Content uploaded by Pierre Chymkowitch
Author content
All content in this area was uploaded by Pierre Chymkowitch on Nov 10, 2017
Content may be subject to copyright.
Molecular Cell, Vol. 16, 187–197, October 22, 2004, Copyright 2004 by Cell Press
Selective Regulation of Vitamin D
Receptor-Responsive Genes by TFIIH
al., 1996). While essential for DNA-repair activity of
TFIIH, the XPD helicase serves a structural rather than an
enzymatic role in transcription. It links the cdk-activating
Pascal Drane
´, Emmanuel Compe, Philippe Catez,
Pierre Chymkowitch, and Jean-Marc Egly*
Institut de Ge
´ne
´tique et de Biologie Mole
´culaire
et Cellulaire kinase (CAK) complex to the core of TFIIH (Rossignol
et al., 1997 and references therein). CAK is a trimericBP 163; 67404 Illkirch Cedex
France complex containing the cdk7 kinase, cyclin H, and
MAT1, which phosphorylates the C-terminal domain of
the largest subunit of RNA pol II, thereby facilitating the
promoter clearance.Summary
Mutations in the XPD subunit cause Xeroderma Pig-
mentosum (XP) and Trichothiodystrophy (TTD), whichMutations in the XPD subunit of the transcription/
repair factor TFIIH cause the Xeroderma pigmentosum are primarily defined as DNA repair disorders (Lehmann,
2001). XP is characterized by pigmentations abnormali-disorder. We show that in some XP-D deficient cells,
transactivation by the vitamin D receptor (VDR) is se- ties of the skin with an elevated risk of skin cancers
over the general population. The principal hallmark oflectively inhibited for a subset of responsive genes,
such as CYP24, and that the XPD/R683W mutation TTD is sulfur-deficient, brittle hair accompanied by men-
tal retardation, bone abnormalities, ichthyotic skin, andprevents VDR recruitment on its promoter. Contrary
to other nuclear receptors, VDR, which lacks a func- dwarfism. While all the XPD mutations are detrimental
for the XPD helicase activity explaining the NER defect,tional A/B domain, is not phosphorylated and conse-
quently not regulated by the cdk7 kinase of TFIIH. they also result in distinct transcriptional defects de-
pending on the site of mutations (Taylor et al., 1997;In fact, we demonstrate that the VDR transactivation
defect resides in Ets1, another activator that cannot Dubaele et al., 2003).
The nuclear receptors are hormone-dependent tran-be phosphorylated by TFIIH in XP-D cells. Indeed, the
phosphorylated Ets1 seems to promote the binding of scription factors that play essential roles in develop-
ment, differentiation, and metabolism by controlling theVDR to its responsive element and trigger the subse-
quent recruitment of coactivators and RNA pol II. We expression of specific networks of genes (Dilworth and
Chambon, 2001; Glass and Rosenfeld, 2000). Memberspropose a model in which TFIIH regulates the activity
of nuclear receptors by phosphorylating either their of this family, including receptors for retinoids (RAR and
RXR), vitamin D (VDR), steroid (ER, AR, GR) and thyroidA/B domain or an additional regulatory DNA binding
partner. hormones (TR) are characterized by a conserved struc-
tural and functional organization (Kumar and Thompson,
1999), a heterogeneous N-terminal A/B domain that canIntroduction
contain a ligand-independent transactivation domain
(AF-1 domain), a highly conserved DNA binding domain,Gene expression is regulated by activators that bind to
specific sequences and whose efficiency depends on a homo- and heterodimerization domain, and a large
C-terminal ligand binding domain that harbors a ligand-physiological signals. An activator’s potential could be
provided either by one specific activator or by following inducible transactivation function (AF-2 domain).
The molecular communication between TFIIH and nu-the combined and sophisticated action of several DNA
binding proteins responding to diverse signaling path- clear receptors, including RAR␣, AR, and ER␣, is dis-
rupted in cells from a particular group of XP and TTDways, leading to a unique and specific regulation of a
given gene (Brivanlou and Darnell, 2002). In order to patients. Indeed, mutations in the C-terminal third of
XPD result in a drop in the ability of certain nuclearhave the right protein in the right amount at the right
time as development proceeds, the cell has set up a receptor AF-1 domain to be phosphorylated by TFIIH
and to mediate transactivation (Keriel et al., 2002), ex-coordinated action between sequence specific DNA
binding proteins and the basal transcription machinery, plaining, at least partially, the developmental defects in
the XPD-deficient patients.including RNA pol II. Such coordination is achieved
through modular complexes of coactivators that bridge Indeed, certain symptoms found in patients carrying
a mutation in XPD could be due to a defect in the vitaminactivators and RNA pol II (Naar et al., 2001; Rachez and
Freedman, 2001). The activators and the basal transcrip- D (vitD) pathway, a hormone known for the maintenance
of mineral homeostasis and skeleton architecture (Jonestion machinery can also be directly linked. For instance,
the general transcription factor TFIIH has been found to et al., 1998). We therefore investigated how VDR trans-
activates in XPD-deficient cells, and we found that thetarget and phosphorylate different activators, including
nuclear receptors (Chen et al., 2000; Lu et al., 1997; vitD-mediated response can be selectively impaired.
Rochette-Egly et al., 1997).
TFIIH is a multienzymatic complex essential for tran- Results
scription initiation and nucleotide excision repair (NER).
The XPB subunit is involved in the formation of a melted The Vitamin D Response Is Selectively Impaired
region around the transcription start site (Holstege et by XPD/R683W Mutation
To investigate whether TFIIH, carrying a mutated XPD
subunit, affects vitamin D (vitD) response, we used
*Correspondence: egly@igbmc.u-strasbg.fr
Molecular Cell
188
Figure 1. VitD Response Is Altered by XPD/R683W Mutation
VitD-mediated response of CYP24 (A) and osteopontin (B), two VDR-target genes, in XPD-deficient cells. RT-PCR analysis was performed 8
hr and 16 hr after vitD addition (10
⫺
7
M). The results are the mean of three independent experiments and represent the relative expression
level of the two genes versus 18S. (C) UV survival of XPD-complemented cells. XPJCLO (XPD/R683W) (closed diamonds), XPJCLO⫹LXPDN
(open squares), and GM3348D (NHF) (open circles) cells were (10
4
cells/well) cultured overnight and UV irradiated (254nm) at the indicated
dose (J.m
⫺
2
). Cells were stained two-weeks later using crystal violet dye. Results of three independent experiments are expressed as percentage
of cell survival relative to nonirradiated cells. (D) VitD-mediated response of CYP24 in XPD complemented cells. The kinetic of vitD-mediated
CYP24 expression was measured in cells depicted in (C).
Ag10032, TTD8PV, and TTD12PV fibroblasts derived blasts: the XPJCLO cells, derived from a UV-sensitive XP
patient and bearing the XPD/R683W mutation, and thefrom XP and TTD patients bearing the most frequent
XPD mutations at position R683W, R112H and R722W, XPJCLO⫹LXPDN cells, complemented for XPD activity
after transfection with a retrovirus carrying the wild-typerespectively (http://www.xpmutations.org). We mea-
sured by real-time quantitative RT-PCR the vitD-medi- XPD gene (Carreau et al., 1995). In XPJCLO⫹LXPDN,
expression of the wild-type XPD gene restores UV resis-ated activation of CYP24 and osteopontin, two VDR-
target genes (Akeno et al., 1997; Li et al., 1998). After tance (Figure 1C). The CYP24 response to vitD was next
measured in parental (XPJCLO) and rescue cellsvitD addition, osteopontin is similarly activated in the
three cell lines when compared to normal human fibro- (XPJCLO⫹LXPDN). At first, we did not observe any dif-
ference in basal expression of CYP24 between bothblasts (GM03348D; NHF) (Figure 1B). In contrast, the
CYP24 response to vitD is altered in cells harboring isogenic cells. However, the vitD-mediated CYP24 in-
duction defect observed in XPJCLO is circumventedthe XPD/R683W mutation (Figure 1A, left). In Ag10032,
CYP24 expression is much weaker than in NHF at 8 or upon reintroduction of wild-type XPD gene (Figure 1D).
