The mechanism of release of P-TEFb and HEXIM1 from the 7SK snRNP by viral and cellular activators includes a conformational change in 7SK.

Abstract

The positive transcription elongation factor, P-TEFb, is required for the production of mRNAs, however the majority of the factor is present in the 7SK snRNP where it is inactivated by HEXIM1. Expression of HIV-1 Tat leads to release of P-TEFb and HEXIM1 from the 7SK snRNP in vivo, but the release mechanisms are unclear.
We developed an in vitro P-TEFb release assay in which the 7SK snRNP immunoprecipitated from HeLa cell lysates using antibodies to LARP7 was incubated with potential release factors. We found that P-TEFb was directly released from the 7SK snRNP by HIV-1 Tat or the P-TEFb binding region of the cellular activator Brd4. Glycerol gradient sedimentation analysis was used to demonstrate that the same Brd4 protein transfected into HeLa cells caused the release of P-TEFb and HEXIM1 from the 7SK snRNP in vivo. Although HEXIM1 binds tightly to 7SK RNA in vitro, release of P-TEFb from the 7SK snRNP is accompanied by the loss of HEXIM1. Using a chemical modification method, we determined that concomitant with the release of HEXIM1, 7SK underwent a major conformational change that blocks re-association of HEXIM1.
Given that promoter proximally paused polymerases are present on most human genes, understanding how activators recruit P-TEFb to those genes is critical. Our findings reveal that the two tested activators can extract P-TEFb from the 7SK snRNP. Importantly, we found that after P-TEFb is extracted a dramatic conformational change occurred in 7SK concomitant with the ejection of HEXIM1. Based on our findings, we hypothesize that reincorporation of HEXIM1 into the 7SK snRNP is likely the regulated step of reassembly of the 7SK snRNP containing P-TEFb.

Full text

Page 1

The Mechanism of Release of P-TEFb and HEXIM1 from
the 7SK snRNP by Viral and Cellular Activators Includes a
Conformational Change in 7SK
Brian J. Krueger1, Katayoun Varzavand2, Jeffrey J. Cooper2, David H. Price1,2*
1Molecular and Cellular Biology Program, University of Iowa, Iowa City, Iowa, United States of America, 2Biochemistry Department, University of Iowa, Iowa City, Iowa,
United States of America
Abstract
Background: The positive transcription elongation factor, P-TEFb, is required for the production of mRNAs, however the
majority of the factor is present in the 7SK snRNP where it is inactivated by HEXIM1. Expression of HIV-1 Tat leads to release
of P-TEFb and HEXIM1 from the 7SK snRNP in vivo, but the release mechanisms are unclear.
Methodology/Principal Findings: We developed an in vitro P-TEFb release assay in which the 7SK snRNP
immunoprecipitated from HeLa cell lysates using antibodies to LARP7 was incubated with potential release factors. We
found that P-TEFb was directly released from the 7SK snRNP by HIV-1 Tat or the P-TEFb binding region of the cellular
activator Brd4. Glycerol gradient sedimentation analysis was used to demonstrate that the same Brd4 protein transfected
into HeLa cells caused the release of P-TEFb and HEXIM1 from the 7SK snRNP in vivo. Although HEXIM1 binds tightly to 7SK
RNA in vitro, release of P-TEFb from the 7SK snRNP is accompanied by the loss of HEXIM1. Using a chemical modification
method, we determined that concomitant with the release of HEXIM1, 7SK underwent a major conformational change that
blocks re-association of HEXIM1.
Conclusions/Significance: Given that promoter proximally paused polymerases are present on most human genes,
understanding how activators recruit P-TEFb to those genes is critical. Our findings reveal that the two tested activators can
extract P-TEFb from the 7SK snRNP. Importantly, we found that after P-TEFb is extracted a dramatic conformational change
occurred in 7SK concomitant with the ejection of HEXIM1. Based on our findings, we hypothesize that reincorporation of
HEXIM1 into the 7SK snRNP is likely the regulated step of reassembly of the 7SK snRNP containing P-TEFb.
Citation: Krueger BJ, Varzavand K, Cooper JJ, Price DH (2010) The Mechanism of Release of P-TEFb and HEXIM1 from the 7SK snRNP by Viral and Cellular
Activators Includes a Conformational Change in 7SK. PLoS ONE 5(8): e12335. doi:10.1371/journal.pone.0012335
Editor: Mikhail V. Blagosklonny, Roswell Park Cancer Institute, United States of America
Received April 30, 2010; Accepted August 1, 2010; Published August 23, 2010
Copyright: ? 2010 Krueger et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by National Institutes of Health grants GM35500 to D.H.P. The funders had no role in study design, data collection and
analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: david-price@uiowa.edu
Introduction
Transcription elongation by RNA polymerase II (RNAPII) is a
highly regulated process resulting from the concerted effort of both
negative and positive elongation factors. After initiation engaged
RNAPII molecules come under the control of negative factors
including NELF and DSIF that limit the elongation potential of
the polymerases and trap them in promoter proximal positions
[1,2,3]. These polymerases are poised for release into productive
elongation that ultimately generates mRNAs. The positive
transcription elongation factor, P-TEFb, is a cyclin dependent
kinase that phosphorylates the negative factors and RNAPII
leading to the transition into productive elongation [4,5]. It is clear
that P-TEFb plays a key role in the regulation of gene expression
because a number of genome wide studies have determined that
many genes in Drosophila and most genes in mammals are
occupied by poised polymerases [6,7,8,9,10,11].
The kinase activity of P-TEFb is required for the generation of
mRNAs; however, the majority of P-TEFb is sequestered within
the 7SK snRNP where it is inactivated through association with
HEXIM1 or HEXIM2 proteins [12,13,14,15]. A mechanism must
exist to extract the kinase from this inhibitory complex and direct
the function of P-TEFb to specific genes. Release of P-TEFb from
the 7SK snRNP could occur through post-translational modifica-
tion of P-TEFb or components of the 7SK snRNP. One study
showed that dephosphorylation of the T-loop of P-TEFb by PP1a
and PP2B results in its release from the 7SK snRNP [16].
Similarly, activation of the PI3K/Akt pathway through treatment
of cells with HMBA results in the phosphorylation of the cyclin T1
binding region of HEXIM1 and leads to the global release of P-
TEFb in vivo [17]. Finally, an analysis of active P-TEFb in the cell
showed that free low molecular weight P-TEFb is acetylated, while
P-TEFb bound to the 7SK snRNP is not, suggesting that
acetylation could cause release of P-TEFb from the 7SK snRNP
[18]. In addition to post-translational modifications, there is
evidence that cellular proteins recruit P-TEFb to sites of active
transcription. These include the p65/RELA subunit of NF-kB
[19], CIITA [20,21], Myc [22,23], MyoD [24,25], the androgen
PLoS ONE | www.plosone.org1August 2010 | Volume 5 | Issue 8 | e12335

