Molecular Cell 21, 227–237, January 20, 2006 ª2006 Elsevier Inc.DOI 10.1016/j.molcel.2005.11.024
of hSpt5 C-Terminal Repeats Is Critical
for Processive Transcription Elongation
Tomoko Yamada,1,3Yuki Yamaguchi,1,2,3
Naoto Inukai,1,4Sachiko Okamoto,1Takashi Mura,1
and Hiroshi Handa1,*
1Graduate School of Bioscience and Biotechnology
Tokyo Institute of Technology
Japan Science and Technology Agency
Human DSIF, a heterodimer composed of hSpt4 and
hSpt5, plays opposing roles in transcription elonga-
tion by RNA polymerase II (RNA Pol II). Here, we de-
scribe an evolutionarily conserved repetitive hepta-
peptide motif (consensus = G-S-R/Q-T-P) in the
C-terminal region (CTR) of hSpt5, which, like the
C-terminal domain (CTD) of RNA Pol II, is highly phos-
phorylated by P-TEFb. Thr-4 residues of the CTR re-
peats are functionally important phosphorylation
sites. In vitro, Thr-4 phosphorylation is critical for the
elongation activation activity of DSIF, but not to its
elongation repression activity. In vivo, Thr-4 phos-
phorylation is critical for epidermal growth factor
(EGF)-inducible transcription of c-fos and for efficient
progression of RNA Pol II along the gene. We consider
this phosphorylation to be a switch that converts DSIF
CTD’’ hypothesis, in which phosphorylated CTR is
thought to function in a manner analogous to phos-
phorylated CTD, serving as an additional code for ac-
tive elongation complexes.
It has been believed that transcription is largely con-
trolled at the step of preinitiation complex assembly on
promoter DNA. However, a number of recent studies us-
ing chromatin immunoprecipitation (ChIP) have shown
that the degree of preinitiation complex assembly
does not correlate well with the level of transcription
(Andrulis et al., 2000; Barboric et al., 2001; Eberhardy
and Farnham, 2002; Sawado et al., 2003; Soutoglou
and Talianidis, 2002; Wang et al., 2005). Significant
amounts of RNA Pol II are sometimes associated with
promoter regions of transcriptionally inactive genes.
These findings suggest that a postassembly step, per-
haps transcription elongation, is rate limiting and impor-
tant for transcriptional regulation.
We and others have shown biochemically that tran-
scription elongation is controlled both positively and
negatively by a number of transcription elongation fac-
tors (reviewed in Conaway et al.  and Hartzog
). Among these, DSIF and NELF bind to RNA Pol
II together and repress transcription elongation in the
promoter-proximal region (Wada et al., 1998a; Yamagu-
chi et al., 1999a). P-TEFb, a protein kinase composed
of Cdk9 and a cyclin subunit, reverses this repression
by phosphorylating the RNA Pol II CTD (Wada et al.,
1998b; Yamaguchi et al., 1999a). CTD phosphorylation
leads to the dissociation of NELF, after which DSIF is in-
stead able to activate transcription elongation by an as
yet unknown mechanism (Wada et al., 1998a; Guo
et al., 2000). These factors have been implicated in the
regulation ofadiversearrayofgenes:somethat arerap-
idly induced by extracellular stimuli (Ainbinder et al.,
2004; Aiyar et al., 2004; Andrulis et al., 2000; Barboric
et al., 2001; Eberhardy and Farnham, 2002; Kaplan
et al., 2000; Lis et al., 2000; Wu et al., 2003), some that
are controlled during development (Guo et al., 2000;
Jennings et al., 2004; Simone et al., 2002), and some
that are encoded by pathogenic viruses (Wei et al.,
1998; Yamaguchi et al., 2001).
The RNA Pol II CTD in humans consists of the hepta-
peptide motif Y-S-P-T-S-P-S repeated 52 times. During
ylated by different protein kinases, such as P-TEFb and
TFIIH (Komarnitsky et al., 2000), and this phosphoryla-
tion is known to play multiple roles in mRNA synthesis,
tional repression and stimulation of mRNA processing
(reviewed in Orphanides and Reinberg ). In fact,
the phosphorylated CTD can stimulate capping, splic-
sibly by serving as a scaffold for other factors (reviewed
in Hirose and Manley ).
As mentioned above, DSIF plays opposing roles in
transcription elongation. In vitro, DSIF is clearly capable
of both repressing and activating transcription under
different assay conditions (Wada et al., 1998a; Guo
et al., 2000). Recent studies on Drosophila heat-shock
genes have indicated that the idea also holds true in
vivo (Andrulis et al., 2000; Kaplan et al., 2000; Wu
et al., 2003). It may be that DSIF acts as a repressor
and an activator in promoter-proximal and -distal re-
gions, respectively, undergoing a transition from a re-
pressor to an activator during the course of transcrip-
tion. It is not known, however, what might trigger such
a transition. DSIF is composed of two subunits, Spt4
and Spt5, both of which are widely conserved among
eukaryotes (Guo et al., 2000; Kaplan et al., 2000; Pei
and Shuman, 2002; Shim et al., 2002; Swanson et al.,
1991; Wada et al., 1998a). Similar to RNA Pol II, Spt5 is
highly phosphorylated by P-TEFb. Although the exact
phosphorylation sites and their functional roles remain
elusive, several laboratories have shown that the CTR
(a.k.a. CTR1 and CTD) of Spt5 is a phosphorylation tar-
get (Ivanov et al., 2000; Kim and Sharp, 2001; Lavoie
etal., 2001;Pei andShuman, 2003). Inthe presentstudy,
3These authors contributed equally to this work.
we show that P-TEFb phosphorylates Thr residues in
a series of evolutionarily conserved pentapeptide re-
peats within the CTR and that this phosphorylation is
critical for processive transcription elongation.
Reexamining the CTR of hSpt5
Although it has been proposed that the Spt5 CTR has
a repeat structure, due to a strongly biased amino acid
composition in the CTR, the literature is inconsistent
as to what the repeating unit consists of (Guo et al.,
2000; Ivanov et al., 2000; Wada et al., 1998a). We there-
ing on its evolutionary conservation and determined the
consensus to be G-S-R/Q-T-P (Figure 1A, see also Fig-
ure S1 available in the Supplemental Data with this arti-
cle online). About seven repeats (including variants) are
present in Spt5 proteins of fission yeast, fungi, and ver-
tebrates. This consensusmotif is also found in Spt5 pro-
teins of insects and plants, whereas in budding yeast,
there are three variants but no consensus motif. The
pentapeptide motif of the Spt5 CTR is similar to the hep-
tapeptide motif of the RNA Pol II CTD in that both con-
tain Ser, Thr, and Pro residues. Differences include the
facts that the CTR repeats are not contiguous and that
these repeats are not located at the extreme C terminus.
