Báo cáo khoa học: 15-Deoxy D12,14-prostaglandin J2 suppresses transcription by promoter 3 of the human thromboxane A2 receptor gene through peroxisome proliferator-activated receptor c in human erythroleukemia cells ppt

15-Deoxy D
12,14
-prostaglandin J
2
suppresses transcription
by promoter 3 of the human thromboxane A
2
receptor
gene through peroxisome proliferator-activated receptor c
in human erythroleukemia cells
Adrian T. Coyle, Martina B. O’Keeffe and B. Therese Kinsella
Department of Biochemistry, Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Ireland
The cyclopentanone prostaglandin 15-deoxy-D
12,14
-
prostaglandin (PG) J
2
(15d-PGJ
2
), a dehydration prod-
uct of cyclooxygenase (COX)-derived PGD
2
present in
inflammatory exudates, is elevated during the resolu-
tion phase of inflammation and was initially identified
as a high affinity natural ligand for peroxisome prolif-
erator-activated receptors (PPAR)c [1,2] although a
number of PPARc independent effects have recently
been reported [3]. The nuclear hormone receptor
PPARc classically up-regulates gene expression by
binding as a heterodimer with the retinoic X receptor

(RXR) to specific response elements consisting of one
Keywords
thromboxane receptor; promoter;
peroxisome proliferator-activated receptor c;
15-deoxy D
12,14
-prostaglandin J
2
; isoforms
Correspondence
B. T. Kinsella, Department of Biochemistry,
Conway Institute of Biomolecular and
Biomedical Research, University College
Dublin, Belfield, Dublin 4, Ireland
Fax: +353 1 2837211
Tel: +353 1 7166727
E-mail:
(Received 15 June 2005, revised 28 July
2005, accepted 29 July 2005)
doi:10.1111/j.1742-4658.2005.04890.x
In humans, thromboxane (TX) A
2
signals through two receptor isoforms,
thromboxane receptor (TP)a and TPb, which are transcriptionally regula-
ted by distinct promoters, Prm1 and Prm3, respectively, within the single
TP gene. The aim of the current study was to investigate the ability of the
endogenous peroxisome proliferator-activated receptor (PPAR)c ligand
15-deoxy-D
12,14
-prostaglandin J

2
(15d-PGJ
2
) to regulate expression of the
human TP gene and to ascertain its potential effects on the individual TPa
and TPb isoforms. 15d-PGJ
2
suppressed Prm3 transcriptional activity and
TPb mRNA expression in the platelet progenitor megakaryocytic human
erythroleukemia (HEL) 92.1.7 cell line but had no effect on Prm1 or Prm2
activity or on TPa mRNA expression. 15d-PGJ
2
also resulted in reductions
in the overall level of TP protein expression and TP-mediated intracellular
calcium mobilization in HEL cells. 15d-PGJ
2
suppression of Prm3 tran-
scriptional activity and TPb mRNA expression was found to occur through
a novel mechanism involving direct binding of PPARc–retinoic acid X
receptor (RXR) heterodimers to a PPARc response element (PPRE) com-
posed of two imperfect hexameric direct repeat (DR) sequences centred at
)159 and )148, respectively, spaced by five nucleotides (DR5). These data
provide direct evidence for the role of PPARc in the regulation of human
TP gene expression within the vasculature and point to further critical dif-
ferences in the modes of transcriptional regulation of TPa and TPb in
humans. Moreover, these data highlight a further link between enhanced
risk of cardiovascular disease in diabetes mellitus associated with increased
synthesis and action of thromboxane A
2
(TXA

2
).
Abbreviations
AP-1, activator protein-1; 15d-PGJ
2
, 15-deoxy-D
12,14
-prostaglandin J
2
; CHIP, chromosomal immunoprecipitation; COX, cyclooxygenase;
DR, direct repeat; d ⁄ s, double stranded; EMSA, electromobility shift assay; HEK, human embryonic kidney; HEL, human erythroleukemia;
PG, prostaglandin; PPAR, peroxisome proliferator-activated receptor; PPRE, peroxisome proliferator-activated receptor response element;
Prm, promoter; RLU, relative luciferase units; RAR, retinoic acid receptor; RXR, retinoic acid X receptor; TZD, thiazolidinedione;
TP, thromboxane receptor; TXA
2
, thromboxane A
2
; VSMC, vascular smooth muscle cell.
4754 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
or more copies of the hexameric DNA consensus
sequence AGGTCA arranged as a direct repeat (DR)
spaced by one nucleotide, hence termed DR1 [4,5].
PPARc-transcriptional activation may also involve the
recruitment of various coactivators to specific target
genes such as p300 [6], the SRC-1 coactivators [7–9],
PGC-1 and PGC-2 [8,9], ARA70 [10] and DRIP205
(or TRA220) [11]. In addition to activating trans-
cription, PPARc can also negatively regulate gene
expression through mechanisms involving either the
trans-repression (negative cross-talk) of activating tran-