Page 2

receptor [26,27], the estrogen receptor [28], and the bromodo-
main containing protein Brd4 [29,30]. However, it is not known if
any of these enzymatic modifications or protein interactions
liberates P-TEFb directly from the 7SK snRNP.
At least four viral proteins have been shown to recruit P-TEFb
to their promoters through an interaction with cyclin T1. These
include EBV E2, HSV VP16, HTLV Tax and HIV-1 Tat
[31,32,33,34]. Early work on the functional domains of Tat
showed that a cysteine rich region is required for Tat transactiva-
tion [35,36]. It was originally thought that this region was
important for Tat dimerization, but it is now known that it is the
P-TEFb binding domain [34,37,38]. The ability of Tat to bind to
the TAR element through its basic RNA binding domain is also
important for viral replication, because loss of this region results in
a significant reduction in HIV Tat transactivation [36,39]. Also of
interest, the bulge sequence that Tat binds to in TAR (AUCUG) is
repeated 3 times in the first 100 bases of 7SK RNA and seems
unlikely to be a mere coincidence considering the fact that Tat has
a greater affinity for 7SK than TAR in vitro [40]. It has been
shown that Tat can compete for HEXIM1 binding to P-TEFb and
that the cysteine rich P-TEFb binding region of Tat is required for
this to occur in vitro and in vivo [40,41]. Although details of the
interaction between Tat and P-TEFb have been recently revealed
by structural studies [38], it is not known if Tat can extract P-
TEFb directly from the 7SK snRNP, or if this release is mediated
by other proteins.
Since HIV Tat is capable of recruiting P-TEFb to a specific
genomic location and viral proteins typically mimic the function of
cellular proteins, it seems likely that endogenous proteins perform
a similar function in recruiting P-TEFb to specific sites of
transcription. One potential candidate is bromodomain-contain-
ing protein 4 (Brd4). Brd4 has been shown to interact with many
different chromatin and transcription related proteins by binding
to acetylated lysines, and most importantly its C-terminal helical
region binds directly to cyclin T1 of P-TEFb [42,43]. In fact, most
of the P-TEFb that is not in the 7SK snRNP is found complexed
with Brd4 when the proteins are extracted from nuclei at high salt
and then the salt is removed [29]. Recent evidence also shows that
Brd4, P-TEFb, and HEXIM1 localize to nuclear speckles and this
may represent a region of P-TEFb activity or a switch between
inactive and active P-TEFb [44]. There is also evidence that Brd4
and P-TEFb appear simultaneously at sites of activation [30,45];
however, the direct role that Brd4 plays in this process is not fully
understood. Because Brd4 associates with a large fraction of P-
TEFb and also binds to active chromatin, it is possible that it is
part of a general mechanism for functional recruitment of P-
TEFb.
The 7SK snRNA plays a critical role in the regulation of P-
TEFb. Its discovery as a major structural component of a P-TEFb
inhibitory complex was novel [46,47]. Further characterization of
the regions of 7SK required for P-TEFb inhibition were
performed and it was found that the 1–100 region of 7SK is
bound specifically by HEXIM1 and that the 39 stem loop is bound
to cyclin T1 [13,48]. Additionally, HEXIM1 photo-crosslinks
specifically to U30 of 7SK further underscoring the importance of
the 1–100 region in P-TEFb inhibition [49]. The protein
composition of the 7SK snRNP is dynamic; however, the La
related protein LARP7 is constitutively associated with 7SK
[50,51,52,53]. 7SK was found to form multiple RNPs with
hnRNPA1, A2/B1, R and Q along with RNA Helicase A [52,54].
The authors of these studies speculated that hnRNP proteins are
important for releasing P-TEFb from the 7SK snRNP by
competing with HEXIM1 for 7SK binding. This might be