Because Ser-2 and Thr-4 of the pentapeptide are poten-
tial phosphorylation sites, we constructed a series of
hSpt5 mutants bearing amino acid substitutions at
Ser-2 and/or Thr-4 (Figure 1B). hSpt5 and its mutants
were expressed in bacteria, purified (Figure 1C), and
used for the following experiments.
CTR Thr-4 Is Required for Activation Activity,
but Not for Repression Activity, of DSIF
The repression and activation activities of DSIF can
be measured by depletion and add-back experiments
using HeLa nuclear extracts (NE) or DSIF-depleted
NE (NEDDSIF). Anti-hSpt5 antibody immunodepletes
both subunits of DSIF from NE (Kim et al., 2003). To ex-
amine the repression activity of DSIF, we carried out
Figure 1. Structure of the hSpt5 CTR and Its
motif of the RNA Pol II CTD is also shown for
(B)Partialsequencesof wild-type andmutant
hSpt5 proteins used in this study. Schematic
structure of hSpt5 is shown at the top. Boxes
in amino acid sequence indicate the penta-
(C) hSpt5 and its mutants that were ex-
pressed in bacteria and purified extensively.
transcription reactions in the presence of DRB by using
a DNA template that produces relatively short (380 nt)
transcripts (Figure 2A). DRB, a P-TEFb kinase inhibitor,
was included in the reactions such that DSIF-mediated
repression is not reversed by endogenous P-TEFb. As
reported previously (Wada et al., 1998a), endogenous
DSIF in the NE strongly repressed transcription in the
presence of DRB (lanes 1, 2, 9, and 10), and depletion
of DSIF restored transcription to a level similar to that
of controls (without DRB; lanes 3, 4, 11, and 12). The ad-
dition of wild-type hSpt5 to NEDDSIF together with
hSpt4 resulted in a marked repression, and all of the
hSpt5 mutants also caused repression (lanes 5–8 and
13–18). Thus, neither Ser-2 nor Thr-4 is involved in tran-
scriptional repression by DSIF.
To examine the activation activity of DSIF, we carried
out transcription reactions in the absence of DRB by
using a DNA template that produces long transcripts
containing double G free cassettes (Figure 2B). RNase
T1 treatment of the transcripts allows simultaneous
Figure 2. CTR Thr-4 Is Required for Activation Activity, but Not for Repression Activity, of DSIF
proceed for 10 minin the presenceorabsence of DRB. ‘‘DSIF’’ represents an equimolar mixture of hSpt4 and one of the hSpt5 proteins. An arrow
indicates a full-length 380 nt transcript.
(B and D) pSLG402 was used as a template in place of pTF3-6C2AT. Arrows indicate promoter-proximal (40–124) and -distal (1512–1888) frag-
ments of transcripts. In (D), P-TEFb and DSIF were incubated in the presence of 60 mM ATP at 30ºC for 10 min in separate reactions, which were
then combined with mixtures containing NEDDSIF and the template prior to addition of nucleotides, as shown in the diagram.
(C) Transcription reactions were carried out essentially as in (B), using [g-32P] ATP instead of [a-32P] UTP. hSpt5 was immunoprecipitated, re-
solved by SDS-polyacrylamide gel electrophoresis (PAGE), and autoradiographed.
Roles of hSpt5 Phosphorylation by P-TEFb
quantification of promoter-proximal and -distal regions.
As reported previously (Guo et al., 2000), promoter-dis-
tal transcripts were generated efficiently in NE (lanes 1–
or hSpt5-SA was added back to NEDDSIF together with
hSpt4, promoter-distal transcripts were generated effi-
ciently (lanes 7–9 and 13–18). Strikingly, however,
hSpt5-STA and hSpt5-TA affected transcription negligi-
bly (lanes 10–12 and 19–21). Thus, Thr-4, but not Ser-2,
is critical to transcriptional activation by DSIF.
CTR Thr-4 Is Phosphorylated by P-TEFb In Vitro
In vitro, hSpt5 is rapidly phosphorylated by P-TEFb with
a concomitant shift in electrophoretic mobility (Figure
3A, lanes 1–4 and 17–20). Close examination of silver-
stained gels shows that the mobility shift occurs in at
least two steps: a rapid shift occurring within 0.5 min
and a slow shift occurring within 3 min. hSpt5-SA was
phosphorylated as efficiently as wild-type hSpt5 (lanes
9–12). In contrast, hSpt5-STA and hSpt5-TA show re-
duced levels of phosphorylation (lanes 5–8 and 13–16)
and appear to undergo only a mobility shift with slow ki-
netics (Figure 3B). These results indicate that P-TEFb
rapidly phosphorylates the CTR Thr-4 and slowly phos-
phorylates other parts of hSpt5.
It is still possible that Thr-4 residues are not direct tar-
gets for phosphorylation, but rather, their substitution
may affect phosphorylation of other residues. To ad-
dress this issue, we carried out a phosphoamino acid
analysis with glutathione S transferase (GST)-tagged
CTR as a substrate. After phosphorylation, the32P la-
beled CTR peptide was acid hydrolyzed and analyzed
on a TLC plate. As shown in Figure 3C, phosphorylation
occurred exclusively on Thr residues. We therefore con-
Thr-4 Phosphorylation Is Critical to DSIF-Mediated
Transcriptional Activation In Vitro
The concordance between the amino acid residues re-
quired for the activation activity and those phosphory-
lated by P-TEFb strongly suggests that Thr-4 phosphor-
ylation is critical to DSIF-mediated transcriptional
activation. To further substantiate this idea, we con-
ducted the following experiment. The hSpt5 proteins
used in this study are bacterial recombinant proteins
and are therefore originally in the unphosphorylated
form. We assume that in Figure 2B, hSpt5 becomes
phosphorylated upon the addition of nucleotides and
then functions as an activator. Indeed, a fraction of
reaction (Figure 2C). Thus, if Thr-4 phosphorylation is
functionally important and if this phosphorylation step
is rate limiting, then prephosphorylation of hSpt5 should
result in hyperactivation of transcription elongation. As
expected, hSpt5 prephosphorylated by P-TEFb showed
a significantly greater activation activity than untreated
hSpt5 (Figure 2D). Thus, we conclude that Thr-4 phos-
phorylation by P-TEFb is critical to DSIF-mediated tran-
scriptional activation in vitro.
