Oncogene (2003) 22, 6204–6213
& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00
www.nature.com/onc
Histone deacetylase 1 represses the small GTPase RhoB expression in
human nonsmall lung carcinoma cell line
Shaowen Wang1, Yan Yan-Neale1, Denise Fischer1, Marija Zeremski1, Richard Cai1, Jian Zhu1,
Fred Asselbergs1, Garret Hampton2 and Dalia Cohen*,1
1
Department of Functional Genomics, Novartis Pharmaceutical Corporation, East Hanover, Summit, NJ 07901, USA; 2Genomics
Institute of the Novartis Research Foundation, San Diego, CA 92121, USA
The dynamic balance between histone acetylation and
deacetylation plays a significant role in the regulation of
gene transcription. Much of our current understanding of
this transcriptional control comes from the use of HDAC
inhibitors such as trapoxin A (TPX), which leads to
hyperacetylated histone, alters local chromatin architecture and transcription and results in tumor cell death. In
this study, we treated tumor cells with TPX and HDAC1
antisense oligonucleotides, and analysed the transcriptional consequences of HDAC inhibition. Among other
genes, the small GTPase RhoB was found to be
significantly upregulated by TPX and repressed by
HDAC1. The induction of RhoB by HDAC inhibition
was mediated by an inverted CCAAT box in the RhoB
promoter. Interestingly, measurement of RhoB transcription in B130 tumor-derived cell lines revealed low
expression in almost all of these samples, in contrast to
RhoA and RhoC. Accumulating evidence indicates that
the small GTPase Rho proteins are involved in a variety of
important processes in cancer, including cell transformation, survival, invasion, metastasis and angiogenesis. This
study for the first time demonstrates a link between
HDAC inhibition and RhoB expression and provides an
important insight into the mechanisms of HDACmediated transcriptional control and the potential therapeutic benefit of HDAC inhibition.
Oncogene (2003) 22, 6204–6213. doi:10.1038/sj.onc.1206653
Keywords: HDAC1; HDAC inhibitor; RhoB; expression profile; transcription regulation
Introduction
The eukaryotic genome is compacted into nucleosomes
to form chromatin and dynamic alterations in chromatin structure can facilitate or suppress access of
transcription factors to DNA, leading to transcription
regulation. Changes in the acetylation of nucleosomal
core histones play a pivotal role in this regulation.
*Correspondence: D Cohen; E-mail: [email protected]
Received 12 March 2003; revised 31 March 2003; accepted 14 April 2003
Typically, increased levels of histone acetylation are
associated with active transcription, whereas decreased
acetylation is associated with transcription repression
(Grunstein, 1997; Strahl and Allis, 2000).
The acetylation status of histones is dynamically
controlled by histone acetyltransferases (HATs) and
histone deacetylases (HDACs). A large number of
HATs have been identified as transcriptional coactivators, among them GCN5 (general control nonderepressible), (CBP)/p300 (cAMP response element binding
protein-binding protein), PCAF (p300/CBP-associated
factor), SRC-1 (steroid receptor coactivator) and
TAF250 (TATA box binding protein associated factor)
(Chen et al., 2001a; Nakatani, 2001). A total of 18
mammalian HDACs described to date are found to be
associated with transcription repression and have been
classified into three classes (Imai et al., 2000; Gray and
Ekstrom, 2001; Khochbin et al., 2001). Class I,
encompassing HDAC1-3, 8 and 11 are related to Rpd3
(a first class of yeast histone deacetylase). Class II,
including HDAC4-7, 9 and 10 are related to Hda1 (a
second class of yeast histone deacetylase) (Bertos et al.,
2001; Zhou et al., 2001). Finally, Class III, also known
as the Sir2 (silent information regulator 2) family,
consists of seven genes related to yeast Sir2, and possess
nicotinamide-adenine dinucleotide (NAD þ )-dependent
deacetylase activity (Vaziri et al., 2001).
HDACs are usually found in large multiprotein
complexes that regulate gene transcription. A number
of transcriptional repressors and corepressors, such as
Sin3, SMRT (silencing mediator of retinoid acid and
thyroid hormone receptor) and N-CoR (nuclear receptor corepressor), have been shown to recruit HDAC
containing complexes to promoter regions (Imai et al.,
2000; Gray and Ekstrom, 2001; Khochbin et al., 2001).
Evidence is accumulating that HDACs differ in cellular
localization and presence in transcription complexes
regulating a variety of genes, suggesting that different
HDACs may have different biological activities (Jones
et al., 2001). In addition to deacetylation of histones,
HDACs have been shown to regulate gene expression by
deacetylating transcription factors, such as p53, the
erythroid kruppel-like factor (EKLF) and the erythroid
cell differentiation factor GATA-1 (Pazin and Kadonaga, 1997; Liu et al., 1999).
