Using MCF-7 cells as a model, we found that Jumonji AT-rich interactive domain 1A JARID1A; KDM5A⁄ RBP2, an enzyme that removes the activating H3K4 di- and trimethylation marks, was invol
Trang 1receptor expression
Antje Stratmann1,2and Bernard Haendler1
1 Therapeutic Research Group Oncology, Bayer Schering Pharma AG, Berlin, Germany
2 Institute of Chemistry and Biochemistry, Free University Berlin, Germany
Introduction
The progesterone receptor (PR) plays a fundamental
role in reproduction [1,2] and its expression is under
estrogen control [3–5] Several motifs that mediate the
estrogen response have been identified in the human
PR gene, including estrogen response element (ERE)
half-sites, Sp1- and AP-1-binding sites [3,6,7] Two PR
isoforms arising from a single gene have been identi-fied in humans [3,4,8] They differ in the length of their N-terminal domain and the extension found in PRB endows it with specific properties, possibly linked to the differential recruitment of cofactors [2,9] Both PR isoforms are expressed in the breast and several reports
Keywords
ChIP; estrogen; histone demethylation;
JARID1; progesterone receptor
Correspondence
B Haendler, Therapeutic Research Group
Oncology, Bayer Schering Pharma AG,
Mu¨llerstr 178, D-13342 Berlin, Germany
Fax: +49 30 468 18069
Tel: +49 40 468 12669
E-mail: bernard.haendler@bayer.com
(Received 12 October 2010, revised 21
January 2011, accepted 21 February 2011)
doi:10.1111/j.1742-4658.2011.08058.x
Transcriptional control of the progesterone receptor gene by estrogen is a complex mechanism It involves estrogen receptor a which uses several enzymes that locally modify histone tails as cofactors Using MCF-7 cells
as a model, we found that Jumonji AT-rich interactive domain 1A (JARID1A; KDM5A⁄ RBP2), an enzyme that removes the activating H3K4 di- and trimethylation marks, was involved in the fine-tuning of pro-gesterone receptor gene expression Reduction of JARID1A led to enhanced progesterone receptor expression, at both the basal and estrogen-stimulated levels Conversely, overexpression of JARID1A wild-type, but not the enzymatically inactive mutant, suppressed progesterone receptor promoter activity Chromatin immunoprecipitation experiments showed JARID1A to bind in a ligand-independent manner to a progesterone recep-tor gene upstream region that contains an estrogen response element half-site as well as the CCGCCC sequence, which is potentially recognized by JARID1A Estrogen treatment led to RNA polymerase II recruitment to this region and to increased estrogen receptor a binding to the PR enhan-cer region 1 In addition, elevation of H3K4 trimethylation was detected at the estrogen response element half-site region Reduction of JARID1A expression was followed by higher H3K4 trimethylation in this region Analysis of MDA-MB-231 cells, which do not express the progesterone receptor, indicated that H3K4 trimethylation did not take place in the reg-ulatory regions examined Taken together, the results underscore the importance of epigenetic modifications for regulation of progesterone receptor expression They suggest that H3K4 tri- and dimethylation play
an important role and that JARID1A is the histone-demethylating enzyme responsible for removal of this mark
Abbreviations
ChIP, chromatin immunoprecipitation; ER, estrogen receptor; ERE, estrogen response element; JARID1, Jumonji AT-rich interactive domain 1A; Jmj, Jumonji; PR, progesterone receptor; siRNA, small interfering RNA; tsp, transcription start point.
Trang 2indicate that the ratio between PRA and PRB is
altered during tumorigenesis, such that PRA
predomi-nates [9–11]
PR gene transcription is controlled by estrogen in a
complex way and is mainly driven by estrogen receptor
(ER) a [5] Recent studies indicate that ERa interacts
with histone-modifying enzymes to locally remodel
chromatin, thus leading to activation or repression of
gene expression CARM1 and PRMT1, two enzymes
involved in the methylation of arginine residues,
represent important coactivators for several steroid
receptors, including ERa [12] Methylating enzymes
that target lysines of histone 3 (H3) such as SMYD3
[13,14], MLL2 [15], G9a [16], RIZ1 [17], EZH2 [18] and
NSD1 [19] have all been shown to be ERa coactivators
Far less is known about the role of demethylating
enzymes in modulating ER function The lysine-specific
demethylase 1, which belongs to the amine oxydase
class, controls the expression of a subset of
ER-regu-lated genes by counteracting the repressive function of
H3K9 methylases [20] A second group of histone
demethylases is the Jumonji (Jmj) C domain-containing
family of which the Jumonji AT-rich interactive
domain 1 (JARID1) proteins represent a subgroup that
specifically removes dimethyl and trimethyl marks from
H3K4 [21,22] Four members have been identified in
humans and mice, based on the conservation of the
JmjC, JmjN, ARID and plant homeodomain domains,
and of a C5H2C zinc finger [21,22] JARID1A
(KDM5A⁄ RBP2) was originally described as a protein
that binds to the retinoblastoma protein, leading to the
promotion of cellular differentiation [23] More recently,