DSIF and CTR Thr-4 Are Important for EGF-Inducible
To investigate the roles of hSpt5 and of its phosphory-
lation in living cells, we developed an inducible
knockdown system (Figure 4A). First, a HeLa cell line
expressing tetracycline-repressible Flag-hSpt5 was es-
tablished. Then, the cells were transduced with a lenti-
viral vector expressing short hairpin (sh) RNA against
hSpt5 to generate F-WT cells. Because the Flag-hSpt5
sequence contains silent mutations at the target site of
the shRNA, only endogenous hSpt5 is knocked down
Figure 3. CTR Thr-4 Is Phosphorylated by
P-TEFb In Vitro
(A) P-TEFb and one of the hSpt5 proteins
were incubated in the presence of [g-32P]
ATP for the indicated times, resolved by
SDS-PAGE, and visualized by silver staining
(B) To quantify the radioactivity incorporated
into hSpt5, protein bands were excised and
subjected to liquid scintillation counting. Re-
action rates were calculated and plotted
(C) Phosphoamino acid analysis using GST-
CTR as a substrate. A32P labeled CTR pep-
tide was acid hydrolyzed, electrophoresed
on a TLC plate, and autoradiographed (lane
2). Lane 1 shows ninhydrin-stained phos-
phoamino acid standards. An asterisk indi-
cates phosphopeptides generated by partial
hydrolysis. Abbreviations: pS, phosphoser-
ine; pT, phosphothreonine; pY, phosphotyro-
sine; and ori, origin.
by the shRNA (Figure 4B). The addition of tetracycline to
F-WT cells silences exogenous hSpt5 expression and
results in >90% knockdown of total hSpt5 (Figure 4B).
In knockdown cells, the protein level of hSpt4 is also
low, indicating that free hSpt4 is unstable and degraded
ity, because five days after the addition of tetracycline,
signs of apoptotic cell death can be recognized and al-
most all the cells eventually die (N.I., S.O., T.Y., Y.Y.,
and H.H., unpublished data). Thus, the subsequent ex-
periments were performed before the onset of apopto-
sis. Compared to the conventional method of introduc-
ing siRNA at each experiment, the knockdown in this
system is efficient, reproducible, and easy to perform.
Moreover, confusion between the siRNA’s main effect
and its ‘‘off-target’’ effects is avoided because shRNA
is constitutively expressed in this system.
c-fos gene expression has long been thought to be
regulated attheelongation step,although theregulatory
mechanism remains poorly understood (Collart et al.,
1991; Fort et al., 1987; Ryser et al., 2001). Hence, we
used EGF-inducible c-fos transcription as a model
and examined its mRNA levels by real-time RT-PCR
(Figure 4C). Whereas EGF stimulated c-fos expres-
sion by more than 20-fold in HeLa and F-WT cells, stim-
ulation was significantly attenuated in knockdown cells.
GAPDH expression levels were similar under all the con-
ditions examined. Thus, DSIF contributes to efficient
We established another cell line, F-TA, which ex-
presses Flag-hSpt5-TA and low levels of endogenous
hSpt5, as described above. We performed real-time
RT-PCR by using primer sets that distinguish between
endogenous and exogenous hSpt5 sequences and se-
lected F-WT and F-TA clones that express similar levels
ogenous hSpt5 were also similar between these cells
(data not shown). In addition, when Flag-hSpt5 was im-
munoprecipitated from F-WT and F-TA cells, similar
amounts of endogenous hSpt4 were coprecipitated
Figure 4. DSIF and CTR Thr-4 Are Important
for EGF-Inducible c-fos Transcription
(A) Experimental design. In F-WT cells, en-
dogenous hSpt5 is constitutively knocked
down, and shRNA-resistant Flag-hSpt5 is
conditionally expressed from a tetracycline-
repressible promoter (TRE).
(B) Immunoblot analysis for hSpt5. Knock-
down (KD) cells are F-WT cells that have
been cultured for 4 days in the presence of
2 mg/ml tetracycline. Anti-hSpt5 and anti-
Flag antibodies detect total and exogenous
hSpt5, respectively. Topoisomerase I (topo
I) serves as a loading control.
and GAPDH mRNAs. Total RNA was pre-
pared from cells that were stimulated for the
indicated times with 0.1 mg/ml EGF. The re-
sults shown are means 6 SEM from three in-
dependent experiments. Values derived from
unstimulated F-WT cells are set to 1.
(D) Real-time RT-PCR analysis for hSpt5
mRNA. hSpt5 expression was quantified by
using primer sets that distinguish between
Roles of hSpt5 Phosphorylation by P-TEFb
(data not shown). Therefore, the stability and integrity of
DSIF are not significantly affected by the mutations. As
shown in Figure 4E, EGF-inducible c-fos transcription
was impaired in F-TA cells as strongly as in knockdown
duction and for transcriptional activation by DSIF in vivo.
DSIF and CTR Thr-4 Are Important for Progression
of RNA Pol II along c-fos Gene after Induction
Tounderstand the role of DSIFin inducible gene expres-
sion further, we examined the distribution of RNA Pol II
and hSpt5 over the c-fos gene by ChIP with anti-RNA
Pol II and anti-Flag antibodies, respectively (Figure
5A). When control IgG was used, no significant ChIP sig-
nal was detected on c-fos (data not shown), whereas,
using anti-RNA Pol II and anti-Flag antibodies, no signif-
icant ChIP signal was detected on a control, intergenic
region. Furthermore, with anti-Flag antibody, no signifi-
cant ChIP signal was detected on c-fos in HeLa and
knockdown cells, which express little or no Flag-tagged
The following conclusions can be drawn from the data
Figure 5. DSIF and CTR Thr-4 Are Important
for Progression of RNA Pol II along c-fos
Gene after Induction
(A) Indicated cell types were stimulated with
EGF or not and subjected to ChIP with anti-
RNA Pol II or anti-Flag antibody. Promoter,
coding, and 30regions of c-fos and a control,
intergenic region on chromosome 2 were am-
plified by real-time PCR. Each bar represents
a mean 6 SEM from three independent ex-
periments. In the diagram of the c-fos gene,
open and closed boxes represent introns
and exons, respectively.
(B) Cell lysates were immunoblotted with an-
tibodies against Thr-202- and Tyr-204-phos-
phorylated ERK, ERK, Ser-383-phosphory-
lated Elk-1, and topoisomerase I (topo I).
(C) ChIP was performed as in (A), using anti-
TBP antibody (Santa Cruz, sc-273). Each
bar represents a mean 6 SEM from three in-
functionally replace endogenous hSpt5. This is based
on the finding that the distribution of RNA Pol II in F-
WT cells is very similar to that in HeLa cells. Also consis-
tent with this is the observation that ChIP results for en-
dogenous hSpt5, obtained with anti-hSpt5 antibody, are
similar to those for Flag-hSpt5 with F-WT cells (data not
shown). Second, c-fos transcription is regulated at
a postrecruitment step. Prior to induction, a substantial
amount of RNA Pol II is associated with the promoter re-
gion, but not with the coding and 30regions. In EGF-
stimulated HeLa and F-WT cells, the distribution of
RNA Pol II is extended to downstream regions. Third,
DSIF and RNA Pol II probably track along the c-fos
gene together. This is based on the observation that
the distribution of Flag-hSpt5 is very similar to that of
RNA Pol II in F-WT cells before and after induction.
that DSIF and RNA Pol II are stably associated with each
other in solution (Wada et al., 1998a; Zhang et al., 2004).
after induction. In the induced state, hSpt5 knockdown
results in only a weak (w20%) reduction in the amount
of RNA Pol II associated with the promoter region but
has a more pronounced effect on the amount of RNA
hSpt5 knockdown has no detectable effect on the distri-
bution of RNA Pol II in the uninduced state. Fifth, the
CTR Thr-4 contributes to the progression of RNA Pol II
after induction. In F-TA cells, RNA Pol II and Flag-
hSpt5-TA remain mostly localized to the promoter re-
gion even after induction. It seems that Thr-4 mutations
do not compromise the association between hSpt5 and
RNA Pol II (see also Figure S2). This finding does not
support the idea that hSpt5 phosphorylation may be im-
portant for its continued association with the elongation
complex, which was recently proposed by Zhou et al.