scription factors, e.g. NF-jB and activator protein-1
(AP-1) [12,13] or the sequestration of limiting amounts
of coactivator molecules such as CBP [14].
The beneficial insulin-sensitizing actions of 15d-PGJ
2
and the thiazolidinediones (TZDs) as PPARc agonists
are widely recognized, such as in the treatment of type
II diabetes [15]. Moreover, while the inhibitory effects
of PPARc play a prominent role in the resolution
of inflammation [15], they are also thought to be
important within the vasculature where they offer a
cardioprotective effect during myocardial ischemia,
reperfusion and atherosclerosis and hence it is rea-
soned that PPARc agonists may help to alleviate the
adverse cardiovascular events associated with diabetes
mellitus [16,17]. For example, PPARc activators inhi-
bit matrix metalloproteinase-9 expression in vascular
smooth muscle cells (VSMCs) [18] and thrombin
induced endothelin-1 production in endothelial cells
[19]. Moreover, it has recently been established that
the expression of a number of other key vascular genes
are suppressed in response to PPARc activation inclu-
ding those encoding the inducible cyclooxgenase
(COX)II [20] and nitric oxide synthase [14], the rat
thromboxane (TX)A
2
synthase [21] and the rat throm-
boxane A
2
(TXA

2
) receptor (thromboxane receptor,
TP) [22].
The COX-derived TXA
2
is a potent biologically act-
ive eicosanoid primarily released from activated plate-
lets, monocytes and damaged vessel walls and plays a
central role in the dynamic regulation of vascular hae-
mostasis [23]. Alterations in the level of TXA
2
, TXA
2
synthase or the TXA
2
receptor (TP) are widely
implicated in a variety of vascular diseases including
thrombosis, unstable angina, systemic and pregnancy-
induced hypertension [24–26]. TXA
2
is also known to
play a pathophysiological role in inflammatory diseases
such as in atherosclerosis [27], glomerulonephritis [28]
and diabetic nephropathy [29]. TXA
2
signals through
the TXA
2
receptor, or TP, a G-protein coupled recep-
tor primarily coupled to phospholipase (PL) Cb activa-

tion [23]. In humans, but not in other nonprimates,
TXA
2
signals through two TP isoforms, namely TPa
and TPb, that arise through differential splicing and
differ exclusively within their carboxyl terminal
domains [30,31]. Whilst the biologic significance for
the existence of two TP isoforms in humans has not
been fully elucidated, there is extensive evidence that
they may be physiologically distinct thereby greatly
adding to the complexity of TXA
2
signalling in
humans [32]. For example, while both TPa and TPb
identically couple to PLCb, they differentially regulate
other secondary effectors including adenylyl cyclase
and tissue transglutaminase [33,34]; they undergo dif-
ferential homologous and heterologous desensitization
[35–38] and are also differentially expressed in a range
of cell ⁄ tissue types [39]. Moreover, recent studies have
established that TPa and TPb expression are actually
transcriptionally regulated by distinct promoters within
the single human TP gene located on chromosome 19
[40,41]. Whilst the originally identified promoter (Prm)
1 directs TPa expression, a novel promoter (Prm3) was
identified within the human TP gene that exclusively
directs TPb expression [40].
In view of the observations that PPARc activation
is associated with suppression of a number of key dis-
ease-associated genes within the vasculature including