mediated by hnRNP binding to a hairpin loop in the middle of
7SK, which when deleted prevents the release of cyclin T1 from
exogenously expressed RNAs [52]. HEXIM1 has been shown to
be a promiscuous double stranded RNA binding protein and binds
to 7SK in vitro even in the absence of P-TEFb [55], so it is odd
that HEXIM1 leaves the 7SK snRNP when P-TEFb is released in
vivo. It is possible that HEXIM is displaced by the binding of
hnRNP proteins that occurs upon P-TEFb release [51,52,54].
Results
Release of P-TEFb from the 7SK snRNP by HIV Tat in vitro
Overexpression of Tat in vivo results in the loss of P-TEFb from
the 7SK snRNP in the absence of any other HIV proteins, RNA
or DNA and in vitro Tat can bind to 7SK RNA and compete with
HEXIM1 for binding [40]. However, the ability of Tat to directly
extract P-TEFb from the 7SK snRNP has not been explored. To
test this, Tat was expressed and purified from E. coli. The 7SK
snRNP was then isolated from HeLa cell lysates using affinity
purified antibodies to LARP7 covalently attached to paramagnetic
beads. The isolated complexes were washed, aliquoted, and
recombinant protein was then incubated with the 7SK snRNP for
15 minutes before the beads were concentrated and analyzed by
Western blot for LARP7 as a control for loading and for Cdk9 to
detect release of P-TEFb. As a positive control, RNase A was
added to one reaction. Degradation of 7SK destroys the link
between LARP7 and P-TEFb and as expected RNase caused
essentially all Cdk9 to be released from the complex (Figure 1A).
Addition of 10, 30, 100, or 300 ng of Tat for 15 minutes resulted
in the graded release of P-TEFb from the 7SK snRNP (Figure 1A).
The kinetics of Tat release was also analyzed by exposure of the
7SK snRNP to 100 ng of Tat for 3, 10, or 30 minutes. More P-
TEFb was released as exposure time increased. Evidently, Tat can
cause the release of P-TEFb directly from the 7SK snRNP in the
absence of soluble cellular factors.
The effect of the mutation of the P-TEFb binding or RNA
binding domains of Tat on P-TEFb release was also determined.
The first 48 amino acids of Tat forms extensive contacts with P-
TEFb by directly interacting with mainly cyclin T1 and with two
Zn atoms coordinated by the complex [34,38]. The remainder of
the 86 amino acid protein is basic and interacts with TAR RNA
[56]. Both domains are required for recruitment of P-TEFb to
TAR and for Tat transactivation in vivo. Of interest, the bulge
sequence that Tat binds to in TAR (AUCUG) is repeated 3 times
in the first 100 bases of 7SK RNA. It was hypothesized that, in
addition to the ability of Tat to bind to P-TEFb, a competition
with HEXIM1 for binding to 7SK may also be important for Tat
mediated P-TEFb release [40]. To determine the effect of the loss
of the P-TEFb binding region, cysteine 22 that is critical for Zn
coordination was mutated to glycine (Tat C22G). An RNA
binding mutant was generated by changing two arginines that are
essential for RNA binding region to alanines (Tat R52/53A).
The effect of the loss of the RNA binding activity of Tat on P-
TEFb release was tested by titrating Tat R52/53A into reactions
containing the 7SK snRNP. Although addition of increasing
amounts of the RNA binding mutant resulted in a graded loss of
Cdk9 from the 7SK snRNP, the relative amount of release was less
than that caused by wild type Tat (Figure 1B). The same kinetic
experiment as before was performed and a similar, although
slower loss of Cdk9 was observed (Figure 1B). The role of the P-
TEFb binding domain of Tat in P-TEFb release was then
explored. Addition of increasing amounts of Tat C22G or
incubation of the 7SK snRNP with 100 ng of Tat C22G for 30
minutes abrogated the P-TEFb release activity seen with the
wildtype protein indicating that the P-TEFb binding region is
Release of P-TEFb and HEXIM1
PLoS ONE | www.plosone.org2 August 2010 | Volume 5 | Issue 8 | e12335