its function as a histone-demethylating enzyme was
evi-denced [24,25] Unlike other histone-modifying enzymes,
JARID1A recognizes a specific CCGCCC DNA sequence
via its ARID domain, thus regulating the expression of
numerous genes [26] A genome-wide analysis showed
that two groups of genes are controlled by JARID1A,
some of which are involved in cellular differentiation,
whereas others are implicated in mitochondrial function
and RNA⁄ DNA metabolism [27] Very recently,
JARID1A was found to be required for the
establish-ment of a reversible drug-tolerant phenotype in cancer
cell subpopulations [28] JARID1B is upregulated in
breast and prostate cancer and may favour tumour
pro-gression [24,29,30] Microarray studies have identified
100 genes controlled by JARID1B, several of which
are involved in the cell cycle and in signal transduction
[31] JARID1C is encoded by the SMCX gene on
chromosome X and escapes X inactivation [32] It plays
an important role in brain function and was found to
be mutated in families with X-linked mental retardation
[32] JARID1D is encoded by the Y chromosome and
forms a complex with a protein involved in meiosis during spermatogenesis [33]
Here, we analysed the role of the JARID1 family in transducing ERa function using the estrogen-con-trolled progesterone receptor (PR) gene as a model
We found that in the breast cancer cell line MCF-7, JARID1A acted as a repressor of PR expression Chromatin immunoprecipitation (ChIP) analysis showed the presence of JARID1A in a region covering
an ERE half-site located downstream of the PRB transcription start point (tsp) We furthermore found JARID1A to control H3K4 trimethylation at this location and this modification was much increased after estrogen treatment Conversely, H3K4 trimethyla-tion was not observed in the PR-negative cell line MDA-MB-231 Taken together, our results strongly suggest that H3K4 methylation in the region around the ERE half-site is important for modulation of PR expression by ERa, and JARID1A is involved in the removal of this mark
Results
JARID1A is involved in the regulation of PR expression in MCF-7 cells
In order to find out whether the JARID1 demethylases played a role in PR expression, two small interfering RNA (siRNA) pairs with selectivity for each JARID1 family member were obtained and validated MCF-7 breast cancer cells were treated with each siRNA pair and the respective JARID1 mRNA levels were deter-mined at different time-points after transfection, using quantitative PCR for measurement A robust and spe-cific reduction of expression was observed for JAR-ID1A, -1B and -1C two days post treatment, when compared with cells treated with lipids only or trans-fected with unrelated siRNAs (Fig S1) The selectivity
of the siRNAs was ascertained by determining the expression levels of the nontargeted family members and no significant change was observed (Fig S1) JARID1D is not expressed in MCF-7 cells because it
is encoded by the Y chromosome, and was therefore not analysed Western blot analysis was performed after siRNA transfection to confirm that the reduction
in transcript levels also impacted the respective protein levels A strong decrease in JARID1A, -1B and -1C protein levels was seen 3 days after transfection (Fig 1) No change in the GAPDH levels was observed in the treated cells compared with the differ-ent controls used
Next, we looked at the effects of JARID1 knock-down on the induction of PR expression by estrogen in
Trang 3MCF-7 cells As reported previously [34], we found that
the predominant PR isoform expressed in these cells is
PRB A time-course experiment showed that the
great-est PR stimulation by 1 nm great-estrogen took place at 16
and 24 h, as evidenced by quantitative PCR (Fig 2A)
Next, MCF-7 cells were transfected with siRNAs
direc-ted against the different JARID1 family members
Three days after transfection when the protein was
depleted, the cells were treated or not with estrogen and
total RNA was purified 24 h later Quantitative PCR
and western blot analysis showed that knocking-down
the expression of JARID1A led to a strong increase in
PR mRNA and protein levels (Fig 2B–D) The effect
was more pronounced in the absence of estradiol
(Fig 2B,C) than in its presence (Fig 2B,D) Reducing
the expression of JARID1B or -1C had far less effect
on basal PR levels (Fig 2C) and no effect on
estrogen-stimulated PR levels (Fig 2D) No change in GAPDH
levels was observed (Fig 2E,F) Because PR expression
was mostly affected by JARID1A knockdown, the
underlying mechanisms were studied in more detail
Overexpression of JARID1A represses PR
promoter activity
We then analysed whether JARID1A enzymatic activity
was involved in regulating the PR promoter Because
estrogen selectively increases the levels of transcripts
derived from promoter B in breast cancer cells [4], a
reporter plasmid containing the PRB promoter region
from)711 to +29 (see Fig 4A) placed upstream of the
Luc reporter was devised This region does not contain
