Molecular Cell 24, 257–266, October 20, 2006 ª2006 Elsevier Inc. DOI 10.1016/j.molcel.2006.10.001
Template Misalignment in Multisubunit
RNA Polymerases and Transcription Fidelity
Ekaterina Kashkina,1Michael Anikin,1
Florian Brueckner,2Richard T. Pomerantz,3,4,5
William T. McAllister,1Patrick Cramer,2
and Dmitry Temiakov1,*
1Department of Cell Biology
School of Osteopathic Medicine
University of Medicine and Dentistry of New Jersey
42 East Laurel Road
Stratford, New Jersey 08084
2Department of Chemistry and Biochemistry
Gene Center Munich
Ludwig-Maximilians-Universita ¨t Mu ¨nchen
3Department of Microbiology and Immunology and
4Graduate Program in Molecular and Cellular Biology
The State University of New York Downstate
450 Clarkson Avenue
Brooklyn, New York 11203
Recent work showed that the single-subunit T7 RNA
polymerase (RNAP) can generate misincorporation
errors by a mechanism that involves misalignment
of the DNA template strand. Here, we show that the
tion by the multisubunit yeast RNAP II and bacterial
RNAPs. Fluorescence spectroscopy reveals a reorga-
nization of the template strand during this process,
and molecular modeling suggests an open space
above the polymerase active site that could accommo-
date a misaligned base. Substrate competition assays
indicate that template misalignment, not misincorpo-
ration, is the preferred mechanism for substitution
errors by cellular RNAPs. Misalignment could account
for data previously taken as evidence for additional
NTP binding sites downstream of the active site. Anal-
ysis of the effects of different template topologies on
misincorporation indicates that the duplex DNA imme-
diately downstream of the active site plays an impor-
tant role in transcription fidelity.
RNA polymerase (RNAP) is a central enzyme in gene ex-
pression in all organisms. In both prokaryotes and eu-
karyotes, RNAPs are multisubunit enzymes that exhibit
high sequence and structural homology from species
three phases: promoter binding during initiation, proc-
essive synthesis of RNA during elongation, and release
of the transcript and dissociation of RNAP during termi-
nation. During elongation, incorporation of a single-sub-
strate nucleoside triphosphate (NTP) into the transcript
is repeated many times. In the so-called nucleotide ad-
dition cycle, extension of RNA involves two sites in the
RNAP active center (Steitz, 1998). At the beginning of
the cycle, the 30end of the RNA occupies the product
site and the incoming nucleotide is bound in the sub-
strate (insertion) site. Formation of the phosphodiester
bond is followed by movement of the newly formed
30end of the RNA into the product site (translocation)
and release of pyrophosphate (Landick, 2004).
Whereas most enzymes discriminate only between
a single cognate substrate and its low-affinity deriva-
tives, RNAPs use a DNA template to select among four
similar yet distinct substrates. Based mainly on studies
of DNA polymerases, the fidelity of polymerases is gen-
active site to the base pair formed between the template
base and the incoming NTP that occupies the inser-
tion site. Binding of the cognate, complementary NTP
provides the correct geometry required for catalysis
(Doublie and Ellenberger, 1998; Johnson et al., 2003),
whereas a noncomplementary (mismatched) base does
not lead to correct geometry, and results in a decreased
efficiency of phosphodiester bond formation (Johnson
and Beese, 2004).
recognition and fidelity after the NTP has entered the
active site, recent studies with T7 RNAP indicated that
there is an earlier level of substrate discrimination that
occurs in a ‘‘preinsertion site’’ prior to delivery of the in-
coming NTP into the substrate (insertion) site (Temiakov
et al., 2004). It has been suggested that a similar mech-
anism of fidelity in multisubunit RNAPs may also involve
template-dependent recognition of the incoming NTP
(Holmes et al., 2006; Kettenberger et al., 2004; Svetlov
et al., 2004; Temiakov et al., 2004). Other studies have
suggested that additional substrate binding sites down-
stream of the active site may also contribute to the fidel-
ity of multisubunit RNAPs (Gong et al., 2005; Nedialkov
et al., 2003), but structural evidence to support this
model is thus far lacking (Landick, 2005).
The fidelity of transcription may be further improved
by proofreading mechanisms that act subsequent to
the formation of the phosphodiester bond. A number
of studies have indicated that RNAPs may carry out
proofreading by means of an intrinsic cleavage activity
or as a result of binding of auxiliary factors such as
multisubunit RNAPs (1 in 105) and the effectiveness of
selection of ribonucleotide substrates over deoxyribo-
nucleotides are similar to those of the single-subunit
T7 RNAP (Blank et al., 1986; Huang et al., 2000; Ninio,
1991; Ozoline et al., 1980; Remington et al., 1998; Rose-
nberger and Hilton, 1983; Shaw et al., 2002; Springgate
and Loeb, 1975; Svetlov et al., 2004). The principal
mechanisms that determine fidelity may therefore be
similar in cellular and single-subunit RNAPs.
5Present address: Laboratory of DNA Replication, Howard Hughes
Medical Institute, The Rockefeller University, 1230 York Avenue,
New York, New York 10021.