Page 3

required (Figure 1C). Furthermore, since the RNA binding region
of Tat was still intact in this mutant, this experiment also shows
that the RNA binding region is not sufficient to release P-TEFb
from the 7SK snRNP. The P-TEFb release experiments were
repeated a number of times and the results quantitated. The
average percent of P-TEFb remaining in the complex after
treatment with wildtype and mutant Tats is plotted in Figure 1D.
It is clear that the P-TEFb binding domain of Tat is required for
release of P-TEFb from the 7SK snRNP and that the RNA
binding domain plays a lesser, but not insignificant role in the
release.
Because the preceding results suggested that the RNA binding
region may play role in P-TEFb release from the 7SK snRNP and
because Tat binds preferentially to 7SK RNA over TAR RNA in
an in vitro assay [40], we tested the effect of adding RNAs to the
release assay. This was done by pre-incubating 100 ng of Tat with
increasing amounts of HIV TAR or 7SK RNA for 15 minutes
before the 7SK snRNP was added to these reactions. Interestingly,
TAR RNA had no effect on the ability of Tat to release P-TEFb;
however, 7SK RNA inhibited it, again indicating that Tat binding
to 7SK RNA may be important for its ability to release P-TEFb
and confirming that it has a greater affinity for 7SK than TAR
(Figure 1E). To determine the effect of the loss of the RNA binding
domain on P-TEFb release, the same experiment was performed
except this time Tat R52/53A was used. In contrast to wildtype
Tat, release of P-TEFb by Tat R52/53A was not inhibited by the
presence of 7SK RNA (Figure 1F). These results suggest that Tat
binding to 7SK snRNA may play role in P-TEFb release from the
7SK snRNP, but the RNA binding region is neither essential nor
alone sufficient to cause release.
Release of P-TEFb from the 7SK snRNP by the C-terminal
region of Brd4
Because the viral protein Tat was able to release P-TEFb directly
from the 7SK snRNP, we hypothesized that an endogenous cellular
protein might possess a similar activity. Brd4 was screened as a
potential P-TEFb release factor because under some extraction
conditions Brd4 is associated with the majority of non-7SK bound
P-TEFb in the cell and it has a known P-TEFb binding domain
[29,30,43]. The P-TEFb binding region was originally identified to
be in an N-terminal region containing one of the two bromo
domains [29], but this was later attributed to a cloning error [43].
The P-TEFb binding region is actually encoded by helical region 3
in the extreme C-terminal region of Brd4 [43]. Additionally, two
labs have shown that over expression of the C-terminal tail of Brd4
can inhibit Tat transactivation of the HIV LTR [42,43] and block
Figure 1. Release of P-TEFb from the 7SK snRNP by HIV Tat. A) P-TEFb release reactions were performed as described in Materials and Methods.
IndicatedamountsofTatwereincubatedfortheindicatedtimeswithimmunoprecipitated7SKsnRNPs.B)SameasinAexcepttheRNAbindingdeficient
mutant of Tat (Tat R52/53A) was used. C) Same as in A except the P-TEFb binding deficient mutant of Tat (Tat C22G) was used. D) Tat release was
quantified from three independent experiments for wild type Tat and the RNA binding mutant and two independent experiments for the P-TEFb
binding mutant. Mean values for the percent of Cdk9 left in the complex are plotted and all error bars represent standard error. E) 100 ng of Tat was
incubated with the indicated amounts of TAR or 7SK RNA before being added to the 7SK snRNP. F) Same as in E except TatR52/53A was used.
doi:10.1371/journal.pone.0012335.g001
Release of P-TEFb and HEXIM1
PLoS ONE | www.plosone.org3August 2010 | Volume 5 | Issue 8 | e12335

