
4712–4721.
30. Holstege,F.C.P. and Timmers,H.T.M. (1997) Analysis of open complex
formation during RNA polymerase II transcription initiation using
heteroduplex templates and potassium permanganate probing. Methods,
12, 203–211.
31. Samkurashvili,I. and Luse,D.S. (1998) Structural changes in the RNA
polymerase II transcription complex during transition from initiation to
elongation. Mol. Cell. Biol., 18, 5343–5354.
32. Nudler,E., Goldfarb,A. and Kashlev,M. (1994) Discontinuous mechanism
of transcription elongation. Science, 265, 793–796.
33. Lee,D.N. and Landick,R. (1992) Structure of RNA and DNA chains in
paused transcription complexes containing Escherichia coli RNA
polymerase. J. Mol. Biol., 228, 759–777.
34. Holstege,F.C.P., Tantin,D., Carey,M., van der Vliet,P.C. and Timmers,H.T.M.
(1995) The requirement for the basal transcription factor IIE is determined by
the helical stability of promoter DNA. EMBO J., 14, 810–829.
35. Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A
Laboratory Manual, 2nd Edn. Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, NY.
36. Timmers,H.T.M. and Sharp,P.A. (1991) The mammalian TFIID protein is
present in two functionally distinct complexes. Genes Dev., 5, 1946–1956.
37. Holstege,F.C.P., Fiedler,U. and Timmers,H.T.M. (1997) Three transitions
in the RNA polymerase II-transcription complex during initiation. EMBO
J., 16, 7468–7480.
38. Uptain,S.M., Kane,C.M. and Chamberlin,M.J. (1997) Basic mechanisms of
transcript elongation and its regulation. Annu. Rev. Biochem., 66, 117–172.
39. Wind,M. and Reines,D. (2000) Transcription elongation factor SII.
Bioessays, 22, 327–336.
40. Conaway,J.W., Shilatifard,A., Dvir,A. and Conaway,R.C. (2000) Control
of elongation by RNA polymerase II. Trends Biochem. Sci., 25, 375–380.
41. Cipres-Palacin,G. and Kane,C.M. (1994) Cleavage of the nascent
transcript induced by TFIIS is insufficient to promote read-through of
intrinsic blocks to elongation by RNA polymerase II. Proc. Natl Acad. Sci.
USA, 91, 8097–8091.
42. Thomas,M.J., Platas,A.A. and Hawley,D.K. (1998) Transcriptional fidelity
and proofreading by RNA polymerase II. Cell, 93, 627–637.
43. Krummel,B. and Chamberlin,M.J. (1992) Structural analysis of ternary
complexes of Escherichia coli RNA polymerase. J. Mol. Biol., 225, 221–237.
44. Awrey,D.E., Shimasaki,N., Koth,C., Weilbaecher,R., Olmsted,V.,
Kazanis,S., Shan,X., Arello,J., Arrowsmith,C.H., Kane,C.M. and
Edwards,A.M. (1998) Yeast transcription elongation factor (TFIIS),
structure and function. J. Biol. Chem., 273, 22595–22605.
45. Ishiguro,A., Nogi,Y., Hisatake,K., Muramatsu,M. and Ishihama,A. (2000)
The Rpb6 subunit of fission yeast RNA polymerase II is a contact target of the
transcription elongation factor TFIIS. Mol. Cell. Biol., 2000, 1263–1270.
46. Zaychikov,E., Denissova,L. and Heumann,H. (1995) Translocation of the
Escherichia coli transcription complex observed in the registers 11 to 20:
“jumping” of RNA polymerase and assymmetric expansion and contraction
of the “transcription bubble”. Proc. Natl Acad. Sci. USA, 92, 1739–1743.
47. Parvin,J.D., Timmers,H.T.M. and Sharp,P.A. (1992) Promoter specificity
of basal transcription factors. Cell, 68, 1135–1144.