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6205
HDAC inhibitors are potent inducer of growth arrest,
differentiation and apoptotic cell death in a variety of
transformed cells in culture and in tumor-bearing
animals (Weidle and Grossmann, 2000; Marks et al.,
2001). Hence, HDAC inhibitors are being evaluated as
potent anticancer agents. Trapoxin A (TPX) is a potent,
noncompetitive HDAC inhibitor, which has been used
to explore the mechanisms by which histone deacetylation alters transcription (Furumai et al., 2001). The
action of HDAC inhibitors on gene expression appears
to be selective, altering the transcription of only a
limited number of genes in cultured tumor cells (Van
Lint et al., 1996). The basis for this gene selectivity of
HDAC inhibitors is unknown.
The biochemical events subsequent to HDAC inhibition that lead to cell cycle arrest, cellular differentiation
and apoptosis have not yet been identified. We have
used a genomic approach to identify the transcriptional
consequences of HDAC inhibition and to elucidate
molecular mechanisms of HDAC-mediated transcriptional repression. Previously, we and others found that
the expression of p21Waf1 was induced following inhibition of HDAC by TPX (Van Lint et al., 1996;
Sambucetti et al., 1999). In addition, TGF-b type II
receptor gene (TbRII) (Park et al., 2002), multidrug
resistance gene 1 (MDR1) (Jin and Scotto, 1998) and
thioredoxin binding protein-2 (TBP-2) (Butler et al.,
2002) were also found to be upregulated following
treatment of tumor cells with HDAC inhibitor. In this
study, we used DNA microarrays to identify additional
genes that were modified by HDAC inhibition and
found that RhoB was one of the most significantly
upregulated gene in H1299 cells. It has been shown that
the small GTPase Rho proteins are involved in a variety
of tumorigenic processes, including cell transformation,
survival, invasion, metastasis and angiogenesis (Chen
et al., 2000; Prendergast, 2001; Adnane et al., 2002;
Sahai and Marshall, 2002). The altered expression or
activity of Rho proteins might be crucial to cancer
progression and therapeutic responses. Hence, we focus
our investigation on revealing the molecular mechanism
of RhoB induction by HDAC inhibition.
Rho proteins have been reported to possess diverse
cellular functions, including regulation of cytoskeletal
actin organization, adhesion, motility, vesicle transport,
cell-cycle progression, cytokinesis and transcription
(Prendergast, 2001). RhoB belongs to the Rho subgroup
of the Ras superfamily, and shares over 85% homology
with RhoA and RhoC, two other members of the Rho
subgroup. However, each protein plays a distinct
cellular biological role (Prendergast, 2001; Takai et al.,
2001; Sahai and Marshall, 2002). There are several
features that distinguish RhoB from other Rho proteins.
RhoB is an immediate early response gene that is rapidly
induced by various growth and stress stimuli, especially
DNA-damage reagents (Jahner and Hunter, 1991; Fritz
et al., 1995; Engel et al., 1998) and is turned over much
more rapidly than other Rho proteins. It was also shown
that RhoB has an important role in growth inhibition
and apoptosis in transformed cells (Chen et al., 2000;
Liu, 2001; Adnane et al., 2002).
In this study, for the first time a molecular mechanism
of RhoB induction by HDAC inhibition was deciphered. The suppression of RhoB transcription by
HDAC was found to be mediated by an inverted
CCAAT element in its promoter. Furthermore, microarray analysis of RhoB gene expression in transformed
cells demonstrates almost consistently low expression,
strongly suggesting that low levels of RhoB expression
might be involved in high proliferation. These results
support the model that RhoB has a growth inhibitory
function and may be a negative modifier in cancer.
Results
RhoB expression is induced following HDAC inhibition
To identify transcriptional events that occur as a result
of HDAC inhibition, we used DNA microarray analysis
to monitor gene expression in response to HDAC
inhibition by TPX treatment. By comparing changes in
TPX-treated versus untreated human nonsmall lung
carcinoma, H1299 cells, a gene expression pattern that
correlated with HDACs inhibition was determined.
Among several genes, the small GTPase RhoB was
found to be one of the most significantly upregulated
genes, about eightfold, in response to TPX treatment in
these cells (Figure 1a). p21waf1, a gene previously
Figure 1 Induction of RhoB expression by TPX. (a) Fold changes
of RhoA, RhoB, RhoC and p21waf1 mRNA following TPX
treatment in microarray analysis. (b) Northern blots were
performed to confirm the changes in transcription as measured
by microarray analysis, b-Actin was used as loading control to
ensure equal loading. (c) Induction of RhoB protein by TPX was
examined by Western blot analysis. Lower panel shows membrane
staining by Ponceau S as a loading control
Oncogene
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6206
Figure 2 Expression profile of RhoB, RhoA, and RhoC in tumor-derived cells. Affymetrix U95a GeneChip microarrays were used to
profile the expression of RhoB, RhoA and RhoC genes in 136 cell lines. Columns, RhoA, RhoB and RhoC genes; rows, individual
samples. Each cell represents the expression level of a single transcript in a single sample. Black represents an intensity value of 200.
