Download Physical interaction between pRb and cdk9/cyclinT2 complex

ã
Oncogene (2002) 21, 4158 ± 4165
2002 Nature Publishing Group All rights reserved 0950 ± 9232/02 $25.00
www.nature.com/onc
Physical interaction between pRb and cdk9/cyclinT2 complex
Cristiano Simone1,2,5, Luigi Bagella1,3,5, Cristiana Bellan1,3 and Antonio Giordano*,4
1
Department of Pathology, Anatomy and Cell Biology, Thomas Je€erson University, Philadelphia, Pennsylvania, PA 19107 USA;
Department of Internal Medicine and Public Medicine, Division of Medical Genetics, University of Bari, Bari 70124 Italy;
3
Institute of Pathologic Anatomy and Histology, University of Siena, Siena 53100 Italy; 4Sbarro Institute for Cancer Research and
Molecular Medicine, Temple University, Philadelphia, Pennsylvania, PA 19122 USA
2
Cyclin-dependent kinase 9 (cdk9) is a multifunctional
kinase with roles in di€erent cellular pathways such as
transcriptional elongation, di€erentiation and apoptosis.
Cdk9/cyclin T di€ers functionally from other cdk/cyclin
complexes that regulate cell cycle progression, but
maintains structural anity with those complexes. In
addition, previous reports have demonstrated that the
cdk9 complex is able to phosphorylate p56/pRb in vitro.
In this report we show in vitro and in vivo interaction
between cdk9/cyclinT2 and the protein product of the
retinoblastoma gene (pRb) in human cell lines. The
interaction involves the region composed of residues
129 ± 195 of cdk9, cyclinT2 (1 ± 642 aa) and the Cterminal domain of pRb (835 ± 928 aa). We located the
minimal region of cdk9 phosphorylation on the Cterminus of pRb, by identifying the residues between
793 and 834. This region contains at least three prolinedirected serines (sp), S795, S807 and S811, which have
been reported to be phosphorylated in vivo and which
could be targeted by the cdk9 complex. These data
suggest that, in logarithmically growing cells, cdk9/
cyclin T2 and pRb are located in a nuclear multiprotein
complex probably involved in transduction of cellular
signals to the basal transcription machinery and that one
of these signals could be the cdk9 phosphorylation of
pRb.
Oncogene (2002) 21, 4158 ± 4165. doi:10.1038/sj.onc.
1205511
Keywords: cdk9; cyclin T2; pRb; phosphorylation
Introduction
Cyclin-dependent kinase 9 (cdk9) is a cdc2-related
serine/threonine kinase that is isolated using degenerate
oligonucleotide primers derived from conserved regions
found in cdc2 and cdk2 (Grana et al., 1994). The cdk9
gene is widely expressed in human and murine tissues,
*Correspondence: A Giordano, Sbarro institute for Cancer Research
and Molecular Medicine, College of Science and Technology, Temple
University, BioLife Science Bldg. Suite 333, 1900 N 12th Street,
Philadelphia, PA 19122, USA; E-mail: [email protected]
5
The ®rst two authors contributed equally to this study.
Received 19 June 2001; revised 13 March 2002; accepted 21 March
2002
with higher levels found in terminally di€erentiated cells,
and its promoter activity parallels the protein levels (De
Luca et al., 1997; Bagella et al., 1998, 2000). The
regulatory units of cdk9 are the T-family cyclins (T1,
T2a and T2b) (Wei et al., 1998; Peng et al., 1998) and
cyclin K (Edwards et al., 1998). The elevated levels of
cdk9 and its regulatory subunits in terminally di€erentiated cells, together with the fact that cdk9/cyclin T
complexes are not cell cycle-regulated, distinguish cdk9
from the other cdks (MacLachlan et al., 1995; De Falco
and Giordano, 1998). Moreover, unlike the other cdks,
which regulate cell cycle progression and phosphorylate
histone H1, cdk9 fails to phosphorylate H1 (Grana et
al., 1994). Cyclin T levels are not cell cycle-regulated
and, for this reason, cdk9 kinase activity does not
change during the di€erent phases of the cell cycle. Cdk9
is instead implicated in the regulation of transcriptional
elongation via phosphorylation of the carboxyl-terminal
domain (CTD) of RNA polymerase II (RNApolII) and
forms the positive transcription factor b (PTEF-b) with
cyclin T (Dahmus, 1996; Marshall et al., 1996; Zhu et
al., 1997; De Falco and Giordano, 2002). Kinase activity
corresponding to cdk9 has been detected in the HIV Tatassociated kinase (TAK) complex, where cdk9 is
required for Tat-dependent stimulation of transcriptional elongation (Yang et al., 1996; Zhu et al., 1997;
Mancebo et al., 1997). CycT1 was the ®rst cyclin
discovered to associate with cdk9 to form a complex
interacting with HIV-1-Tat protein in the TAK complex
(Wei et al., 1998). Human cycT1 stimulates Tat
activation, while cycT2a and T2b fail to form a complex
with Tat bound to the transactivation response RNA
element (TAR) (Kwak et al., 1999; Wimmer et al., 1999).
These data indicate functional di€erences between
cycT1 and cycT2, even if cycT2 function remains less
understood. The catalytic activity of the cdk9 immunocomplexes results only from cdk9 and it is abolished
using a dominant negative form (kinase inactive mutant,
cdk9dn) (De Falco et al., 2000), which consists of a
point mutation within the ATP-binding domain that
changes the Asp to Asn at residue 167.
