Content uploaded by A. Rachid Rahmouni
Author content
All content in this area was uploaded by A. Rachid Rahmouni
Content may be subject to copyright.
Francine Toulme
Â
, Christine Mosrin-Huaman,
Jason Sparkowski
1
, Asis Das
1
, Marc Leng
and A.Rachid Rahmouni
2
Centre de Biophysique Mole
Â
culaire, CNRS, rue Charles Sadron,
45071 Orle
Â
ans ce
Â
dex 2, France and
1
Department of Microbiology,
University of Connecticut Health Center, Farmington, CT 06030, USA
2
Corresponding author
e-mail: rahmouni@cnrs-orleans.fr
The GreA and GreB proteins of Escherichia coli show
a multitude of effects on transcription elongation
in vitro, yet their physiological functions are poorly
understood. Here, we investigated whether and how
these factors in¯uence lateral oscillations of RNA
polymerase (RNAP) in vivo, observed at a protein
readblock. When RNAP is stalled within an (ATC/
TAG)
n
sequence, it appears to oscillate between an
upstream and a downstream position on the template,
3 bp apart, with concomitant trimming of the tran-
script 3¢ terminus and its re-synthesis. Using a set of
mutant E.coli strains, we show that the presence of
GreA or GreB in the cell is essential to induce this
trimming. We show further that in contrast to a tern-
ary complex that is stabilized at the downstream pos-
ition, the oscillating complex relies heavily on the
GreA/GreB-induced `cleavage-and-restart' process to
become catalytically competent. Clearly, by promot-
ing transcript shortening and re-alignment of the cata-
lytic register, the Gre factors function in vivo to rescue
RNAP from being arrested at template positions
where the lateral stability of the ternary complex is
impaired.
Keywords: elongation factors/GreA/GreB/RNA
polymerase/transcription elongation complex
Introduction
RNA polymerases (RNAPs) carry out transcript elonga-
tion in a discontinuous manner. Frequently, DNA
sequences through which the polymerase must pass
to complete an RNA chain lead the enzyme to stop
RNA synthesis for varying lengths of time before
resuming elongation. This process of RNAP pausing is
physiologically signi®cant for both positive and nega-
tive control of transcription elongation (Das, 1993;
Henkin, 1996; Landick et al., 1996; Uptain et al.,
1997; Artsimovitch and Landick, 2000). In some
instances in vitro, a fraction of the paused polymerases
eventually transforms into an arrested state in which
the ternary complex neither elongates nor dissociates
(Davenport et al., 2000 and references therein).
According to current models, this loss of catalytic
activity results from a backward translocation of RNAP
along the DNA and RNA chains (Landick, 1997; von
Hippel, 1998; Nudler, 1999). The upstream movement
shifts the transcription bubble, the RNA±DNA hybrid
and the catalytic center away from the 3¢ end of the
transcript, which becomes extruded out of the poly-
merase (Reeder and Hawley, 1996; Komissarova and
Kashlev, 1997a,b; Nudler et al., 1997; Samkurashvili
and Luse, 1998; Sidorenkov et al., 1998). Thus, unless
the sliding process is reversed, resumption of RNA
synthesis requires trimming of the transcript to gener-
ate a new 3¢ hydroxyl group in register with the
catalytic center of the enzyme.
A special class of elongation factors, which includes
GreA and GreB for the bacterial RNAP and SII for the
eukaryotic RNAP II, stimulates transcript cleavage
within ternary complexes that are arti®cially halted
by NTP starvation (Borukhov et al., 1992, 1993;
Reines, 1992, 1994; Reines et al., 1992; Kassavetis
and Geiduschek, 1993). These in vitro observations
raised the attractive possibility that the Gre/SII family
of elongation factors may play a key role in vivo (i) to
mitigate pausing and suppress elongation arrest and
(ii) to enhance transcription ®delity (Erie et al., 1993;
Jeon and Agarwal, 1996; Thomas et al., 1998). To
date, however, evidence that GreA and GreB as well
as SII act as cleavage factors inside the cells is still
lacking. Ternary transcription complexes made with a
variety of puri®ed RNAPs have been shown to possess
substantial endonucleolytic cleavage activity in the
absence of the known stimulatory factors (Surratt et al.,
1991; Hagler and Schuman, 1992; Wang and Hawley,
1993; Rudd et al., 1994; Tschochner, 1996; Sastry and
Ross, 1997; Chedin et al., 1998). In particular,
experiments with Escherichia coli RNAP puri®ed
from a greA
±
greB
±
double mutant indicated that the
cleavage reaction is inherent to the polymerase and
that Gre factors simply enhance this intrinsic property
(Orlova et al., 1995). Nevertheless, the biological
function of the transcript trimming activity of any
RNAP has yet to be revealed.