Red and green represent intensity levels above and below 200, respectively. Color intensity is proportional to the magnitude of the
difference from the intensity value 200. The arrows indicate normal epithelial cells
reported to be induced in response to HDAC inhibition
(Sambucetti et al., 1999) was found to be upregulated by
about sixfold. Interestingly, the expression of two other
Rho GTPase subfamily members, RhoA and RhoC,
which share over 85% homology with RhoB, was
unchanged following the TPX treatment. We performed
Northern and Western blots to analyse RhoB, providing
independent validation, as well as extending the results
of the DNA microarray experiments (Figure 1b, 1c).
Increased levels of RhoB protein could be detected as
early as 6 h, with maximum levels observed by 24–48 h
after TPX treatment (Figure 1c).
RhoB is differentially expressed in transformed cell lines
The use of HDAC inhibitors in the clinic as new cancer
therapeutics and the induction of RhoB in response to
an HDAC inhibitor promoted us to survey the status of
RhoB expression in tumor cell lines. We used DNA
microarray analysis to characterize RhoB, RhoA and
Oncogene
RhoC expression in 136 cell lines. RhoB levels were low
in most transformed cell lines, compared with the
expression levels of RhoA and RhoC. Moreover, RhoB
expression levels were lower in transformed cells than in
corresponding normal cells (Figure 2). In addition, low
RhoB expression was also detected in several tumor
tissues relative to corresponding normal tissues (data
not shown).
Human and mouse RhoB promoters
To understand the induction of RhoB following TPXmediated inhibition of HDACs, and to determine
whether this induction could be mapped to the RhoB
promoter, we cloned the region of the human RhoB
gene from 1224 to 253 (relative to the translation
initiation site, þ 1 ATG) upstream of a luciferase
reporter gene (Figure 3a). On the basis of EST and
transcript data analysis, a transcriptional initiation site
was predicted at position 428 (G). Comparison of the
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6207
sequences of human and mouse RhoB 50 upstream
regions (EMBL Accession number Y09248) (Fritz
and Kaina, 1997) revealed that the two promoters are
highly conserved around the potential transcription
initiation site. Furthermore, these promoters also share
many putative transcription factors binding sites
(Figure 3b).
An inverted CCAAT box in the RhoB promoter is crucial
for RhoB induction
The effect of TPX on the transcriptional activity of the
RhoB promoter was determined in H1299 cells trans-
fected with RhoB promoter–reporter constructs. Luciferase activity driven by the RhoB promoter was
induced to about 8–10-fold following TPX treatment
(Figure 3c). To define the region required for transcriptional induction, we transfected a series of RhoB
promoter 50 deletion constructs into H1299 cells. All
these constructs generated were induced similarly as
shown by luciferase activity (Figure 3c).
Examination of the RhoB promoter sequences that
were encompassed in the shorter RhoB promoter
luciferase construct (497 and 392) and were also
induced by TPX, revealed the presence of two potential
transcription factor binding sites. A putative TATA box
Figure 3 Identification of TPX responsive sequences in the RhoB promoter (a) Sequence of the 50 -noncoding region of human RhoB gene. The
enlarged G in bold at position 428 indicates the putative transcription start site. Putative Sp1 sites (blue), CCAAT boxes (yellow) and TATA
box (pink) are indicated. (b) Human and mouse RhoB promoters are schematically depicted. The putative binding sites for various transcription
factors are indicated. þ 1 ATG is the translation start site. (c) Relative activity (TPX/DMSO) of the RhoB promoter deletion constructs. A
schematic representation of RhoB promoter deletion constructs is shown on the left. (d) Percent activity of the RhoB promoter mutations
constructs. The activity of the pGL3RhoB5-WT was set as 100%. Point mutations and insertion are indicated in red
Oncogene
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6208
Figure 3
Continued
at position 438, and an inverted CCAAT box (50 ATTGG-30 ) at position 451 with a five nucleotide
space between these two DNA elements (Figure 3a). To
define which, if any, of these elements were required for
TPX induction, we examined the activation of luciferase
constructs containing point mutations in the CCAAT
and TATA boxes, as well as a construct containing an
insertion of three nucleotides that changed the distance
between these two boxes. The response of RhoB
promoter to TPX was significantly reduced when the
inverted CCAAT box (50 -ATTGG-30 ) was mutated to
50 -TTTAG-30 . However, mutation of the TATA box (50 TATATTAA-30 ) to 50 -CAGATCAA-30 had no significant effect on RhoB promoter induction. In addition,
the insertion of three nucleotides (AAA) between the
CCAAT and TATA boxes had no effect. These data
indicated that the inverted CCAAT box is critical for the
Oncogene
induction of the RhoB promoter following TPXmediated inhibition of HDACs (Figure 3d).
Nuclear protein specifically binds to the inverted CCAAT
element in the RhoB promoter
To ensure that nuclear proteins bind to the inverted
CCAAT element, which mediated RhoB promoter
activity in response to TPX treatment, gel shift assays
were performed. Nuclear extracts were prepared from
TPX-treated or untreated H1299 cells, and RhoB
promoter sequence from 497 to 392 was used as
probe. As shown in Figure 4, DNA/protein complex was
observed. Moreover, the complex was only competed by
excess unlabeled probes containing the wild-type
CCAAT element, but not by excess unlabeled oligonucleotides containing the mutant CCAAT element. TPX
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6209
treatment had no obvious effect on complex formation
(data not shown).