Previous studies have already demonstrated that
cdk9 is able to phosphorylate p56/pRb (a deletion
mutant maintaining the A/B pocket and the C-terminal
domain of pRb, 379 ± 928 aa) in vitro (Grana et al.,
1994; De Luca et al., 1997). Phosphopeptide analysis of
p56/pRb after phosphorylation by cdk9, compared to
cdk9/cyclinT2 binds pRb
C Simone et al
that after cdk2 and cdc2 phosphorylation, indicates
that, at least in vitro, the three cdk complexes share
several target phosphoresidues, but cdk9 kinase activity
involves only serine residues (De Luca et al., 1997).
During preliminary studies of cdk9/cyclin T complexes, we performed immunoprecipitations (Ips) with
anity puri®ed anti-cdk9 antibody from nuclear
extracts of S35-labeled NIH3T3 cells. By comparing
the speci®c pattern of immunoprecipitation with an Ip
performed using pre-immune rabbit serum, we identi®ed several bands that could represent cdk9-interacting
proteins. We started to consider bands with predicted
molecular weights potentially corresponding to known
proteins that may interact with cdk9. The predicted
protein structure anity with the other cdks prompted
us to consider of interest a band at 110 kD. Cdk4/
cyclin D, cdk2/cyclin E and cdk1/cyclin A complexes
all bind and/or phosphorylate pRb during di€erent
phases of the cell cycle (Ezhevsky et al., 1997;
Hatakeyama et al., 1994; Lundberg and Weinberg,
1998; Pan et al., 2001). The retinoblastoma gene codes
for a 105 kD nuclear protein that is regulated by
phosphorylation. Depending on the amount of phosphorylated residues, pRb shifts from an active
hypophosphorylated state towards an inactive hyperphosphorylated state (Mittnacht, 1998; Stiegler and
Giordano, 2001). The basic structure of pRb consists
of the N-terminal, the pocket region (spanning the A
and B domains, separated by a spacer) and the Cterminal region. The N-terminal region is the least
understood in its functional role. The central region,
which folds into a pocket, o€ers a docking site for
several interacting proteins. Several residues within the
pocket are responsible for the interaction with proteins
containing a LXCXE docking site, such as E1A, E7,
SV40 LTag, HDACs and cyclin D (Whyte et al., 1988;
DeCaprio et al., 1998; Lee et al., 1998; MagnaghiJaulin et al., 1998; Ewen et al., 1993; Dowdy et al.,
1993). The pocket region is also the binding site for
other cellular factors lacking the LXCXE motif, such
as E2F transcription factors (Stiegler and Giordano,
2001). Mutations in these pocket residues render pRb
inactive or heavily impaired in controlling the cell
cycle. The function to arrest the cell cycle is separate
and independent from an intact LXCXE binding site,
demonstrating that additional sequences besides the
pocket are important for pRb function. The C-terminal
region of pRb is involved in association with important
cellular factors including E2F, Mdm2 and c-Abl (Qian
et al., 1992; Xiao et al., 1995; Welch and Wang, 1993).
The C-terminal region spanning residues 793 ± 928 is
the smallest domain of pRb that is eciently
phosphorylated by cdk4/cycD1, cdk2/cycE and cdk2/
cycA (Pan et al., 1998; Adams et al., 1999). Very
recently, examination of the pRb C-terminus revealed
that it contains sequence elements related to ZRXL,
which is the E2F- and p21-conserved cyclin/cdkbinding motif (Adams et al., 1999). These ®ndings
separate the C-terminal region into two subdomains,
one, the putative cyclin/cdk-binding domain spanning
residues 830 ± 928, and the other (780 ± 826 aa) contain-
ing the S/TP phosphorylation sites for cdk/cyclin
complexes (Adams et al., 1999). In addition to RIL
(830 ± 2 aa), RVL (857 ± 9), KPLKKL (870 ± 5) and
KHL (889 ± 91) cdk/cyclin-binding domains (Adams et
al., 1999), a new domain surrounding residue L901 was
found to be important for the interaction with the
cdk4/cyclin D1 complex (Pan et al., 2001).
pRb has been identi®ed to interact with several
cellular factors that in¯uence gene expression. Many of
these transcription factors, like E2F1, E2F2 and E2F3,
transcribe RNApolII-dependent genes. This is another
link between cdk9/cyclin T (P-TEFb) and pRb that
suggests a possible functional/physical interaction
during gene transcription.
In this study, we demonstrate that cdk9/cycT2 binds
to pRb, involving residues 129 ± 195 of cdk9, the region
of cycT2 comprising amino acids 1 ± 642 and the Cterminal region of the retinoblastoma protein (835 ±
928) and that cdk9 complexes phosphorylate the pRb
region spanning amino acids 793 ± 834.