that of the rat TP [22] coupled to the fact that there is
no significant sequence homology between the rat TP
promoter and human Prm1 or Prm3 sequences [22,41],
the aim of the current study was to investigate the
effect of 15d-PGJ
2
on expression of the human TP
gene within the platelet progenitor megakaryocytic
human erythroleukemic (HEL) 92.1.7 cell line. More-
over, in view of the existence of two independently
expressed TP isoforms in humans, it was also sought
to investigate whether 15d-PGJ
2
regulates TPa and ⁄ or
TPb expression in an isoform-dependent manner. It
was established that 15d-PGJ
2
suppresses both Prm3-
directed luciferase reporter gene expression and TPb
mRNA expression without affecting Prm1-directed
gene expression or TPa mRNA levels in HEL cells.
Moreover, we describe a novel mechanism of 15d-
PGJ
2
⁄ PPARc-mediated suppression of gene expression
involving the direct binding of the activated PPARc–
RXR heterodimer to a PPARc response element
(PPRE)–RXR response element within Prm3 resulting
in selective down-regulation of TPb mRNA expression.
These data provide further evidence for the role of

PPARc in the regulation of TP gene expression within
the vasculature and point to further critical differences
into the modes of regulation of TPa and TPb in
humans. Moreover, in view of the critical link between
the enhanced risk of cardiovascular disease in patients
with diabetes mellitus and in animal models of diabetes
mellitus associated with increased synthesis and TXA
2
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4755
action [42–45], these data point to an added benefit to
the current use of PPARc agonists in the treatment of
cardiovascular disease-associated diabetes.
Results
Analysis of the effect of 15d-PGJ
2
on Prm1-,
Prm2- and Prm3-directed gene expression
Previous studies have identified the presence of three
promoter regions, designated Prm1, Prm2 and Prm3,
within the single human TXA
2
receptor gene located
on chromosome 19p13.3 [40,41]. A schematic of the
human TP gene highlighting the positions of Prm1,
Prm2 and Prm3 relative to its translational start site
(ATG, designated +1) is presented in Fig. 1. Initially
the effect of the endogenous PPARc ligand 15d-PGJ
2
on Prm1-, Prm2- and Prm3-directed reporter gene

expression in transfected human erythroleukemic
92.1.7 (HEL) cells and, as a negative control, human
embryonic kidney (HEK) 293 cells was investigated.
Consistent with previous reports [39], Prm1, Prm2 and
A
B
Fig. 1. Effect of 15d-PGJ
2
on Prm1, Prm2 and Prm3-directed luciferase expression. (A and B) A schematic of the human TXA
2
receptor (TP)
genomic region spanning nucleotides )8500 to +786 encoding Prm1, Prm2 and Prm3 in addition to exon (E) 1, E1b and E2 are illustrated
above each panel where nucleotide +1 corresponds to the translational start site (ATG) and nucleotides 5¢ of that site are given a –designa-
tion. Recombinant pGL3Basic plasmids encoding Prm1 ()8500 to )5895), Prm2 ()3308 to )1979), Prm3 ()1394 to +1) or, as a control,
pGL2Control were transiently cotransfected along with pRL-TK into HEL 92.1.7 (A) and HEK 293 (B) cells. Thirty-six h post-transfection, cells
were incubated with either 15d-PGJ
2
(10 lM) or the vehicle [0.1% (v ⁄ v) dimethylsulfoxide] for 16 h. Mean firefly relative to renilla luciferase
activity is expressed in arbitrary relative luciferase units (RLU ± SEM; n ¼ 5). The asterisk (*) indicates that Prm3-directed luciferase activity
in HEL cells was significantly reduced in 15d-PGJ
2
treated cells relative to vehicle treated cells; *P < 0.05.
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4756 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
Prm3 each directed luciferase activity in both HEL
and HEK 293 cells, albeit at significantly different
levels relative to each other (Fig. 1). Pre-incubation of
HEL cells with 10 lm 15d-PGJ
2
for 16 h resulted in a

1.5-fold reduction in Prm3-directed luciferase expres-
sion (P<0.05) but had no significant effect on either
Prm1- or Prm2-directed luciferase expression (Fig. 1A).
Moreover, 15d-PGJ
2
suppressed Prm3-directed luci-
ferase expression in a concentration- (Fig. 2C) and
time-dependent (Fig. 2D) manner but had no signifi-
cant effect on Prm1- (Fig. 2A,B) or Prm2-directed
(data not shown) reporter gene expression regardless
of the concentration (0–40 lm) or incubation time
(0–24 h). In control HEK 293 cells, 15d-PGJ
2
had no
significant effect on either Prm1-, Prm2- or Prm3-
directed luciferase expression (Fig. 1B).
Effect of 15d-PGJ
2
on TPa and TPb mRNA and
protein expression in HEL cells
As previously stated, the TPa and TPb isoforms of the
human TXA
2
receptor (TP) are under the transcrip-
tional control of distinct promoters, namely Prm1 and
Prm3, respectively [40]. Hence, in view of the finding
herein that 15d-PGJ
2
significantly suppressed Prm3-
but not Prm1-directed reporter gene expression the