(2004). Taken together, these results indicate that
upon EGF induction, DSIF directly facilitates efficient
translocation of RNA Pol II from promoter-proximal to
downstream regions of c-fos and that the CTR Thr-4 is
critical for this transition.
fects other steps leading to c-fos activation. EGF in-
duces phosphorylation of ERK MAP kinase and its nu-
clear translocation. ERK then phosphorylates Elk-1,
a transcription factor that activates c-fos through a se-
rum-response element. Immunoblot analysis demon-
strated that EGF-induced phosphorylation of ERK and
Elk-1 occurs to similar extents in F-WT and knockdown
cells (Figure 5B). We also examined TBP occupancy at
the c-fos promoter. ChIP analysis showed that the TBP
occupancy is similar between F-WT and knockdown
cells before and after induction (Figure 5C). Thus,
hSpt5 knockdown has little effect on the EGF-inducible
signal transduction pathway and on preinitiation com-
plex assembly on the c-fos promoter.
CTR Thr-4 Is Phosphorylated by P-TEFb In Vivo
Anti-hSpt5 antibody detects protein bands of w160 and
w180 kDa in immunoblot analyses with HeLa cell ly-
sates (Figure 6A). Consistent with an earlier study
(Lavoie et al., 2001), the species with slower mobility is
sensitive to prior treatment of cells with Flavopiridol,
phorylated by P-TEFb in HeLa cells. The last question
addressed in this study is whether or not the CTR Thr-
4 is an in vivo phosphorylation target of P-TEFb. We car-
ried out tryptic phosphopeptide mapping on hSpt5 im-
munoprecipitated from HeLa cells, HeLa cells treated
with Flavopiridol, F-WT cells, and F-TA cells. The
hSpt5 proteins from HeLa and F-WT cells yielded indis-
tinguishable phosphopeptide maps, with strong signals
in a cluster (Figure 6B, circles). However, these signals
were markedly reduced by Thr-4 mutations, demon-
strating that they were derived from tryptic phospho-
peptides of the CTR (note that the CTR is digested into
multiple short peptides because of Arg residues within
the repeats). If P-TEFb is responsible for Thr-4 phos-
phorylation, then inhibiting it with Flavopiridol should
also reduce those signals. As expected, treatment of
HeLa cells with Flavopiridol resulted in reduced signals
in the cluster, as well as affecting certain other spots.
Thus, P-TEFb phosphorylates the CTR Thr-4 and per-
haps other parts of hSpt5 in vivo.
Many researchers have shown that hSpt5 is phosphory-
lated by P-TEFb (Ivanov et al., 2000; Kim and Sharp,
2001; Lavoie et al., 2001; Pei and Shuman, 2003). In
Figure 6. CTR Thr-4 Is Phosphorylated by P-TEFb In Vivo
(A) HeLa cells were treated with various concentrations of Flavopir-
idol for 1 hr, and total cell lysates were immunoblotted with anti-
(B) In vivo phosphopeptide mapping. hSpt5 was immunoprecipi-
tated from cells that were metabolically labeled with32P, resolved
by SDS-PAGE, isolated, digested with trypsin, and analyzed on
TLC plates. Where indicated, Flavopiridol was added to a final con-
centration of 0.5 mM 1 hr prior to harvest. The dot and the circle of
each panel mark the origin and the signals corresponding to the
CTR phosphopeptides, respectively.
Roles of hSpt5 Phosphorylation by P-TEFb
addition, using deletion mutants, Ivanov et al. (2000)
showed that the CTR is important for Tat-dependent
transactivation of the HIV-1 promoter in vitro. However,
preciserolesfor theCTRandits phosphorylation remain
largely unknown. In this study, we established that P-
ical for transcriptional activation by DSIF.
Mechanisms of Elongation Control by DSIF
We reported previously that a point mutation in zebra-
fish Spt5 at Val-1012 (corresponding to Val-1014 of
hSpt5) selectively abolishes the DSIF repression activity
(Guo et al., 2000). In the present study, we showed that
activation activity. These findings together indicate that
present which one(s) of the seven Thr residues is func-
Based on our findings, we propose the following
switch model for CTR phosphorylation (Figure 7). Un-
phosphorylated DSIF and NELF bind to unphosphory-
lated RNA Pol II to induce promoter-proximal pausing.
As we have shown previously, P-TEFb phosphorylates
the RNA Pol II CTD to cause the dissociation of NELF,
thereby ‘‘turning off’’ the DSIF repression activity
(Wada et al., 1998b; Yamaguchi et al., 1999a). As we
have shown here, P-TEFb also phosphorylates the
hSpt5 CTR, thereby ‘‘turning on’’ the DSIF activation ac-
tivity. We consider that these two phosphorylation
events fully convert DSIF from a repressor to an activa-
tor, although when and where CTR phosphorylation oc-
curs are still unknown. A number of recent reports have
shown that P-TEFb is recruited to promoter regions of
target genes and thereby phosphorylates the RNA Pol
II CTD in response to various stimuli (Barboric et al.,
2001; Eberhardy and Farnham, 2002; Lis et al., 2000).