A recent study with T7 RNAP demonstrated a novel
mechanism of error production due to template (T)-
Molecular Cell]). RNA primer-extension studies showed
by T7 RNAP, resulting in an extension of the RNA by two
nucleotides (nt). This involves a temporary flipping-out
of the n DNA base and a misalignment of the T strand,
enabling the n+1 DNA base to pair with the NTP and
direct its incorporation (Figure 1, right). The T strand
then realigns, resulting in a mismatched RNA end that
is further extended.
is involved in misincorporation errors by cellular, multi-
subunit RNAPs from both prokaryotes (E. coli and
T. thermophilus) and eukaryotes (S. cerevisiae). These
findings provide new insights into understanding the
fidelity of transcription and have implications for mech-
anistic models of substrate selection and incorporation.
Our data also reveal a critical role of the nontemplate
(NT) strand in the downstream DNA in RNAP fidelity.
Figure 1. Possible Pathways of n+1 NTP Misincorporation in Multisubunit RNAPs
The template base that is paired with the 30end of the RNA is designated as n-1, and the template base in the substrate site as n; template bases
further downstream are designated n+1, n+2, etc.The noncognate substrate complementary to the n+1 base is termed the n+1 NTP. Note that in
the alternative nomenclature (Landick, 2005), the n-1 and n bases are designated i and i+1, correspondingly. In the misincorporation model (left),
the n+1 NTP (CTP) is misincorporated opposite the n base, resulting in a mismatch at the 30terminus of the RNA; this is followed by mismatch
extension by correct incorporation of CTP opposite the n+1 base. In the misalignment model (right), misalignment of the T strand allows incor-
poration of CTP opposite the n+1 base; this is followed by realignment of the primer/template and mismatch extension as above. In both cases,
the result is a substitution error in the RNA product.
Results and Discussion
Misincorporation by Template Misalignment
To monitor nucleotide misincorporation, we utilized
a primer-extension assay in which elongation com-
plexes (ECs) were assembled on nucleic acid scaffolds
that contained an 8 to 9 nt RNA primer annealed to
a DNA T strand that extended 14 to 15 nt downstream.
NT strand downstream of the primer (Figures 2A and
In the presence of the correct NTP, the 8 nt RNA
primer was rapidly extended to 9 nt. A small amount of
bation, presumably as a result of misincorporation (Fig-
ure 2A, lanes 2–7). Extension of the primer in the pres-
ence of noncognate NTPs was less efficient. However,
the rate of incorporation of UTP, which is the substrate
that is complementary to the n+1 base on this template
(the n+1 NTP)was notably higher andthe primerwas ex-
tended by 2 nt, rather than by 1 nt, as observed for other
noncognate substrates and for the correct substrate
(Figure 2A, lanes 20–25). A similar pattern of preferential
NTP was previously observed for T7 RNAP (Pomerantz
et al., 2006) and was attributed to a T strand misalign-
ment mechanism (see Figure 1). Preferential misincor-
poration of the n+1 NTP and extension by 2 nt were
also observed when ECs were assembled on a scaffold
that contained a complementary NT strand downstream
of the RNA primer (Figure 2B). In the latter case, how-
ever, the efficiency of extension was lower than in the
Figure 2. The n+1 NTP Is a Preferable Substrate during Misincorporation by Bacterial RNAP
(A and B) Misincorporation of noncognate NTPs by E. coli RNAP. ECs were assembled on scaffolds having single-stranded (A) or double-
stranded downstream DNA (B) and 50radiolabeled RNA primer as described in Experimental Procedures. ECs were incubated in the presence
of 0.5 mM substrate NTPs at 37?C for the times indicated, and the products of the reactions were resolved using 20% PAGE containing 6 M urea
and quantified by PhosphorImager. The fraction of the primer extended is plotted as a function of time. Each data point represents an average of
at least three independent experiments.
(C) Substrate misincorporation in different sequence context. ECs were incubated with 0.5 mM UTP for the time indicated, and analyzed as
Template Misalignment in Multisubunit RNA Polymerases
the primer with the n+1 NTP was still greater than ex-
tension due to misincorporation of other noncognate
substrates (Figure 2B).
Although the preferential extension of the RNA in the
presence of the n+1 UTP shown in Figures 2A and 2B
may be accounted for by the misalignment mechanism
(Figure 1, right), an alternative explanation is that UTP
may be preferentially misincorporated in the particular
sequence context utilized in this experiment. To deter-
mine whether the preferential incorporation of the n+1
NTP is a general phenomenon, we utilized templates in
which each of the four possible bases was present at
the n+1 position in the T strand. In each case, the primer
was preferentially extended in the presence of the n+1
NTP, and always by 2 nt (see Figure S1A in the Supple-
mental Data available with this article online).
The misalignment mechanism (Figure 1, right) pro-
poses that preferential incorporation of the n+1 NTP oc-
curs as the result of forming a base pair with the n+1
DNA base in the T strand. We examined this by using
a series of templates in which each possible base was
present at the n+1 position but the template base at po-
sition n remained the same (Figure 2C). Preferential ex-
tension of the primer in the presence of UTP was ob-
served only when the complementary base (adenine)
was in the n+1 position (Figure 2C, lanes 1–6). This dem-
onstrates that the multisubunit bacterial RNAPs carry
out misincorporation via a T strand misalignment mech-
anism similar tothatobservedfor T7 RNAP. Importantly,
we observed an identical pattern of misincorporation on
these templates when using yeast RNAP II, suggesting
that the misalignment mechanism is universal for all
DNA-dependent RNAPs (Figure S1B).
In the presence of the n+1 NTP, the RNA primer was
always extended by 2 nt, with little accumulation of an
intermediate product extended by only 1 nt. This is con-
sistent with the notion that the rate-limiting step is the
initial incorporation event based on template misalign-
ment, and that extension of the RNA having a mismatch
30end is fast. We directly examined extension of mis-
matched RNA primer as shown in Figure S2 and ob-
served that bacterial RNAP from T. thermophilus and
E. coli as well as yeast RNAP II were able to readily
extend the mismatched primer.