a canonical ERE, but has previously been shown to be strongly induced by estrogen in cell-based transactiva-tion assays [3–5] This was confirmed in our experi-ments in which we measured a dose-dependent induction in MCF-7 cells co-transfected with an ERa expression plasmid and treated with increasing estrogen concentrations for 24 h (Fig 3B) The impact of overexpressed JARID1A either as a wild-type form or
as a mutated form in which the iron-binding site was modified by replacing histidine at position 483 with alanine, as described [24], was compared The corre-sponding mutation in JARID1B is known to destroy enzymatic activity [35] Overexpression of the wild-type and mutated JARID1A forms after transfection of the corresponding plasmids in MCF-7 cells was confirmed
at the protein level by western blot analysis (Fig 3A) Either plasmid was cotransfected with the ERa expres-sion vector and the reporter vector containing the PRB promoter into MCF-7 cells After 5 h, the cells were treated with different estrogen concentrations for an additional 24 h Determination of Luc activity revealed that overexpressing JARID1A wild-type, but not the enzymatically inactive mutant, entirely prevented estro-gen induction of the PRB promoter, even at the highest hormone concentration tested (Fig 3B)
JARID1A binds to the PR regulatory regions and influences the H3K4 methylation status
Genome-wide surveys indicate that ER-binding sites are often located far away from the tsp of estrogen-regulated genes [36–40] The precise regions involved
in the estrogen control of the PR gene have not all been characterized and multiple EREs exist, as shown
in several studies [41,42] A half-palindromic ERE located 571 bp downstream of the PRB tsp and adja-cent to two Sp1-binding sites has been identified (Fig 4A) Both the ERE half-site and the Sp1 elements are implicated in the estrogen responsiveness of this regulatory region [43] In addition, two ER-binding sites located 168 kb and 206 kb upstream of the PRB tsp were identified by whole-genome cartography [42]
In order to find out whether JARID1A binds to these different regulatory sequences and whether JARID1A and the H3K4 methylation status played a role in estrogen response, ChIP was performed using primers specific for each region As a control, we used primers directed against an intron of the unrelated F-box and leucine-rich repeat protein 11 gene, a region which is known to contain ERE motifs that are not bound by the ER [42] We first took an antibody directed against H3 and obtained similar signals at all analysed regions
Fig 1 Identification of siRNAs specific for JARID1 family
mem-bers MCF-7 cells were transfected with two siRNA pairs (#1 and
#2) specific for each JARID1 family member, as indicated In the
controls, cells were transfected with unrelated siRNAs (siCo), with
lipid only or left untreated (untr) JARID1 protein levels were
deter-mined by western blot analysis 3 days later GAPDH protein levels
were determined as loading control.
Trang 4(data not shown) RNA polymerase II (Pol II)
occu-pancy was then determined It was not observed in the
absence of hormone, but increased significantly at the
ERE half-site after estrogen treatment (Fig 4B)
Load-ing of ERa was not observed in the absence of
estro-gen Conversely, it was seen at enhancer 1 after
hormone treatment (Fig 4C) JARID1A binding was
near control levels in the absence and presence of
estrogen at enhancer region 1, and slightly more elevated at enhancer region 2 (Fig 4D) Much higher levels were seen at the ERE half-site region, indepen-dent of the hormone treatment (Fig 4D) Interestingly, sequence analysis indicated that two JARID1A-bind-ing sites were present in this region We then examined the H3K4 methylation status using antibodies specific for each methylation pattern Mono- and
dimethyla-Fig 2 Effects of JARID1 knockdown on PR expression (A) MCF-7 cells were treated with 1 n M 17b-estradiol for the indicated times, total RNA was prepared and the PR transcript levels were determined by quantitative PCR (B–F) MCF-7 cells were transfected with siRNAs spe-cific for each JARID1 (siJ1A: JARID1A; siJ1B: JARID1B; siJ1C: JARID1C) In the controls, cells were transfected with unrelated siRNAs (siCo), with lipid only (lip) or left untreated (untr) On day 3 post transfection, MCF-7 cells were additionally treated with vehicle or with 1 n M
17b-estradiol for 24 h (B) PR protein levels were determined by western blot analysis 4 days post transfection GAPDH protein levels were measured as loading control (C,D) PR transcript levels were determined by quantitative PCR 4 days post transfection (C) MCF-7 cells were additionally treated with vehicle (D) MCF-7 cells were additionally treated with 1 n M 17b-estradiol (E,F) GAPDH transcript levels were deter-mined by quantitative PCR 4 days post transfection (E) MCF-7 cells were additionally treated with vehicle (F) MCF-7 cells were additionally treated with 1 n M 17b-estradiol The fold inductions compared with siCon are given The asterisk denotes a significant difference (P < 0.01) between siRNA-treated cells and the corresponding untreated control, as analysed by Student’s t-test.