Thus, similar to CTD phosphorylation, it is likely that
CTR phosphorylation also occurs in such a spatio-tem-
porally regulated manner.
are probably present in the elongation complex and
track along DNA together, (2) the CTR and the CTD are
structurally similar and phosphorylated by P-TEFb,
and (3) their phosphorylation leads to upregulation of
mRNA synthesis, we propose that the CTR acts as
a mini-CTD, stimulating transcription elongation by
a mechanism similar to that of the CTD (Figure 7). Phos-
phorylated CTD has been considered a code or a mark
for active elongation complexes, playing pivotal roles
in capping, splicing, and 30end processing through re-
cruitment of various mRNA processing factors (re-
viewed in Hirose and Manley  and Orphanides
and Reinberg ). Analogously, the phosphorylated
CTR may also serve as an additional code for active
elongation complexes and exert its function through re-
cruitment of other proteins having elongation activation
activity, e.g., mRNA processing factors. In support of
this view, it has been shown that U snRNPs are capable
of stimulating transcription elongation in vitro (Fong and
Zhou, 2001). Another candidate is the prolyl isomerase
Pin1, which is known to interact with various phospho-
proteins, including hSpt5 and RNA Pol II (Lavoie et al.,
CTR phosphorylation, like CTD phosphorylation,
might play a general role in mRNA synthesis. Accumu-
lating evidence suggests that Spt4 and Spt5 may regu-
late capping, splicing, 30end processing, and export of
mRNA (Bucheli and Buratowski, 2005; Burckin et al.,
2005; Cui and Denis, 2003; Lindstrom et al., 2003; Pei
and Shuman, 2002; Wen and Shatkin, 1999). Remark-
ably, the capping enzyme directly interacts with the
Spt5 CTR, although the role, if any, of CTR phosphoryla-
tion in this interaction has not been determined (Pei and
Shuman, 2002; Wen and Shatkin, 1999).
Figure 7. ModelsShowingthe Role of DSIF in
EGF-Inducible c-fos Transcription
(A) The distribution of RNA Pol II (filled circle)
over the c-fos gene. In the induced state, the
DSIF knockdown results in only a weak re-
duction in RNA Pol II associated with the pro-
moter region, but it has a more pronounced
effect on RNA Pol II associated with down-
(B) Upon EGF stimulation, P-TEFb-mediated
phosphorylation of the RNA Pol II CTD and
the hSpt5 CTR occurs and converts DSIF
from a repressor to an activator. The phos-
phorylated CTR may function in a manner
analogous to that of the phosphorylated
tivating transcription elongation.
Transcriptional Regulation of c-fos
Regulation of transcription elongation in vivo has been
extensively studied by using the Drosophila hsp70
gene, and several transcription elongation factors,
such as P-TEFb, DSIF, NELF, Spt6, FACT, ELL, Elongin,
and TFIIS, have been shown to participate in the regula-
tion (Adelman et al., 2005; Andrulis et al., 2000; Gerber
et al., 2001, 2005; Kaplan et al., 2000; Lis et al., 2000; Sa-
unders et al., 2003; Wu et al., 2003). In humans, although
it has long been thought that genes such as c-myc and
c-fos are regulated at the elongation step (Bentley and
Groudine, 1986; Collart et al., 1991; Fort et al., 1987;
Strobl and Eick, 1992), the regulatory mechanisms are
still poorly understood. In this study, we used c-fos as
a model to investigate the role of DSIF in vivo. Our
ChIP analysis showed that elongation was robustly
blocked in the promoter-proximal region of c-fos. This
finding is fully consistent with previous data from nu-
clear run-on assays (Collart et al., 1991; Fort et al.,
1987; Ryser et al., 2001). Moreover, we demonstrated
that DSIF is important for the release of RNA Pol II
from the promoter region and for its efficient progres-
sion downstream after induction. Thus, this study sheds
light on the regulatory mechanism of c-fos transcription.
Interestingly, the requirement for hSpt5 in c-fos tran-
scription was not seen in the uninduced state. If the
low-level, uninduced transcription is mechanistically
similar to the EGF-induced transcription in that DSIF
contributes to efficient transcription elongation, hSpt5
knockdown may well result in attenuation of c-fos tran-
scription and associated changes in RNA Pol II’s ChIP
signalsintheuninduced state.Alternatively, ifDSIFisin-
volved in promoter-proximal pausing, hSpt5 knock-
down may enhance c-fos transcription in the uninduced
state. In fact, however, the knockdown did not, prior to
induction, significantly affect RNA Pol II distribution
tional significance is unclear was observed (Figure 4C).
We offer four possible explanations for the above is-
sue. First, DSIF, although associated with RNA Pol II in
the promoter-proximal region, may not actually play
any role in c-fos transcription in the uninduced state.
Regulation of c-fos may be different from that of Dro-
sophila hsp70, which is thought to involve promoter-
proximal pausing by DSIF and NELF (Wu et al., 2003).
to the transcriptional block prior to induction. In the for-
mer case, the reason that hSpt5 knockdown does not
enhance c-fos transcription could be because the
knockdown also abolishes the DSIF activation activity,
which is important for the progression of RNA Pol II (Fig-
ure 5A). In the latter case, alternatively, in addition to the
presence of ‘‘negative’’ elongation factors, the absence
of ‘‘positive’’ elongation factors, such as Elongin and
TFIIS, on the c-fos gene may prevent efficient transcrip-
tion elongation in the uninduced state, as reported for
Drosophila hsp70 (Adelman et al., 2005; Gerber et al.,
2005). Fourth, residual hSpt5 after knockdown may ob-
scure true phenotypic effects. Hence, if hSpt5 is elimi-
nated completely, the defect in transcription elongation
might become significant.
In this study, we have shown that DSIF and its phos-
phorylation play a positive role in transcription elonga-
tion in vitro and in vivo. The next step is to understand
its role on a genome-wide basis. Moreover, we have
shown that the hSpt5 CTR and the RNA Pol II CTD
Identification of the CTD-like domain in a component of
the transcription elongation complex may provide anew
basis for future studies on eukaryotic gene expression.
Point mutations within the CTR were introduced by PCR using mu-
tagenic primers and pBS-Flag-hSpt5 (Yamaguchi et al., 1999b) as
a template. For bacterial expression of hSpt5 mutants, the inserts
were subcloned into pET-14b. For the bacterial expression of
GST-CTR, a cDNA sequence encoding 759–883 amino acids of
hSpt5 was amplified by PCR and cloned into pGEX-6P1. For mam-
malian expression of hSpt5, an insert of pCMV-Flag-hSpt5 (Yama-
guchi et al., 1999b) was subcloned into pTRE. To confer resistance
to RNAi, two silent mutations, a G-to-C mutation at position 411
and a C-to-G mutation at position 414, were introduced into the
hSpt5 cDNA to give rise to pTRE-Flag-hSpt5(RNAir). For the expres-
sion of shRNA, the mouse U6 promoter (positions 2315 to +5) was
cloned into pBluescript SK+. A double-stranded oligonucleotide
was inserted downstream of the promoter so as to express the fol-
lowing RNA: 50-GAACUGGGCGAGUAUUACAuucaagagaUGUAAUA
CUCGCCCAGUUCuu-30. The sequence in uppercase corresponds
to positions 406–424 of the hSpt5 mRNA. The expression cassette
was subcloned into pLenti6 to generate pLenti-U6-hSpt5-RNAi#1.
Histidine-tagged hSpt5 and hSpt4 were expressed in Escherichia
coli strain BL21(DE3) and purified as described (Yamaguchi et al.,
1999b). Cdk9 and cyclin T1 subunits of P-TEFb were coexpressed
in Sf9 cells with baculoviral vectors and purified as described
(Peng et al., 1998).