Template Misalignment Involves
an Extrahelical Base
The results of the above experiments indicate that mis-
incorporation of the n+1 NTP in multisubunit RNAPs oc-
curs by a misalignment mechanism (Figure 1, right). In
DNAPs, template misalignment requires that one base
in the DNA template occupies an extrahelical ‘‘flipped-
out’’ position (Bebenek and Kunkel, 1990; Garcia-Diaz
et al., 2006; Kobayashi et al., 2002). To verify that
template misalignment in bacterial RNAP also involves
a similar rearrangement, and to provide direct evidence
for the misalignment mechanism, we monitored the
chemical environment of the n base in the T strand by
fluorescence spectroscopy. We assembled ECs on
a scaffold in which the template base at position n (cyti-
dine) was substituted by its fluorescent analog pyrrolo-
cytidine (pyrrolo-C) (Figure 3). Importantly, the fluores-
cence quantum yield of pyrrolo-C is very sensitive to
the local chemical environment, in particular to the rel-
ative position of the neighboring bases (Martin et al.,
2003). Thus, the quantum yield is low when the base is
within an undistorted DNA duplex, but is high when the
DNA is melted and the bases are unstacked (Liu and
Martin, 2002; Martin et al., 2003). An even higher quan-
tum yield is expected when the base is extrahelical
(Kobayashi et al., 2002).
An E. coli RNAP EC formed with a scaffold containing
pyrrolo-C showed a strong increase in fluorescence
emission as opposed to the scaffold alone (data not
shown), indicating that the n template base is not in-
volved in base-stacking interactions in the EC, consis-
tent with similar studies of the n template base in DNAP
complexes (Kobayashi et al., 2002). Addition of the cog-
nate substrate (GTP) to the pyrrolo-C-containing EC re-
duced the fluorescence quantum yield. This quenching
is apparently due to incorporation of the GMP into the
transcript and stacking of the resulting pyrrolo-C-
duplex (Figure 3). To monitor changes in the template
base position due to substrate binding, we used non-
hydrolyzable NTP analogs that cannot be incorporated
into the RNA (Figures 3A and 3B). The use of natural
NTPs in these experiments was not possible since bac-
terial RNAP and yeast RNAP II can extend mismatched
RNAs (Figure S2), preventing the formation of defined
halted complexes under these conditions. Addition of
the n+1 AMPcPP increased fluorescence emission, as
predicted from a misalignment mechanism that involves
flipping-out of the pyrrollo-C to adopt an extrahelical
position. In contrast, addition of UTP or UMPcPP, which
can base pair with neither the n nor the n+1 template
bases, did not affect the quantum yield of pyrrolo-C,
providing a negative control (Figure 3). Similar results
were obtained with yeast RNAP II (data not shown).
These observations are consistent with findings for
DNA polymerase IV, where misalignment resulted in in-
creased fluorescence emission in the presence of the
n+1 dNTP (Kobayashi et al., 2002). We conclude that
the n template base apparently adopts an extrahelical
position during T-strand misalignment, allowing the
n+1 base to occupy the site required to direct cognate
Inspection of the crystal structure of an RNAP II EC
(Kettenberger et al., 2004) suggests that there would
be space sufficient to accommodate a flipped-out base
structure of the misaligned intermediate, we performed
modeling using the DNA primer-template duplex with
a flipped-out base in the misaligned T strand that was
observed in the structure of DNA polymerase l (PDB
code 2BCQ [Garcia-Diaz et al., 2006]). Superposition of
this duplex onto the RNA:DNA hybrid duplex in the
RNAP II EC in various registers shows that an extrahel-
ical base can be accommodated above the RNAP II pro-
tween registers –1 and +1. The modeling thus suggests
that an extrahelical T base can be accommodated in
the active center cleft of yeast RNAP II above the active
site. The modeling, however, has limitations, since
superposition of the B form DNA-DNA duplex onto the
near A form RNA:DNA hybrid can only be approximate.
Therefore, the exact position of an extrahelical base
The n and n+1 Bases Likely Compete for the Same or
Overlapping Binding Sites during Misalignment
The above model predicts that the flipping-out of the
template base and template misalignment can be sup-
pressed in the presence of cognate, nonreactive NTP
analogs, which should stabilize the normal T-strand
conformation (Figure 5). We found that preincubation
of the bacterial EC with the cognate GTP analog
GMPcPP inhibits misalignment and misincorporation
of the n+1 NTP (Figure 5A, lanes 26–31, C), as well as
misincorporation of CTP (Figure 5A, lanes 32–37). To
verify this effect, we used a scaffold with a different
downstream DNA sequence (Figure 5D) and two nonre-
active NTP analogs, AMPcPP (cognate) and GMPcPP
(noncognate), and monitored misalignment with UTP
(n+1 NTP). Again, the presence of the correct nonhydro-
lyzable NTP efficiently inhibited misalignment, while the
presence of an incorrect NTP did not (Figure 5D). Based
on these results and the modeling, we speculate that
both the n template base in a substrate-bound EC and
the n+1 base in an EC with a misaligned template may
bind to overlapping sites. This implies that the same
NTP binding site could be utilized by RNAP during
both normal incorporation and misincorporation due to
Misalignment versus NTP Line-Up
In previous work, stimulation of substrate incorporation
in the presence of the n+1 NTP was taken as evidence
for an additional NTP binding site that does not overlap
Figure 3. The n Template Base Adopts an Extrahelical Position in the Presence of the n+1 NTP
(A) Fluorescence emission of pyrrolo-C at the n position in the presence of different substrate NTPs. Fluorescent probe, pyrrolo-C, was placed in
the T strand at the n position in R9/TS42pC scaffold. E. coli core RNAP was added to the scaffold in equimolar concentration (200 nM) and in-
cubated for 10 min at room temperature. Fluorescence emission of the assembled ECs was measured as a function of time with a 1 s interval
following addition of cognate (GTP) or noncognate substrates (UTP, UMPcPP, and AMPcPP).