We have recently shown that a transcription elonga-
tion complex, halted within an (ATC/TAG)
n
sequence
due to a readblock imposed by the lac repressor in
E.coli, oscillates between an upstream and a down-
stream position on the template, with accompanying
transcript cleavage and re-synthesis (Toulme
Â
et al.,
1999). We have now investigated the in¯uence of
GreA and GreB on RNAP oscillations in this system.
Results reported below provide direct evidence for the
transcript cleavage reaction mediated by GreA and
GreB inside the cell. They also show that the Gre
factor-induced transcript shortening is essential for
RNAP to read through template sequences that impair
the lateral stability of the ternary complex.
GreA and GreB proteins revive backtracked RNA
polymerase in vivo by promoting transcript
trimming
The EMBO Journal Vol. 19 No. 24 pp. 6853±6859, 2000
ã European Molecular Biology Organization
6853
Results
GreA and GreB induce transcript cleavage within
an oscillating ternary complex in vivo
To investigate the cellular function of the Gre factors, we
analyzed in vivo the structural behavior of an RNAP
elongation complex readblocked on a reporter plasmid
(Figure 1). On this template, transcription is initiated from
a constitutive promoter and the elongation complex is
stalled by a physical readblock, imposed by the lac
repressor bound to its operator. By a combination of in situ
DNA footprinting and RNA 3¢ end mapping experiments,
we have shown previously (Toulme
Â
et al., 1999) that the
ternary complex halted within an (ATC/TAG)
n
sequence,
which potentially generates an unstable RNA±DNA
hybrid, is distributed between a downstream (±6) and an
upstream (±9) translocated position (Figure 1, plasmid
pATC6a). In contrast, a template sequence that should
yield a stable RNA±DNA hybrid holds the ternary
complex ®xed in the downstream location (plasmid
pATC6b).
We envisaged that the multiple forms of the ternary
complex observed within the (ATC/TAG)
n
stretch are in a
dynamic equilibrium, where the readblocked polymerase
oscillates between two positions on the template with
accompanying transcript cleavage and re-synthesis.
According to this scenario, the polymerase ®rst elongates
the transcript to the point where the operator-bound lac
repressor becomes a physical barrier. Within such a
readblocked complex, the catalytic center of the enzyme is
in register with position ±6 (the location of the 3¢ terminal
RNA nucleotide) with respect to the upstream edge of the
operator motif. When the RNA±DNA hybrid within the
complex is relatively strong (5 rC±dG bp in the putative
8 bp heteroduplex), the polymerase remains ®xed at this
single site on the template. However, when the complex
contains a weak hybrid (3 rC±dG bp in the heteroduplex),
the polymerase becomes prone to slide backward. The
backsliding of RNAP relocates the catalytic center in
register with position ±9, where transcript cleavage occurs.
This generates the upstream translocated complex. Re-
elongation of the transcript from the new 3¢ end brings
RNAP back to the downstream, readblocked location.
Conceivably, transcript cleavage within the backslided
ternary complex is an intrinsic activity of polymerase, or
Fig. 1. Schematic models of a stable and an oscillating elongation
complex halted in vivo by the lac repressor bound to its operator.
The middle section of the ®gure shows the relevant part of plasmids
pATC6a and pATC6b with the non-template strand sequence of the
repeats. These plasmids, which have been described previously
(Toulme
Â
et al., 1999), are pKK232-8 derivatives that carry the
b-lactamase (Amp) and chloramphenicol acetyltransferase (CAT)
genes. The transcription of the cat gene initiated at the constitutive
hisR promoter goes through the repeats and the operator sequences that
are inserted at position +40 relative to the transcription start site.
Positions ±6 and ±9 relative to the upstream edge of the operator motif
are also indicated, as the mRNA start site (right-angled arrow). See text
for details.
Fig. 2. Nuclease S1 mapping of the 3¢ ends of RNAs produced from
pATC6a in wild-type (Wt), greA
±
, greB
±
or greA
±
greB
±
(Dble) mutant
strains. (A) Autoradiogram showing the distribution of the RNA 3¢
ends. In this and subsequent ®gures, the + or ± signs at the top of the
lanes denote the addition or omission, respectively, of the inducer
(IPTG). The arrowheads indicate the ±6 and ±9 positions of the 3¢ ends
of the transcript relative to the upstream edge of the operator motif.
(B) Densitometric scans of the S1-protected bands obtained in the
absence of IPTG, quanti®ed and processed on a PhosphorImager
(Molecular Dynamics) with Imagequant software Version 3.3 for data
processing. The data reported in the graph point to variations in the
intensity of the ±6 and ±9 signals in the greA
±
, greB
±
and greA
±
greB
±
mutant strains compared with the wild-type strain.