In Vitro Transcription Assays
Transcription assays were performed with HeLa cell nuclear ex-
tracts and plasmid DNA templates as described (Yamaguchi et al.,
activity of DSIF, pTF3-6C2AT was used as a template, which pro-
duces 380 nt transcripts under the control of the adenovirus E4 pro-
moter. Transcription was allowed to proceed for 10 min in the pres-
DRB. To observe its activation activity, pSLG402 was used as a tem-
plate, which produces transcripts under the control of the adenovi-
rus major-late promoter. Transcription was allowed to proceed for
various times in the presence of 30 mM ATP, 30 mM GTP, 300 mM
CTP, and 2.5 mM UTP.
Analysis of hSpt5 Phosphorylation
In Figure 3A, P-TEFb (12 ng) and one of the hSpt5 proteins (80 ng)
were incubated in 10 ml reactions containing 60 mM ATP and 1.5
mCi of [g-32P] ATP for various times at 30ºC, resolved by SDS-
PAGE, and autoradiographed. To quantify the radioactivity incorpo-
rated into hSpt5, protein bands were excised from dried gels and
subjected to liquid scintillation counting. The radioactivity (in cpm),
representing the amount of product produced, was plotted against
time, and the time-course data were fitted to exponential functions
by KaleidaGraph (Synergy Software). Then, derivatives of these
functions with respect to time (in cpm/min), representing the reac-
tion rates, were calculated and plotted against time (Figure 3B).
Whenasingle phosphorylationsite isconsidered,thereactionstarts
at the rate determined by initial substrate concentrations and Km
and Vmax values for that particular amino acid. The reaction rate
then decreases, eventually to zero, mainly due to substrate con-
sumption. In reactions containing multiple phosphorylatable amino
acid residues, the overall reaction rate decreases in a complex man-
thus is a useful measure to obtain a quick view of the whole reaction.
ose was incubated with P-TEFb and [g-32P]ATP for 30 min at 30ºC,
Roles of hSpt5 Phosphorylation by P-TEFb
washed, and further incubated with PreScission Protease (Amer-
sham Bioscience) in the presence of phosphatase inhibitors for
8 hr at 4ºC to liberate the CTR portion. A free CTR was precipitated
with acetone, resolved by SDS-PAGE, and transferred to a PVDF
membrane (Millipore). Phosphoamino acid analysis was performed
as described (Kim and Sharp, 2001).
of [32P] orthophosphate in phosphate-free medium for 16 hr. hSpt5
was immunoprecipitated and subjected to tryptic phosphopeptide
mapping (Nagahara et al., 1999). Samples were electrophoresed
with 1% ammonium carbonate (pH 8.9) in the first dimension and
chromatographed with isobutyric acid buffer (62.5% isobutyric
acid, 1.9% butanol, 4.8% pyridine, and 2.9% acetic acid) in the sec-
HeLa Tet-Off cells (BD Biosciences), which express the tetracycline-
repressible transactivator tTA, were cotransfected with pTRE-Flag-
puromycin, single colonies were isolated, and clones expressing
similar levels of exogenous Flag-hSpt5 were selected. To knock
down endogenous hSpt5, HeLa/Flag-hSpt5(RNAir) and HeLa/Flag-
pressing shRNA against hSpt5 and selected in the presence of 4 mg/
ml blasticidin to give rise to F-WT and F-TA cells, respectively. The
lentiviruses were produced by cotransfection of 293FT cells with
RNA Preparation and Real-Time RT-PCR
harvest. Cells were cultured for 18 hr in the presence of 0.2% serum
and stimulated for 15–60 min with 0.1 mg/ml EGF (Peprotech) or left
untreated. Total RNA was prepared, and real-time RT-PCR was per-
formed with SuperScript III reverse transcriptase (Invitrogen) and
QuantiTect SYBR Green PCR master mix (Qiagen). The following
primers were used: GAPDH-forward (F), 50-CTGGCGTCTTCACCA
CCATGG-30; GAPDH-reverse (R), 50-CATCACGCCACAGTTTCCC
GG-30; c-fos-F (+2237), 50-CATGGAGCTGAAGACCGAGC-30; c-fos-R
(+2587), 50-AGCAGCGTGGGTGAGCTGAG-30; hSpt5 endo-F, 50-
AGCGAGAAGAAGAACTGGGC-30; hSpt5 exo-F, 50-AGCGAGAA
GAAGAACTCGGG-30; and hSpt5-R, 50-CTTCACGTGGGTCTGCTT
GT-30. Numbers in parentheses represent positions relative to the
transcription initiation site of c-fos.
ChIP was performed essentially as described (Oshima et al., 2004).
Cells were stimulated with 0.1 mg/ml EGF for 7.5 min or not, and
then were crosslinked with 1% formaldehyde for 5 min. Nuclei
were isolated from crosslinked cells, and chromatin was sonicated
to an average size of 500 bp by using Bioruptor UCW-201 (Cosmo
Bio). Aliquots of soluble chromatin were diluted 10-fold and sub-
jected to immunoprecipitation with anti-RNA Pol II (Santa Cruz, N-
20) or anti-Flag antibody. In the latter case, anti-Flag M2 agarose
(Sigma) was used. Genomic DNA fragments in inputs and immuno-
precipitates were purified and used for real-time PCR. The following
primers were used: promoter-F (272), 50-TTGAGCCCGTGACGTTTA
CACTC-30; promoter-R (+176), 50-GTTGAAGCCCGAGAACATCATC
G-30; coding-F (+1083), 50-AACTTCATTCCCACGGTCACTGC-30,
coding-R (+1396), 50-AGTGGCTTCATCCTCTGTACTGG-30; down-
stream-F (+2646), 50-AGCTGGTGCATTACAGAGAGGAG-30, down-
stream-R (+2879), 50-GCCTGGCTCAACATGCTACTAAC-30; con-
50-GGAAGGCACTGTTAAAGTTGAG-30. Numbers in parentheses
represent positions relative to the transcription initiation site of
c-fos. The control primers are directed against an intergenic region
on chromosome 2. For absolute quantification, four serial dilutions
of inputs were assayed concurrently with immunoprecipitates, and
data were analyzed by SDS software (Applied Biosystems).
Supplemental Data include two figures and can be found with this
article online at http://www.molecule.org/cgi/content/full/21/2/227/
David Price for providing P-TEFb expression vectors, Yasunori Tsu-
boi for technical support, and David Gilmour and Tetsu Yung for dis-
cussions and comments on the manuscript. This work was sup-
ported in part by a Grant-in-Aid for Scientific Research on Priority
Areas and by the Grant of the 21st Century COE Program from the
Ministry of Education, Culture, Sports, Science, and Technology of
Japan. This work was also supported by a grant from the New En-
ergy and Industrial Technology Development Organization, Japan.