Indeed, in XPJCLO⫹LXPDN, CYP24 expression reaches16 hr posttreatment. No such defect is observed with
the other two XPD mutated TTD8PV and TTD12PV, sug- the same level than in NHF. Altogether, our results dem-
onstrate that a XPD mutation can modulate the vitD-gesting that impairment of CYP24 induction is restricted
to cells bearing the R683W mutation. In absence of vitD, mediated response in a gene-specific manner. More-
over, this selective effect seems to be restricted to cellswe only found minor differences in basal expression of
CYP24 and osteopontin (Figures 1A and 1B). As ob- bearing the XPD/R683W mutation.
served with Ag10032, the induction of the CYP24 gene
is also impaired in HD2, in comparison to HeLa (Figure CYP24 Expression Defect Is Associated
with Impairment in VDR Recruitment1A, right). The HD2 cell line is a result of the fusion
between human fibroblasts (harboring a XPD/R683W In order to understand how the XPD/R683W mutation
discriminates between CYP24 and osteopontin, we setmutation) and HeLa cells (Johnson et al., 1985).
To address whether the impairment of CYP24 induction up a chromatin immunoprecipitation (ChIP) assay. An
antibody directed against VDR was used to immunopre-is due to the XPD mutation, we used two primary fibro-
TFIIH and Vitamin D Receptor
189
ined on the CYP24 promoter after vitD addition. We
noticed a modest increase of TFIIH recruitment in HeLa
(with a peak 1 hr posttreatment), while a slight decrease
is detected at the same time in HD2 (Figure 2B). We
especially found that RNA pol II recruitment is at once
delayed and attenuated in HD2 as compared to HeLa
(Figure 2B). In fact, the RNA pol II recruitment is detected
0.5 and 2 hr after vitD addition in HeLa and HD2, respec-
tively. Beyond 2 hr posttreatment, the amount of RNA
pol II bound to CYP24 promoter in HD2 represents only
30%–40% of those observed in HeLa. Note that the
increase of vitD-mediated RNA pol II recruitment repre-
sents both initiating and elongating RNA pol II on CYP24
(Figure 2B). Finally, we investigated the recruitment of
TRAP220 and TIF2 coactivators, both interacting di-
rectly with VDR (Rachez et al., 1999; Takeyama et al.,
1999). In HeLa, TIF2 occupancy parallels that of RNA
pol II, with two peaks 2 and 16 hr postvitD addition,
while ligand-induced TRAP220 recruitment is detected
only 2 hr after treatment. In contrast, the recruitment of
both coactivators remains low and almost unchanged
after vitD treatment in HD2 (Figure 2B). It is worthwhile
to notice that in mock-treated HeLa and HD2, VDR and
the other transcriptional components are constitutively
and similarly associated with both CYP24 and osteopon-
tin promoters (data not shown). Collectively, our data
demonstrate that the XPD/R683W mutation prevents
both the binding of liganded VDR on CYP24 promoter
and the proper assembly of the transcriptional machin-
ery on this promoter.
XPD Mutation Does Not Compromise VDR DNA
Binding Activity
We next examined whether XPD/R683W mutation com-
promises the intrinsic ability of VDR to bind to CYP24.
Figure 2. XPD/R683W Mutation Affects VDR Recruitment on
Firstly, immunoblotting experiments show that VDR ac-
CYP24 Promoter
cumulates in response to vitD in HeLa and HD2 (Figure
(A) ChIP analysis on the CYP24 (left) and osteopontin (right) promot-
3A), as previously reported for other cell lines (Li et al.,
ers in HeLa and HD2. Soluble chromatin was prepared from cells
1999). The pattern of VDR accumulation is similar in both
treated with vitD (10
⫺
7
M) for the indicated time and immunoprecipi-
cells. Hela and HD2 also contain equivalent amounts of
tated with an antibody against VDR. The genomic DNAs were ana-
lyzed by quantitative PCR.
TFIIH, RNA pol II, TIF2, TRAP220, as well as RXR, the
(B) Similar ChIP assays were performed to investigateTFIIH (XPB),
partner of VDR (Figure 3A), suggesting that the recruit-
RNA pol II, TRAP220, and TIF2 recruitment on CYP24 .
ment impairments on CYP24 observed in HD2 are not
a bias imputable to a default of expression of these
components. Secondly, we performed an electropho-cipitate VDR-bound genomic DNA fragments, which
were further analyzed by quantitative PCR. Our assay retic mobility gel shift assay (EMSA) to study the ability
of VDR from XPD-deficient HD2 cell extracts to bindis specific since no VDR is detected 6kb upstream of
the CYP24 transcription start site nor on the GAPDH to a double-stranded oligonucleotide encompassing
VDRE1, one CYP24-vitD-responsive element (Figure 3B)promoter (data not shown).
In HeLa and XPD/R683W-mutated HD2, VDR is simi- (Zierold et al., 1995). The HeLa nuclear extract (NE) was
first incubated with the radiolabeled VDRE1 before beinglarly recruited on the osteopontin promoter over a period
of 16 hr post-vitD treatment (Figure 2A, right). In both resolved by electrophoresis (Figure 3B, lane 3). We de-
tected a nucleoprotein complex (NC1) that contains acells, recruitment on the osteopontin promoter reaches
a peak 2 hr after vitD addition and appears to be cyclic, VDR/RXR heterodimer, since it is (1) inhibited by an
antibody directed against the VDR DNA binding domainas previously reported for the recruitment of the estro-
gen receptor on the promoters of cathepsin D or pS2 (lane 1), and (2) supershifted by an anti-RXR monoclonal
antibody (lane 2). The specificity of this VDR/RXR com-gene (Metivier et al., 2003; Shang et al., 2000). In con-
trast, VDR occupancy in the CYP24 promoter is quite plex is further supported by its inhibition by a 50-fold
molar excess of cold VDRE1 (lane 4) and not by a non-different: in HeLa, VDR recruitment increases with time
to reach a 10-fold stimulation 8 hr after vitD addition, specific oligonucleotide (lane 5). HeLa and HD2 NE, from
mock-treated cells or cells treated with vitD for 4, 8, orwhile it remains low and unchanged in HD2, irrespective
of the treatment duration (Figure 2A, left). The recruit- 16 hr were next incubated with VDRE1. By using these
cell extracts, we noticed that the amount of VDR/RXRment of other transcription components was next exam-
Molecular Cell
190
complexes increases similarly with time, reflecting the
vitD-mediated accumulation of VDR previously ob-
served (Figure 3B, lanes 7–14 and Figure 3A).
It has been reported that in certain cases, EMSA does
not reflect the in vivo DNA binding capacity of a tran-
scription factor. Therefore, we investigate the effects of
the XPD mutation on VDR activity in a more physiological
context. We found that pGL3-VDRE1 containing the lu-
ciferase-reporter gene placed under the control of
VDRE1 alone is similarly activated in HeLa and HD2
cells (Figure 4A). In contrast, a retinoic acid receptor-
responsive element (RARE) cloned in the same promoter
environment gives a reduced activity in HD2 cells (Figure
4A). These data demonstrate that in opposition to RAR,
VDR activity is not compromised by the XPD mutation.