F.Toulme
Â
et al.
6854
the RNA trimming is induced by Gre factors. To obtain
evidence to support these possibilities, we extended the
footprinting analyses with E.coli cells that lack functional
GreA, GreB or both. The pATC6a plasmid that harbors the
oscillation-inducing template sequence was moved into a
set of four isogenic strains: wild type, greA::KanR,
DgreB::CamR and the greA
±
greB
±
double mutant. We
®rst mapped the 3¢ ends of the truncated transcripts
generated by the readblocked ternary complex.
Cellular RNAs were extracted from the transformed
cells and the RNA 3¢ ends were determined by S1 nuclease
protection experiments (see Materials and methods). In
agreement with previous results (Toulme
Â
et al., 1999), the
RNAs extracted from wild-type cells grown in the absence
of the inducer isopropyl-b-
D-thiogalactopyranoside
(IPTG) contained short transcripts with 3¢ ends distributed
primarily between positions ±6 and ±9 with respect to the
lac operator (Figure 2A). This signature of the apparent
`cleavage-and-restart' process within the oscillating com-
plex was also observed for the RNAs extracted from either
the greA
±
or the greB
±
mutants. In striking contrast, a
major band corresponding to the 3¢ end at position ±6 was
detected for the RNAs produced in the double-mutant
strain. The intensity of this band is approximately equal to
the sum of the two signals (positions ±6 and ±9) observed
for either the wild type or single mutants (Figure 2B).
These results are consistent with the hypothesis that the
RNAP readblocked within the (ATC/TAG)
n
tract back-
slides, and that it resumes elongation upon transcript
shortening promoted by GreA and GreB.
The apparent transcript cleavage activity in the read-
blocked ternary complex is more ef®cient with GreA than
with GreB. A quantitative analysis of the S1 nuclease
protection patterns obtained with the single mutants
revealed a reproducible, albeit modest, difference between
the distributions of the RNA 3¢ ends (see densitometric
scans in Figure 2B). Whereas the 3¢ ends in the greB
±
mutant are divided in nearly equal amounts between
positions ±6 and ±9, as was observed with the wild-type
strain, the RNAs produced in the greA
±
mutant have more
3¢ ends at position ±6.
Another diagnostic for the cleavage and re-synthesis
process within the oscillating complex is obtained by
analyzing the methylation status of the G residue at
position ±6 on the template strand. The in situ dimethyl-
sulfate (DMS) probing data, shown in Figure 3, reveal the
readblock-dependent hypermethylation of the G-6 residue
in wild-type cells and single mutants, but clearly not in the
double mutant. The extent of methylation of the G-6
residue in the double mutant is only slightly greater than
that of the surrounding G residues. These results strongly
support our previous suggestion that the hyper-reactivity
of the G-6 residue with DMS is associated with the
dynamic equilibrium, the template base residue being
repeatedly blocked in a highly exposed con®guration
during the re-incorporation of the rC residue into the
transcript (Toulme
Â
et al., 1999). Thus, the absence of both
GreA and GreB renders the stalled elongation complex
transcriptionally quiescent by impairing the dynamic
process of cleavage and re-synthesis between the ±6 and
±9 positions.
Interestingly, these in situ DMS probing experiments
(Figure 3) also show a slightly lower hypermethylation of
the G-6 residue in the greA
±
cells compared with the greB
±
or wild-type strains (densitometric scans not shown). In
agreement with the S1 mapping data described above,
these results emphasize a lower frequency of rC-6 re-
incorporation in the greA
±
strain due to the reduced
ef®ciency of transcript cleavage.
The catalytic competence of the backslided RNA
polymerase requires functional GreA or GreB
The molecular properties of the ternary complex generated
in the greA
±
greB
±
double mutant with plasmid pATC6a
could be readily explained if the readblocked RNAP is no
longer sliding back and forth within the (ATC/TAG)
n
tract, but instead is trapped in the downstream position.
Fig. 3. In situ DMS modi®cation patterns of the template strand of
pATC6a in wild-type (Wt), greA
±
, greB
±
and greA
±
greB
±
(Dble)
mutant strains. The position of the G-6 residue is indicated by an
arrow.
Fig. 4. Primer extension analysis of CAA modi®cation on the non-
template strand of pATC6a in wild-type (Wt) or greA
±
greB
±
(Dble)
mutant strains. The two arrows at positions ±6 and ±18 delimit the
apparent footprint of the complex, which is in equilibrium between
the upstream and downstream conformations.