(B) Schematic model of the T-strand misalignment.
Template Misalignment in Multisubunit RNA Polymerases
with the normal substrate insertion site opposite the n
base (Nedialkov et al., 2003). This model was extended
to suggest that of simultaneous binding of several
NTPs toRNAP, leading toaline-up ofNTPs downstream
before the active site (Gong et al., 2005). We examined
incorporation rate in bacterial EC by preincubating an
EC with AMPcPP, the substrate that is complementary
to the n+1 template base (Figures 5A and 5B). We ob-
served that the rate of CTP misincorporation was not in-
creased in the presence of AMPcPP, showing that the
ration at the n position (Figure 5A, lanes 8–19). When
CTP and ATP were used, efficient ATP misincorporation
by misalignment was still observed, resulting in exten-
sion by AA, whereas little or no extension of the 9-mer
RNA primer by C or CA was detected (compare lanes
8–13 and 20–25, Figure 5A). Failure to detect primer ex-
tension by CMP (or by CMP and AMP) is most likely due
to the much higher rate of misincorporation of the n+1
ATP by misalignment. Alternatively, template rearrange-
ments during misalignment may result in suppression of
misincorporation of another noncognate NTP at the n
position; however, we did not detect a significant differ-
ence insuppression inthe experiments involving nonhy-
drolyzable n+1 NTP analog (Figure 5A, lanes 14–19). A
similar effect of the n+1 NTP (diminished accumulation
of the products of misincorporation of another noncog-
nate NTP) previously observed in experiments with hu-
man RNAP II was interpreted in support of an ‘‘NTP
line-up mechanism’’ (Gong et al., 2005). However, our
data suggest that template misalignment may also
have occurred in this experiment, and lead us to ques-
tion the NTP line-up model. As our findings suggest
that addition of the n+1 NTP to the halted EC may stabi-
lize an alternative, flipped-out conformation of the n
base in the RNAP active site and thus affect cognate
substrate binding and incorporation, itmay be appropri-
the incoming n+1 substrate in light of the misalignment
Implications for Overlapping Substrate
Sites in RNAPs
It has been suggested that incoming NTP substrates
first bind to an entry site (E site) in multisubunit RNAPs,
before their delivery to the incorporation site (Westover
et al., 2004). The E site lacks direct interactions with the
base and ribose moiety of the substrate, and an NTP in
the E site does not base pair with a DNA template base
(Westover et al., 2004). Our observation that only a cog-
nate NTP can prevent misalignment (Figure 5D) sug-
gests that the E site is not involved in incorporation dur-
ing template misalignment. This, however, does not
exclude the possibility of NTP entry via the E site during
normal incorporation. During template misalignment,
and may bind either to the insertion site or to an overlap-
ping, preinsertion-like site (Kettenberger et al., 2004).
not allow us to distinguish between these possibilities.
The Downstream DNA-Binding Site Contributes to
The rates of primer extension by 1 nt due to misincorpo-
ration, and extension by 2 nt due to misalignment, were
both reduced in the presence of a complementary NT
strand immediately downstream of the active site (Fig-
ures 2A and 2B), indicating the importance of the NT
strand in the fidelity of transcription. Nevertheless,
misincorporation of the n+1 NTP by misalignment in
such complexes was always greater than misincorpora-
ing that misalignment may be the predominant mecha-
nism by which substitution errors are made by cellular
To add further evidence that misalignment is a com-
monly occurring phenomenon during transcription by
in the topology of the nucleic acid scaffold on misincor-
poration due to misalignment in scaffolds that either
have or lack a complementary downstream NT strand,
or have an abasic site in the NT strand at position n+1
(Figure 6A). As before, we found that n+1 CTP misincor-
poration by misalignment was decreased when the NT
strand was present, but the lack of a complementary
n+1 base in the NT strand restored efficient misalign-
ment (Figure 6A).
We next examined whether misalignment occurs in
ECs halted downstream of a promoter. Three templates
containing the T7A1 promoter were used: a fully double-
in the NT strand commencing 11 nt downstream of the
start site (middle), and a template with an abasic site in
the NT strand 11 nt downstream of the start site (right).
On these templates, Tth RNAP forms a halted EC 9 nt
downstream of the start site in the presence of ATP,
GTP, and CTP. Upon removal of NTPs by gel filtration,
the complexes were incubated with the n+1 NTP (ATP).
Efficient misalignment was observed when n+1 tem-
plate DNA base was not base-paired, in the case of tem-
plates having either a gap or an abasic site (Figure 6B).
The observed misalignment in promoter-originated
ECs on templates with abasic sites indicates that this
Figure 4. Close-Up of the Active Site of the Yeast RNAP II EC
The protein environment (surface representation, gray) near the
n and n+1 bases of the template DNA strand (blue) and the 30end
of the RNA (red) is shown. The presumed site to accommodate the
extrahelical template base during misalignment is indicated.
is likely to be an important mechanism for producing
errors during transcription by multisubunit RNAPs.