XPD Mutation Prevents Cooperation between VDR
and Ets Proteins
To further understand how the XPD/R683W mutation
selectively affects the vitD-response, we dissected the
CYP24 promoter (Figure 4A). In response to vitD, the
pCYP24wt construct gives a 90- and 38-fold level of
induction of CAT gene expression in HeLa and HD2,
respectively (Figure 4A, top). Similarly, pCYP24(-204)
activity is also impaired in HD2, suggesting that the
XPD mutation defect concerns the truncated promoter
starting at position ⫺204. Besides VDRE1, this part of
the promoter contains an Ets binding site (EBS) (Figure
4A). Interestingly, Ets1, a member of the Ets transcription
factor family, is known to interact with nuclear receptors,
including VDR (Tolon et al., 2000). This observation and
the fact that a reporter gene placed under the control
of VDRE1 alone is similarly activated in both normal and
XPD-deficient cells (Figure 4A) prompted us to investi-
gate a connection between VDRE and EBS in the CYP24
regulation. We hypothesized that the XPD mutation
would prevent a putative cooperation between VDR and
the EBS binding proteins. Accordingly, we designed the
pCYP24mEBS reporter construct in which the EBS motif
is mutated to prevent the binding of Ets factors (Dwivedi
et al., 2000). When transfected in HeLa, pCYP24mEBS
exhibits a reduced CAT activity, as compared to
pCYP24wt (Figure 4A), emphasizing the implication of
the EBS motif in the vitD-mediated regulation of CYP24.
We obtained similar results when using the pCYP24-
(204)mEBS construct instead of pCYP24mEBS (Figure
4A). In HD2, both pCYP24wt and pCYP24mEBS reporter
constructs display a comparable level of CAT expres-
sion. All these results would suggest that XPD mutation
Figure 3. DNA Binding Activity of VDR Is Not Affected by XPD/
compromises EBS participation (and consequently its
R683W Mutation
cognate protein) in the regulation of CYP24 response.
(A) Western blot analyses of VDR, RXR, TFIIH (p62), RNA pol II, TIF2,
This led us to investigate whether the binding of Ets
and TRAP220 from nuclear extracts (NE) of HeLa and HD2 cells
proteins to the EBS motif is involved in VDR recruitment
treated with vitD (10
⫺
7
M). Pol-IIo and Pol-IIA are the hyper- and
on CYP24 promoter. Both pCYP24wt and pCYP24mEBS
hypophosphorylated form of RNA polII, respectively.
(B) VDR from HD2 cells is able to bind to a CYP24-responsive ele-
were transfected in HeLa and HD2, and VDR recruitment
ment. Left: NE prepared from HeLa cells treated for 16 hr with vitD
was monitored by ChIP assays. In contrast to what is
(10
⫺
7
M) was incubated with a
32
P-labeled VDR-CYP24 responsive
observed in HD2, vitD treatment allows a 10-fold recruit-
element (lane 3) and with an antibody against VDR or RXR (lanes 1
ment of VDR on pCYP24wt in HeLa (Figure 4B), in accor-
and 2, respectively), or with a 50⫻molar excess of either specific
dance with results presented in Figure 2A. This default
(S, lane 4) or nonspecific (NS, lane 5) cold oligonucleotides to be
observed in HD2 cells is circumvented upon reintroduc-
analyzed by EMSA. Right: NE from HeLa (lanes 7–10) and HD2 cells
(lanes 11–14) treated with vitD (10
⫺
7
M) were analyzed by EMSA, as
tion of XPDwt. Furthermore, inactivation of the EBS ele-
described in (A). NC1 is a heterodimeric complex containing VDR
ment compromises the vitD induced VDR recruitment
and RXR, and the asterisk indicates nonspecific complexes.
on pCYP24mEBS promoter in HeLa (Figure 4B, left).
TFIIH and Vitamin D Receptor
191
Figure 4. XPD/R683W Mutation Affects Ac-
tivity of an Ets Binding Site
(A) Transfections in HeLa and HD2 of reporter
vectors containing vitD-responsive element 1
and 2 (VDRE1 and 2), and either the wild-type
Ets binding site (EBS) or the mutated one
(mEBS) as indicated at the left of the panel.
Activity of these constructs as well as of
those containing the luciferase gene under
the control of an SV40 promoter (SVE) and
either VDRE1 (pGL3-VDRE1) or a retinoic acid
receptor responsive element RARE (pGL3-
RARE) was measured in cells treated with
either 10
⫺
7
M vitD or 10
⫺
7
M all-trans retinoic
acid, respectively. The results are expressed
as fold activation relative to nontreated cells.
(B) EBS participates in VDR recruitment. The
VDR recruitment on the indicated plasmid
transfected in HeLa and HD2 cells was moni-
tored by ChIP, 8 hr after vitD addition.
(C) The Ets1 activity is impaired in HD2 cells.
The pMdm2 construct was transfected in
HeLa and HD2 cells with an expression vector
encoding Ets1.
Finally, we asked whether the XPD mutation directly nous Ets1 in HeLa, but not in HD2 (lanes 1 and 2). Like-
wise, the labeling of the transfected Ets1 is lower in HD2affects the activity of Ets1, an EBS binding protein. Both
HeLa and HD2 were therefore transfected with the than in HeLa (lanes 3 and 4).
pMdm2 construct containing the luciferase reporter We also test whether Ets1 is a substrate for TFIIH and
gene placed under the control of the Mdm2 promoter, whether the XPD/R683W mutation affects the phosphor-
which has been shown to be specifically regulated by ylation of Ets1 by TFIIH. Infected Sf9 cell extracts con-
Ets1 (Ries et al., 2000). Overexpression of Ets1 stimu- taining Ets1, either alone or in combination with TFIIH/
lates up to 5-fold the activity of pMdm2 in HeLa (Figure XPDwt, /XPD/R683W, or /XPD/R722W, were then re-
4C). In contrast, activation of this promoter is twice low- solved by a two-dimensional gel electrophoresis (2D-
ered in HD2, demonstrating that the Ets1 transactivation PAGE) (Figure 5B). Western blot analysis reveals that
function is impaired in XPD mutated cells. We thus con- Ets1 is resolved in three polypeptide spots, named a to
cluded that the binding of Ets transcription factors to c (panel I). When Ets1 is coexpressed with TFIIH, we
EBS is required for the vitD-mediated VDR recruitment detect an additional more acidic spot d (panel II). Since
on the CYP24 promoter and that the XPD/R683W muta- phosphorylation induces a shift toward acidic pH, the
tion likely alters the cooperation between Ets1 and VDR. extract from cells infected by baculoviruses expressing
both TFIIH and Ets1 was pretreated with calf-intestine
phosphatase before 2D-PAGE (panel V). Spot d disap-
TFIIH Phosphorylates Ets1 In Vivo
pears under these conditions, showing that Ets1 is phos-
Knowing that TFIIH can regulate gene expression by
phorylated in vivo by TFIIH. We should note that phos-
targeting directly the transcription factors (Rochette-
phatase treatment also results in the disappearance of
Egly et al., 1997), we therefore compared the phosphory-
spot c, indicating that Ets1 is constitutively phosphory-
lation status of Ets1 proteins in HeLa versus HD2. Since
lated in sf9 cells. Interestingly, spot d is absent when
the endogenous Ets1 levels are low in these cell-types
Ets1 is coexpressed with TFIIH/XPD/R683W (compare
(Bradford et al., 1997; Figure 5A, lower panels), investi-
panels III and II) and reappears in presence of TFIIH/
gations were also performed with Ets1 transfected cells.