In vivo function of GreA and GreB
6855
Indeed, similar characteristics were found previously in
wild-type cells for a ternary complex that is stably
positioned at the downstream location (as in the case of
the pATC6b variant). Alternatively, we imagine that in the
absence of GreA and GreB, the ternary complex in
pATC6a is still repeatedly switching between the up-
stream and downstream positions without breaking the 3¢
tip of the transcript. A distinguishing feature between the
oscillating and non-oscillating elongation complexes is the
apparent size of the transcription bubble. Whereas in the
laterally stabilized complex the DNA bubble is limited to
8 bp, the oscillating complex is characterized by an
appropriately larger DNA bubble (Toulme
Â
et al., 1999).
Therefore, we extended our comparative studies by
analyzing the accessibility of the non-template strand to
the single-strand-speci®c chemical probe chloroacetalde-
hyde (CAA).
Escherichia coli cells bearing pATC6a were treated
with CAA, and following plasmid DNA extraction the
modi®ed bases on the non-template strand were revealed
by primer extension. As shown in Figure 4, the CAA
reactivity patterns of pATC6a modi®ed in the wild-type
and double-mutant strains are virtually identical, although
the modi®cations at the upstream and downstream margins
of the footprint differ slightly between the two strains. The
CAA reactivities are spread over a region of ~14 bases
located upstream from position ±6. This apparent acces-
sibility to the probe of a large region of the non-template
strand must result from the superposition of the different
forms of the complex that are in equilibrium. As shown
above (Figure 2), the ternary complexes of pATC6a in the
greA
±
greB
±
double mutant have a single 3¢-ended
transcript (position ±6). Thus, these CAA probing results
indicate that in the absence of functional Gre factors, the
ternary complexes readblocked within the (ATC/TAG)
n
stretch do slide back and forth between the upstream and
downstream locations; however, they do not shorten the 3¢
terminal part of the transcript. Notably, the slight shift of
the CAA footprint towards the upstream positions
observed in the double mutant might re¯ect the fact that
the oscillating ternary complexes spend more time at the
backward location when GreA and GreB are absent
(Figure 4). In the view where the backward sliding of
the polymerase extrudes the RNA 3¢ tip from the active
center, the upstream form of the complex should be
catalytically inactive. Evidence supporting this notion is
presented below.
The operator-bound lac repressor is expected to cause
only a transient block to the emergent elongation complex.
Once the two complexes are in close apposition on the
DNA, the duration of the block is a function of the
dwelling time of the lac repressor on its binding site.
However, the ef®ciency with which the polymerase will
read through the operator motif, following repressor
dissociation, depends upon whether the ternary complex
is able to escape rapidly from this position before the lac
repressor re-binds DNA. Thus, the level of transcription
readthrough under repressing conditions should be related
to the catalytic competence of the halted ternary complex.
Therefore, we measured the amounts of downstream cat
mRNA produced in the different strains harboring
pATC6a. A
32
P end-labeled oligonucleotide complemen-
tary to the early part of the cat message was used to
quantitate mRNA by primer extension with reverse
transcriptase. A second oligonucleotide was also included
to detect the plasmid-encoded b-lactamase transcript (bla)
for normalization. Under repressing conditions, a fraction
of RNAPs did transcribe beyond the operator motif, and
this level of readthrough was clearly affected by the
absence of both GreA and GreB (Figure 5A). Curiously,
the gre mutations showed a defect in cat gene expression
that is independent from the lac repressor impediment,
since a reduction in cat mRNA was also observed under
inducing conditions. We have not addressed the molecular
basis of this defect. We note, however, that this could be
due to a reduced frequency of RNA chain initiation in the
gre mutants (Hsu et al., 1995; A.Das et al., manuscript in
preparation). Alternatively, in the absence of GreA and
GreB, RNA chain elongation could have been delayed at
some natural pause sites within the early part of the cat
Fig. 5. cat mRNA quantitation in wild-type (Wt), greA
±
, greB
±
and
greA
±
greB
±
(Dble) mutant cells transformed with either pATC6a or
pATC6b. (A) Autoradiogram of the extension products obtained after
in vitro reverse transcription of the RNAs with
32
P primers hybridizing
to either the cat or the bla transcripts. The asterisk indicates a non-
speci®c site of arrest during reverse transcription of the cat mRNA.
(B) Quanti®cation of the extension products shown in (A) was
performed and processed based on densitometric scans as described in
Figure 2. For each lane, the value plotted indicates the relative level of
cat mRNA synthesis, calculated as explained in the text. The error bars
re¯ect the standard deviation from a mean of three independent
experiments. The 1.0 value for the wild type is an arbitrary unit. On the
right-hand side of the histogram, the data obtained with the pATC6b
plasmid introduced in the wild-type and the double-mutant strains is
also reported to underline the Gre factor-independent readthrough when
the ternary complex is stabilized in the downstream position
(experimental data not shown).
F.Toulme
Â
et al.