4159
Results
Cdk9 interacts in vitro with pRb through 129 ± 195 aa
To understand if the 110 kD band (Figure 1)
previously discussed is the protein product of the
retinoblastoma protein, we performed an in vitro GST
pull-down assay. To test whether a direct physical
interaction between cdk9 and pRb occurs, we ®rst
investigated if the proteins could interact in vitro. The
in vitro translated (IVT) full-length cdk9 was wellprecipitated by GST-pRb (379 ± 928), while no interaction was found with GST alone Figure 2d, lanes 5
and 6). To map the pRb binding site on the cdk9
protein, three deletion mutants, cdk9 (129 ± 372), cdk9
Figure 1 Immunoprecipitations from nuclear extracts of S35labeled NIH3T3 cells. Immunoprecipitation (Ip) was performed
with anity puri®ed anti-cdk9 antibody from nuclear extracts of
S35-labeled NIH3T3 cells and compared to an Ip with preimmune rabbit serum (NRS). Bands corresponding to cdk9, cyclin
T and a band at 110 kD were indicated
Oncogene
cdk9/cyclinT2 binds pRb
C Simone et al
4160
Figure 2 Cdk9 interacts in vitro with pRb through the region corresponding to 129 ± 195 aa. (a) Summary of in vitro translated
(IVT) S35-labeled cdk9 proteins and their ability to interact with GST-pRb (379 ± 928). (The PSTAIRE-like domain is referred to as
(PITALRE)). (b) Summary of GST-fusion proteins and their ability to coprecipitate cdk9 (1 ± 372). (A: A domain; B: B domain; AB: pocket domain; C: C-terminal domain). (c) IVT-cyclin T2 isoforms (a and b). (The two splice isoforms are identical in the 1 ±
642aa region. 1 ± 262: cyclin box region; 642 ± 663 for cycT2a and 642 ± 730 for cyclin T2b: C-terminal domain). (d) The interaction
between GST-pRb (379 ± 928) and di€erent IVT fragments of cdk9 was tested. GST alone was used as a negative control. The GSTfusion proteins were precipitated by glutathione-agarose anity chromatography and the interaction with the in vitro translated S35labeled proteins was visualized by autoradiography
(218 ± 372) and cdk9 (1 ± 195) Figure 2a, lanes 8 and
12), were tested for their ability to interact with GSTpRb (379 ± 928) in a pull-down assay. GST alone was
used as a negative control. An interaction, comparable
to that with the full-length cdk9, was detected with two
deleted forms, cdk9 (129 ± 372) and cdk9 (1 ± 195)
Figure 2a), suggesting that the minimal binding region
involves amino acids 129 ± 195.
Minimal cdk9-binding region of pRb involves 835 ± 928 aa
A similar experiment was performed to map the
minimal cdk9-binding region of pRb and to address
if pRb was also able to interact with cyclin T2, which is
not extensively characterized in the literature. IVT S35labeled cdk9 and cyclin T2 (isoforms a and b) were
tested for their ability to interact with GST-pRb (379 ±
928) Figure 3a) in a pull-down assay, using GST alone
as a negative control. This experiment provided
evidence that either cdk9 or cyclin T2a (and T2b)
could directly interact with pRb and that they could
form a trimeric complex in vitro. Afterwards, we
investigated whether a pRb mutant (379 ± 928,
Oncogene
C706F), unable to interact through its pocket domain
(Magnaghi-Jaulin et al., 1998), could still bind the
cdk9/cyclin T2 complex. Under the same experimental
conditions, the GST-pRbC706F mutant was still able
to recruit both cdk9 and cyclin T2 (a and b), as the
wild type does Figure 3b). These data indicate that the
possible region of interaction could involve the Cterminus of pRb. To address this point, we tested three
deletion mutants of pRb. IVT S35-labeled cdk9 and
cyclin T2 (isoforms a and b) were tested with GST-pRb
(793 ± 928), GST-pRb (768 ± 834) and GST-pRb (835 ±
928) Figure 3c ± e) in a pull-down assay, using GST
alone as a negative control. Like GST-pRb (379 ± 928),
GST-pRb (793 ± 928) and GST-pRb (835 ± 928) were
able to precipitate cdk9/cycT2 proteins; the third
deletion mutant, GST-pRb (768 ± 834), did not interact
with PTEF-b components Figure 3c ± e).
These data indicate that the regions mediating the
binding between cdk9/cycT2 and pRb encompass the
C-terminal domain of pRb (835 ± 928 aa), cyclinT2 (1 ±
642) and the central region (129 ± 195 aa) of cdk9
surrounding the ATP-binding kinase domain (see
Figures 2 and 3).
cdk9/cyclinT2 binds pRb
C Simone et al
4161
Figure 3 Minimal cdk9/cycT2-binding region of pRb involves the region containing 835 ± 928 aa. (a) IVT S35-labeled proteins used
in the GST pull-down experiments. (Lane 1: cdk9; lane 2: reticulocyte lysate; lane 2:cyclin T2a; lane 4: cyclin T2b). (b ± f) The GSTpRb fusion proteins indicated in Figure 2b were tested for their ability to coprecipitate IVT-cdk9, IVT-cycT2a and IVT-cycT2b.
GST alone was used as a negative control. The GST-fusion proteins were precipitated by glutathione-agarose anity
chromatography and the interaction with the in vitro translated S35-labeled proteins was visualized by autoradiography
Cdk9, cyclin T2 and pRb form a multimeric nuclear
complex in human cells
Cdk9 complexes phosphorylate pRb on the C-terminal
domain
To verify that the interaction described above was
present in vivo, we performed a GST pull-down of
human nuclear cell extracts. GST-pRb was able to
co-precipitate cdk9 from nuclear extracts of Jurkat
and HeLa cells Figure 4a), while GST alone, used
as a negative control, was not. As a positive
control, the same ®lter was blotted with antiHDAC1 antibody, revealing the interaction as well
(data not shown). Increasing the time of the
binding reaction from 2 h to overnight incubation,
GST-pRb co-precipitated more cdk9, as was
demonstrated by Western blot with anity puri®ed
anti-cdk9 antibody Figure 4a, compare lanes 2 and
4, and 7 and 9), while GST alone continued to
not interact with cdk9 Figure 4a, lanes 1, 3, 8,
and 10).