XPD/R722W (compare panels IV and II). This shows that
Cells were then incubated with [
32
P]orthophosphate, and
XPD/R683W, in contrast to the XPD/R722W mutation,
immunoprecipitated Ets1 was resolved by SDS-PAGE
(Figure 5A). We first detected a weak labeling of endoge- affects the TFIIH-dependent phosphorylation of Ets1.
Molecular Cell
192
Figure 5. Physical and Functional Interaction between TFIIH and Ets1
(A) In vivo phosphorylation of endogenous Ets1 (lanes 1 and 2), overexpressed Ets1 wt (lanes 3 and 4) or Ets1 T38A (lanes 5 and 6) in HeLa
and HD2, after immunoprecipitation. Western blot (WB), and autoradiography (Auto.). Note that the exposure of the autoradiograph of Ets1
T38A (lanes 3 and 4) was six times longer than the one of Ets1.
(B) The Sf9 cells were coinfected with baculoviruses encoding Ets1 alone (panel I) or in combination with TFIIHwt (II and V) or with TFIIH
containing either a XPD/R683W (III) or a XPD/R722W (IV) mutated subunit. Extracts from infected cells were pretreated (V) with calf intestinal
phosphatase. Proteins were fractionated by 2D-PAGE, and Ets1 was detected by immunobloting. The position of the Ets1 isoforms is indicated
by arrows (a–d).
(C) 1 g of purified Ets1 (lanes 3–5), Ets1/T38A (lanes 6–8), RAR␣(lanes 9–11), or VDR (lanes 12–14) was incubated HeLa TFIIH or recombinant
CAK. Lanes 1 and 2 show CAK and TFIIH, respectively, incubated alone in the reaction buffer. Blue staining of the gels (stain.) is shown to
compare level of each protein.
(D) The Sf9 cells were coinfected with baculoviruses encoding all the TFIIH subunits and Ets1. Immunoprecipitation was done by using an
antibody directed against TFIIH/p44 subunit (p44) or a control antibody (C). The bound proteins were analyzed by Western blot by using
antibodies against three subunits of TFIIH (cdk7, p62 and XPD) or Ets1. The position of the antibody heavy chain is shown (H). The input
lanes (1–4) represent 5% of the total volume of extract used in each immunoprecipitation.
(E) Interaction between TFIIH and VDR was investigated by using a baculovirus encoding VDR as described in (D).
(F) Interaction of Ets1 with the core TFIIH. A GST-Ets1 recombinant protein was incubated with extracts from cells infected with baculovirus
encoding the core subunits. Washing were performed with buffer containing 300 mM KCl. The input lanes represent 5% of the total volume
of the extract used for the pulldowns.
TFIIH and Vitamin D Receptor
193
To identify the Ets1 target site, purified recombinant
Ets1 was tested in an in vitro kinase assay in the pres-
ence of either HeLa TFIIH or recombinant CAK (Figure
5C). Ets1 is indeed efficiently phosphorylated by TFIIH,
but not by the CAK subcomplex (lanes 5 and 4, respec-
tively). Recombinant TFIIH that contains a mutated cdk7
in its ATP binding site is unable to phosphorylate Ets1
(data not shown). It is worthwhile to mention, that TFIIH
and CAK were unable to phosphorylate VDR, although
both use RAR␣as a substrate (Figure 5C). Importantly,
Ets1/T38A, in which the T38 residue known to be phos-
phorylated by proline directed kinases (Yang et al., 1996)
is changed into alanine, is not phosphorylated by TFIIH
(Figure 5C, lane 8). In agreement with the above data
and with the fact that the XPD/R683W mutation affects
the in vivo phosphorylation of Ets1 by TFIIH (see below),
we found that the level of transfected Ets1/T38A labeling
is similar in HeLa and HD2 transfected cells (Figure 5A,
lanes 5 and 6), strongly suggesting that T38 is the phos-
phoacceptor residue targeted by TFIIH.
We finally investigated whether the phosphorylation
of Ets1 by TFIIH requires a stable interaction between
both components. Thus, Sf9 insect cells were infected
by baculoviruses encoding Ets1 either alone or in combi-
nation with all the subunits of TFIIH. Under high-salt
Figure 6. Role of Ets1 T38 Phosphorylation in CYP24 Promoter Reg-
concentration (0.3 M KCl), Ab-p44, an antibody directed
ulation
against the p44 subunit of TFIIH, immunoprecipitates
(A) The Ets1 phosphorylation is important for CYP24 promoter activ-
not only TFIIH (as indicated by the presence of the three
ity. Hela and HD2 cells were cotransfected with the indicated pro-
subunits cdk7, p62, and XPD), but also Ets1 (Figure 5D,
moter construct in combination with an empty vector or a vector
upper and lower panels, lane 12). We observed that Ab-
encoding Ets1 or Ets1/T38A. The cells were then treated with vitD
p44 does not retain Ets1 in the absence of TFIIH (lane
(10
⫺
7
M). Promoter activity was checked as described in Figure 4.
8), and Ab-C, the control antibody, does not immunopre-
(B) The Ets1 phosphorylation is crucial for VDR recruitment. The cells
were transfected with the pCYP24wt construct and then treated for
cipitate either component (upper and lower panels, lane
8 hr with vitD, and VDR recruitment is checked by ChIP assays.
11). We examined also the way in which TFIIH interacts
with Ets1. Consistent with the fact that the CAK does
not phosphorylate Ets1, the interaction CAK/Ets1 is
pression of Ets1/T38A does not result in a further activa-
weak (data not shown), suggesting the existence of a
tion of pCYP24wt (Figure 6A), demonstrating that TFIIH-
docking site in Ets1 for the core of TFIIH. Indeed, we
dependent phosphorylation of Ets1 is a prerequisite for
did find a direct interaction, in presence of 300mM KCl,
enhancing VDR-dependent transactivation. In accor-
between Ets1 and XPB, a subunit of the core TFIIH
dance with this, overexpression of Ets1/T38A has no
(Figure 5F). Note that we found no interaction between
significant effect on vitD-mediated VDR recruitment
VDR and TFIIH, using the same experimental approach
(Figure 6B).
(Figure 5E, lane 12).Altogether, our results suggest that
cdk7, as part of the TFIIH transcription factor, phosphor-
Discussion
ylates the T38 residue of Ets1 and that the XPD/R683W
mutation affects the efficiency of TFIIH to phosphorylate
In order to understand the links between nuclear recep-
Ets1 in vivo.
tors and TFIIH, we focused on the regulation of VDR
transactivation function. We discovered that in some
TFIIH-Dependent Phosphorylation Regulates
cases, the vitD-response can be mediated by the inter-
Cooperation between VDR and Ets1
mediate of another activator phosphorylated by TFIIH.
We finally investigated whether the phosphorylation of
Ets1 by TFIIH modulates its association with VDR on
the CYP24 promoter. HeLa and HD2 were then cotrans- TFIIH Coordinates the Action of Two Activators
We investigated the effects of an XPD mutation ontofected with pCYP24wt and an expression vector encod-
ing either Ets1 or Ets1/T38A, in which the T38 residue the vitD-mediated response of several genes, such as
osteopontin,CYP24, and p21/WAF1. The expression ofwas mutated into alanine. Firstly, Ets1 overexpression
causes a further increase in vitD-mediated transactiva- osteopontin and p21/WAF1 (data not shown) is normal
in all the XPD-deficient cells so far tested, while those oftion of pCYP24wt in HeLa, but not in HD2 cells (Figure
6A). Inactivation of the EBS motif abolishes this positive CYP24 is affected in some mutated cells. This selective
defect in the vitD response seems indeed to be re-effect of Ets1 on pCYP24wt activity (Figure 6A). Parallel
ChIP experiments show that the vitD-induced VDR re- stricted to cells harboring the XPD/R683W mutation.