effect of 15d-PGJ
2
on TPa and TPb mRNA expression
in HEL cells was investigated. Consistent with previ-
ous reports [39], RT-PCR followed by Southern blot
analysis confirmed expression of TPa,TPb and gly-
ceraldehyde 3¢ phosphate dehydrogenase (GAPDH)
mRNA in HEL cells (Fig. 3A,B, lanes 1–3, respect-
ively) with an approximately twofold higher level of
TPa relative to TPb mRNA expression. Pre-incubation
with 15d-PGJ
2
had no significant effect on the levels of
either TPa or GAPDH mRNA expression in HEL
cells relative to the vehicle-treated cells (Fig. 3A–C). In
contrast, preincubation with 15d-PGJ
2
resulted in a
1.62-fold reduction in TPb mRNA expression in HEL
cells compared to vehicle-treated cells (Fig. 3A–C).
These data correlate well with the observed effect of
15d-PGJ
2
on Prm3 activity and provides further evi-
dence for a distinct role for Prm3 in the regulation of
TPb expression.
To assess the affect of 15d-PGJ
2
on the overall level
of TP protein expression and function, HEL cells were

preincubated with 10 lm 15d-PGJ
2
for 24 and 48 h
and its affect on TP-radioligand binding and on
AB
CD
Fig. 2. Concentration- and time-dependent effect of 15d-PGJ
2
on Prm1 and Prm3-directed luciferase expression. HEL 92.1.7 cells were tran-
siently cotransfected with pRL-TK along with pGL3b:Prm1 (A and B) or pGL3b:Prm3 (C and D). Thirty-six hours post-transfection, cells were
incubated for 16 h with 0–40 l
M 15d-PGJ
2
(A and C) or for 0–24 h with 10 lM 15d-PGJ
2
(B and D). Mean firefly relative to renilla luciferase
activity is expressed in arbitrary relative luciferase units (RLU ± SEM; n ¼ 5). The asterisks (*) indicate the concentration or time that Prm3-
directed luciferase activity was significantly reduced in 15d-PGJ
2
treated HEL cells relative to vehicle treated cells; ***P £ 0.001.
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4757
TP-mediated intracellular calcium ([Ca
2+
]
i
) mobiliza-
tion, in response to the selective TXA
2
mimetic

U46619, was investigated. In addition, as a control, we
also investigated the effect of 10 lm 15d-PGJ
2
on
signalling by an unrelated receptor, namely the EP
1
subtype of the prostaglandin (PG)E
2
receptor. Pre-
incubation of HEL cells with 15d-PGJ
2
for 24 h signi-
ficantly reduced the overall level of TP expression from
58.8 ± 8.2 fmol [
3
H]SQ29,548Æmg cell protein
)1
(n ¼
8) to 17.0 ± 4.3 fmol [
3
H]SQ29,548Æmg cell protein
)1
(n ¼ 11; P ¼ 0.0001). Moreover, 15d-PGJ
2
preincuba-
tion significantly reduced the overall level of U46619-
mediated [Ca
2+
]
i

mobilization from 23.7 ± 4.2 nm
[Ca
2+
]
i
to 8. 0 ± 0.82 nm (P ¼ 0.01), as illustrated in
Fig. 3D,E. Similar data was obtained following 48 h
incubation with 15d-PGJ
2
(data not shown). On the
other hand, 15d-PGJ
2
(10 lm, 48 h) did not signifi-
cantly affect [Ca
2+
]
i
mobilization by the control EP
1
receptor in response to its agonist 17 phenyl trinor
PGE
2
(compare D[Ca
2+
]
i
¼ 150.9 ± 21.9 nm, n ¼ 3
vs. D[Ca
2+
]