Received: July 20, 2005
Revised: September 12, 2005
Accepted: November 29, 2005
Published: January 19, 2006
Adelman, K., Marr, M.T., Werner, J., Saunders, A., Ni, Z., Andrulis,
E.D., and Lis, J.T. (2005). Efficient release from promoter-proximal
stall sites requires transcript cleavage factor TFIIS. Mol. Cell 17,
Ainbinder, E., Amir-Zilberstein, L., Yamaguchi, Y., Handa, H., and
Dikstein, R. (2004).Elongation inhibition by DRB sensitivity-inducing
factor is regulated by the A20 promoter via a novel negative element
and NF-kB. Mol. Cell. Biol. 24, 2444–2454.
Aiyar, S.E., Sun, J.L., Blair, A.L., Moskaluk, C.A., Lu, Y.Z., Ye, Q.N.,
Yamaguchi, Y., Mukherjee, A., Ren, D.M., Handa, H., and Li, R.
(2004). Attenuation of estrogen receptor a-mediated transcription
through estrogen-stimulated recruitment of a negative elongation
factor. Genes Dev. 18, 2134–2146.
Andrulis, E.D., Guzman, E., Doring, P., Werner, J., and Lis, J.T.
(2000). High-resolution localization of Drosophila Spt5 and Spt6 at
heat shock genes in vivo: roles in promoter proximal pausing and
transcription elongation. Genes Dev. 14, 2635–2649.
Barboric, M., Nissen, R.M., Kanazawa, S., Jabrane-Ferrat, N., and
Peterlin, B.M. (2001). NF-kB binds P-TEFb to stimulate transcrip-
tional elongation by RNA polymerase II. Mol. Cell 8, 327–337.
Bentley, D.L., and Groudine, M. (1986). A block to elongation is
largely responsible for decreased transcription of c-myc in differen-
tiated HL60 cells. Nature 321, 702–706.
Bucheli, M.E., and Buratowski, S. (2005). Npl3 is an antagonist of
mRNA 30end formation by RNA polymerase II. EMBO J. 24, 2150–
Burckin, T., Nagel, R., Mandel-Gutfreund, Y., Shiue, L., Clark, T.A.,
Chong, J.L., Chang, T.H., Squazzo, S., Hartzog, G., and Ares, M.,
Jr. (2005). Exploring functional relationships between components
of the gene expression machinery. Nat. Struct. Mol. Biol. 12, 175–
Collart, M.A., Tourkine, N., Belin, D., Vassalli, P., Jeanteur, P., and
Blanchard, J.M. (1991). c-fos gene transcription in murine macro-
phages is modulated by a calcium-dependent block to elongation
in intron 1. Mol. Cell. Biol. 11, 2826–2831.
Conaway, J.W., Shilatifard, A., Dvir, A., and Conaway, R.C. (2000).
Control of elongation by RNA polymerase II. Trends Biochem. Sci.
Cui, Y., and Denis, C.L. (2003). In vivo evidence that defects in the
transcriptional elongation factors RPB2, TFIIS, and SPT5 enhance
upstream poly(A) site utilization. Mol. Cell. Biol. 23, 7887–7901.
Eberhardy, S.R., and Farnham, P.J. (2002). Myc recruits P-TEFb to
mediate the final step in the transcriptional activation of the cad pro-
moter. J. Biol. Chem. 277, 40156–40162.
Fong, Y.W., and Zhou, Q. (2001). Stimulatory effect of splicing fac-
tors on transcriptional elongation. Nature 414, 929–933.
Fort, P., Rech, J., Vie, A., Piechaczyk, M., Bonnieu, A., Jeanteur, P.,
and Blanchard, J.M. (1987). Regulation of c-fos gene expression in
hamster fibroblasts: initiation and elongation of transcription and
mRNA degradation. Nucleic Acids Res. 15, 5657–5667.
Gerber, M., Ma, J., Dean, K., Eissenberg, J.C., and Shilatifard, A.
(2001). Drosophila ELL is associated with actively elongating RNA
polymerase II on transcriptionally active sites in vivo. EMBO J. 20,
Gerber, M., Tenney, K., Conaway, J.W., Conaway, R.C., Eissenberg,
J.C., and Shilatifard, A. (2005). Regulation of heat shock gene ex-
pression by RNA polymerase II elongation factor, Elongin A. J.
Biol. Chem. 280, 4017–4020.
Guo, S., Yamaguchi, Y., Schilbach, S., Wada, T., Lee, J., Goddard,
A., French, D., Handa, H., and Rosenthal, A. (2000). A regulator of
transcriptional elongation controls vertebrate neuronal develop-
ment. Nature 408, 366–369.
Curr. Opin. Genet. Dev. 13, 119–126.
Hirose, Y., and Manley, J.L. (2000). RNA polymerase II and the inte-
gration of nuclear events. Genes Dev. 14, 1415–1429.
Ivanov, D., Kwak, Y.T., Guo, J., and Gaynor, R.B. (2000). Domains in
the SPT5 protein that modulate its transcriptional regulatory proper-
ties. Mol. Cell. Biol. 20, 2970–2983.
Jennings, B.H., Shah, S., Yamaguchi, Y., Seki, M., Phillips, R.G.,
Handa, H., and Ish-Horowicz, D. (2004). Locus-specific require-
ments for Spt5 in transcriptional activation and repression in Dro-
sophila. Curr. Biol. 14, 1680–1684.
Kaplan, C.D., Morris, J.R., Wu, C., and Winston, F. (2000). Spt5 and
spt6 are associated with active transcription and have characteris-
tics of general elongation factors in D. melanogaster. Genes Dev.
Kim, D., Inukai, N., Yamada, T., Furuya, A., Sato, H., Yamaguchi, Y.,
Wada, T., and Handa, H. (2003). Structure-function analysis of hu-
man Spt4: evidence that hSpt4 and hSpt5 exert their roles in tran-
scriptional elongation as parts of the DSIF complex. Genes Cells
Kim, J.B., and Sharp, P.A. (2001). Positive transcription elongation
minal domain independently of cyclin-dependent kinase-activating
kinase. J. Biol. Chem. 276, 12317–12323.
Komarnitsky, P., Cho, E.J., and Buratowski, S. (2000). Different
phosphorylated forms of RNA polymerase II and associated
mRNA processing factors during transcription. Genes Dev. 14,
Lavoie, S.B., Albert, A.L., Handa, H., Vincent, M., and Bensaude, O.
(2001). The peptidyl-prolyl isomerase Pin1 interacts with hSpt5
phosphorylated by Cdk9. J. Mol. Biol. 312, 675–685.