In these cells and in the absence of vitD, the level ofcruitment is enhanced upon overexpression of wild-type
Ets1 only in HeLa cells (Figure 6B). Secondly, overex- expression of CYP24 and the recruitment of VDR to the
Molecular Cell
194
promoter are normal. However, after addition of vitD, the Phosphorylated Ets1 Promotes VDR Binding
The recruitment of liganded VDR on the CYP24 promoteroccupancy of VDR in the CYP24 promoter is affected. As
a result, the recruitment of TFIIH, TRAP220, and TIF2 requires not only Ets1, but also its phosphorylation. We
investigated the way in which the phosphorylation ofcoactivators, as well as of RNA pol II is impaired, ex-
plaining the lower rate of CYP24 mRNA synthesis. The Ets1 affects the binding of VDR to CYP24. There is, in
fact, a direct interaction of VDR with Ets1 that inducesdefault in recruitment of VDR is due to the XPD/R683W
mutation itself, since XPDwt transfection restores both a conformational change in both VDR and Ets1, which
can be detected by an increased resistance to trypticthe VDR binding and the transactivation process.
Our results rule out that the XPD mutation per se digestion (Tolon et al., 2000, and our unpublished data).
However, the phosphorylation of Ets1 by TFIIH does notimpairs the intrinsic ability of VDR to bind to a CYP24-
responsive element. We next identified EBS, a proximal modify the hydrolysis pattern of either Ets1 or VDR when
interacting with each other. We concluded that the inter-DNA binding site for Ets proteins, as a key element in
the regulation of CYP24 (Dwivedi et al., 2000). Indeed, action of a liganded VDR with a phosphorylated Ets1 is
not sufficient to regulate the CYP24 expression unlessmutations of EBS prevent the vitD-induced VDR recruit-
ment in XPDwt cells. In addition, overexpression of Ets1, additional partner(s) are recruited. We exclude WINAC,
a chromatin-remodeling complex that mediates the re-a member of the Ets family, causes a further increase
of liganded VDR activity. Along this line, Ets proteins cruitment of unliganded VDR to promoter, as a potential
candidate (Kitagawa et al., 2003), since we show hereare known to regulate the expression of genes involved
in several cellular functions, including differentiation, se- that in absence of vitD, the binding of VDR to the CYP24
promoter is normal in XPD mutated cells. In contrast,nescence, or apoptosis, through interactions with other
transcription factors, such as AP-1, NF-B, and PAX the histone acetyltransferases CBP/p300 could be good
candidates since they (1) interact with both Ets1 andfamily members, on composite DNA binding sites (Li et
al., 2000; Oikawa and Yamada, 2003). VDR (Jayaraman et al., 1999; Sierra et al., 2003), and (2)
were shown to associate with Ets1 to form a stable DNA-
p53/Ets1 transcriptional complex (Xu et al., 2002). Thus,
to gain insight into our model, additional experimentsEts1 Is Phosphorylated by TFIIH
TFIIH is known to phosphorylate several activators are required to identify partner(s) that can help to the
formation of the vitD-dependent transcriptional complex.(Chen et al., 2000; Lu et al., 1997; Rochette-Egly et al.,
1997). The serine or threonine targeted within these acti- Previous studies have shown that the activation of
the Ras pathway also leads to the phosphorylation ofvators is adjacent to a proline and located in a proline-
rich region, consistent with the fact that cdk7 is a serine/ T38 and increases Ets1 activity (Yang et al., 1996). The
Ras effect is mediated by the ERK1/2 MAPK. It is notproline directed kinase. Interestingly, the short VDR A/B
domain, to which no transactivation function has been surprising to find that two pathways lead to the phos-
phorylation of the same residue. For instance, the S118ascribed, lacks such putative cdk7 phosphorylation site,
making its Ets1 partner a good candidate for being tar- residue of ER␣is phosphorylated not only by TFIIH, but
also by ERK1/2 (Chen et al., 2002). We propose thatgeted by TFIIH. Indeed, we found that Ets1 is phosphory-
lated in vitro by TFIIH at position T38. When we analyzed the way in which Ets1 interacts with MAPK and TFIIH
depends upon both the initiating signal and its timing.the in vivo modification status of Ets1, we showed that
this protein is underphosphorylated in XPD/R683W- Indeed, Ets1 would be phosphorylated in response to
a mitogenic signal that activates the MAPK pathway,deficient cells. By using infected insect cells, we also
demonstrated that (1) Ets1 is indeed targeted by TFIIH while its phosphorylation by TFIIH would occur after the
formation of the DNA-VDR transcriptional complex thatin vivo, and (2) the defect in Ets1 phosphorylation is due
to the XPD/R683W mutation itself. Interestingly, in vivo puts TFIIH (within the preinitiation complex) and Ets1 in
very close proximity.labeling with [
32
P]orthophosphate, also indicate that the
level of phosphorylation of Ets1/T38A (in which the thre-
onine 38 was changed into alanine) is similar in HD2 Selective Impairment of VitD Response
versus HeLa transfected cells, suggesting that the T38 The present work also challenges the hypothesis that
residue is also phosphorylated in vivo by TFIIH. XP is a repair syndrome, while most of the TTD clinical
In contrast to what is observed with other nuclear features were considered to be a consequence of tran-
receptors such as RAR␣(Keriel et al., 2002), only the scription deficiency. Thus, the XP-D patients, bearing a
XPD/R683W mutation (and not the XPD/R722W) pre- XPD mutation at position R683W, should show vitD-
vents the phosphorylation of Ets1 by TFIIH. In fact, RAR␣related abnormalities and would possess XPD-specific
and Ets1 differently contact TFIIH, since RAR␣interacts phenotypes not found within the other XP groups. At
with both the core of TFIIH and the CAK subcomplex the present stage of our investigations, it is hazardous
(Bastien et al., 2000), whereas Ets1 contacts the XPB to evaluate this. The clinical documents at our disposal
subunit of the core of TFIIH. In addition, TFIIH phosphor- do not report XP-D phenotypes associated with a vitD
ylates Ets1 whereas RAR␣is the substrate of both TFIIH defect (Berneburg and Lehmann, 2001), perhaps be-
and CAK (Rochette-Egly et al., 1997). It thus seems that cause no physicians were aware of such a specific de-
the nature of the activators (RAR versus Ets1) deter- fect that concerns a part of the XP patients. However,
mines the way in which they interact with TFIIH, ex- it results from our study that XPD/R683W patients could
plaining why mutations in the XPD subunit might selec- be predisposed to a vitD intoxication and its adverse
tively impair their phosphorylation and consequently pathological complications (Chiricone et al., 2003). In-
deed, vitD, in the form of vitD
3
, is made in the skin bytheir transactivation capability.
TFIIH and Vitamin D Receptor
195
al., 1988) or VDR (Lemon et al., 1997) were incubated for 4 hr at
exposure to UV light, and then undergoes an activation
4⬚C with protein A-Sepharose crosslinked with Ab-p44 monoclonal
process to produce the biologically active compound
antibody in buffer A (20 mM Tris-HCl [pH 7.9], 150 mM KCl, 5 mM
1,25-(OH)
2
D
3
(Jones et al., 1998). The initiation of the
MgCl
2
, 20% glycerol, 0.2 mM EDTA, 1 mM DTT, and 0.05% NP-40).
degradation of 1,25-(OH)
2
D
3
is next catalyzed by the
After extensive washing, bound proteins were resolved by SDS-
24R-hydroxylase, the product of CYP24 (Henry, 2001).