To con®rm the interaction, we performed Ips from
nuclear extracts of proliferating HeLa and Jurkat cells
using anti-cdk9 anity puri®ed antibody. Western blot
with anti-pRb antibody revealed the presence of the
retinoblastoma gene product in association with cdk9
in the nucleus of human proliferating cells Figure 4b).
Further analysis of the a-cdk9 antibody immunoprecipitates revealed the presence of cdk9 and cycT2 (data
not shown).
Previous reports indicate that p56/pRb is a good
substrate in vitro for the cdk9 complex (Grana et al.,
1994; De Luca et al., 1997). Using the reported GSTpRb fusion proteins as substrates Figure 2b), we
performed a kinase assay with immunoprecipitated
cdk9 complexes from HeLa nuclear extracts Figure 5)
(pre-immune rabbit serum used as a negative control,
data not shown). Cdk9 complexes phosphorylated
pRb (793 ± 928) as well as p56/pRb (379 ± 928),
con®rming that, at least in vitro, this complex
behaves the same as other cdk/cyclin complexes.
Very interestingly, neither of the two deleted forms of
the pRb C-terminus, pRb (835 ± 928) and pRb (768 ±
834), was phosphorylated. The other cdk/cyclin
complexes are unable to phosphorylate the region
794 ± 829 in the absence of recognition motifs
contained between pRb residues 829 ± 928 (Adams et
al., 1999). Our data suggest that this is true also for
cdk9 complexes. In fact, cdk9/cyclin T2 binds pRb
(835 ± 928), phosphorylates the C-terminal domain
(793 ± 928), but fails to phosphorylate the two parts
that compose it. The region of pRb corresponding to
793 ± 834 aa contains at least three proline-directed
serines (sp), S795, S807 and S811, which have been
reported to be phosphorylated in vivo (Hollingsworth
Oncogene
cdk9/cyclinT2 binds pRb
C Simone et al
4162
Figure 4 Cdk9, cyclin T2 and pRb form a multimeric nuclear complex in human cells. (a) GST pull-down was performed with
human nuclear cell extracts. GST-pRb (378 ± 928) was able to coprecipitate cdk9 from nuclear extracts of Jurkat and HeLa cells.
The same ®lter was blotted with anti-HDAC1 antibody as a positive control (immunoblot not shown). GST alone was used as a
negative control. (b) Jurkat and HeLa nuclear cell extracts were precleared and incubated with anity puri®ed anti-cdk9 and preimmune serum (NRS) (used as a negative control). After extensive washes, the samples were resuspended in Laemmli bu€er and
loaded on acrylamide gel. The associated proteins were detected by immunoblot by using anti-pRb, anti-cycT2 and anti-cdk9
antibodies (cdk9 and cycT2 immunonoblots are not shown)
the C-terminus of pRb to phosphorylate serine
residues present between 793 and 834 aa.
Discussion
Figure 5 Cdk9 immunocomplexes phosphorylate pRb in vitro
on the C-terminal domain. Ips were performed with anti-cdk9 in
logarithmically growing HeLa cells. (Pre-immune rabbit serum
used as a negative control, data not shown). The immunocomplexes were extensively washed, incubated in kinase assay reaction
bu€er with 0.5 mg of the reported GST-fusion proteins and
resolved on 13% gel
et al., 1993) and could be targeted by cdk9
complexes. From our data, we cannot exclude that
in the pRb amino acid sequence there are other
serines targeted by cdk9 complexes and that this
phosphorylation has a speci®c role in vivo; in our
model, cdk9 complexes use the recognition sites of
Oncogene
Cdk9 protein levels and kinase activity are high in
terminally di€erentiated cells, as are cyclin T family
protein levels (De Luca et al., 1997; Bagella et al.,
1998; Wei et al., 1998; Peng et al., 1998). PTEF-b
kinase activity is constant throughout the cell cycle,
depending on cdk9 and cyclin T protein levels, which
do not change during the di€erent phases (Grana et al.,
1994). In addition, cdk9 complexes fail to phosphorylate histone H1, but exert their kinase activity on the
C-terminal domain of RNA polymerase II (Dahmus,
1996; Marshall et al., 1996; Zhu et al., 1997) and
MyoD, a bHLH transcription factor part of the
Muscle Regulatory Factor (MRF) family (Simone et
al., 2002; Simone and Giordano, 2001). Despite these
relevant functional di€erences with the other cdk/cyclin
complexes, cdk9 shares 41 ± 43% identity (61 ± 65%
similarity) with human cdc2, cdk2, cdk3, cdk5 and
38% with cdk4 (Grana et al., 1994). Similarly, cyclin T
has an amino-terminal cyclin box motif sharing 39%
identity with the human cyclins and a carboxyl-
cdk9/cyclinT2 binds pRb
C Simone et al
terminal PEST sequence (709 ± 726 aa), which is
frequently found in G1 cyclins to regulate their
turnover by cellular ubiquitination and proteolytic
pathways (Wei et al., 1998).