Trang 5tion were visible at all locations, with no effect of the
estrogen treatment (Fig 4E,F) Interestingly,
monome-thylation was highest at the enhancer 2 region and
dimethylation at the half-ERE region Trimethylation
was only observed at the half-ERE region and
estro-gen treatment led to a significant increase of the signal
(Fig 4G)
Because the modification in H3K4 trimethylation
did not parallel a change in JARID1A binding, we
looked whether the cellular JARID1A levels affected
H3K4 methylation at the PR gene upstream region
MCF-7 cells were transfected with JARID1A-specific
siRNAs and the PR regulatory regions analysed by
ChIP, as before No effect on H3K4 monomethylation
was observed (Fig 5A) Concerning H3K4
dimethyla-tion, a significant increase was seen at all sites
exam-ined (Fig 5B) The most dramatic changes were
observed for H3K4 trimethylation at the ERE half-site
region Here, a significant increase was observed
fol-lowing JARID1A knockdown both in the non- and
hormone-stimulated cells (Fig 5C) By contrast, no
trimethylation of H3K4 was detected at the far
upstream enhancers 1 or 2
To further characterize the importance of H3K4 methylation in the control of PR expression, we used the ERa-negative cell line MDA-MB-231 These cells
do not express PR [44] We compared JARID1A pro-tein levels by western blot analysis and found compa-rable signals in MDA-MB-231 and MCF-7 cells (Fig S2A) We then wanted to find out whether the local H3K4 methylation patterns in the PR gene regu-latory region differed between these two cell lines ChIP analysis indicated H3 levels to be comparable at all sites examined (data not shown) No Pol II occu-pancy was detected at any analysed site (Fig S2B) H3K4 mono- and dimethylation were seen at the half-ERE region (Fig S2C,D) but no trimethylation was seen (Fig S2E) In order to determine whether JARID1A was responsible for the lack of histone trimethylation and therefore possibly involved in the silencing of the PR gene in MDA-MB-231 cells, bind-ing of this demethylase to the PR regulatory regions was determined ChIP results showed that there was
no occupancy (Fig S2F) Taken together these data further underline the importance of H3K4 methylation
in controlling PR gene expression and indicate that JARID1A controls PR gene transcription only in the context of ERa expression
Discussion
The role of the H3K4 methylation status in controlling gene transcription is documented by several studies [40,41,45], but how far nuclear receptors use this his-tone mark for the regulation of downstream target genes has not been extensively analysed Here we stud-ied the role of the H3K4 demethylating enzymes of the JARID1 family in modulating estrogen control of the human PR gene transcriptional activity Using the breast cancer cell line MCF-7 as a model, we found that JARID1A regulated expression of the PR gene and fine-tuned its control by ERa Reducing JAR-ID1A expression by specific siRNAs dramatically enhanced PR expression at the basal and less so at the estrogen-stimulated levels Conversely, overexpressing the JARID1A wild-type enzyme, but not its inactive mutant, suppressed the activity of the PRB promoter
in transient transfection experiments This prompted
us to determine the role of H3 methylation in regulat-ing PR expression and the potential bindregulat-ing of JARID1A to the PR regulatory regions
The H3K4 methylation pattern of three PR regula-tory regions possibly involved in the control by estro-gen, namely enhancer 1, enhancer 2 and the half-ERE region located 571 bp downstream of the PRB tsp was analysed ChIP experiments showed the overall H3K4
Fig 3 Effects of JARID1A on PR stimulation by estrogen (A)
MCF-7 cells were transfected with JARID1A wild-type (wt), mutant
H483A (H483A) or empty vector (vector) Protein levels were
analy-sed by western blot analysis 2 days later (B) MCF-7 cells were
cotransfected with a PRB promoter reporter vector and an ERa
expression vector (control) Expression vectors for wild-type and
mutant H483A JARID1A were additionally cotransfected to
gener-ate the indicgener-ated curves (wt and H483A, respectively) Five hours
post transfection, stimulation was performed with 0–100 n M
17b-estradiol Luc activity was determined 24 h later The fold
inductions compared with cells not treated with estrogen are given.
Trang 6methylation pattern to fit well with recent
genome-wide studies reporting monomethylation in different
regions of active genes, dimethylation preferentially at
promoters and trimethylation exclusively at promoters
[46,47] The changes we observed in the PR locus upon estrogen stimulation were an increase in Pol II binding
to the half-ERE region and of ERa binding to enhan-cer 1 Interestingly, the enhanenhan-cer 1 region amplified in
Fig 4 ChIP analysis of PR regulatory regions in MCF-7 cells (A) Schematic representation of the PR gene upstream region The tsp for the PRB and PRA transcripts are indicated with arrows Distal ER-binding sites and a proximal half-ERE located 571 bp downstream of the PRB tsp are highlighted with black diamonds The respective DNA regions amplified with the primer pairs used in ChIP are shown above the sites with black bars The white box indicates the PRB promoter region that was subcloned upstream of the Luc reporter construct (B) ChIP assay of Pol II binding (C) ChIP assay of ERa binding (D) ChIP assay of JARID1A binding (E) ChIP assay of H3K4 monomethylation (F) ChIP assay of H3K4 dimethylation (G) ChIP assay of H3K4 trimethylation Chromatin isolated from MCF-7 cells treated for 1 h with 1 n M
17b-estradiol (black bars) or untreated (white bars) was immunoprecipitated with the appropriate antibodies and the bound DNA analysed by quantitative PCR amplification An IgG-specific antibody was used in control experiments (light and dark grey bars) The asterisk denotes a significant difference (P < 0.01) between hormone-treated and the corresponding untreated data sets, as analysed by Student’s t-test.