PAGE and revealed by immunoblotting.
In XPD/R683W cells, the partial inhibition of CYP24 ex-
Electrophoresis Mobility Shift Assay
pression would prevent such degradation and would
The complementary oligonucleotides CTAGCGGCGCCCTCACTCAC
result in the accumulation of vitD in tissues of patients,
CTCGCTGAC TCCAC and TCGAGTGGAGTCAGCGAGGTGAGTGAG
as also observed in CYP24 null mutant mouse (St-
GGCGCCG (the underlined nucleotides correspond to the VDRE1
Arnaud et al., 1996). The resulting phenotypes would be
sequence of CYP24 promoter) were annealed and labeled by fill in.
hardly diagnosed since these XP-D patients are sunlight
10 g of nuclear extract were incubated with 0.4 ng of
32
P-labeled
deprived according to the physician advice.
probe and complexes were resolved on a 5% native acrylamide gel
as described previously (Drane et al., 2001).
Our data shed new light on the molecular communica-
tion between TFIIH and nuclear receptors. At first, the
Transfections and Reporter Assays
transactivation mediated by nuclear receptors appears
HeLa and HD2 cells were transfected using the Fugene 6 (Boehringer
to depend on both their structure and the integrity of
Mannheim). Cells (2.5 ⫻10
5
) were plated in 6-well plates and trans-
TFIIH. Indeed, for receptors such as RAR␣/␥or ER␣,
fected with either CYP24 promoter (2 g) or pGL3 constructs (1 g)
the TFIIH-dependent phosphorylation of their A/B (AF-1)
and pCH110 (1 g) (plasmid from Invitrogen coding for -galactosi-
domain modulates their activity, and mutations in the
dase).Each transfection was repeated four times. Seven hours later,
the cells were refed with medium without red phenol and containing
C-terminal part of XPD (including R683W and R722W)
10% of charcoal-treated FCS and incubated 16 hr at 37⬚C. Cells
compromise this process (Keriel et al., 2002). In the
were then treated with the carrier or 10
⫺
7
M vitamin D (vitD) and
absence of a functional AF-1 domain, as observed for
harvested 48 hr posttransfection. CAT assays were performed by
VDR, transactivation can be mediated by a second acti-
using the CAT ELISA kit (Roche Diagnostic), whereas -galactosi-
vator that might fill up the role of an AF-1 domain. The
dase and luciferase assays were made as described (Keriel et al.,
latter, when phosphorylated, allows for an optimal li-
2002).
gand-dependent transactivation by nuclear receptors
In Vivo and In Vitro Phosphorylation of Ets1
and defines another level of regulation directed by TFIIH
Cells (1.25 ⫻10
6
) were plated in 10 cm dishes and were transfected
that would act as a genuine coregulator.
with 10 g of either Ets1 or Ets1-T38A expression vector as de-
scribed above. Next, [
32
P]orthophosphate labeling and immunopre-
Experimental Procedures cipitation were done as described (Keriel et al., 2002).
His fusion proteins were produced in E. coli strain BL21 and puri-
Cell Lines fied on NI-NTA agarose (Qiagen) as described by the manufacturer.
HeLa and HD2 cells were grown in Dulbecco’s Eagle medium Equal amounts (1 g) of recombinant proteins were incubated with
(DMEM)/HamF10 (v/v) containing 10% Fetal Calf Serum (FCS) and either purified recombinant CAK complex or TFIIH (hydroxylapatite
40 g/ml gentamycin. The normal human fibroblast (NHF) fraction prepared from HeLa cells) in the presence of (␥-
32
P)ATP, as
GM03348D was from Coriell Institute (Camden). Human XPD-defi- previously described (Rossignol et al., 1997).
cient fibroblasts TTD8PV and TTD12PV (Botta et al., 1998) were
isolated from TTD patients, whereas XPJCLO (Taylor et al., 1997)
Retrotranscription and Real-Time Quantitative PCR
and Ag10032 (purchased from Coriell Institute) were isolated from
cDNA synthesis was performed by using random hexanucleotides
XP patients. XPJCLO⫹LXPDN cells were obtained by retroviral-
and AMV reverse transcriptase (Sigma). Real-time quantitative PCR
mediated transduction of the wild-type XPD gene in XPJCLO paren-
was done with the FastStart DNA Master SYBR Green kit and the
tal cells (Carreau et al., 1995).
Lightcycler apparatus (Roche Diagnostic). PCR were performed with
the oligonucleotide pairs TATGATGGCCGAGGTGATAG and AGGTG
Antibodies
ATGTCCTCGTCTGTA for osteopontin, TCTTGACAAGGCAACAGTTC
The monoclonal antibodies against p44, p62, cdk7, XPD, the TFIIH
and AAGCCAACGTTCAGGTCTAA for CYP24, and CGGACAGGATT
subunits (Dubaele et al., 2003), Ets1, RNA pol II, RXR, and TIF2 were
GACAGATTG and TGCCAGAGTCTCGTTCGTTA for 18S. Results are
produced by the IGBMC facility. The polyclonal antibodies against
normalized to 18S.
XPB (S-19), TRAP220 (M-255), and VDR (sc-9164) were purchased
from SantaCruz.
Chromatin Immunoprecipitation
15 cm dishes of subconfluent cells were treated with vitD (10
⫺
7
M)
Vector Constructs
and chromatin immunoprecipitation experiments were next per-
Substitution of the threonine 38 residue to alanine was introduced
formed as described (Soutoglou and Talianidis, 2002), except that
into the cDNA of chicken Ets1 by using the Quick Change Site-
immunoprecipitated DNA were quantified by real-time quantitative
Directed Mutagenesis kit (Stratagene). Specific PCR products in-
PCR by using the oligonucleotide pairs CGAAGCACACCCGGTGA
cluding the entire coding sequence of Ets1 and VDR were inserted
ACT and CCAATGAGCACGCAGAGGAG for CYP24 and TGTGCTAA
into the NdeI site of pET15b (Novagen) to allow production of his-
GCATTGCTAGT and GTTCTGAATTCCGCTGTGT for osteopontin.
Ets1 and his-VDR, respectively. The pCYP24wt plasmid has been
Alternatively, cells (1.25 ⫻10
6
) were plated in 10 cm dishes and
previously described (Zierold et al., 1995). The pCYP24(-204) was
were transfected by using 5 gofCYP24 promoter construct and
made by deleting the 5⬘sequence of CYP24 promoter up to an AgeI
0.5 g of Ets1 or XPD expression vector. Immunoprecipitated DNA
site. Site-direct mutagenesis of the Ets binding site (EBS) was done
was analyzed with the primer pair GCTTGGCATTCCGGTACTGT and
as described above. The pGL3-RARE was as described (Keriel et
TCTCCAGCGGTTCCATCTTC.
al., 2002), whereas the pGL3-VDRE1 contains the VDRE1 oligonucle-
otide described below. The pMdm2 plasmid contains 283 bp of the
mouse internal promoter of Mdm2 gene inserted upstream of the Two-Dimensional Gel Electrophoresis
Sf9 extracts were prepared as described (Dubaele et al., 2003),luciferase gene (Ries et al., 2000).