We also analyzed the effect of the NT strand on mis-
alignment efficiency in the yeast RNAP II system by ex-
amining RNA primer extension by n+1 ATP on scaffolds
that differed by the presence or absence of a comple-
mentary NT strand downstream of the active site
(Figure 6C). In contrast to bacterial RNAP, the rate of
n+1 NTP misincorporation by yeast RNAP II was not
suppressed in scaffolds having a complementary NT
strand in the downstream region (Figure 6C),suggesting
that misalignment in yeast RNAP II may occur even in
a natural double-stranded context. The observed dif-
ference in the efficiency of misalignment on double-
stranded templates between RNAP II and bacterial
RNAPs may indicate that during elongation RNAP II
melts the n+1 base pair in the downstream duplex while
bacterial RNAP does not (Holstege et al., 1997; Komis-
sarova and Kashlev, 1998). Taken together, the above
experiments suggest that misalignment is a common
mechanism by which transcription errors are generated
by different multisubunit RNAPs.
It has long been known that there is a DNA-binding
utes to the stability of the EC (Arndt and Chamberlin,
1990; Nudler et al., 1996). Structural studies of core
RNAP II ECs revealed contacts of the DNA T strand
1–3 nt downstream of the n site (Gnatt et al., 2001; West-
over et al., 2004). The structure of the complete RNAP II
EC, which contains duplex DNA in the proximate down-
stream region, revealed some additional DNA-polymer-
ase contacts that could be responsible for the observed
Figure 5. NTP-Stabilized Misalignment
(A) Effect of substrate NTPs on misincorporation and misalignment. E. coli ECs were preincubated with nonhydrolyzable NTP analogs (1 mM) for
5 min prior to the addition of the substrate indicated (0.5 mM each).
(B) Effect of the n+1 NTP on misincorporation. The fraction of the primer extended in lanes 2–25 (A) is plotted as a function of time.
(C)Effectof the cognate NTP on misalignment andmisincorporation. The fraction of theprimer extended inlanes 1–13 and26–37 (A)is plotted as
a function of time.
(D)The misaligned and correctly alignedtemplate bases bind to overlapping sites. E.coli EC(R8/TS35) was preincubated with correct(AMPcPP,
1 mM) or noncognate (GMPcPP, 1 mM) substrates for 5 min at room temperature prior to the addition of n+1 UTP (0.5 mM).
Template Misalignment in Multisubunit RNA Polymerases
stabilizing effect (Kettenberger et al., 2004). The data
presented here suggest that the proximate downstream
DNA grip is important to prevent misalignment errors
and to position the templating base accurately in the
RNAP active site. Loosening the downstream grip, as
duction due to misincorporation and template misalign-
Comparison with Misalignment in DNAPs
erantz et al., 2006), suggest that template misalignment
is a universal mechanism for generation of transcription
errors by both single-subunit and multisubunit RNAPs.
Our studies also reveal two fundamental differences to
the mechanism of misalignment in DNAPs (Bebenek
and Kunkel, 1990; Bell et al., 1997; Garcia-Diaz et al.,
2006; Kobayashi et al., 2002). First, in DNAPs, as re-
vealed by recent structural data (Garcia-Diaz et al.,
2006), the extrahelical base can be positioned between
the n-1 and the n-2 base upstream of the active site,
whereas our data so far reveal only that a position be-
Second, in DNAPs, the extrahelical base can apparently
persist and be translocated along the growing template-
primer duplex, resulting in a deletion mutation (Bebenek
and Kunkel, 1990);but in RNAPs,our data sofar indicate
that the extrahelical base is not translocated into the
next position but undergoes realignment, leading to
a substitution mutation. The modeling shows that an ex-
trahelical base between positions n-2 and n-1 clashes
with two protein regions in yeast RNAP II, switch 2 and
switch 3 (Cramer et al., 2001). This steric constraint ex-
plains why an extrahelical base accommodated just be-
fore that position during misalignment is apparently not
translocated but is rather realigned, resulting in the for-
mation of a mismatched base pair at the end of the
Figure 6. Effect of the NT Strand on Template-Strand Misalignment in Different Types of Multisubunit RNAPs
(A) Misincorporation by misalignment is enhanced in the absence of the NT strand and in the presence of an abasic site at the n+1 position in
NT DNA strand in E. coli ECs. Nucleic acid scaffolds used are indicated. ECs were incubated with the n+1 CTP (0.5 mM) at 37?C for the times
(B) The absence of the complementary n+1 base in the NT strand stimulates misincorporation by misalignment in Tth ECs halted downstream of
the promoter. ECs were incubated with the n+1 ATP (0.5 mM) at 60?C for the times indicated.