Page 4

Tat binding to P-TEFb [43]. cDNAs encoding the P-TEFb binding
region (amino acids 1209–1362) and a P-TEFb binding mutant
lacking helical region 3 (1209–1362 D1329–1345) of Brd4 were
obtained from the Verdin Lab (Figure 2A). Both proteins were
cloned into pET21a with a C-terminal histidine tag and then
expressed and purified from E. coli (Figure 2B). Addition of
increasing amounts of Brd4 1209–1362 resulted in the obvious
releaseof Cdk9 from the 7SKsnRNP(Figure 2C). Kinetic release of
Cdk9 with 200 ng of Brd4 was also observed implicating Brd4 as a
cellular P-TEFb release factor. Finally, the Brd4 mutant lacking the
C-terminal helical domain 3 of Brd4 was used in the release assay
and had no effect on Cdk9 release either by titration or time course
(Figure 2D). The Brd4 release data quantified from three
independent experiments shows that the only the intact P-TEFb
binding domain of Brd4 causes a significant release of P-TEFb from
the 7SK snRNP (Figure 2E).
To determine if the P-TEFb binding domain of Brd4 would
cause the release of P-TEFb from the 7SK snRNP in vivo, HeLa
cells were transfected with plasmids that led to the expression
FLAG tagged versions of either the wildtype P-TEFb binding
domain of Brd4 (1209–1362) or the mutant domain (1209–1362
D1329–1345). Transfection efficiencies were similar for the two
proteins and greater than 50% as evidenced by immunofluores-
cence microscopy using antibodies against the FLAG epitope
(Figure 3A). High levels of both proteins were found in the nucleus
and cytoplasm (Figure 3A). Whole cell lysates were generated from
Figure 2. Release of P-TEFb from the 7SK snRNP by the P-TEFb binding domain of Brd4. A) A schematic of Brd4 constructs used. BD1,
Bromodomain 1; BD2, Bromodomain 2; ET, Extraterminal domain; H1, H2 and H3, Helical domains 1, 2 and 3. Brd4 1209–1362 contains three helical
regions, Brd4 1209–1362 D1329–1345 is missing helical region 3 which is required for P-TEFb binding. B) Recombinant protein expressed and purified
from E. coli. A contaminating band can be seen in the Brd4 mutant preparation and this was confirmed to be an E. coli protein by Western blot (data
not shown), not full length Brd4. M, Marker; 1, Brd4 1209–1362; 2, Brd4 1209–1362 D1329–1345. C) Indicated amounts of Brd4 were added into the
release reaction for the indicated times. D) Same as in C except the Brd4 mutant missing helical domain 3 was used for the reactions. E) Brd4 release
was quantified from three independent experiments. Two independent experiments were done to calculate the mean for the Brd4 helical domain 3
mutant. The y-axis is a measure of percent of Cdk9 left in the complex. All error bars represent standard error.
doi:10.1371/journal.pone.0012335.g002
Release of P-TEFb and HEXIM1
PLoS ONE | www.plosone.org4 August 2010 | Volume 5 | Issue 8 | e12335