6856
gene. To evaluate speci®cally the readblock-dependent
effects of the Gre factors, we normalized, for each strain,
the level of cat mRNA produced under repressing
conditions to that obtained under inducing conditions.
This analysis revealed a clear defect of RNAP in
overcoming the readblock in the absence of GreA and
GreB (Figure 5B). Whereas the ef®ciency of readthrough
was almost unaffected in the greB
±
mutant compared with
the wild-type cells, the greA
±
strain displayed a 30%
reduction of cat mRNA. By comparison, the readthrough
was reduced by 60% in the double mutant.
We infer that the ternary complex in pATC6a slides
back and forth between a catalytically dormant (upstream)
and an active form (downstream). Upon repressor dis-
sociation, only the active form escapes before the lac
repressor re-binds. By promoting transcript trimming and
realignment of the catalytic register, the Gre factors
reactivate the backtracked ternary complex more ef®-
ciently than do the oscillations without cleavage. Our
conclusion is clearly supported by the results with the
pATC6b variant (Figure 5B). In this case, where the
ternary complex is ®xed in the downstream location
(active form), the ef®ciency of transcription readthrough is
virtually unaffected by the absence of the two Gre factors.
Discussion
Transcription processivity, essential for gene expression
and control, relies on the perplexing grips of RNAP that
hold the template DNA and the nascent transcript in a
stable but ¯exible manner. The stability of these grips
prevents the random dissociation of the ternary complex,
while their ¯exibility enables the catalytic center to rapidly
reposition itself (15±50 ms) at every cycle of nucleotide
addition. During typical elongation, the polymerase has
little or no reason to march backward. Yet, RNAP does
slide backward at certain template sites, when it is
deprived of nucleotides, or when RNAP encounters
some natural pause sites or a physical barrier imposed
by a protein bound on the DNA. The backsliding event
renders RNAP catalytically incompetent owing to the out-
of-register catalytic center. As a result, transcription stalls
for varying periods of time until RNAP regains catalytic
activity. The experiments presented in this paper show that
elongation factors GreA and GreB play a signi®cant role in
the cell to revive RNAP from its backslided, catalytically
incompetent state. They do so by promoting transcript
cleavage, which generates a new 3¢ terminus of the
transcript in register with the catalytic center (Figure 1).
We have shown that the backslided RNAP fails to trim
the RNA 3¢ terminus in vivo when both GreA and GreB are
absent. This contrasts with previous in vitro results
demonstrating that both E.coli RNAP and eukaryotic
RNAP II possess intrinsic transcript cleavage activity
(Rudd et al., 1994; Orlova et al., 1995). Notably, in these
in vitro experiments, transcript hydrolysis at internal
positions was induced by speci®c conditions, such as NTP
starvation and high pH or the presence of pyrophosphate
activity. These studies suggested that the endonucleolytic
cleavage reaction is probably catalyzed by the same active
center that performs RNA synthesis. Our ®ndings do not
exclude this possibility, but they clearly indicate that
within the cellular environment of E.coli, the interaction of
GreA or GreB with the ternary complex is essential for the
cleavage reaction to occur. We suspect that this rule is
likely to hold true for eukaryotes as well.
GreA and GreB are closely related proteins that share a
substantial amino acid sequence homology, have a similar
structural organization and are conserved in the bacterial
kingdom (Koulich et al., 1997). However, the two proteins
appear to act differently within ternary complexes in vitro.
Whereas GreA-induced transcript cleavage produces short
RNAs, two or three nucleotides long, GreB (like SII)
produces 2±18 nucleotide RNA fragments (Borukhov
et al., 1992, 1993; Reines, 1992; Reines et al., 1992).
Indeed, our in vivo results presented here reveal a
differential sensitivity of RNAP for the Gre factors, and
they provide evidence that this differential speci®city is
biologically signi®cant. While the two factors clearly
substitute for each other in reviving backtracked RNAP in
our system, GreA is more ef®cient than GreB. Apparently,
GreA is more specialized for modulating RNAP where
backtracking is limited over a short distance. The
molecular mechanism underlying this specialization is
still unclear. Recent experiments suggest that the func-
tional property of the two Gre proteins is determined in
part by the size of the respective basic patch residing on
the N-terminal coiled-coil domain (Kulish et al., 2000).
The `Zipping' model of transcript elongation (Reeder
and Hawley, 1996; Komissarova and Kashlev, 1997a,b;
Nudler et al., 1997; Toulme
Â
et al., 1999) holds that the
lateral stability of the ternary complex at each template
position is governed by the direct competition between the
DNA±DNA and RNA±DNA base pairs that can form at
the upstream and downstream branching points of the
transcription bubble. Accordingly, we expect that the
back-and-forth oscillations of the ternary complex over
short distances, such as that operational in our experimen-
tal system, are relatively common under the dynamic
conditions of transcription in vivo. RNAP oscillations
should be induced at template sequences where the
competition between the base pairs at the two branching
points of the bubble is unfavorable to elongation.