Trang 7our ChIP experiments is directly adjacent to a motif
recently found to represent a bona fide ERE [48], and
therefore likely to be detected under our conditions In
addition, an increase in H3K4 trimethylation of the
half-ERE region was also associated with hormone
stimulation This change in H3K4 trimethylation
cor-roborates previous findings showing the presence of
this mark in the 5¢ region of genes to be essential
for gene activity [45,47] Concerning JARID1A, a hormone-independent binding to the half-ERE was evidenced Sequence analysis showed that this region also contains two DNA binding motifs (CCGCCC) known to be recognized by the ARID domain of JARID1A It should however be added that other DNA motifs are also bound by JARID1A [49] so that only a comprehensive mutational analysis will allow to determine which PR regulatory region is involved Despite the fact that JARID1A levels were not affected by estrogen treatment, this demethylase is likely to be involved in controlling the H3K4 methyla-tion status of the PR gene, as shown by our knock-down ChIP experiments in MCF-7 cells, which led to
an increase of di- and especially of trimethylation, mainly at the half-ERE region The requirement of the demethylase activity for the repressive effects of JARID1A on PR gene expression was furthermore underlined by the luciferase assay which showed that overexpression of JARID1A wild-type but not of the active site mutant resulted in suppression of PRB promoter activity
Our results show that both JARID1A and ERa are involved in the regulation of the PR promoter activity but that their binding to the three regulatory regions examined did not follow the same pattern ERa binds
to the enhancer region in a hormone-dependent man-ner and is recruited following estrogen stimulation By contrast, the binding of JARID1A to the region close
to the tsp does not depend on the hormonal status However, the activity of JARID1A is influenced by the presence of the hormone-activated ERa A possible scenario is that in the absence of estrogen, JARID1A
is needed to shut down PR gene expression and that following hormone treatment, long-range interactions between distantly spaced cis-acting elements take place, thus allowing crosstalk between ERa, JARID1A and the transcriptional machinery at the tsp Indeed numerous long-range chromatin interactions have been detected for ERa [37,50], including interactions with the promoter-bound transcription factor Sp1 [51] These interactions might lead to conformational modi-fications of JARID1A and changes in protein–protein interactions that regulate its enzymatic activity
The analysis of MDA-MB-231 cells which do not express PR or ERa revealed that Pol II did not bind
to the regulatory regions examined and that only the half-ERE region was methylated at H3K4 Impor-tantly however, no enrichment in trimethylation was observed, in line with the fact that PR is not expressed
in these cells Previous data generated in MDA-MB-231 cells showed that DNA methylation of the PR pro-moter at a specific CpG island led to repression and
Fig 5 Influence of JARID1A levels on H3K4 methylation of the PR
regulatory regions in MCF-7 cells JARID1A-specific siRNAs (mixed
pair #1 and #2) were used for expression knockdown (light and
dark grey bars) An unrelated siRNA pair was used in the control
experiments (white and black bars) Four days post transfection,
chromatin isolated from MCF-7 cells and treated for 1 h with 1 n M
17b-estradiol (black and dark grey bars) or left untreated (white and
light grey bars) was immunoprecipitated with antibodies specific for
H3K4 methylation Bound DNA was analysed by quantitative PCR
amplification An IgG-specific antibody was used as negative
con-trol (striped bars) (A) ChIP assay of H3K4 monomethylation (B)
ChIP assay of H3K4 dimethylation (C) ChIP assay of H3K4
trime-thylation The asterisk denotes a significant difference (P < 0.01)
between the unrelated siRNA control and the JARID1A siRNA,
hor-mone-treated and untreated respectively, as analysed by Student’s
t-test.
Trang 8that reactivation can be achieved by the DNMT
inhib-itor 5-aza-2¢-deoxycytidine [44] However, when
treat-ing the cells with a combination of
5-aza-2¢-deoxycytidine and antiestrogen, the unmethylated PR
gene remains in the repressed state [44] Our data
strongly suggest that absence of H3K4 trimethylation
is an important factor preventing PR expression in
MDA-MB-231 cells and indicate a possible role of
JARID1A in the silencing process in this cell line
Because no binding of JARID1A to the PR regulatory
region could be detected, the role of this demethylase
may be that of an early regulator responsible for the
active removal of positive marks For the later
mainte-nance of the silenced status, DNA methylation and the
absence of ER might be sufficient Future experiments