In past years no functional reason has been found to
justify cdk9 phosphorylation of pRb in vitro, probably
due to the complexity of an in vivo study of pRb
phosphorylation and regulation. In this study, we
characterize the physical interaction between cdk9,
cyclin T2 and pRb, demonstrating that the C-terminus
of pRb (835 ± 928) directly interacts with the region
surrounding the kinase domain of cdk9 (129 ± 195) and
with both isoforms of cyclin T2 (a and b), suggesting
that the region of interaction involves the common
sequence comprising 1 ± 642 aa Figure 2a ± c). In
addition, cdk9 complexes phosphorylate the C-terminus
of pRb (793 ± 928) in vitro and our data suggest that the
phosphorylated region spans amino acids 793 ± 834
Figure 6). This evidence is intriguing because it con®rms
the division of the pRb C-terminus into two subdomains. In fact, several groups have reported that pRb Cterminus (793 ± 928) could be separated into two
subdomains with speci®city for cdk/cyclin complexes.
The ®rst region (793 ± 829) contains the cdk phosphorylation sites and the second (829 ± 928) is the recognition/binding site for cdk/cyclin complexes Figure 6).
This subdivision has been proven to be important for
cdk2/cycE, cdk2/cycA and cdk4/cycD1 (Pan et al., 1998,
2001; Adams et al., 1999). During the cell cycle, pRb
phosphorylation mediated by cdk/cyclin complexes is
required for the cell to progress through the di€erent
phases. pRb shifts from a hypophosphorylated active
form to a hyperphosphorylated inactive form and allows
E2F transcription factors to be free from pRb control
and bind to speci®c gene promoters.
Several pieces of evidence might help to elucidate the
meaning of cdk9-mediated phosphorylation of pRb in
vivo. First, pRb has been identi®ed to functionally
interact with factors which in¯uence RNA polymerase
II-dependent gene transcription (De Luca et al., 1998;
Fanciulli et al., 2000). This important e€ect of pRb on
RNApolII transcription could represent a functional
link with PTEF-b. In fact, cdk9 complexes have been
characterized as transcriptional elongation factors
involved in transcription of speci®c genes (Lis et al.,
2000; Kanazawa et al., 2000; Foskett et al., 2001).
Most likely, cdk9 complexes exert their action through
the hyperphosphorylation of the CTD of RNApolII,
allowing passage from the initiation to the elongation
phase of RNA transcription (Lis et al., 2000;
Napolitano et al., 2000). The cdk9 phosphorylation
of pRb could be another signal that allows conversion
from an inactive complex into a fully active complex
that is ready to induce RNA transcription.
The second point of interest is the involvement of
pRb during cellular di€erentiation and in terminally
di€erentiated cells. Several pieces of evidence have
implicated pRb as an essential cofactor during muscle
di€erentiation. Muscle cells derived from pRb7/7
mice fail to irreversibly exit the cell cycle (Schneider et
al., 1994), express reduced levels of late di€erentiation
markers and display impaired fusion into multinucleated myotubes (Novitch et al., 1996). Several
hypotheses have been proposed to justify pRb
involvement in di€erentiation, including a physical
binding with MyoD (Gu et al., 1993) and the inhibition
of the anti-myogenic activity of E2F (Shin et al., 1995;
Corbeil et al., 1995; Puri et al., 1997). More recently,
pRb has been shown to promote functional synergism
between MyoD and MEF2 protein, a coactivator of
speci®c gene transcription (Novitch et al., 1999).
However, the molecular mechanisms through which
pRb promotes myogenic transcription remain poorly
understood. We found that cdk9/cycT2 complex
activates MyoD-mediated transcription and that cdk9
enzymatic activity is necessary for the myogenic
program. In fact, cells expressing the kinase-inactive
mutant of cdk9 (which behaves as a dominant
negative) fail to di€erentiate (Simone et al., manuscript
submitted). Immunoprecipitated cdk9 complexes from
C2C12 mouse myoblasts induced to di€erentiate have
an increasing kinase activity on p56/pRb with a peak
of phosphorylation at 96 h of the myogenic program
(Bagella et al., 1998). During muscle di€erentiation,
pRb is present in the active hypophosphorylated form,
especially due to down-regulation of cyclins A, E and
D1 and the up-regulation of cdk inhibitors. On the
other hand, MyoD is inactivated during the cell cycle
by cdk/cyclin complexes through direct phosphorylation or physical binding (Puri and Sartorelli, 2000),
while the kinase activity of cdk9 is a signal that triggers
myogenic transcription (Simone and Giordano, 2002).
It is possible that cdk9/cycT2 activity is involved in the
basal phosphorylation of the retinoblastoma protein
and that pRb and cdk9/cycT2 cooperate to support
MyoD-mediated myogenic transcription. Future studies are awaited to explain the relevance of cdk9/
cycT2-pRb interaction and to clarify the importance of
pRb phosphorylation in di€erent pathways other than
the cell cycle, ideally by the in vivo characterization of
the phosphoresidues targeted by cdk/cyclin complexes.
4163
Materials and methods
Figure 6 C-terminal domain of pRb protein. The putative
proline-directed serines and the RXL domains are in bold and
underlined
Cell lines
Jurkat and HeLa cells were grown, respectively, in RPMI
1640 and in DMEM supplemented with 10% FBS, LOncogene
cdk9/cyclinT2 binds pRb
C Simone et al
4164
glutamine and antibiotics. All cell lines were obtained from
ATCC.