Furthermore, misincorporation, which hinders further
catalysis, should favor a backward slide of RNAP. Our
results show clearly that the oscillating complex relies
heavily on the `cleavage-and-restart' mechanism mediated
by Gre factors to escape from the protein readblock. In the
absence of GreA and GreB, the ternary complex spends
time switching between active and inactive isoforms.
Thus, the Gre proteins are processivity factors that
preclude RNAP from being retarded at positions where
the lateral stability of the ternary complex is impaired.
They probably function to prevent RNAP from extensive
backtracking that would lead to an irreversible arrest.
Finally, the Gre factors must also play a signi®cant role in
editing the misincorporated RNAs in the cell. It would be
surprising if bacteria do not utilize GreA and GreB to
control speci®c genes by suppression of transcription
arrest and RNA editing.
The disruption of the two gre genes in E.coli is not lethal
under normal growth conditions, but renders the cells
thermosensitive (Orlova et al., 1995). It is tempting to
speculate that the lateral oscillations of the ternary
complex are aggravated by high temperature, which in
turn makes the cell more dependent on the Gre factors for
In vivo function of GreA and GreB
6857
proper gene expression and control (A.Das et al., manu-
script in preparation). We imagine that the high tempera-
ture defect might be ampli®ed by some genetic alterations
in RNAP that favor its propensity for backward sliding,
either directly or indirectly. This would certainly explain
why certain RNAP mutants confer high temperature
lethality that is rescued by an elevated level of GreA
(Sparkowski and Das, 1990, 1992).
Materials and methods
Enzymes, chemicals and oligonucleotides
Restriction enzymes and T4 polynucleotide kinase were obtained from
New England Biolabs. The Klenow fragment of DNA polymerase I was
purchased from Amersham. Superscript RNase H
±
Reverse Transcriptase
from M-MLV, S1 nuclease and T4 DNA ligase were from Life
Technologies. DMS and most chemicals, including antibiotics, were
from Sigma. CAA was bought from Fluka Chemie and double distilled
before use (boiling point 78±80°C). Unlabeled dNTPs were from
Boehringer Mannheim, whereas [a-
32
P]dATP, [g-
32
P]ATP and all the
oligonucleotides were from Amersham.
Bacterial strains and plasmids
The four E.coli recipient strains used are isogenic to W3110 (E.coli
Genetic Stock Center). The AD8782, AD8786 and AD8775 strains harbor
the greA
±
(greA::KanR), the greB
±
(DgreB::CamR) and the greA
±
greB
±
double mutations, respectively (A.Das et al., manuscript in preparation).
In all subsequent in vivo experiments, the strains were grown with the
appropriate antibiotics at the following concentrations: kanamycin,
18 mg/ml; chloramphenicol, 15 mg/ml.
The pATC6a and pATC6b plasmids described previously (Toulme
Â
et al., 1999) are pKK232-8-based plasmids. These constructs contain a
constitutive promoter that drives transcription of the cat gene through the
(ATC/TAG)
n
repeats and the lac operator motif (Figure 1). The
introduction of either of these plasmids into the strains described above
is via selection on agar plates containing 100 mg/ml ampicillin in addition
to the aforementioned antibiotics. A second plasmid (pAC184IQ) that
overproduces lac repressor was then introduced in these transformants by
selecting tetracycline (16 mg/ml) and ampicillin resistance. The
pAC184IQ plasmid was created by subcloning the lacIQ gene from
pAC177IQ (Toulme
Â
et al., 1999) into the PvuII (position 515) and StyI
(position 2863) restriction sites of pACYC184 (New England Biolabs),
thus deleting the cat gene. In general, all DNA manipulations and
transformations into E.coli cells were performed according to standard
procedures (Sambrook et al., 1989).
In situ DNA footprinting and RNA analyses
The in situ DNA probing with DMS or CAA and the subsequent analyses
of the modi®cations by hot piperidine cleavage (DMS) or primer
extension with the Klenow fragment of DNA polymerase I (CAA) were
carried out as described (Gue
Â
rin et al., 1996; Toulme
Â
et al., 1999). The
RNA extraction and the 3¢ end mapping with S1 nuclease protection
experiments were also performed exactly as reported previously (Toulme
Â
et al., 1999).