except that 1 mM sodium orthovanadate, 10 mM NaF and 30 mM
sodium pyrophosphate were added in lysis buffer. Extracts wereCoimmunoprecipitation
Cell extract from Sf9 cells infected with baculoviruses encoding the diluted in 200 l of 7 M urea, 2 M thiourea, 4% CHAPS, 50 mM DTT,
and 0.5% biolytes (pH 3–10) (Biorad). ReadyStrips (pH 4–7) wereTFIIH subunits (Dubaele et al., 2003), and either Ets1 (Boulukos et
Molecular Cell
196
loaded with the extracts and allowed to rehydrate during 16 hr. A J.M., et al. (2002). Phosphorylation of human estrogen receptor
alpha at serine 118 by two distinct signal transduction pathwaysstep gradient of 50–8000 V was applied to the strips followed by
constant 8000 V, with focusing complete after 40,000 V-h. Prior to revealed by phosphorylation-specific antisera. Oncogene 21, 4921–
4931.the second dimension, strips were incubated 10 min in 6 M urea,
0.375 M Tris (pH 8.8), 2% SDS, 20% glycerol, first with 2% (w/v) Chiricone, D., De Santo, N.G., and Cirillo, M. (2003). Unusual cases
DTT and second with 2.5% (w/v) iodoacetamide. Equilibrated strips of chronic intoxication by vitamin D. J. Nephrol. 16, 917–921.
were inserted onto a 12% SDS-PAGE. After electrophoresis, pro- Dilworth, F.J., and Chambon, P. (2001). Nuclear receptors coordi-
teins were transferred onto a nitrocellulose membrane and probed nate the activities of chromatin remodeling complexes and coactiva-
with an antibody directed against Ets1. tors to facilitate initiation of transcription. Oncogene 20, 3047–3054.
Drane, P., Bravard, A., Bouvard, V., and May, E. (2001). Reciprocal
Acknowledgments down-regulation of p53 and SOD2 gene expression-implication in
p53 mediated apoptosis. Oncogene 20, 430–439.
We acknowledge J. Ghysdael for providing the cDNA and the Ets1
Dubaele, S., Proietti De Santis, L., Bienstock, R.J., Keriel, A., Ste-
baculovirus, and H. DeLuca, M. Gamble, E. May, N. Rochel, and A.
fanini, M., Van Houten, B., and Egly, J.M. (2003). Basal transcription
Sarasin for providing some vector constructs and cell lines. We are
defect discriminates between xeroderma pigmentosum and tri-
grateful to C. Braun for her high technical expertise and to I. Kolb-
chothiodystrophy in XPD patients. Mol. Cell 11, 1635–1646.
Cheynel and J.L. Weickert for production of recombinant baculovi-
ruses. These studies were supported by CNRS/INSERM/ULP, and Dwivedi, P.P., Omdahl, J.L., Kola, I., Hume, D.A., and May, B.K.
grants from the Association pour la Recherche sur le Cancer, the (2000). Regulation of rat cytochrome P450C24 (CYP24) gene expres-
European Communities (QLG1-1999 and QLRT-1999-02002), the sion. Evidence for functional cooperation of Ras-activated Ets tran-
Research ministery ACI Biologie Cellulaire et Structurale (3-2-535), scription factors with the vitamin D receptor in 1,25-dihydroxyvita-
and Institut des maladies Rares (A03098MS) and the Commissarriat min D(3)-mediated induction. J. Biol. Chem. 275, 47–55.
a
`l’Energie Atomique. P.D. is a recipient of a fellowship from the Glass, C.K., and Rosenfeld, M.G. (2000). The coregulator exchange
Fondation Lefoulon-Delalande and was supported by the prix Des- in transcriptional functions of nuclear receptors. Genes Dev. 14,
cartes awarded to J.-M.E. by EEC in 2000. E.C. is a recipient of 121–141.
Association pour la Recherche sur le Cancer fellowship. Henry, H.L. (2001). The 25(OH)D(3)/1alpha,25(OH)(2)D(3)-24R-
hydroxylase: a catabolic or biosynthetic enzyme? Steroids 66,
Received: April 8, 2004 391–398.
Revised: August 5, 2004 Holstege, F.C., van der Vliet, P.C., and Timmers, H.T. (1996). Opening
Accepted: August 10, 2004 of an RNA polymerase II promoter occurs in two distinct steps
Published: October 21, 2004 and requires the basal transcription factors IIE and IIH. EMBO J.
15, 1666–1677.
References
Jayaraman, G., Srinivas, R., Duggan, C., Ferreira, E., Swaminathan,
S., Somasundaram, K., Williams, J., Hauser, C., Kurkinen, M., Dhar,
Akeno, N., Saikatsu, S., Kawane, T., and Horiuchi, N. (1997). Mouse
R., et al. (1999). p300/cAMP-responsive element-binding protein
vitamin D-24-hydroxylase: molecular cloning, tissue distribution,
interactions with ets-1 and ets-2 in the transcriptional activation of
and transcriptional regulation by 1alpha,25-dihydroxyvitamin D3.
the human stromelysin promoter. J. Biol. Chem. 274, 17342–17352.
Endocrinology 138, 2233–2240.
Johnson, R.T., Squires, S., Elliott, G.C., Koch, G.L., and Rainbow,
Bastien, J., Adam-Stitah, S., Riedl, T., Egly, J.M., Chambon, P.,
A.J. (1985). Xeroderma pigmentosum D-HeLa hybrids with low and
and Rochette-Egly, C. (2000). TFIIH interacts witht eh retinoic acid
high ultraviolet sensitivity associated with normal and diminished
receptor gamma and phosphorylates its AF-1-activating domain
DNA repair ability, respectively. J. Cell Sci. 76, 115–133.
through cdk7. J. Biol. Chem. 275, 21896–21904.
Jones, G., Strugnell, S.A., and DeLuca, H.F. (1998). Current under-
Berneburg, M., and Lehmann, A.R. (2001). Xeroderma pigmentosum
standing of the molecular actions of vitamin D. Physiol. Rev. 78,
and related disorders: defects in DNA repair and transcription. Adv.
1193–1231.
Genet. 43, 71–102.
Keriel, A., Stary, A., Sarasin, A., Rochette-Egly, C., and Egly, J.M.
Botta, E., Nardo, T., Broughton, B.C., Marinoni, S., Lehmann, A.R.,
(2002). XPD mutations prevent TFIIH-dependent transactivation by
and Stefanini, M. (1998). Analysis of mutations in the XPD gene in
nuclear receptors and phosphorylation of RARalpha. Cell 109,
Italian patients with trichothiodystrophy: site of mutation correlates
125–135.
with repair deficiency, but gene dosage appears to determine clini-
cal severity. Am. J. Hum. Genet. 63, 1036–1048. Kitagawa, H., Fujiki, R., Yoshimura, K., Mezaki, Y., Uematsu, Y.,
Matsui, D., Ogawa, S., Unno, K., Okubo, M., Tokita, A., et al. (2003).
Boulukos, K.E., Pognonec, P., Begue, A., Galibert, F., Gesquiere,
The chromatin-remodeling complex WINAC targets a nuclear recep-
J.C., Stehelin, D., and Ghysdael, J. (1988). Identification in chickens
tor to promoters and is impaired in Williams syndrome. Cell 113,
of an evolutionarily conserved cellular ets-2 gene (c-ets-2) encoding
905–917.
nuclear proteins related to the products of the c-ets proto-onco-
gene. EMBO J. 7, 697–705. Kumar, R., and Thompson, E.B. (1999). The structure of the nuclear
hormone receptors. Steroids 64, 310–319.
Bradford, A.P., Wasylyk, C., Wasylyk, B., and Gutierrez-Hartmann,
A. (1997). Interaction of Ets-1 and the POU-homeodomain protein Lehmann, A.R. (2001). The xeroderma pigmentosum group D (XPD)
GHF-1/Pit-1 reconstitutes pituitary-specific gene expression. Mol. gene: one gene, two functions, three diseases. Genes Dev. 15,
Cell. Biol. 17, 1065–1074. 15–23.