will show whether for the regulation of PR expression
a cross-talk exists between H3K4 methylation and
DNA methylation or histone acetylation, as recently
reported for other genes [52–54]
A recent study showed that the H3K4
methyltrans-ferase SMYD3 directly interacts with ERa and
stimu-lates its activity in presence of estrogen [13]
Examination of different estrogen-regulated genes such
as pS2, CTSD and GREB1 indicates that this is linked
to elevated H3K4 di- and trimethylation at
ERE-con-taining regulatory regions [13] It is thus possible that
the respective levels of SMYD3 and JARID1A, as well
as their interaction with liganded ERa, govern the
local H3K4 methylation levels at estrogen-regulated
genes, and thereby their expression levels
Materials and methods
Plasmid construction
For the design of full-length human JARID1A cDNA
(NM_001042603.1), digestion of
pCMV-SPORT6-JAR-ID1A (RZPD, IRAKp961B10255Q) was performed with
RsrII and the 5¢-terminal 3257 bp were obtained The
3¢-terminal 1868 bp were amplified from a cDNA brain pool
(QUICK-Clone cDNA, BD Biosciences Clontech, San
Jose, CA, USA) using the forward and reverse primers 5¢-C
ATTGTTACAGGTGCTGAGCC-3¢ and 5¢-CTAACTGGT
CTCTTTAAGATCCTC-3¢, transferred into the pCR2.1-T
OPO (Invitrogen, Carlsbad, CA, USA) and partially
digested with RsrII The two resulting fragments were
ligated using the Rapid DNA Ligation Kit (Roche
Diag-nostics GmbH, Mannheim, Germany), transferred into the
pCR2.1-TOPO plasmid and PCR-amplified with the
primers 5¢-ATATGTCGACAATGGCGGGCGTGGGG-3¢
and 5¢-CGATGCGGCCGCCTAACTGGTCTCTTTAAG
ATCC-3¢ The resulting product was cloned between the
SalI and NotI sites of the pCMV-HA expression vector
(BD Biosciences Clontech) The H483A inactivating mutation was introduced into pCMV-HA-JARID1A by site-directed mutagenesis using the Quik-Change kit (Strata-gene, La Jolla, CA, USA) and the primers: 5¢-CTTCTCTT CTTTTTGCTGGGCCATTGAGGATCACTGGAG-3¢ and 5¢-CTCCAGTGATCCTCAATGGCCCAGCAAAAAGAA GAGAAG-3¢ The PRB promoter luciferase (Luc) reporter construct was generated by amplifying the)711 to +29 bp region from Human Genomic DNA (Clontech, Heidelberg, Germany) using the primers 5¢-GGATCCATTTTATAAG CTCAAAGATAATTAC-3¢ and 5¢-CTTACCCCGATTAG TGACAGCTGTGGAC-3¢ The amplified fragment was then transferred into the pCR2.1-TOPO and subcloned into the KpnI and XhoI sites of the pGL4.26[luc2⁄ minP ⁄ Hygro] vector (Promega, Madison, WI, USA) Insertion of the human ERa cDNA into the pSG5 plasmid (Stratagene) was performed using standard procedures All plasmids were veri-fied by DNA sequencing of the inserts and flanking regions
Cell culture
Cell lines were obtained from the American Tissue Culture Collection (LGC Promochem, Wesel, Germany) MCF-7 cells were grown at 37C in a 5% CO2atmosphere in phe-nol-red RPMI-1640 supplemented with 10% fetal bovine serum (both from Biochrom AG, Berlin, Germany), and
100 UÆmL)1 penicillin and 100 lgÆmL)1 streptomycin
(Gib-co, Invitrogen, Karlsruhe, Germany) For estrogen stimula-tion experiments, cells were cultured in estrogen-free medium (phenol-red-free RPMI-1640) supplemented with 10% charcoal-stripped fetal bovine serum and 2 mm l-glu-tamine (Gibco) for 2-3 days prior to treatment with 1 nm estrogen This concentration was chosen based on previous experience 17b-estradiol (ZK5018) was synthesized in-house and dissolved in ethanol MDA-MB-231 cells were maintained in Dulbecco’s modified Eagle’s medium (Gibco) with 10% fetal bovine serum, 2 mm l-glutamine, 1· nones-sential amino acids (NEA; Biochrom AG), and 100 UÆmL)1 penicillin and 100 lgÆmL)1streptomycin (Gibco)
Knockdown experiments
MCF-7 cells were transfected with JARID1 siRNAs (Stealth siRNA, Invitrogen) at a final concentration of
40 nm by using Lipofectamine 2000 (Invitrogen) Stealth RNAi Negative Universal Control Medium (Invi-trogen) was transfected in the control experiments Specific siRNAs sequences for each JARID1 family member were selected For JARID1A: si#1 5¢-ATACTAACCAGCCAC CCTAGAGCTC-3¢, si#2 5¢-CAGCCTCCATTTGCCTG TGAAGTAA-3¢; for JARID1B: si#1 5¢-CACGTATCCA GAGACTGAATGAATT-3¢, si#2 5¢-GCCTTCTTGTTTG CCTGCATCATGT-3¢; for JARID1C: si#1 5¢-AGGCCC AGACGAGAGTGAAACTGAA-3¢, si#2 5¢-CGGTTTCCC
Trang 9TGTCAGTGACAGTAAA-3¢ Total RNA was extracted
using the RNeasy Mini Kit (Qiagen, Hilden, Germany) and
reverse-transcription performed using the SuperScript III
First-Strand (Invitrogen) and random primers Specific
transcript levels were determined by real-time PCR using
the following assays from Applied Biosystems (Darmstadt,
Germany): JARID1A Hs00231908_m1, JARID1B
Hs00366783_m1, JARID1C Hs00188277_m1, PGR
Hs00172183_m1 and GAPDH 4326317E Human
cyclophi-lin A levels were determined as control using the primers
hu Cyc 4326316E Protein extracts were prepared using the
Mammalian Protein Extraction Reagent (M-Per, Pierce