Quantitative analyses of the cat and bla transcripts were carried out
simultaneously with two
32
P end-labeled primers that anneal between
positions 4850 and 4868 for the bla gene and between positions 248 and
265 for the cat gene (the positions are those of the original vector
pKK232-8). Twenty micrograms of total RNAs were mixed in water with
0.1 pmol of each labeled primer in a volume of 10 ml. The samples were
heat denatured at 90°C for 2 min and chilled on ice. Four microliters of
53 reverse transcription buffer (250 mM Tris±HCl pH 8.3, 375 mM
KCl, 15 mM MgCl
2
), 2 ml of 100 mM dithiothreitol and 5 U of RNase
inhibitor (RNAguardÔ from Amersham) were then added prior to
hybridization at 45°C for 20 min. cDNA synthesis was initiated by the
addition of a mixture containing 4 ml of dNTPs (5 mM each) and 100 U
of reverse transcriptase from M-MLV, and incubation was continued at
45°C for 1 h. Nucleic acids were then ethanol precipitated, dried, and
®nally resolved on a 6% denaturating polyacrylamide gel. Note that the
cat primer also hybridizes to the chromosome-encoded cat transcript
produced in the AD8786 and AD8775 strains. However, the resulting
cDNAs are easily identi®ed by their smaller size and are not present on
the part of the gel shown in Figure 5A.
Acknowledgements
This article is dedicated to the memory of Marc Leng who died on May 7,
2000. We are grateful to Marc Boudvillain and Martine Gue
Â
rin for helpful
discussions. This work was supported in part by l'Association de la
Recherche sur le Cancer, la Ligue Contre le Cancer (comite
Â
du Loiret)
and by the USPHS Grant (GM28946) from the National Institutes of
Health to (A.D.).
References
Artsimovitch,I. and Landick,R. (2000) Pausing by bacterial RNA
polymerase is mediated by mechanistically distinct classes of
signals. Proc. Natl Acad. Sci. USA, 97, 7090±7095.
Borukhov,S., Polyakov,A., Nikiforov,V. and Goldfarb,A. (1992) GreA
protein: a transcription elongation factor from Escherichia coli. Proc.
Natl Acad. Sci. USA, 89, 8899±8902.
Borukhov,S., Sagitov,V. and Goldfarb,A. (1993) Transcript cleavage
factors from E.coli. Cell, 72, 459±466.
Chedin,S., Riva,M., Schultz,P., Sentenac,A. and Carles,C. (1998) The
RNA cleavage activity of RNA polymerase III is mediated by an
essential TFIIS-like subunit and is important for transcription
termination. Genes Dev., 12, 3857±3871.
Das,A. (1993) Control of transcription termination by RNA-binding
proteins. Annu. Rev. Biochem., 62, 893±930.
Davenport,R.J., Wuite,G.J.L., Landick,R. and Bustamante,C. (2000)
Single-molecule study of transcriptional pausing and arrest by E.coli
RNA polymerase. Science, 287, 2497±2500.
Erie,D.A., Hajiseyedjavadi,O., Young,M.C. and von Hippel,P.H. (1993)
Multiple RNA polymerase conformations and GreA: control of the
®delity of transcription. Science, 262, 867±873.
Gue
Â
rin,M., Leng,M. and Rahmouni,A.R. (1996) High resolution
mapping of E.coli transcription elongation complex in situ reveals
protein interactions with the non-transcribed strand. EMBO J., 15,
5397±5407.
Hagler,J. and Schuman,S. (1992) Stability of ternary transcription
complexes of vaccinia virus RNA polymerase at promoter-proximal
positions. J. Biol. Chem., 267, 7644±7654.
Henkin,T.M. (1996) Control of transcription termination in prokaryotes.
Annu. Rev. Genet., 30, 35±57.
Hsu,L.M., Vo,N.V. and Chamberlin,M.J. (1995) Escherichia coli
transcript cleavage factors GreA and GreB stimulate promoter
escape and gene expression in vivo and in vitro. Proc. Natl Acad.
Sci. USA, 92, 11588±11592.
Jeon,C. and Agarwal,K. (1996) Fidelity of RNA polymerase II
transcription controlled by elongation factor TFIIS. Proc. Natl Acad.
Sci. USA, 93, 13677±13682.
Kassavetis,G.A. and Geiduschek,E.P. (1993) RNA polymerase marching
backward. Science, 259, 944±945.
Komissarova,N. and Kashlev,M. (1997a) Transcriptional arrest:
Escherichia coli RNA polymerase translocates backward, leaving
the 3¢ end of the RNA intact and extruded. Proc. Natl Acad. Sci. USA,
94, 1755±1760.
Komissarova,N. and Kashlev,M. (1997b) RNA polymerase switches
between inactivated and activated states by translocating back and
forth along the DNA and the RNA. J. Biol. Chem., 272, 15329±15338.
Koulich,D., Orlova,M., Malhotra,A., Sali,A., Darst,S.A. and
Borukhov,S. (1997) Domain organization of Escherichia coli
transcript cleavage factors GreA and GreB. J. Biol. Chem., 272,
7201±7210.