(C)The rateof misincorporationbymisalignment isnot affected bythe presence ofthe NTstrand inyeastRNAP IIEC.ECs wereassembledusing
yeast RNAP II and the scaffolds indicated and incubated with the n+1 ATP (0.5 mM) for 4–45 min at 30?C.
ration errors by RNAPs likely result only in phenotypic
mutations in the cell, as they do not change the genetic
material and are therefore not inherited. It is known
that errors at the transcription level can give rise to the
synthesis of toxic proteins as demonstrated for the var-
iants of b-amyloid precursor protein and ubiquitin B
found in neurons of Alzheimer’s and Down’s syndrome
scription of nonmutated genes has been termed ‘‘mis-
reading’’ and results in frameshift mutations (van Leeu-
wen et al., 2000). It remains to be seen whether
erroneous RNA resulting from template misalignment,
which may be facilitated at certain DNA sequences,
To understand whether and how RNAPs can generate
deletion errors via a misalignment mechanism requires
Our data on cellular multisubunit RNAPs, together with
recent studies of T7 RNAP (Pomerantz et al., 2006), sug-
gest that template misalignment is a universal mecha-
nism by which substitution errors are generated during
transcription. In vitro, errors due to T-strand misalign-
mentappear todominate over errors due to misincorpo-
ration, independent of the nucleic acid scaffold or the
nature of the EC. Template-strand misalignment should
therefore be considered as a factor that may affect ex-
periments that involve kinetics of NTP binding, incorpo-
ration, and misincorporation. In vivo, it is likely that mis-
alignment is enhanced when abasic sites or lesions are
encountered by RNAP before the DNA-repair machinery
fixes the damages. Although the presence of the correct
NTP inhibits misalignment, the high efficiency of this
mechanism as opposed to misincorporation suggests
thataconsiderablenumber ofsubstitution errorsinRNA
might be generated by misalignment. Indeed, experi-
ments in vivo demonstrated that when substrates are
present in nonequimolar concentrations, misincorpora-
tion by DNAP due to misalignment is increased 7- to
15-fold (Bebenek and Kunkel, 1990). An important aim
for the future will be to investigate the role of template
misalignment in the fidelity of cellular RNAPs in vivo.
Purification of RNAP and Polymerase Activity Assay
WT Tth core and holo RNAP were purified from cell biomass ob-
tained from HB8 strain as described (Vassylyeva et al., 2002). WT
E. coli core polymerase was purified from cell biomass obtained
from MRE600 strain as described (Nudler et al., 2003). The purity
of these RNAPs was greater then 99.8% as judged by SDS-PAGE.
Polymerase activity was measured by the ability to extend a
32P-labeled RNA primer by 1 nt in a reaction where the concentra-
tions of nucleic acid scaffold (R8/TS35/NT35) and RNAP were equi-
molar (see transcription conditions below). Yeast RNAP II was
prepared as described (Armache et al., 2003).
RNA and DNA Oligonucleotides
The following synthetic oligonucleotides were used (all sequences
are 50to 30): RNA oligomers (Dharmacon), (R8) GCGGCGAU and
(R9) GCGGCGAUA; DNA oligomers (IDT DNA), (TS35) CCTGTC
TGAATCGATATCGCCGC, (TS36) CCTGTCTGAATCGATATCGCCG
(TS42pC, Cpindicates pyrrolo-C) CCTGTCTGACTATCpTATCGCC
GG, (TS40) CCTGTCTGAATCGTCATCGCCGC, (TS44G) CCTGTCTG
ATTCAGACAGG, (NT38) ATCGATTCAGACAGG, (NT36) CTGTCCGA
GGAGATGACACTCGATTCAGACAGG, and (NTA36, ‘‘-’’ indicates
abasic site) CTGTCCGAGGAGATGACACTC-ATTCAGACAGG.
For promoter-based experiments, the following DNA oligos con-
taining T7A1 promoter wereused: EK33 (Tstrand,TTCCCCGTGTAC
CT), EK32 (NT strand, AGTATTGACTTAAAGTCTAACCTATAGGATA
CTTACAGCCATCGAGAGGTACACGGGGAA), EK36 (NT strand with
abasic [-] site) AGTATTGACTTAAAGTCTAACCTATAGGATACTTAC
AGCCATCGAGAGGT-CACGGGGAA), EK34 (NT strand, AGTATTGA
CTTAAAG TCTAACCTATAGGATACTTACAGCCATCGAGAGG), and
EK35 (NT strand, CACGGGGAA)
Nonhydrolyzable NTP analogs were from Sigma (AMPcPP), Jena
Scientific (GMPcPP), and BIOLOG Life Science Institute (UMPcPP).
All other NTP substrates were from GE Health.
Assembly of ECs and Transcription Conditions
Nucleic acid scaffolds were assembled by annealing equimolar con-
centrations of complementary RNA and DNA oligomers as previ-
ously described (Temiakov et al., 2002). RNA and DNA oligomers
were labeled at their 50ends using g-32P ATP and polynucleotide
kinase (New England Biolabs) prior to assembly. To assemble
ECs, core RNAP (0.2–1 mM) was incubated with an equimolar con-
centration of scaffold for 5 min at room temperature in 10 ml of
transcription buffer (40 mM Tris [pH 7.9] at 25?C, 100 mM NaCl,
5 mM MgCl2, and 5 mM 2-mercaptoethanol). Primer extension
was achieved by incubation of complexes with substrate NTPs
(500 mM) for 1–60 min at 60?C (Tth RNAP), 37?C (E. coli RNAP), or
30?C (yeast RNAP II).
ECs halted downstream of a promoter were formed by incubating
Tth holoRNAP (40 nM) with linearDNAtemplate containing the T7A1
promoter (40 nM), ApU (200 mM), and appropriate NTPs (50 mM) on
ice in 10 ml of transcription buffer. The reactions were incubated at
60?C for 10 min, and NTPs were removed by gel filtration on Quick
Spin columns (G50, Roche). Reactions were terminated by the addi-
tion of 90% formamide and 50 mM EDTA solution, and products
were resolved by electrophoresis in 20% PAGE in the presence of
6 M urea and visualized by PhosphorImager (GE Health).
Fluorescence measurements were carried out using a Fluoromax-3
fluorometer (HORIBA Jobin Yvon) equipped with an F-3004 thermo-
length Electronics). Quartz cells (30 ml) with 1.5 3 1.5 mm optical
length paths (Hellma) were used. Samples were excited at 350 nm,
and fluorescence emission was monitored at 440 nm with both exci-
tation and emission slits set at 5 nm. The nucleic acid scaffold car-
rying a fluorescent probe (pyrrolo-C) was prepared as described
above by annealing together oligonucleotides R9 and TS42pC.