Brivanlou, A.H., and Darnell, J.E., Jr. (2002). Signal transduction and Lemon, B.D., Fondell, J.D., and Freedman, L.P. (1997). Retinoid X
the control of gene expression. Science 295, 813–818. receptor:vitamin D3 receptor heterodimers promote stable preinitia-
tion complex formation and direct 1,25-dihydroxyvitamin D3-depen-
Carreau, M., Quilliet, X., Eveno, E., Salvetti, A., Danos, O., Heard,
dent cell-free transcription. Mol. Cell. Biol. 17, 1923–1937.
J.M., Mezzina, M., and Sarasin, A. (1995). Functional retroviral vector
for gene therapy of xeroderma pigmentosum group D patients. Hum. Li, J.J., Kim, R.H., Zhang, Q., Ogata, Y., and Sodek, J. (1998). Charac-
teristics of vitamin D3 receptor (VDR) binding to the vitamin D re-Gene Ther. 6, 1307–1315.
sponse element (VDRE) in rat bone sialoprotein gene promoter. Eur.
Chen, D., Riedl, T., Washbrook, E., Pace, P.E., Coombes, R.C., Egly,
J. Oral Sci. Suppl. 106, 408–417.
J.M., and Ali, S. (2000). Activation of estrogen receptor alpha by
S118 phosphorylation involves a ligand-dependent interaction with Li, R., Pei, H., and Watson, D.K. (2000). Regulation of Ets function
by protein - protein interactions. Oncogene 19, 6514–6523.TFIIH and participation of CDK7. Mol. Cell 6, 127–137.
Chen, D., Washbrook, E., Sarwar, N., Bates, G.J., Pace, P.E., Thiru- Li, X.Y., Boudjelal, M., Xiao, J.H., Peng, Z.H., Asuru, A., Kang, S.,
Fisher, G.J., and Voorhees, J.J. (1999). 1,25-Dihydroxyvitamin D3nuvakkarasu, V., Taylor, J., Epstein, R.J., Fuller-Pace, F.V., Egly,
TFIIH and Vitamin D Receptor
197
increases nuclear vitamin D3 receptors by blocking ubiquitin/pro- enhances the transactivation activities of c-Ets1 and c-Ets2. Mol.
Cell. Biol. 16, 538–547.teasome-mediated degradation in human skin. Mol. Endocrinol.
13, 1686–1694. Zierold, C., Darwish, H.M., and DeLuca, H.F. (1995). Two vitamin D
response elements function in the rat 1,25-dihydroxyvitamin D 24-
Lu, H., Fisher, R.P., Bailey, P., and Levine, A.J. (1997). The CDK7-
hydroxylase promoter. J. Biol. Chem. 270, 1675–1678.
cycH-p36 complex of transcription factor IIH phosphorylates p53,
enhancing its sequence-specific DNA binding activity in vitro. Mol.
Cell. Biol. 17, 5923–5934.
Metivier, R., Penot, G., Hubner, M.R., Reid, G., Brand, H., Kos, M.,
and Gannon, F. (2003). Estrogen receptor-alpha directs ordered,
cyclical, and combinatorial recruitment of cofactors on a natural
target promoter. Cell 115, 751–763.
Naar, A.M., Lemon, B.D., and Tjian, R. (2001). Transcriptional coacti-
vator complexes. Annu. Rev. Biochem. 70, 475–501.
Oikawa, T., and Yamada, T. (2003). Molecular biology of the Ets
family of transcription factors. Gene 303, 11–34.
Rachez, C., and Freedman, L.P. (2001). Mediator complexes and
transcription. Curr. Opin. Cell Biol. 13, 274–280.
Rachez, C., Lemon, B.D., Suldan, Z., Bromleigh, V., Gamble, M.,
Naar, A.M., Erdjument-Bromage, H., Tempst, P., and Freedman,
L.P. (1999). Ligand-dependent transcription activation by nuclear
receptors requires the DRIP complex. Nature 398, 824–828.
Ries, S., Biederer, C., Woods, D., Shifman, O., Shirasawa, S., Sasa-
zuki, T., McMahon, M., Oren, M., and McCormick, F. (2000). Oppos-
ing effects of Ras on p53: transcriptional activation of mdm2 and
induction of p19ARF. Cell 103, 321–330.
Rochette-Egly, C., Adam, S., Rossignol, M., Egly, J.M., and Cham-
bon, P. (1997). Stimulation of RAR alpha activation function AF-1
through binding to the general transcription factor TFIIH and phos-
phorylation by CDK7. Cell 90, 97–107.
Rossignol, M., Kolb-Cheynel, I., and Egly, J.M. (1997). Substrate
specificity of the cdk-activating kinase (CAK) is altered upon associ-
ation with TFIIH. EMBO J. 16, 1628–1637.
Shang, Y., Hu, X., DiRenzo, J., Lazar, M.A., and Brown, M. (2000).
Cofactor dynamics and sufficiency in estrogen receptor-regulated
transcription. Cell 103, 843–852.
Sierra, J., Villagra, A., Paredes, R., Cruzat, F., Gutierrez, S., Javed,
A., Arriagada, G., Olate, J., Imschenetzky, M., Van Wijnen, A.J., et
al. (2003). Regulation of the bone-specific osteocalcin gene by p300
requires Runx2/Cbfa1 and the vitamin D3 receptor but not p300
intrinsic histone acetyltransferase activity. Mol. Cell. Biol. 23, 3339–
3351.
Soutoglou, E., and Talianidis, I. (2002). Coordination of PIC assembly
and chromatin remodeling during differentiation-induced gene acti-
vation. Science 295, 1901–1904.
St-Arnaud, R., Candeliere, G.A., and Dedhar, S. (1996). New mecha-
nisms of regulation of the genomic actions of vitamin D in bone cells:
Interaction of the vitamin D receptor with non-classical response
elements and with the multifunctional protein, calreticulin. Front.
Biosci. 1, d177–d188.
Takeyama, K., Masuhiro, Y., Fuse, H., Endoh, H., Murayama, A.,
Kitanaka, S., Suzawa, M., Yanagisawa, J., and Kato, S. (1999). Selec-
tive interaction of vitamin D receptor with transcriptional coactiva-
tors by a vitamin D analog. Mol. Cell. Biol. 19, 1049–1055.
Taylor, E.M., Broughton, B.C., Botta, E., Stefanini, M., Sarasin, A.,
Jaspers, N.G., Fawcett, H., Harcourt, S.A., Arlett, C.F., and Lehmann,
A.R. (1997). Xeroderma pigmentosum and trichothiodystrophy are
associated with different mutations in the XPD (ERCC2) repair/tran-
scription gene. Proc. Natl. Acad. Sci. USA 94, 8658–8663.
Tolon, R.M., Castillo, A.I., Jimenez-Lara, A.M., and Aranda, A. (2000).
Association with Ets-1 causes ligand- and AF2-independent activa-
tion of nuclear receptors. Mol. Cell. Biol. 20, 8793–8802.
Xu, D., Wilson, T.J., Chan, D., De Luca, E., Zhou, J., Hertzog, P.J.,
and Kola, I. (2002). Ets1 is required for p53 transcriptional activity
in UV-induced apoptosis in embryonic stem cells. EMBO J. 21, 4081–
4093.
Yang, B.S., Hauser, C.A., Henkel, G., Colman, M.S., Van Beveren,
C., Stacey, K.J., Hume, D.A., Maki, R.A., and Ostrowski, M.C. (1996).
Ras-mediated phosphorylation of a conserved threonine residue