i
¼ 176.0 ± 9.8 nm, n ¼ 5 in vehicle- and
15d-PGJ
2
-treated cells, respectively; P ¼ 0.255).
Examination of the role of PPARc2, PPARc3 and
RXRa in 15d-PGJ
2
mediated inhibition of Prm3
activity
PPARc ⁄ retinoic X receptor (RXR) heterodimerization
has been shown to be an important step in mediating
the effect of 15d-PGJ
2
in a number of cell types.
Hence, to investigate whether PPARc ⁄ RXR hetero-
dimers might be involved in regulating Prm3, the effect
of expression of either PPARc2 (Fig. 4A) or PPARc3
(Fig. 4B) alone or coexpression of PPARc2 ⁄ PPARc3
along with RXR a on 15d-PGJ
2
regulation of Prm3-
directed reporter gene expression was investigated
AB C
E
D
Fig. 3. Effect of 15d-PGJ
2
on TPa and TPb mRNA expression and TP-mediated intracellular signalling. (A and B) RT-PCR analysis of RNA iso-
lated from HEL cells preincubated for 16 h with the vehicle 0.1% (v ⁄ v) dimethylsulfoxide (lanes 1–3) or 10 l

M 15d-PGJ
2
(lanes 4–6) using
primers to amplify TPa (lanes 1 and 4), TPb (lanes 2 and 5) and GAPDH (lanes 3 and 6) mRNA sequences. (B) Southern blot analysis of the
RT-PCR products (lanes 1–6) coscreened using
32
P-radiolabelled oligonucleotide probes specific for TPa ⁄ TPb mRNA and GAPDH mRNA
sequences. (C) Mean levels of TPa,TPb and GAPDH mRNA expression in 15d-PGJ
2
-treated HEL cells were represented as a percentage of
their expression in vehicle-treated cells (Relative expression,% ± SEM, n ¼ 4). The asterisks (*) indicate that the level of TPb mRNA expres-
sion in HEL cells was significantly reduced in 15d-PGJ
2
treated cells relative to vehicle treated cells; **P £ 0.02. (D and E) HEL 92.1.3 cells
were preincubated for 24 h with the vehicle 0.1% dimethylsulfoxide (D) or with 10 l
M 15d-PGJ
2
(E) prior to stimulation with 1 lM U46619,
added at the times indicated by the arrows. Data presented are representative profiles from at least four independent experiments and are
plotted as changes in intracellular Ca
2+
mobilization (D[Ca
2+
]
i
,nM) as a function of time (second, s). Actual mean changes in U46619-medi-
ated [Ca
2+
]
i

mobilization (nM ± SEM) were as follows: D[Ca
2+
]
i
¼ 23.7 ± 4.2 nM for vehicle treated cells (n ¼ 6); D[Ca
2+
]
i
¼ 8. 0 ± 0.82 nM
for 10 lM 15d-PGJ
2
treated cells (n ¼ 4).
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4758 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
(Fig. 4). Consistent with previous data, preincubation
of HEL 92.1.7 cells with 15d-PGJ
2
resulted in a 1.5-
fold suppression of Prm3 activity (Fig. 4). Using
anova one way comparisons, it was established that
relative to cells transfected with the empty vector,
expression of either PPARc2 or RXRa alone
(Fig. 4A), or PPARc3 or RXRa alone (Fig. 4B) had
no significant effect on the ability of 15d-PGJ
2
to sup-
press Prm3-activity (P ¼ 0.2196 and p ¼ 0.2235; Fig. 4,
respectively). However, coexpression of PPARc2 with
RXRa significantly augmented 15d-PGJ
2

-suppression
of Prm3 activity relative to cells cotransfected with the
vector alone or vector encoding PPARc2(P<0.0014)
or encoding RXRa (P<0.0014) yielding a 2.2-fold
reduction in luciferase expression relative to vehicle
treated cells (Fig. 4A). Similarly, coexpression of
PPARc3 with RXRa augmented 15d-PGJ
2
-suppression
of Prm3 activity relative to cells cotransfected with the
vector alone or vector encoding PPARc2(P<0.0009)
or encoding RXRa (P<0.0005) yielding a 2.4-fold
reduction in luciferase expression relative to vehicle
treated cells (Fig. 4B). Western blot analysis confirmed
over-expression of PPARc and RXRa in transfected
HEL cells (data no shown). Hence, these data are
suggestive that both PPARc and RXRa transcription
factors may be required to mediate 15d-PGJ
2
-inhibi-
tion of Prm3-directed gene expression.
Localization of the site of action of 15d-PGJ
2
within Prm3 by 5¢ deletion analysis
Thereafter, 5¢ deletion analysis of Prm3 ()1394 to +1)
was used to localize the key regulatory domains direct-
ing 15d-PGJ
2
-inhibition of Prm3 within HEL cells.
Consistent with previous reports [46], 5¢ deletion of