The constructs pcDNA3-cdk9wt-HATag expressing fulllength wild type cdk9 were previously described (De Falco
et al., 2000). The construct pcDNA3-cdk9wt-HATag was
cleaved with HindIII or KpnI and the fragments were recircularized with T4 ligase generating the construct pcDNA3cdk9 (129 ± 372)-HATag and pcDNA3-cdk9 (218 ± 372)HATag, respectively. The fragments of 582 bp obtained after
cleavage of pcDNA3-cdk9wt-HATag with KpnI were subcloned in the KpnI site of pcDNA3 vector (Invitrogen)
generating pcDNA3-cdk9 (1 ± 195). The constructs to generate the GST-Rb fusion proteins GST-Rb (379 ± 928), GSTRb (793 ± 928), GST-Rb (768 ± 834) and GST-Rb (835 ± 928)
were obtained by PCR ampli®cation using, respectively, the
following couples of primers: GGATCCATGAACACTATCCAA, GAGCTCTCATTTCTCTTCCTTG; GATCCCCTAGTTCACCCTTAC, GAGCTCTCATTTCTCTTCCTTG;
GATCCATTTTGCAGTATGCTTC,
GAGCTCGATACTAAGATTCTTG;
GGATCCATTGGTGAATCATTCG,
GGATCCTCATTTCTCTTCCTTG.
The PCR products were cloned into pGEX 4T-1 or pGEX
2T (Pharmacia) to obtain, respectively, pGEX-4T-1 Rb-D(1 ±
378), pGEX-4T-1 Rb-D(1 ± 792), pGEX-4T-1 Rb-D(1 ± 767,
835 ± 928) and pGEX-2T Rb-D(1 ± 834).
The construct pGEX-4T-1 Rb-C706F-D(1 ± 378) to generate GST-Rb (379 ± 928) C706F was performed using a
QuickChange Site-Directed Mutagenesis Kit (Promega). The
following primers were used: GGACCAAATTATGATGTTTTCCATGTATGGCATATG and CATATGCCATACATGGAAAACATCATAATTTGGTCC. The constructs
expressing HA-cyclin T2a and HA-cyclin T2b were previously described (Peng et al., 1998).
determined by Bradford assay (Biorad, CA, USA), following
the manufacturer's instructions and by using BSA as a
standard.
GST-pRb (379 ± 928), GST-pRb (379 ± 928) C706F, GSTpRb (793 ± 928), GST-pRb (768 ± 834) and GST-pRb (835 ±
928) were expressed in bacteria and puri®ed as described
above (De Falco et al., 2000). Nuclear cell extracts were
incubated with the reported GST fusion protein for 2 h and/
or overnight at 48C. Bond proteins were recovered by
centrifugation at 14 000 r.p.m. for 30 s at 48C and washed
three times with Lysis Bu€er (50 mM Tris, 5 mM EDTA,
250 mM NaCl, 50 mM NaF, 0.1% Triton, 0.1 mM Na3VO4,
and 10 mg/ml aprotinin, leupeptin and phenylmethylsulphonyl ¯uoride (PMSF)). Immunoprecipitations were performed
after preclearing the extracts with pre-immune rabbit serum.
Anity puri®ed anti-cdk9 (1 : 50; De Falco et al., 2000)
antibody incubations were performed overnight, at 48C,
followed by the addition of protein A-sepharose for 60 min,
rocking, at 48C. Beads were washed three times with Lysis
Bu€er and twice with Lysis Bu€er containing 400 mM NaCl,
and loading bu€er (50 mM Tris/HCl pH 6.8, 2% SDS, 10%
glycerol) with 5% b-mercaptoethanol was added. The
samples were resolved in 8 or 10% SDS-PAA gels and were
transferred to a Hybond-ECL nitrocellulose ®lter (Amersham
Life Science Inc.) at 48C and at 70 V for 2 h. 0.5% Ponceau
Red was used to ensure equal transferring. The blots were
blocked with TBST containing 5% non-fat dry milk. Anticdk9 (1 : 200), anti-cycT2 (1 : 2000; Peng et al., 1998) and antiHDAC1 (1 : 200, Santa Cruz Biotechnology, CA, USA)
polyclonal antibodies and anti-pRb (1 : 500, Pharmingen)
monoclonal antibody were used in TBST containing 5% nonfat dry milk, according to the Western blot conditions
suggested by Santa Cruz (Santa Cruz Biotechnology, CA,
USA). Anti-rabbit and anti-mouse peroxidase conjugated
(1 : 20000) (Amersham, IL, USA) and ECL detection system
(NEN, Du Pont, MS, USA) were used for detection.
In vitro binding
Kinase assay
The TNT coupled reticulocyte kit was used for in vitro
translation (Promega, WI, USA), according the manufacturer's instructions. All the samples were labeled using 35SMethionine (Amersham, IL, USA). The labeled samples were
incubated with GST-fusion proteins, GST-pRb (379 ± 928) or
GST-pRb (379 ± 928) C706F or GST-pRb (793 ± 928) or
GST-pRb (768 ± 834) or GST-pRb (835 ± 928), using GST
as a negative control, for 2 h at 48C, rocking, for in vitro
binding. These fusion proteins were precipitated by glutathione-agarose anity chromatography and the immunoprecipitates were washed extensively in CPA bu€er (20 mM
Tris-HCl pH 7.6, 250 mM NaCl, 5 mM MgCl2, 0.2 mM
EDTA, 10% glycerol, 0.2% NP-40) containing fresh
inhibitors and 1 mM DTT. Afterwards, the immunoprecipitates were resolved on 8 or 15% SDS ± PAGE and subjected
to autoradiography.