Elevated levels of acetylated histone H3 and H4 in TPX
responsive region of RhoB promoter by TPX treatment
To investigate whether RhoB promoter transcriptional
activation correlated with acetylation of Histone H3 or
H4, chromatin immunoprecipitations were performed
with TPX-treated and untreated H1299 cells using
antibodies against acetylated histone H3 or H4. The
immunoprecipitated DNAs were analysed by PCR using
RhoB-specific primers. As shown in Figure 5, TPX
treatment for 12 or 24 h increased histone H3 and H4
acetylation associated with RhoB promoter region from
557 to 412, which encompassed TPX-responsive
sequences. Interestingly, TPX did not effect the acetylation level of RhoB promoter region between 823
and 536 located upstream of the TPX-responsive
sequences.
The RhoB promoter is associated with and repressed by
HDAC1
Figure 4 Gel shift analysis. An aliquot of 10 mg nuclear extracts from
H1299 cells was analysed for DNA binding activity to the 32p-labeled
0.1 kb RhoB promoter fragment (495 to 392) by gel retardation.
For competition experiments, the unlabeled fragment was added in 50fold molar excess. The arrow indicates the shifted bands
To determine whether the induction of the RhoB
promoter by TPX is related to in vivo recruitment of
HDAC complexes to the promoter, we investigated the
association of HDACs with the RhoB promoter by
chromatin immunoprecipition assay. H1299 cells were
fixed in 1% formaldehyde, and chromatin lysates were
immunoprecipited with specific antibodies against
HDAC1, 2 and 6. RhoB promoter fragment from
557 to 412 was found to be associated with HDAC1
and HDAC2, but not with HDAC6. As a control,
chromatin was immunoprecipitated in the absence of
specific antibodies and the same RhoB promoter
fragment was not detected (Figure 6a).
As shown in Figure 3, the change in the induction of
luciferase activity following TPX treatment was
mediated mainly by the inverted CCAAT box in the
RhoB promoter. To investigate the direct effect of
HDACs on the RhoB promoter, the effects of ectopic
expression of HDAC1 and HDAC6 on RhoB promoter
constructs were tested. The activity of RhoB promoter
was repressed with overexpression of HDAC1, but not
Figure 5 Inhibition of HDAC by TPX leads to increased histone H3 and H4 acetylation in the TPX responsive region of RhoB
promoter. CHIP assays were performed using chromatin isolated from H1299 cells either treated with TPX for 12 or 24 h or 0.1%
DMSO. Antibodies aganist acetylated histones H3 and H4 were used to isolate acetylated chromatin. PCR primers for two regions,
one of them encompassing the TPX-responsive region, were used to amplify the RhoB promoter fragments. Input: input DNAs were
used to show the relative locations of the PCR products. M: DNA 1 kb ladder. þ 1: translation start site
Oncogene
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6210
Figure 7 Induction of RhoB by blocking of HDAC1 expression.
(a) Analysis of HDAC1 RNA levels by real-time RT–PCR after
transfection with HDAC1 antisense or control oligonucleotides.
The RNA level of HDAC1 in mock-transfected cells was set as
100%. (b) Analysis of HDAC1 and RhoB protein levels by Western
blot. Lower panel shows membrane staining by Ponceau S as a
loading control. Mock: mock transfection; NAS 5596, unrelated
luciferase control oligonucleotides; NAS 7618, HDAC1 corresponding mismatch oligonucleotides, NAS 6887, HDAC1 antisense
oligonucleotides
Figure 6 RhoB promoter is associated with and inhibited by
HDAC1. (a) Association of HDAC on RhoB promoter was
investigated by chromatin immunoprecipition (CHIP) assay. The
upper panel shows the product of a PCR reaction of RhoB
promoter fragment (557 to 412) from CHIP using indicated
antibodies and controls. Schematic representation of PCR analysis
is shown on the lower panel. þ 1: translation start site. (b) The
RhoB promoter was repressed by HDAC1 but not by HDAC6.
The cells were transfected with indicated constructs and luciferase
activities were determined. The activity of the vector control was
set as 100%. Lower panel shows the expression levels of HDAC1
and HDAC6 in the corresponding cell lysates
by HDAC6 (Figure 6b). However, HDAC1 overexpression had no effect on the construct containing the
mutation in the inverted CCAAT box. These results
demonstrated that RhoB promoter activity is specifically
regulated by HDAC1, but not by HDAC6 via the
inverted CCAAT box (Figure 6b).