The ECs were prepared by preincubating 200 nM scaffold with an
equimolar amount of E. coli RNAP or yeast RNAP II in transcription
buffer containing 5 mM MgCl2for 10 min at 23?C. The resulting
fluorescent complexes were incubated with hydrolyzable (0.5 mM)
or nonhydrolyzable (1–2 mM) NTPs for 5 min at 23?C, and
emission was monitored for 5–10 min as a function of time. All mea-
surements were performed at 23?C. The effect of protein and NTP
absorption on the emission of the pyrrolo-C was determined to be
Supplemental Data include two figures and can be found with
this article online at http://www.molecule.org/cgi/content/full/24/2/
We thank Dr. Laurie K. Read, SUNY Buffalo School of Medicine, for
the generous giftof UMPcPP.Thiswork wassupportedbyNIHgrant
GM38147 (to W.T.M.) and by a UMDNJ Foundation grant (to D.T.).
P.C. is supported by the Deutsche Forschungsgemeinschaft.
Template Misalignment in Multisubunit RNA Polymerases
Received: July 28, 2006 Download full-text
Revised: September 27, 2006
Accepted: October 3, 2006
Published: October 19, 2006
Armache, K.J., Kettenberger, H., and Cramer, P. (2003). Architecture
of initiation-competent 12-subunit RNA polymerase II. Proc. Natl.
Acad. Sci. USA 100, 6964–6968.
Arndt, K.M., and Chamberlin, M.J. (1990). RNA chain elongation by
Escherichia coli RNA polymerase. Factors affecting the stability of
elongating ternary complexes. J. Mol. Biol. 213, 79–108.
Bebenek, K., and Kunkel, T.A. (1990). Frameshift errors initiated by
Bell, J.B., Eckert, K.A., Joyce, C.M., and Kunkel, T.A. (1997). Base
miscoding and strand misalignment errors by mutator Klenow poly-
merases with amino acid substitutions at tyrosine 766 in the O helix
of the fingers subdomain. J. Biol. Chem. 272, 7345–7351.
Blank, A., Gallant,J.A., Burgess, R.R.,andLoeb, L.A. (1986). An RNA
polymerase mutant with reduced accuracy of chain elongation. Bio-
chemistry 25, 5920–5928.
Cramer, P., Bushnell, D.A., and Kornberg, R.D. (2001). Structural ba-
sis of transcription: RNA polymerase II at 2.8 angstrom resolution.
Science 292, 1863–1876.
Doublie, S., and Ellenberger, T. (1998). The mechanism of action of
T7 DNA polymerase. Curr. Opin. Struct. Biol. 8, 704–712.
Erie, D.A., Hajiseyedjavadi, O., Young, M.C., and von Hippel, P.H.
(1993). Multiple RNA polymerase conformations and GreA: control
of fidelity of transcription. Science 262, 867–873.
Garcia-Diaz, M., Bebenek, K., Krahn, J.M., Pedersen, L.C., and Kun-
kel, T.A. (2006). Structural analysis of strand misalignment during
DNA synthesis by a human DNA polymerase. Cell 124, 331–342.
Gnatt, A.L., Cramer, P., Fu, J., Bushnell, D.A., and Kornberg, R.D.
(2001). Structural basis of transcription: an RNA polymerase II elon-
gation complex at 3.3 A resolution. Science 292, 1876–1882.
Gong, X.Q., Zhang, C., Feig, M., and Burton, Z.F. (2005). Dynamic
error correction and regulation of downstream bubble opening by
human RNA polymerase II. Mol. Cell 18, 461–470.
Holmes, S.F., Santangelo, T.J., Cunningham, C.K., Roberts, J.W.,
and Erie, D.A. (2006). Kinetic investigation of Escherichia coli RNA
polymerase mutants that influence nucleotide discrimination and
transcription fidelity. J. Biol. Chem. 281, 18677–18683.
Holstege, F.C., Fiedler, U., and Timmers, H.T. (1997). Three transi-
tions in the RNA polymerase II transcription complex during initia-
tion. EMBO J. 16, 7468–7480.
Huang, J., Brieba, L.G., and Sousa, R. (2000). Misincorporation by
wild-type and mutant T7 RNA polymerases: identification of interac-
incompetent open conformation. Biochemistry 39, 11571–11580.
Johnson, S.J., and Beese, L.S. (2004). Structures of mismatch repli-
cation errors observed in a DNA polymerase. Cell 116, 803–816.
Johnson, S.J., Taylor, J.S., and Beese, L.S. (2003). Processive DNA
synthesis observed in a polymerase crystal suggests a mechanism
for the prevention of frameshift mutations. Proc. Natl. Acad. Sci.
USA 100, 3895–3900.
Kettenberger, H., Armache, K.J., and Cramer, P. (2004). Complete
RNA polymerase II elongation complex structure andits interactions
with NTP and TFIIS. Mol. Cell 16, 955–965.