Cell extracts (200 mg) prepared in Lysis Bu€er were used for
immunoprecipitations with a polyclonal anti-cdk9 antibody
(negative control: pre-immune rabbit serum). The immunocomplexes were pulled down with protein A-sepharose and
washed three times with Lysis Bu€er and twice with Lysis
Bu€er containing 400 mM NaCl. The complexes were also
equilibrated in kinase assay bu€er (minus ATP) (20 mM
HEPES pH 7.4, 10 mM Mg Acetate). The kinase assay was
performed in a volume of 20 ml, using 5 mCi/sample of g-ATP
(Amersham, IL, USA). 0.1 mg of GST alone, GST-pRb
(379 ± 928), GST-pRb (793 ± 928), GST-pRb (768 ± 834) and
GST-pRb (835 ± 928) were added to the samples and
incubated for 30 min at 308C. The reactions were stopped
by adding 56 Laemmli Bu€er and the samples were resolved
on a 13% SDS ± PAGE and subjected to autoradiography.
Plasmids
GST pull-down, immunoprecipitations and Western blot
Jurkat and HeLa nuclear cell extracts were used for Western
blot and Ips. Nuclear extracts were obtained by lysing the
cells ®rst in RSB Bu€er (10 mM HEPES, 10 mM KCl,
1.5 mM MgCl2, phenylmethylsulphonyl ¯uoride (PMSF) and
DTT); then, after centrifugation and three washes with RSB
bu€er, the nuclei were lysed with Bu€er C ((20 mM HEPES,
0.2 mM EDTA, 420 mM NaCl, 1.5 mM MgCl2, 25%
Glycerol, fresh aprotinin, leupeptin, DTT and phenylmethylsulphonyl ¯uoride (PMSF)). The protein concentration was
Oncogene
Abbreviations
cdk, cyclin-dependent kinase; GST, glutathione S-transferase; pRb, retinoblastoma protein
Acknowledgments
We thank Marie Basso for her editorial assistance in the
preparation of the manuscript and Dr Gaia Gallo for her
technical support. We thank Dr David Price for generously
cdk9/cyclinT2 binds pRb
C Simone et al
providing reagents. This work was supported in part by
grants from the Sbarro Institute for Cancer Research and
Molecular Medicine (to Dr Giordano). Drs Bellan and
Bagella are supported by a `Dottorato di Ricerca in
Patologia Diagnostica Quantitativa' from the University
of Siena. Finally, we thank Letizia Lavermicocca for her
encouragement and support. We dedicate this work to the
memory of Professor Angelo Carbonara.
4165
References
Adams PD, Li X, Sellers WR, Baker KB, Leng X, Harper
JW, Taya Y and Kaelin WG. (1999). Mol. Cell. Biol., 19,
1068 ± 1080.
Bagella L, MacLachlan TK, Buono RJ, Pisano MM,
Giordano A and De Luca A. (1998). J. Cell. Physiol.,
177, 206 ± 213.
Bagella L, Stiegler P, De Luca A, Siracusa LD and Giordano
A. (2000). J. Cell. Biochem., 78, 170 ± 178.
Corbeil H, Whyte P and Branton P. (1995). Oncogene, 11,
909 ± 920.
Dahmus ME. (1996). J. Biol. Chem., 271, 19009 ± 19012.
DeCaprio JA, Ludlow JW, Figge J, Shew JY, Huand CM,
Lee WH, Marsilio E, Paucha E and Livingston DM.
(1998). Cell, 54, 257 ± 283.
De Falco G and Giordano A. (1998). J. Cell. Physiol., 177,
501 ± 506.
De Falco G and Giordano A. (2002). Cancer Biology and
Therapy, in press.
De Falco G, Bagella L, Claudio PP, De Luca A., Fu Y,
Calabretta B, Sala A and Giordano A. (2000). Oncogene,
19, 373 ± 379.
De Luca A, Esposito V, Baldi A, Claudio PP, Fu Y, Caputi
M, Pisano M, Baldi F and Giordano A. (1997). J. Cell.
Physiol., 172, 265 ± 273.
De Luca P, Majello B and Lania L. (1998). J. Cell Biochem.,
70, 281 ± 287.
Dowdy SF, Hinds PW, Louie K, Reed SI, Arnold A and
Weinberg R. (1993). Cell, 73, 499 ± 511.
Edwards MC, Wong C and Elledge SJ. (1998). Mol. Cell.
Biol., 18, 4291 ± 4300.
Ezhevsky S, Naggahara H, Vocero-Akbani A, Gius D, Wei
M and Dowdy S. (1997). Proc. Natl. Acad. Sci. USA, 94,
10699 ± 10704.
Ewen ME, Sluss HK, Sherr CJ, Matsushime H, Kato J-Y
and Livingston. (1993). Cell, 73, 487 ± 497.
Fanciulli M, Bruno T, Di Padova M, De Angelis R, Iezzi S,
Iacobini C, Floridi A and Passananti C. (2000). FASEB J.,
14, 904 ± 912.