Biotechnology, Rockford, IL, USA) supplemented with a
protease inhibitor mix (Roche) and benzonase (Merck,
Darmstadt, Germany) They were then analysed by western
blot (see below)
Cell-based transactivation
For the transactivation assays, MCF-7 cells were seeded
into 96-well plates at a concentration of 8000 cellsÆ80 lL)1
per well Transfection was carried out 44 h later using
FuGENE HD Transfection Reagent (Roche Diagnostics
GmbH) in Opti-MEM Expression plasmids for wild-type
JARID1A or mutant form (30 ngÆwell)1) were cotransfected
together with a Luc-based reporter vector harbouring the
PRB promoter (40 ngÆwell)1) and a pSG5-based
ERa-expressing plasmid (80 ngÆwell)1) Hormone induction was
performed 4–5 h later by adding estrogen up to 100 nm
Measurement of Luc activity was carried out after 24 h in
a Victor multilabel counter (Perkin–Elmer, Wellesley, MA,
USA), following the addition of 100 lL of SteadyLite Plus
Reagent (Perkin–Elmer) For all points the average value of
six wells treated in parallel was taken
Western blot analysis
Total protein extracts prepared from transfected cells (see
above) were analysed by western blot First, the protein
lysates were separated using NuPAGE gradient gels and
then transferred onto polyvinylidene fluoride membranes
(both from Invitrogen) The membranes were blocked and
incubated overnight with the primary antibody The
anti-bodies were G-20, sc-544 (Santa Cruz Biotechnology, Santa
Cruz, CA, USA) for ERa, A300-897A (Bethyl
Laborato-ries, Montgomery, TX, USA) for JARID1A, ab50958
(abcam, Cambridge, UK) for JARID1B, A301-035A
(Beth-yl Laboratories) for JARID1C, MS-298 Ab-8 (Thermo
Scientific, Rugby, UK) for PR and MAb 6C5 (Advanced
ImmunoChemical, Long Beach, CA, USA) for GAPDH
After incubation with the appropriate secondary antibodies
(rabbit NA934V or mouse NA931V antibody from GE
Healthcare, Chalfont St Giles, UK), the membranes were
developed using the Western Lightning chemiluminescence
kit (Perkin–Elmer)
Chromatin immunoprecipitation
MCF-7 and MDA-MB-231 cells were grown in 150-mm dishes in estrogen-free medium for 3 days They were then treated with 1 nm estrogen for 1 h and dual cross-linked at room temperature by a 45-min incubation in 2 mm disucc-inimidyl glutarate, followed by treatment with 1% formal-dehyde for 10 min The reaction was stopped by adding glycine to a final concentration of 125 nm After 5 min, the cells were washed twice with cold NaCl⁄ Pi, detached by scraping into cold NaCl⁄ Pi and pelleted by centrifugation
at 4C for 5 min at 180 g Chromatin was isolated after cell lysis in 5 mm Pipes pH 8, 85 mm KCl, 0.5% Noni-det P40, supplemented with protease inhibitors (Roche Diagnostics GmbH) This was followed by nuclear lysis using RIPA-buffer (10 mm Tris⁄ HCl pH 8, 1 mm EDTA, 0.5 mm EGTA, 1% Triton X-100, 0.1% SDS, 0.1% Na-de-oxycholate, 140 mm NaCl and protease inhibitors) DNA was sheared using the Biorupter (Diagenode, Lie`ge, Bel-gium) by sonicating for 2· 10 min with a 30-s interval at high level The samples were then centrifuged at 15 700 g for 10 min and at 4C In order to pre-clear the chromatin solution, 60 lL of Dynabeads Protein A⁄ G (Invitrogen) were added to the cell lysates and incubated for 1 h while rotating at 4C For immunoprecipitation, we used 3–4 lg antibody specific for ERa (HC-20X, sc-543X, Santa Cruz Biotechnology), for JARID1A (A300-897A, Bethyl Labora-tories), for RNA Pol II (ab5408, abcam), for H3 (ab1791 abcam), for H3K4me1 (ab8895, abcam), for H3K4me2 (ab7766, abcam) or for H3K4me3 (ab1012, abcam) Rabbit IgG (PP64, Millipore, Bedford, MA, USA) was used as control The antibodies were first pre-bound to 40 lL Dyn-abeads Protein A⁄ G in NaCl ⁄ Pi by incubating them for 2-3 h while rotating at 4C After removing an input sam-ple, the cell lysate was added to the prebound antibodies and incubated overnight at 4C Washing was performed with buffer I (20 mm Tris⁄ HCl pH 8, 2 mm EDTA,
150 mm NaCl, 1% Triton X-100, 0.1% SDS), buffer II (20 mm Tris⁄ HCl pH 8, 2 mm EDTA, 500 mm NaCl, 1% Triton X-100, 0.1% SDS) and buffer III (10 mm Tris⁄ HCl
pH 8, 1 mm EDTA, 250 mm LiCl, 1% Nonidet-P40, 1% sodium deoxycholate) followed by two washes with TE (10 mm Tris⁄ HCl pH 8, 1 mm EDTA) The immunocom-plexes were eluted with 0.1 m NaHCO3, 1% SDS and the cross-link was reverted at 65C overnight The eluates were sequentially treated for 1 h with RNase H and then with proteinase K (both from Qiagen) The DNA was then purified using the DNeasy Blood and Tissue kit (Qiagen) The immunoprecipitated DNA was analysed by quantita-tive PCR using QuantiFast SYBR Green PCR Kit (Qiagen) The reaction was performed at a combined annealing⁄ extension temperature of 60 C for 40 cycles using primers designed to specifically amplify the sequence containing the +571 ERE half-site [3] and two regions cov-ering ER-binding sites named enhancer 1 and enhancer 2