Expression of RhoB is induced by blocking HDAC1
synthesis
To determine the effect of specific HDAC1 blockade on
RhoB expression, we used antisense oligonucleotides
designed against the HDAC1 sequence. Anti-HDAC
antisense oligonucleotides, as well as mismatch oligonucleotide as controls, were synthesized and transfected
Oncogene
into H1299 cells. One antisense oligonucleotide in
particular, NAS 6887, was found to significantly inhibit
HDAC1 mRNA levels, and was subsequently used in
the experiment. Samples from NAS 6887-treated and
mismatch (NAS 7618)-treated cells were analysed by
real time RT–PCR and by Western blot for HDAC1
and RhoB expression levels. The expression of HDAC1
was blocked by NAS 6887, and RhoB protein level was
significantly induced. In contrast, no effects on RhoB
expression were found in cells transfected with the NAS
7618 (mismatch) and NAS 5596 (unrelated luciferase)
oligonucleotides (Figure 7).
Discussion
HDAC inhibitors have shown considerable potential as
new antitumor agents due to their ability to induce
differentiation and apoptosis via transcriptional modulation. Several HDAC inhibitors are presently being
evaluated in clinical trials (Marks et al., 2001). To
elucidate the biological consequences of HDAC inhibition, we determined the global gene expression profiles
of H1299 cells following inhibition of HDAC with TPX.
Among other genes, the small GTPase RhoB was one of
the most significantly upregulated in response to TPX.
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6211
The increased levels of RhoB mRNA, protein and
promoter activity suggested that the RhoB gene was
transcriptionally regulated by HDACs, thus linking
RhoB expression and HDAC inhibition. In contrast, the
transcription of two other closely related Rho GTPase
subfamily members, RhoA and RhoC, was not altered
in response to TPX-mediated HDAC inhibition. Chromatin IP experiments indicated that HDAC1 and
HDAC2, but not HDAC6, were associated with the
RhoB promoter and consistent with this observation,
overexpression of HDAC1 could specifically repress the
RhoB promoter, whereas HDAC6 had no effect on
RhoB expression. In addition, antisense inhibition of
HDAC1 induced RhoB gene expression, providing
independent evidence that HDAC1 mediates RhoB
transcriptional repression.
We investigated the specific mechanism of transcriptional regulation of RhoB by HDAC through determining the effect of TPX inhibition on a series of RhoB
promoter deletions. Interestingly, induction of expression of all these deletions was similar following TPX
treatment. To define the TPX responsive sequences,
mutations were introduced into the CCAAT and the
TATA boxes in the proximal region of the RhoB
promoter. Only promoter constructs harboring mutations in the inverted CCAAT box lost the ability to be
induced following TPX treatment, indicating the involvement of this sequence in the repression of RhoB
promoter activity by HDAC. Furthermore, mutation in
this CCAAT box reduced the ability of HDAC1 to
induce repression of RhoB promoter activity. In
addition, chromatin immunoprecipitation assays revealed that TPX increased the level of chromatin
acetylation in the TPX-responsive region of the RhoB
promoter containing the inverted CCAAT box. Gel shift
assays indicated that NF-Y complex bind to this
inverted CCAAT box in RhoB promoter in vitro (data
not shown), suggesting that NF-Y might be a component in the HDAC1/CCAAT complex.
These findings represent the first evidence that HDAC
activity is involved in the transcriptional regulation of
RhoB by a specific element in its promoter.
It has been reported that the RhoB promoter can also
be induced by a variety of growth and stress stimuli
(Jahner and Hunter, 1991; Fritz et al., 1995; Engel et al.,
1998) and studies have shown that the CCAAT box in
the RhoB promoter is involved in transcription activation by genotoxicants (Fritz and Kaina, 1997, 2001b).
Previously, p21waf1 was reported also to be induced by
TPX, and an SP1 site in its promoter was shown to
mediate this response (Sambucetti et al., 1999). Interestingly, the Sp1 site in the RhoB promoter was not
found to mediate the effect of HDAC inhibition. Thus,
supporting a model of at least two classes of genes
regulated by HDAC inhibition, genes that required Sp1
sites and genes that required CCAAT sites as suggested
recently by Richon’s group (Butler et al., 2002).
Recent findings indicated that altered expression or
activity of Rho proteins might be crucial to cancer
progression and therapeutic responses (Prendergast,
2001; Adnane et al., 2002; Sahai and Marshall, 2002).
As an important step toward understanding RhoB
function, and the findings that RhoB induction is
mediated by HDAC inhibition, we performed a survey
of RhoB expression in transformed cells. Significantly
low levels of RhoB expression were observed in most of
the cells, suggesting that RhoB suppression is associated
with high proliferation rates. The activity of Rho family
members has been reported to be determined by GTP/
GDP binding (Prendergast, 2001; Takai et al., 2001) and
post-translation modifications (isoprenylation) (Liu and
Prendergast, 2000). However, recent studies indicated
that the expression levels of Rho family members are
also important in cancer. For example, the expression of
RhoB was suppressed in invasive head and neck cancer
(Adnane et al., 2002), and the expression of RhoA and
RhoC play an essential role in tumor invasion and
metastasis (Yoshioka et al., 1999; Clark et al., 2000).