Prm3 to generate a )404 subfragment did not signifi-
cantly affect the level of basal (nonstimulated) luci-
ferase activity in HEL cells (Fig. 5A). However, 5¢
deletion of Prm3 from a )404 to a )320 bp fragment
yielded an approximately twofold increase in basal lu-
ciferase activity while 5¢ deletion of the )320 bp to a
)154 bp fragment did not yield a further alteration
in basal luciferase expression. These data suggest that
the )404 to )320 region contains negative regulatory
element(s), the removal of which results in increased
basal Prm3 activity whilst nucleotides located between
)320 and )154 do not significantly affect basal Prm3
activity [46].
Pre-incubation with 15d-PGJ
2
resulted in approxi-
mately 1.4–1.6-fold reductions in luciferase activity
directed by Prm3 (P<0.05) and the )404 (P<0.05)
and )320 (P<0.01) subfragments (Fig. 5A). How-
ever, 15d-PGJ
2
did not significantly affect luciferase
activity directed by the )154 subfragment of Prm3
(Fig. 5A). Hence, these data indicate that the )320 bp
subfragment contains 15d-PGJ
2
regulatory domain(s)
A
B
Fig. 4. The effect of co-expression of RXRa with either hPPARc2

or hPPARc3 on 15d-PGJ
2
-mediated inhibition of Prm3-directed luci-
ferase expression. HEL 92.1.7 cells were transiently cotransfected
with pGL3b:Prm3 (1 lg) together pSG5-hPPARc2 plus pSG5 (1 lg
each), pSG5-mRXRa plus pSG5 (1 lg each), or pSG5-hPPAR c 2 plus
pSG5-mRXRa (1 lg each) in the presence of 200 ng pRL-TK (A).
Alternatively, HEL cells were transiently cotransfected with
pGL3b:Prm3 (1 lg) together pcDNA3-hPPARc3 plus pcDNA3 (1 lg
each), pSG5-mRXRa plus pSG5 (1 lg each), or pcDNA3-hPPARc3
plus pSG5-mRXR a (1 lg each) in the presence of 200 ng pRL-TK
(B). Thirty-six hours post-transfection, cells were incubated for 16 h
with 10 l
M 15d-PGJ
2
(Panels A and B). Mean firefly relative to renil-
la luciferase activity is expressed in arbitrary relative luciferase units
(RLU ± SEM; n ¼ 5). The asterisks (*) indicate that Prm3-directed
luciferase activity in HEL cells was significantly reduced in 15d-
PGJ
2
treated cells relative to vehicle treated cells; *P £ 0.05,
**P £ 0.02, ***P £ 0.001, ****P £ 0.0001, respectively.
ANOVA one
way analysis was carried out to determine differences due to over-
expression of plasmids encoding hPPARc2 plus RXRa, or hPPARc2
plus RXRa relative to their expression alone or along with the
empty vector.
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4759

and that the site of action of 15d-PGJ
2
may be located
between )320 and )154. Consistent with the former,
the effect of 15d-PGJ
2
on luciferase expression directed
by the )320 bp subfragment was concentration- and
time-dependent (Fig. 5B,C).
Bioinformatic analysis of Prm3, using the matin-
spector
TM
programme [47], for transcription factor
elements between )320 and )154 identified the pres-
ence of 4 putative retinoic acid X receptor (RXR)
half sites centred at )300, )268, )175 and )148 and
two putative PPARc half sites at )182 and )159,
respectively (Fig. 6A). Hence, further 5¢ deletion ana-
lysis was carried out to investigate if any of the latter
sites might be involved in mediating 15d-PGJ
2
-inhibi-
tion of Prm3-directed gene expression. Progressive 5¢
deletion of Prm3 from the )320 bp fragment to yield
)276, ) 229 and )192 subfragments did not affect
their ability to direct basal luciferase expression in
HEL cells (Fig. 6A). While, consistent with previous
data, 15d-PGJ
2
did not significantly affect luciferase