Foskett SM, Ghose R, Tang DN, Lewis DE and Rice AP.
(2001). J. Virol., 75, 1220 ± 1228.
Grana X, De Luca A, Sang N, Fu Y, Claudio PP, Rosenblatt
J, Morgan DO and Giordano A. (1994). Proc. Natl. Acad.
Sci., 91, 3834 ± 3838.
Gu W, Schneider JW, Condorelli G, Kaushal S, Mahadavi S
and Nadal-Ginard B. (1993). Cell, 72, 309 ± 324.
Hatakeyama M, Brill J, Fink G and Weinberg R. (1994).
Genes Dev., 8, 1759 ± 1771.
Hollingsworth RE, Hensey CE and Lee W-H. (1993). Curr.
Opin. Cell. Biol., 3, 55 ± 62.
Kanazawa S, Okamoto T and Peterlin BM. (2000).
Immunity, 12, 61 ± 70.
Kwak YT, Ivanov D, Guo J, Nee E and Gaynor RB. (1999).
J. Mol. Biol., 288, 57 ± 69.
Lee O-J, Russo AA and Pavletich NP. (1998). Nature, 391,
859 ± 865.
Lis JT, Mason P, Peng J, Price DH and Werner J. (2000).
Genes Dev., 11, 2633 ± 2644.
Lundberg AS and Weinberg RA. (1998). Mol. Cell. Biol., 18,
753 ± 761.
MacLachlan TK, Sang N and Giordano A. (1995) Crit. Rew.
Eukariot. Gene Expr., 5, 127 ± 156.
Magnaghi-Jaulin L, Groisman R, Naguibneva I, Robin P,
Lorain S, Le Villain JP, Troalen F, Trouche D and HarelBellan A. (1998). Nature, 391, 601 ± 605.
Mancebo H, Lee G, Flygare J, Tomassini J, Luu P, Zhu Y,
Blau C, Hazuda D, Price D and Flores O. (1997). Genes
Dev., 11, 2633 ± 2644.
Marshall N, Peng J, Xie Z and Price D. (1996). J. Biol.
Chem., 271, 27176 ± 27183.
Mittnacht S. (1998). Curr. Opin. Genet. Dev., 8, 21 ± 27.
Napolitano G, Majello B, Licciardo P, Giordano A and
Lania L. (2000). Gene, 254, 139 ± 145.
Novitch BG, Mulligan GJ, Jacks T and Lassar AB. (1996). J.
Cell. Biol., 135, 441 ± 456.
Novitch BG, Spicer DB, Kim PS, Cheung WL and Lassar
AB. (1999). Curr. Biol., 9, 449 ± 459.
Pan W, Sun T, Hoess RH and Grafstrom RH. (1998).
Carcinogenesis, 19, 765 ± 769.
Pan W, Cox S, Hoess RH and Grafstrom RH. (2001). Cancer
Res., 61, 2885 ± 2891.
Peng J, Zhu Y, Milton JT and Price DH. (1998). Genes Dev.,
12, 755 ± 762.
Puri PL, Balsano C, Burgio V, Chirillo P, Natoli G, Ricci L,
Mattei E, Graessmann A and Levrero M. (1997).
Oncogene, 14, 1171 ± 1184.
Puri PL and Sartorelli V. (2000). J. Cell. Physiol., 185, 155 ±
173.
Qian Y, Luckey C, Horton L, Esser M and Templeton DJ.
(1992). Mol. Cell. Biol., 12, 5363 ± 5372.
Schneider J, Gu W, Zhu L, Mahdavi V and Nadal-Ginard B.
(1994). Science, 264, 1467 ± 1471.
Shin E, Shin A, Paulding C, Scha€hausen B and Yee A.
(1995). Mol. Cell. Biol., 15, 2252 ± 2262.
Simone C and Giordano A. (2001). Frontiers in Bioscience, 6,
1073 ± 1082.
Simone C, Stiegler P, Bagella L, Pucci B, Bellan C, De Falco
G, De Luca A, Guanti G, Puri PL and Giordano A.
(2002). Oncogene, 21, 4137 ± 4148.
Stiegler P and Giordano A. (2001). Crit. Rev. Eukariot. Gene
Expr., 11, 59 ± 76.
Wei P, Gender ME, Fang SM, Fisher WH and Jones KA.
(1998). Cell, 92, 451 ± 462.
Welch PJ and Wang JYJ. (1993). Cell, 75, 779 ± 790.
Whyte P, Buchkovich KJ, Horowitz JM, Friend SH,
Raybuck M, Weinberg RA and Harlow E. (1988). Nature,
334, 124 ± 129.
Wimmer J, Fujinaga K, Taube R, Cujec TP, Zhu Y, Peng J,
Price DH and Peterlin BM. (1999). Virology, 255, 182 ±
189.
Xiao ZX, Chen J, Levine AJ, Modjtahedi N, Xing J, Sellers
WR and Livingston DM. (1995). Nature, 375, 694 ± 698.
Yang X, Hermann HC and Rice AP. (1996). J. Virol., 70,
4576 ± 4584.
Zhu Y, Pe'ery T, Peng J, Ramanathan Y, Marshall N,
Marshall T, Amendt B, Mathews M and Price D. (1997).
Genes Dev., 11, 2622 ± 2632.
Oncogene