Recent studies have suggested that RhoB can act as
an activator of apoptosis in transformed cells (Fritz and
Kaina, 2000; Liu, 2001). The ability of RhoB to suppress
the cell survival regulator AKT and NF-kB activity
might contribute to the sensitivity of transformed cells
to RhoB-dependent apoptosis (Fritz and Kaina, 2001a;
Liu and Prendergast, 2000). In addition, it was reported
that RhoB might sensitize cells to anticancer agents
including DNA-damaging agents, microtubule-destabilizing drugs and farnesyl transferase inhibitors. Loss or
inactivation of RhoB may also contribute to the
development of radioresistance or chemotherapeutic
resistance in cancer (Prendergast, 2001). In contrast,
RhoA and RhoC play essential roles in transformation,
metastasis and invasion of tumor cells (Yoshioka et al.,
1999; Clark et al., 2000; van Golen et al., 2000; Kamai
et al., 2001).
Our data provide the first global view of the expression
patterns of three Rho subfamily genes in transformed
cells, and further support the hypothesis that RhoB has
growth inhibitory function and might be a negative
modifier in cancer (Du and Prendergast, 1999; Chen et al.,
2000; Liu et al., 2001; Adnane et al., 2002).
In summary, we have used global gene expression
analysis to establish for the first time, a link between
RhoB activity and HDAC inhibition, and have identified a DNA sequence within the RhoB promoter that
mediated this regulation. The approach of mapping
HDAC function to distinct DNA elements in regulatory
regions of specific genes may have the potential of
identifying specific target proteins that mediate promoter repression by HDAC inhibition. The induction of
RhoB following HDAC inhibition offers a novel
mechanism to explain some of the tumor suppression
effects of HDAC inhibitors, enhancing the interest of
developing HDAC inhibitors as anticancer drugs.
Materials and methods
Cell culture, transfection and reporter assay
The human nonsmall lung carcinoma cell line, H1299, and
colon carcinoma cell line, HCT116, were cultured in RPMI
Oncogene
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6212
1640 (Life Technologies, Gaithersburg, MD, USA). HeLa cells
were maintained in DMEM medium (Life Technologies). TPX
(Novartis, Basal, Switzerland) was added directly to the cell
medium at a final concentration of 30 nm. Transfections were
performed using GenePORTER 2 (Gene Therapy Systems,
San Diego, CA, USA) or FuGENE 6 (Roche, Indianapolis,
IN, USA) transfection reagents according to the manufacturer’s instructions. Luciferase and b-gal activity assays were
performed according to the manufacturer’s instructions
(Promega, Madison, WI, USA). For the cotransfection
experiment, luciferase activity was normalized to the pCMVb-gal activity. At least three independent experiments were
performed for each assay.
Microarray analysis
Microarray analysis was performed essentially as described
previously (Welsh, 2001). Poly (A) þ RNA was prepared from
H1299 cells treated with TPX or DMSO for 0, 6 and 24 h.
Total RNA was prepared from 136 different cell lines. Labeled
cDNA was prepared and hybridized to Affymetrix Hu6800 or
U95a GeneChip oligonucleotide arrays (Affymetrix, Santa
Clara, CA, USA). The arrays were scanned on an Affymetrix
confocal scanner and analysed with GENECHIP 3.1 software
(Affymetrix).
Northern and Western analysis
Total RNA was isolated from H1299 cells treated for 6 h with
30 nm TPX or DMSO using Trizol reagent (Life Technologies).
The RNA samples were analysed by Northern blots using
standard protocols. A full-length cDNA of p21waf1 was used as
probe. Specific probes for RhoA, RhoB and RhoC were
amplified by RT–PCR and used for hybridizations. To ensure
equal loading, the b-actin levels were used as a loading control.
The primers used for probe amplification were the following:
RhoA-forward
RhoA-reverse
RhoB-forward
RhoB-reverse
RhoC-forward
RhoC-reverse
50 -GTCCTTTTGACACTGCTCTAAC-30
50 -CTTGAGATGACACTGCTCTAAC-30
50 -TCAGATGTTCGCCCTTCACCAG-30
50 -GTTACAGCGTACAAGTGTGGTCA
G-30
50 -TCACAGGGGTACAGAAATTATCC30
50 -GACCAAATGCAGTGAGAGACAAG30
For Western analysis, total protein was extracted from
H1299 cells at 0, 6, 24 and 48 h following TPX treatment and
hybridized with an antibody against RhoB (Santa Cruz
Biotechnology, Santa Cruz, CA, USA).
Plasmid preparation
To identify the 50 -upstream region of the human RhoB gene,
GenBank and Incyte LifeSeq databases were searched to
identify RhoB cDNAs with the longest 50 -end extensions. The
cDNA was subsequently mapped to the human genome. A
1 kb genomic region upstream of the cDNA start sites was
selected as the putative promoter. A 0.97 kb human RhoB
promoter fragment corresponding to a region from 253 to
1224, relative to the translation initiation site ( þ 1 ATG),
was amplified (primer sequences underlined in Figure 4a). The
series of deletion fragments of the RhoB promoter was made
by amplifying specific 50 -upstream sequences inserted into the
pGL3 basic vector (Promega, Madison, WI, USA), which
contains a luciferase reporter gene. These constructs were
designated as pGL3RhoB1 (1124/253), pGL3RhoB2
Oncogene
(821/253), pGL3RhoB3 (557/253), pGL3RhoB4
(497/392) and pGLRhoB5-WT (497/392). The RhoB5
promoter fragments containing point mutations or insertions
were used to generate pGL3RhoB5-M-CAAT pGL3RhoB5M-TATA, pGL3RhoB5-M-CATA and pGL3RhoB5-M-Ins.
pCMV-b-gal was purchased from Promega.