activity directed by the )154 subfragment of Prm3
(Figs 5A and 6A), it resulted in approximately 1.6
fold reductions in luciferase activity directed by the
)276 (P<0.0009), )229 (P<0.006) and )192
(P<0.007) subfragments (Fig. 6A). These data indi-
cate that the RXR half sites centred at )300 (RXR
I) and )268 (RXR II) do not play a role in 15d-
PGJ
2
-inhibition of Prm3 and that the functional regu-
latory element(s) may be located between nucleotides
)192 and )154 within Prm3 or the surrounding
sequences.
To investigate whether PPARc ⁄ RXR regulation of
Prm3 is mediated by direct nuclear factor binding to
cis-acting elements within the )192 to )154 region of
Prm3, electromobility shift assays (EMSAs) were car-
ried out using a radiolabelled double stranded (ds)
DNA probe spanning nucleotides )198 to )150
(PPARc ⁄ RXR probe A; Kin242) and nuclear extracts
prepared from either vehicle and 15d-PGJ
2
treated
HEL 92.1.7 cells. A diffuse protein:DNA complex was
observed following incubation of the c
32
P-radiolabelled
double stranded (ds) PPARc ⁄ RXR probe A with either
A
BC

Fig. 5. Localization of the site of action of 15d-PGJ
2
within Prm3. (A) Recombinant pGL3 basic plasmids encoding Prm3 ()1394 to +1),
Prm3a ()404 to +1), Prm3ab ()320 to +1) and Prm3aa ()154 to +1) were transiently cotransfected along with pRL-TK into HEL 92.1.7 cells.
Thirty-six h post-transfection, cells were incubated with either 15d-PGJ
2
(10 lM) or the vehicle (0.1% dimethylsulfoxide) for 16 h. Mean fire-
fly relative to renilla luciferase activity is expressed in arbitrary relative luciferase units (RLU ± SEM; n ¼ 5). The asterisks (*) indicate that
luciferase expression in HEL cells was significantly reduced in 15d-PGJ
2
treated cells relative to vehicle treated cells; *P £ 0.05, **P £ 0.02,
respectively. (B and C) HEL 92.1.7 cells were transiently cotransfected with pRL-TK along with pGL3b:Prm3ab. Thirty-six hours post-transfec-
tion, cells were incubated for 16 h with 0–40 l
M 15d-PGJ
2
(B) or for 0–24 h with 10 lM 15d-PGJ
2
(C). Mean firefly relative to renilla luci-
ferase activity is expressed in arbitrary relative luciferase units (RLU ± SEM; n ¼ 5).
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4760 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
nuclear extract prepared from vehicle (Fig. 6B; lane 2)
or 15d-PGJ
2
-treated HEL cells (Fig. 6B; lane 3). The
formation of the latter nuclear factor-DNA complex
was efficiently competed by an excess of the corres-
ponding nonlabelled ds PPARc ⁄ RXR oligonucleotide
using nuclear extracts prepared from both vehicles or
15d-PGJ

2
treated-HEL cells (Fig. 6B; lanes 4 and 5,
respectively). The specificity of nuclear factor binding
to the latter PPARc ⁄ RXR probe A (spanning nucleo-
tides )198 to )150) was also verified by the failure of
excess ds oligonucleotides based on consensus AP-1,
Oct )1 and Sp1 elements to effectively inhibit the
formation of the nuclear factor-DNA complex using
nuclear extract prepared from either vehicle (Fig. 6B;
A
B
Fig. 6. Sub-localization of the site of action of 15d-PGJ
2
within Prm3. (A) A schematic of the TP genomic region encoding Prm3 ()1394 to
+1) in addition to exon (E) 2 spanning nucleotides )1394 to +786 are illustrated. In addition, the relative positions of a two putative PPARc-
responsive elements (PPREs), designated PPARc(a) and PPARc(b), respectively, and four putative retinoic acid X receptor (RXR) responsive
elements, designated RXR I–RXR IV, respectively, are also illustrated. Recombinant pGL2Basic plasmids encoding Prm3ab ()320 to +1),
Prm3aba ()276 to +1), Prm3abb ()229 to +1), Prm3abc ()192 to +1) and Prm3aa ()154 to +1) were transiently cotransfected along with
pRL-TK into HEL 92.1.7 cells Thirty-six hours post-transfection, cells were incubated with either 15d-PGJ
2
(10 lM) or the vehicle [0.1% (v ⁄ v)
dimethylsulfoxide] for 16 h. Mean firefly relative to renilla luciferase activity are expressed in arbitrary relative luciferase units (RLU ± SEM;
n ¼ 5). The asterisks (*) indicate that luciferase expression in HEL cells was significantly reduced in 15d-PGJ
2
treated cells relative to vehicle
treated cells, where *P £ 0.05, **P £ 0.02, ***P £ 0.001, ****P £ 0.0001, respectively. (B) A c
32
P-radiolabelled ds DNA probe correspond-
ing to nucleotides (nucleotides) )198 to )150 of Prm3 (PPARc ⁄ RXR probe A; Kin242 and its complement) was used in EMSAs using nuclear
extracts prepared from HEL 92.1.7 cells preincubated with either 15d-PGJ