Page 5

mock transfected cells as well as the two Brd4 transfected cells and
analyzed by glycerol gradient sedimentation. As expected,
Western blot analysis of LARP7 and Cdk9 from the control cell
lysate indicated that most of the P-TEFb was in the more rapidly
sedimenting 7SK snRNP (Figure 3B, frax 10–12). An identical
pattern was obtained from the Brd4D expressing cells (Figure 3B).
However, expression of Brd4 (1209–1362) had a dramatic effect in
that most of the P-TEFb was now found in slower sedimenting
fractions (Figure 3B frax 4–6). As has been found before [51],
LARP7 sedimentation was unaffected by release of P-TEFb. These
in vivo findings support the in vitro results and strongly implicate
Brd4 in the release of P-TEFb from the 7SK snRNP.
Release of P-TEFb is accompanied by a large
conformational change in 7SK RNA
Re-analysis of the original data used to characterize the structure
of 7SK revealed that 7SK may exist in more than one conformation
in the cell. Based on chemical modification and nuclease sensitivity
studies, Wassarman and Steitz created a best fit structure for 7SK
[57]. They placed uracils 28 and 30, and uracils 66 and 68, which
are in the important HEXIM binding 1–100 region of 7SK, in
double stranded regions even though these bases showed sensitivity
to the uracil modifying agent CMCT (Figure 4A, 7SK W&S).
CMCT should be able to bind to and modify any bases located in
accessible loops because CMCT is a small molecule that reacts
specifically with N-3 of uracil. Since these bases were moderately
sensitive, we hypothesized that 7SK snRNA exists in more than one
conformation in the cell. Although the 7SK snRNP inhibits the
majority of P-TEFb, a significant fraction of this complex is not
normally associated with P-TEFb [51]. This suggests that there are
at least two 7SK snRNPs, one with and the other without P-TEFb,
which may stabilize different structures of 7SK.
To determine if different structures were computationally
possible, the first 100 bases of 7SK snRNA were folded using the
RNA/DNA folding program mFold v2.3 [58]. A number of
structures were generated, but many of the structures with low
energy fell into two ensembles of similar structures. Both groups
were represented in the top two most stable structures. One was
identical to the structure described by Wassarman and Steitz
(Figure 4A, 7SK W&S, DG 240.1 kcal/mole) while the second
folded into an alternate, slightly more stable structure (Figure 4A,
7SK Alt, DG 240.4 kcal/mole). The specific alternative structure
shown is used for illustrative purposes and we do not mean to imply
that the exact structure shown in Figure 4 is present in 7SK snRNP.
Interestingly, in the alternate structures U28, U30, U66, and U68
are found in CMCT accessible loops. A mixture of the two forms
would more adequately explain the Wassarman and Steitz data.
To test if this second alternative structure of 7SK RNA was
physiologically relevant, the structure of 7SK RNA before and after
P-TEFb release by flavopiridol was probed by primer extension of
CMCT treated 7SK snRNPs. This was done by scaling up the
immunoprecipitation protocol used for the release assays to obtain
400 ng of RNA for 7SK structure analysis. Two liters of HeLa cells
were grown in spinner flasks, split in half and treated with either
DMSO carrier as a control or with flavopiridol to release P-TEFb
from the 7SK snRNP. Lysates were made, LARP7 immunoprecip-
itation was performed, isolated complexes were washed, and then
the complexes were treated with CMCT to record the structure of
7SK locked in the 7SK snRNP. The RNA was isolated and then
primer extension was performed using AMV reverse transcriptase
and a radiolabeled DNA primer to the 100–120 region or to the 70–
90 region. Modification of uracil by CMCT blocks reverse
transcription beyond the modification site and results in the
production of an RNA sensitivity ladder. The 70–90 primer was
used to obtain more band separation in the 1–50 region to better
visualize CMCT sensitivity at U28 and U30. Release of P-TEFb
from the 7SK snRNP by flavopiridol treatment of cells resulted in
significant changes in the structure of 7SK. U28, U30 (Figure 4B),
U66, and U68 (Figure 4C) all became more sensitiveto CMCT after
P-TEFb release compared with the untreated control. These results
confirm that the alternative structure is similar to the structure of the
P-TEFb-free form of 7SK RNA (Figure 4A). Additionally, in the
alternative structure, U40/U41, U53/U54, and U76 remain in
single stranded regions and their sensitivities do not change
significantly after P-TEFb release, further supporting the existence
of a conformational switch in 7SK after P-TEFb release (Figure 4C).
Figure 3. The P-TEFb binding region of Brd4 causes a release of P-TEFb from the 7SK snRNP in vivo. HeLa cells were transfected with
plasmids expressing Brd4 1209–1362 (Brd4) or mutant Brd4 1209–1362 D1329–1345 (Brd4D). A) Immunofluorescence microscopy. 48 hours after
transfection cells were fixed and stained for DNA (DAPI) or the FLAG tagged Brd4 constructs (FLAG). B) Glycerol gradient sedimentation analysis. Cell
lysates were prepared 48 hr after transfection and sedimented on 5–45% glycerol gradients as described in Materials and Methods. Western blots of
fractions were probed with antibodies to LARP7 or Cdk9 as indicated.
doi:10.1371/journal.pone.0012335.g003
Release of P-TEFb and HEXIM1
PLoS ONE | www.plosone.org5August 2010 | Volume 5 | Issue 8 | e12335