Gel shift assay
A 0.1 kb. double-stranded oligonucleotide probe selected from
the human RhoB promoter, regions 497 to 392 (RhoB5),
was labeled with [g-32P]ATP and purified with MicroSpin TM
G25 columns (Amersham PharmaciaBiotech). The gel shift
assay was carried out using a Gel Shift Assay Kit (Promega,
Madison, WI, USA). Nuclear extracts (10 mg) from H1299 cells
were incubated with 5 ng of labeled probe in a 20 ml reaction
volume for 20 min at room temperature. For competition test,
50-fold excess (250 ng) of unlabeled wild-type or mutant RhoB
probes (as depicted in Figure 4d) were added to each reaction.
The reaction mixture was subjected to 6% polyacrylamide gel.
Gels were dried and scanned with a Typhoon 9400 Variable
Mode Imager (Molecular Dynamics Inc., Sunnyvale, CA, USA).
Chromatin immunoprecipitation
Chromatin immunoprecipitation analysis was conducted as
described previously (Boyd et al., 1998; Moreno et al., 1999).
The chromatin was immunoprecipitated either with normal
rabbit IgG, anti-acetyl-histone H3, anti-acetyl-histone H4,
anti-HDAC1, anti-HDAC2 and anti-HDAC6 antibodies
(Upstate Biotechnology, Lake Placid, NY, USA, and Santa
Cruz Biotechnology, Santa Cruz, CA, USA), or without
antibody. The supernatant fraction from the reaction lacking
primary antibody was saved as the ‘chromatin input’. DNAs
were purified and used as the PCR templates to amplify the
RhoB promoter regions from 823 to 536 and 557 to
412 relative to the translation initiation site. PCR products
were separated on a 2% agarose gel, stained with ethidium
bromide, and the images were recorded.
Cell transfection with antisense oligonucleotides
Oligonucleotides were synthesized as phosphorothioates (s) in
chimeric form containing partly 20 -O-methoxyethyl-(MOE)modified oligoribonucleotides. H1299 cells were transfected
with HDAC1 antisense oligonucleotides NAS 6887, HDAC1
mismatch oligonucleotide NAS 7618, luciferase mismatch
oligonucleotide NAS 5596 or mock transfected. The sequences
of
the
oligonucleotides
were:
NAS
6887:
50 GCTGTAsCsTsCsCsGsAsCsATGTT-30 , NAS 7618: 50 GCTTTAsCsGsCsCsTsAsCsAGGTT-30 , NAS 5596: 50 CCTTACsCsTsGsCsTsAsGsCTGGC-30 . Transfection was repeated twice on two consecutive days. Cells were harvested
72 h after the first transfection. RNA and protein were isolated
using Trizol reagent (Life Technologies) according to the
manufacturer’s protocols.
Real time RT–PCR
Real time RT–PCR was performed using qPCR Core Kit
(Eurogentec, Belgium) in an ABI PRISM 7700 Sequence
Detector (Applied Biosystems, Forster City, CA, USA)
following the manufacturer’s instructions. HDAC1 forward
primer (50 -GCGATGAGGACGAAGACGAC-30 ), HDAC1
reverse primer (50 -TCACAGGCAATTCGTTTGTCA-30 ) and
HDAC1 TaqMan probe (50 -CTGACAAGCGCATCTCGATCTGCTCC-30 ) containing 50 FAM fluorescent reporter
and 30 TAMRA quencher dyes were used in the reaction at the
Mechanism of HDAC1 repression of RhoB expression
S Wang et al
6213
final concentrations of 200 nm for primers and 100 nm for
fluorescent probe. Relative measurement of the amplified
product was performed using the comparative CTmethod as
described in the manufacturer’s manual (Applied Biosystems,
ABI Prism 7700 sequence detection system, user bulletin 2).
Acknowledgements
We thank R Kramer for reviewing the manuscript, and E
Verdin for providing HDAC6 expression vector (Chen et al.,
2001b).
References
Adnane J, Muro-Cacho C, Mathews L, Sebti SM and MunozAntonia T. (2002). Clin. Cancer Res., 8, 2225–2232.
Bertos NR, Wang AH and Yang XJ. (2001). Biochem. Cell
Biol., 79, 243–252.
Boyd KE, Wells J, Gutman J, Bartley SM and Farnham PJ.
(1998). Proc. Natl. Acad. Sci. USA, 95, 13887–13892.