Kobayashi, S., Valentine, M.R., Pham, P., O’Donnell, M., and Good-
man, M.F. (2002). Fidelity of Escherichia coli DNA polymerase IV.
Preferential generation of small deletion mutations by dNTP-stabi-
lized misalignment. J. Biol. Chem. 277, 34198–34207.
Komissarova, N., and Kashlev, M. (1998). Functional topography of
nascent RNA in elongation intermediates of RNA polymerase.
Proc. Natl. Acad. Sci. USA 95, 14699–14704.
Landick, R. (2004). Active-site dynamics in RNA polymerases. Cell
Landick, R. (2005). NTP-entry routes in multi-subunit RNA polymer-
ases. Trends Biochem. Sci. 30, 651–654.
Liu, C., and Martin, C.T. (2002). Promoter clearance by T7 RNA poly-
merase. Initial bubble collapse and transcript dissociation moni-
tored by base analog fluorescence. J. Biol. Chem. 277, 2725–2731.
Martin, C.T., Ujvari, A., and Liu, C. (2003). Evaluation of fluorescence
spectroscopy methods for mapping melted regions of DNA along
the transcription pathway. Methods Enzymol. 371, 13–33.
Nedialkov, Y.A., Gong, X.Q., Hovde, S.L., Yamaguchi, Y., Handa, H.,
Geiger, J.H., Yan, H., and Burton, Z.F. (2003). NTP-driven transloca-
tion by human RNA polymerase II. J. Biol. Chem. 278, 18303–18312.
Ninio, J. (1991). Transient mutators: a semiquantitative analysis of
the influence of translation and transcription errors on mutation
rates. Genetics 129, 957–962.
Nudler, E., Avetissova, E., Markovtsov, V., and Goldfarb, A. (1996).
Transcription processivity: protein-DNA interactions holding to-
gether the elongation complex. Science 273, 211–217.
Nudler, E., Gusarov, I., and Bar-Nahum, G. (2003). Methods of walk-
ing with the RNA polymerase. Methods Enzymol. 371, 160–169.
Ozoline, O.N., Oganesjan, M.G., and Kamzolova, S.G. (1980). On the
fidelity of transcription by Escherichia coli RNA polymerase. FEBS
Lett. 110, 123–125.
Pomerantz, R.T., Temiakov, D., Anikin, M., Vassylyev, D.G., and
during transcription due to template-strand misalignment. Mol. Cell.
24, this issue, 245–255.
Remington, K.M., Bennett, S.E., Harris, C.M., Harris, T.M., and
Bebenek, K. (1998). Highly mutagenic bypass synthesis by T7 RNA
polymerase of site-specific benzo[a]pyrene diol epoxide-adducted
template DNA. J. Biol. Chem. 273, 13170–13176.
Rosenberger, R.F., and Hilton, J. (1983). The frequency of transcrip-
tional and translational errors at nonsense codons in the lacZ gene
of Escherichia coli. Mol. Gen. Genet. 191, 207–212.
Shaw, R.J., Bonawitz, N.D., and Reines, D. (2002). Use of an in vivo
reporter assay to test for transcriptional and translational fidelity in
yeast. J. Biol. Chem. 277, 24420–24426.
Springgate, C.F., and Loeb, L.A. (1975). On the fidelity of transcrip-
tion by Escherichia coli ribonucleic acid polymerase. J. Mol. Biol.
Steitz, T.A. (1998). A mechanism for all polymerases. Nature 391,
Svetlov, V., Vassylyev, D.G., and Artsimovitch, I. (2004). Discrimina-
tion against deoxyribonucleotide substrates by bacterial RNA poly-
merase. J. Biol. Chem. 279, 38087–38090.
Temiakov, D., Anikin, M., and McAllister, W.T. (2002). Characteriza-
tion of T7 RNA polymerase transcription complexes assembled on
nucleic acid scaffolds. J. Biol. Chem. 277, 47035–47043.
Temiakov, D., Patlan, V., Anikin, M., McAllister, W.T., Yokoyama, S.,
and Vassylyev, D.G. (2004). Structural basis for substrate selection
by T7 RNA polymerase. Cell 116, 381–391.
Thomas, M.J., Platas, A.A., and Hawley, D.K. (1998). Transcriptional
fidelity and proofreading by RNA polymerase II. Cell 93, 627–637.
van Leeuwen, F.W., Fischer, D.F., Kamel, D., Sluijs, J.A., Sonne-
mans, M.A., Benne, R., Swaab, D.F., Salehi, A., and Hol, E.M.
(2000). Molecular misreading: a new type of transcript mutation
expressed during aging. Neurobiol. Aging 21, 879–891.
van Leeuwen, F.W., van Tijn, P., Sonnemans, M.A., Hobo, B., Mann,
D.M., Van Broeckhoven, C., Kumar-Singh, S., Cras, P., Leuba, G.,
Savioz, A., et al. (2006). Frameshift proteins in autosomal dominant
forms of Alzheimer disease and other tauopathies. Neurology 66,
Vassylyeva, M.N., Lee, J., Sekine, S.I., Laptenko, O., Kuramitsu, S.,
Shibata, T., Inoue, Y.,Borukhov, S.,Vassylyev, D.G., and Yokoyama,
S. (2002). Purification, crystallization and initial crystallographic
analysis of RNA polymerase holoenzyme from Thermus thermophi-
lus. Acta Crystallogr. D58, 1497–1500.
basis of transcription: nucleotide selection by rotation in the RNA
polymerase II active center. Cell 119, 481–489.