Kulish,D., Lee,J., Lomakin,I., Nowicka,B., Das,A., Darst,S., Normet,K.
and Borukhov,S. (2000) The functional role of basic patch, a structural
element of Escherichia coli transcript cleavage factors GreA and
GreB. J. Biol. Chem., 275, 12789±12798.
Landick,R. (1997) RNA polymerase slides home: pause and termination
site recognition. Cell, 88, 741±744.
Landick,R., Turnbough,C.L.,Jr and Yanofsky,C. (1996) Transcription
attenuation. In Neidhardt,F.C. (ed.), Escherichia coli and Salmonella
typhimurium. American Society for Microbiology, Washington, DC,
pp. 1263±1286.
Nudler,E. (1999) Transcription elongation: structural basis and
mechanisms. J. Mol. Biol., 288, 1±12.
Nudler,E., Mustaev,A., Lukhtanov,E. and Goldfarb,A. (1997) The
RNA±DNA hybrid maintains the register of transcription by
preventing backtracking of RNA polymerase. Cell, 89, 33±41.
Orlova,M., Newlands,J., Das,A., Goldfarb,A. and Borukhov,S. (1995)
F.Toulme
Â
et al.
6858
Intrinsic transcript cleavage activity of RNA polymerase. Proc. Natl
Acad. Sci. USA, 92, 4596±4600.
Reeder,T.C. and Hawley,D.K. (1996) Promoter proximal sequences
modulate RNA polymerase II elongation by a novel mechanism. Cell,
87, 767±777.
Reines,D. (1992) Elongation factor-dependent transcript shortening by
template-engaged RNA polymerase II. J. Biol. Chem., 267,
3795±3800.
Reines,D. (1994) Nascent RNA cleavage by transcription elongation
complexes. In Conaway,R.C. and Conaway,J.W. (eds), Transcription:
Mechanisms and Regulation. Raven Press, New York, NY, pp.
263±278.
Reines,D., Ghanouni,P., Li,Q.Q. and Mote,J.,Jr (1992) The RNA
polymerase II elongation complex factor-dependent transcription
elongation involves nascent RNA cleavage. J. Biol. Chem., 267,
15516±15522.
Rudd,M.D., Izban,M.G. and Luse,D.S. (1994) The active site of RNA
polymerase II participates in transcript cleavage within arrested
ternary complexes. Proc. Natl Acad. Sci. USA, 91, 8057±8061.
Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A
Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, NY.
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.
Sastry,S.S. and Ross,B.M. (1997) Nuclease activity of T7 RNA
polymerase and the heterogeneity of transcription elongation
complexes. J. Biol. Chem., 272, 8644±8652.
Sidorenkov,I., Komissarova,N. and Kashlev,M. (1998) Crucial role of
the RNA:DNA hybrid in the processivity of transcription. Mol. Cell, 2,
55±64.
Sparkowski,J. and Das,A. (1990) The nucleotide sequence of greA, a
suppressor gene that restores growth of an Escherichia coli RNA
polymerase mutant at high temperature. Nucleic Acids Res., 18, 6443.
Sparkowski,J. and Das,A. (1992) Simultaneous gain and loss of
functions caused by a single amino acid substitution in the b
subunit of Escherichia coli RNA polymerase: suppression of nusA and
rho mutations and conditional lethality. Genetics, 130, 411±428.
Surratt,C.K., Milan,S.C. and Chamberlin,M.J. (1991) Spontaneous
cleavage of RNA in ternary complexes of Escherichia coli RNA
polymerase and its signi®cance for the mechanism of transcription.
Proc. Natl Acad. Sci. USA, 88, 7983±7987.
Thomas,M.J., Platas,A.A. and Hawley,D.K. (1998) Transcriptional
®delity and proofreading by RNA polymerase II. Cell, 93, 627±637.
Toulme
Â
,F., Gue
Â
rin,M., Robichon,N., Leng,M. and Rahmouni,A.R.
(1999) In vivo evidence for back and forth oscillations of the
transcription elongation complex. EMBO J., 18, 5052±5060.
Tschochner,H. (1996) A novel RNA polymerase I-dependent RNase
activity that shortens nascent transcripts from the 3¢ end. Proc. Natl
Acad. Sci. USA, 93, 12914±12919.
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.
von Hippel,P.H. (1998) An integrated model of the transcription
complex in elongation, termination and editing. Science, 281,
660±665.
Wang,D. and Hawley,D.K. (1993) Identi®cation of a 3¢±5¢ exonuclease
activity associated with human RNA polymerase II. Proc. Natl Acad.
Sci. USA, 90, 843±847.
Received September 14, 2000; revised and accepted October 24, 2000
In vivo function of GreA and GreB
6859