Lindstrom, D.L., Squazzo, S.L., Muster, N., Burckin, T.A., Wachter,
K.C., Emigh, C.A., McCleery, J.A., Yates, J.R., 3rd, and Hartzog,
G.A. (2003). Dual roles for Spt5 in pre-mRNA processing and tran-
scription elongation revealed by identification of Spt5-associated
proteins. Mol. Cell. Biol. 23, 368–378.
Lis, J.T., Mason, P., Peng, J., Price, D.H., and Werner, J. (2000). P-
TEFb kinase recruitment and function at heat shock loci. Genes
Dev. 14, 792–803.
D phosphopeptide mapping. Methods Mol. Biol. 112, 271–279.
Orphanides, G., and Reinberg, D. (2002).A unified theory of gene ex-
pression. Cell 108, 439–451.
Oshima, S., Nakamura, T., Namiki, S., Okada, E., Tsuchiya, K., Oka-
moto, R., Yamazaki, M., Yokota, T., Aida, M., Yamaguchi, Y., et al.
(2004). Interferon regulatory factor 1 (IRF-1) and IRF-2 distinctively
up-regulate gene expression and production of interleukin-7 in hu-
man intestinal epithelial cells. Mol. Cell. Biol. 24, 6298–6310.
Pei, Y., and Shuman, S. (2002). Interactions between fission yeast
mRNA capping enzymes and elongation factor Spt5. J. Biol.
Chem. 277, 19639–19648.
Pei, Y., and Shuman, S. (2003). Characterization of the Schizosac-
charomyces pombe Cdk9/Pch1 protein kinase: Spt5 phosphoryla-
tion, autophosphorylation, and mutational analysis. J. Biol. Chem.
Peng, J., Zhu, Y., Milton, J.T., and Price, D.H. (1998). Identification of
multiple cyclin subunits of human P-TEFb. Genes Dev. 12, 755–762.
Ryser, S., Tortola, S., van Haasteren, G., Muda, M., Li, S., and Schle-
gel, W. (2001). MAP kinase phosphatase-1 gene transcription in rat
neuroendocrine cells is modulated by a calcium-sensitive block to
elongation in the first exon. J. Biol. Chem. 276, 33319–33327.
Saunders, A., Werner, J., Andrulis, E.D., Nakayama, T., Hirose, S.,
Reinberg, D., and Lis, J.T. (2003). Tracking FACT and the RNA poly-
merase II elongation complex through chromatin in vivo. Science
Sawado,T., Halow, J., Bender, M.A.,and Groudine, M.(2003). Theb-
globin locus control region (LCR) functions primarily by enhancing
the transition from transcription initiation to elongation. Genes
Dev. 17, 1009–1018.
Shim, E.Y., Walker, A.K., Shi, Y., and Blackwell, T.K. (2002). CDK-9/
cyclin T (P-TEFb) is required in two postinitiation pathways for tran-
scription in the C. elegans embryo. Genes Dev. 16, 2135–2146.
Simone, C., Stiegler, P., Bagella, L., Pucci, B., Bellan, C., De Falco,
G., De Luca, A., Guanti, G., Puri, P.L., and Giordano, A. (2002). Acti-
vation of MyoD-dependent transcription by cdk9/cyclin T2. Onco-
gene 21, 4137–4148.
and chromatin remodeling during differentiation-induced gene acti-
vation. Science 295, 1901–1904.
Strobl, L.J., and Eick, D. (1992). Hold back of RNA polymerase II at
the transcription start site mediates down-regulation of c-myc
in vivo. EMBO J. 11, 3307–3314.
Swanson, M.S., Malone, E.A., and Winston, F. (1991). SPT5, an es-
sential gene important for normal transcription in Saccharomyces
cerevisiae, encodes an acidic nuclear protein with a carboxy-termi-
nal repeat. Mol. Cell. Biol. 11, 3009–3019.
Wada, T., Takagi, T., Yamaguchi, Y., Ferdous, A., Imai, T., Hirose, S.,
Sugimoto, S., Yano, K., Hartzog, G.A., Winston, F., et al. (1998a).
DSIF, a novel transcription elongation factor that regulates RNA
polymerase II processivity, is composed of human Spt4 and Spt5
homologs. Genes Dev. 12, 343–356.
Wada, T., Takagi, T., Yamaguchi, Y., Watanabe, D., and Handa, H.
(1998b). Evidence that P-TEFb alleviates the negative effect of
DSIF on RNA polymerase II-dependent transcription in vitro.
EMBO J. 17, 7395–7403.
Wang, G., Balamotis, M.A., Stevens, J.L., Yamaguchi, Y., Handa, H.,
postrecruitment steps in transcription initiation. Mol. Cell 17, 683–
Wei, P., Garber, M.E., Fang, S.M., Fischer, W.H., and Jones, K.A.
(1998). A novel CDK9-associated C-type cyclin interacts directly
with HIV-1 Tat and mediates its high-affinity, loop-specific binding
to TAR RNA. Cell 92, 451–462.
Wen, Y., and Shatkin, A.J. (1999). Transcription elongation factor
hSPT5 stimulates mRNA capping. Genes Dev. 13, 1774–1779.
Wu, C.H., Yamaguchi, Y., Benjamin, L.R., Horvat-Gordon, M., Wash-
mour, D. (2003). NELF and DSIF cause promoter proximal pausing
on the hsp70 promoter in Drosophila. Genes Dev. 17, 1402–1414.
Yamaguchi, Y., Takagi, T., Wada, T., Yano, K., Furuya, A., Sugimoto,
S., Hasegawa, J., and Handa, H. (1999a). NELF, a multisubunit com-
plex containing RD, cooperates with DSIF to repress RNA polymer-
ase II elongation. Cell 97, 41–51.
Yamaguchi, Y., Wada, T., Watanabe, D., Takagi, T., Hasegawa, J.,
and Handa, H. (1999b). Structure and function of the human tran-
scription elongation factor DSIF. J. Biol. Chem. 274, 8085–8092.
Yamaguchi, Y., Filipovska, J., Yano, K., Furuya, A., Inukai, N., Narita,
T., Wada, T., Sugimoto, S., Konarska, M.M., and Handa, H. (2001).
Stimulation of RNA polymerase II elongation by hepatitis delta anti-
gen. Science 293, 124–127.
Zhang, Z., Wu, C.H., and Gilmour, D.S. (2004). Analysis of polymer-
ase II elongation complexes by native gel electrophoresis. Evidence
for a novel carboxyl-terminal domain-mediated termination mecha-
nism. J. Biol. Chem. 279, 23223–23228.
Zhou, M., Deng, L., Lacoste, V., Park, H.U., Pumfery, A., Kashanchi,
F., Brady, J.N., and Kumar, A. (2004). Coordination of transcription
factor phosphorylation and histone methylation by the P-TEFb ki-
nase during human immunodeficiency virus type 1 transcription.
J. Virol. 78, 13522–13533.
Roles of hSpt5 Phosphorylation by P-TEFb