Trang 10located at)168 and )206 kb in the PR regulatory regions
[42] Primers directed against an unrelated region located in
intron 2 of the F-box and leucine-rich repeat protein 11
gene (Chromosome 11, position 66 656 165 to 66 657 244;
Control 5 in Carroll et al [42]) were used as control This
region contains ERE motifs which are, however, not bound
by the ER [42] For the experiments in which JARID1A
was immunoprecipitated, an expression plasmid containing
the corresponding cDNA needed to be transfected, due to
the low sensitivity of the JARID1A antibody
Statistical analysis
All data represent three independent experiments using cells
from separate cultures and are expressed as the
mean ± sample SD The results between the siRNA
treat-ment and control groups in Figs 2 and 5, and between the
hormone-treated and untreated groups in Fig 4 were
com-pared using the Student’s t-test A P-value < 0.01 was
con-sidered significant
Acknowledgements
We are indebted to Karl Ziegelbauer and Dominik
Mumberg (BSP, Berlin), and to Petra Knaus (Institute
of Chemistry and Biochemistry, Free University,
Berlin) for support throughout the project Many
fruitful discussions with Hortensia Faus and Carlo
Stresemann are gratefully acknowledged
References
1 Ismail PM, Amato P, Soyal SM, DeMayo FJ, Conneely
OM, O’Malley BW & Lydon JP (2003) Progesterone
involvement in breast development and tumorigenesis –
as revealed by progesterone receptor ‘knockout’ and
‘knockin’ mouse models Steroids 68, 779–787
2 Gadkar-Sable S, Shah C, Rosario G, Sachdeva G &
Puri C (2005) Progesterone receptors: various forms
and functions in reproductive tissues Front Biosci 10,
2118–2130
3 Kastner P, Krust A, Turcotte B, Stropp U, Tora L,
Gronemeyer H & Chambon P (1990) Two distinct
estrogen-regulated promoters generate transcripts
encoding the two functionally different human
progesterone receptor forms A and B EMBO J 9,
1603–1614
4 Graham JD, Roman SD, McGowan E, Sutherland RL
& Clarke CL (1995) Preferential stimulation of human
progesterone receptor B expression by estrogen in
T-47D human breast cancer cells J Biol Chem 270,
30693–30700
5 Flototto T, Niederacher D, Hohmann D,
Heimerzheim T, Dall P, Djahansouzi S, Bender HG &
Hanstein B (2004) Molecular mechanism of estrogen receptor (ER)alpha-specific, estradiol-dependent expression of the progesterone receptor (PR) B-isoform
J Steroid Biochem Mol Biol 88, 131–142
6 Petz LN, Ziegler YS, Schultz JR, Kim H, Kemper JK
& Nardulli AM (2004) Differential regulation of the human progesterone receptor gene through an estrogen response element half site and Sp1 sites J Steroid Bio-chem Mol Biol 88, 113–122
7 Petz LN, Ziegler YS, Loven MA & Nardulli AM (2002) Estrogen receptor alpha and activating protein-1 medi-ate estrogen responsiveness of the progesterone receptor gene in MCF-7 breast cancer cells Endocrinology 143, 4583–4591
8 Daniel AR, Knutson TP & Lange CA (2009) Signaling inputs to progesterone receptor gene regulation and promoter selectivity Mol Cell Endocrinol 308, 47–52
9 Conneely OM, Mulac-Jericevic B & Arnett-Mansfield R (2007) Progesterone signaling in mammary gland devel-opment Ernst Schering Found Symp Proc 2007⁄ 1, 45–54
10 Lange CA, Sartorius CA, Abdel-Hafiz H, Spillman
MA, Horwitz KB & Jacobsen BM (2008) Progesterone receptor action: translating studies in breast cancer models to clinical insights Adv Exp Med Biol 630, 94–111
11 Mote PA, Bartow S, Tran N & Clarke CL (2002) Loss
of co-ordinate expression of progesterone receptors A and B is an early event in breast carcinogenesis Breast Cancer Res Treat 72, 163–172
12 Klinge CM, Jernigan SC, Mattingly KA, Risinger KE
& Zhang J (2004) Estrogen response element-dependent regulation of transcriptional activation of estrogen receptors alpha and beta by coactivators and corepres-sors J Mol Endocrinol 33, 387–410
13 Kim H, Heo K, Kim JH, Kim K, Choi J & An W (2009) Requirement of histone methyltransferase SMYD3 for estrogen receptor-mediated transcription
J Biol Chem 284, 19867–19877
14 Hamamoto R, Silva FP, Tsuge M, Nishidate T, Katag-iri T, Nakamura Y & Furukawa Y (2006) Enhanced SMYD3 expression is essential for the growth of breast cancer cells Cancer Sci 97, 113–118
15 Mo R, Rao SM & Zhu YJ (2006) Identification of the MLL2 complex as a coactivator for estrogen receptor alpha J Biol Chem 281, 15714–15720
16 Lee DY, Northrop JP, Kuo MH & Stallcup MR (2006) Histone H3 lysine 9 methyltransferase G9a is a tran-scriptional coactivator for nuclear receptors J Biol Chem 281, 8476–8485
17 Carling T, Kim KC, Yang XH, Gu J, Zhang XK & Huang S (2004) A histone methyltransferase is required for maximal response to female sex hormones Mol Cell Biol 24, 7032–7042
18 Shi B, Liang J, Yang X, Wang Y, Zhao Y, Wu H, Sun L, Zhang Y, Chen Y, Li R et al (2007) Integration