Butler LM, Zhou X, Xu WS, Scher HI, Rifkind RA, Marks
PA and Richon VM. (2002). Proc. Natl. Acad. Sci. USA, 20,
20.
Chen H, Tini M and Evans RM. (2001a). Curr. Opin. Cell
Biol., 13, 218–224.
Chen L, Fischle W, Verdin E and Greene WC. (2001b).
Science, 293, 1653–1657.
Chen Z, Sun J, Pradines A, Favre G, Adnane J and Sebti SM.
(2000). J. Biol. Chem., 275, 17974–17978.
Clark EA, Golub TR, Lander ES and Hynes RO. (2000).
Nature, 406, 532–535.
Du W and Prendergast GC. (1999). Cancer Res., 59,
5492–5496.
Engel ME, Datta PK and Moses HL. (1998). J. Biol. Chem.,
273, 9921–9926.
Fritz G and Kaina B. (1997). J. Biol. Chem., 272, 30637–30644.
Fritz G and Kaina B. (2000). Biochem. Biophys. Res.
Commun., 268, 784–789.
Fritz G and Kaina B. (2001a). J. Biol. Chem., 276, 3115–3122.
Fritz G and Kaina B. (2001b). Nucleic Acids Res., 29, 792–798.
Fritz G, Kaina B and Aktories K. (1995). J. Biol. Chem., 270,
25172–25177.
Furumai R, Komatsu Y, Nishino N, Khochbin S, Yoshida M
and Horinouchi S. (2001). Proc. Natl. Acad. Sci. USA, 98,
87–92.
Gray SG and Ekstrom TJ. (2001). Exp Cell Res., 262, 75–83.
Grunstein M. (1997). Nature, 389, 349–352.
Imai S, Armstrong CM, Kaeberlein M and Guarente L.
(2000). Nature, 403, 795–800.
Jahner D and Hunter T. (1991). Mol. Cell. Biol., 11,
3682–3690.
Jin S and Scotto KW. (1998). Mol. Cell. Biol., 18, 4377–4384.
Jones PL, Sachs LM, Rouse N, Wade PA and Shi YB. (2001).
J. Biol. Chem., 276, 8807–8811.
Kamai T, Arai K, Tsujii T, Honda M and Yoshida K. (2001).
BJU Int., 87, 227–231.
Khochbin S, Verdel A, Lemercier C and Seigneurin-Berny D.
(2001). Curr. Opin. Genet. Dev., 11, 162–166.
Liu A. (2001). Proc. Natl. Acad. Sci. USA, 98, 6192–6197.
Liu A and Prendergast GC. (2000). FEBS Lett., 481, 205–208.
Liu AX, Rane N, Liu JP and Prendergast GC. (2001). Mol.
Cell. Biol., 21, 6906–6912.
Liu L, Scolnick DM, Trievel RC, Zhang HB, Marmorstein R,
Halazonetis TD and Berger SL. (1999). Mol. Cell. Biol., 19,
1202–1209.
Marks PA, Rifkind RA, Richon VM, Breslow R, Miller T and
Kelly WK. (2001). Nat. Rev. Cancer, 1, 194.
Moreno CS, Beresford GW, Louis-Plence P, Morris AC and
Boss JM. (1999). Immunity, 10, 143–151.
Nakatani Y. (2001). Genes Cells, 6, 79–86.
Park SH, Lee SR, Kim BC, Cho EA, Patel SP, Kang HB,
Sausville EA, Nakanishi O, Trepel JB, Lee BI and Kim SJ.
(2002). J. Biol. Chem., 277, 5168–5174.
Pazin MJ and Kadonaga JT. (1997). Cell, 89, 325–328.
Prendergast GC. (2001). Nat. Rev. Cancer, 1, 162.
Sahai E and Marshall CJ. (2002). Nat. Rev. Cancer, 2,
133–142.
Sambucetti LC, Fischer DD, Zabludoff S, Kwon PO,
Chamberlin H, Trogani N, Xu H and Cohen D. (1999). J.
Biol. Chem., 274, 34940–34947.
Strahl BD and Allis CD. (2000). Nature, 403, 41–45.
Takai Y, Sasaki T and Matozaki T. (2001). Physiol. Rev., 81,
153–208.
van Golen KL, Wu ZF, Qiao XT, Bao LW and Merajver SD.
(2000). Cancer Res., 60, 5832–5838.
Van Lint C, Emiliani S and Verdin E. (1996). Gene Expr., 5,
245–253.
Vaziri H, Dessain SK, Ng Eaton E, Imai SI, Frye RA, Pandita
TK, Guarente L and Weinberg RA. (2001). Cell, 107,
149–159.
Weidle UH and Grossmann A. (2000). Anticancer Res., 20,
1471–1485.
Welsh JB. (2001). Cancer Res., 61, 5974–5978.
Yoshioka K, Nakamori S and Itoh K. (1999). Cancer Res., 59,
2004–2010.
Zhou X, Marks PA, Rifkind RA and Richon VM. (2001).
Proc. Natl. Acad. Sci. USA, 98, 10572–10577.
Oncogene