2
(10 lM) or the vehicle (0.1% dimethylsulfoxide) for 16 h, as out-
lined in Experimental procedures. Lane 1, PPARc ⁄ RXR probe A incubated without nuclear extract; lanes 2 and 3, PPARc ⁄ RXR probe A incu-
bated with nuclear extract from vehicle- and 15d-PGJ
2
-treated HEL cells, respectively; lanes 4 and 5, PPARc ⁄ RXR probe A incubated with
nuclear extract from vehicle- and 15d-PGJ
2
-treated HEL cells, respectively, in the presence of excess nonlabelled specific ds competitor
oligonucleotide (Kin242 and its complement); lanes 6 and 7, PPARc ⁄ RXR probe A incubated with nuclear extract from vehicle- and 15d-
PGJ
2
-treated HEL cells, respectively, in the presence of excess nonlabelled ds noncompetitor oligonucleotide (Kin189, corresponding to
nucleotides )32 to )10 of Prm3 containing an AP1 consensus element); lanes 8 and 9, PPARc ⁄ RXR probe A incubated with nuclear extract
from vehicle- and 15d-PGJ
2
-treated HEL cells, respectively, in the presence of excess nonlabelled ds noncompetitor oligonucleotide (Kin195,
corresponding to nucleotides )115 to )92 of Prm3 containing an Oct1 ⁄ 2 consensus element); lanes 10 and 11, PPARc ⁄ RXR probe A incuba-
ted with nuclear extract from vehicle- and 15d-PGJ
2
-treated HEL cells, respectively, in the presence of excess nonlabelled ds noncompetitor
oligonucleotide (Sp1 consensus element, Promega). DNAÆprotein complexes were subject to polyacrylamide gel electrophoresis followed by
autoradiography, as outlined in Experimental procedures.
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4761
lanes 6, 8 and 10, respectively) or 15d-PGJ
2
treated
HEL cells (Fig. 6B; lane 7, 9 and 11, respectively.
Taken together these data demonstrate the specific

binding of nuclear factors within HEL cells to the )198
to )150 region of Prm3 and also show that nuclear
factor–DNA complex formation is independent of
15d-PGJ
2
stimulation.
Identification and characterization of a PPARc
response element within Prm3
Detailed analysis of the )192 ⁄ )154 region of Prm3
reveals the presence of two putative PPAR response
elements (PPREs). The first putative PPRE is com-
posed of the PPARc half site centred at )182
[PPARc(a)] and a RXR half site at )175 (RXR III)
while the second corresponds to a PPARc half site
centred at )159 [PPARc(b)] and a RXR IV half site
at )148. Therefore a combination of site-directed and
deletion mutagenesis was employed to investigate the
functional importance of these putative elements in
directing 15d-PGJ
2
regulation of Prm3 activity. Ini-
tially it was confirmed that neither site-directed muta-
genesis nor deletion of nucleotides between )320 and
)154 significantly affected basal luciferase gene expres-
sion in HEL cells (Fig. 7A). Disruption of the PPARc
(a) half site centred at )182 by site-directed mutagen-
esis (mutation of TTGAGC to TTAGGC, mutated
nucleotides in bold) within the )320 bp subfragment
of Prm3 did not significantly affect the level of
15d-PGJ

2
-inhibition of Prm3 activity (Fig. 7A). More-
over, progressive 5¢deletion of nucleotides surrounding
either the PPARc (a) and RXR III half sites to yield
the )186 and )175 subfragments did not affect
15d-PGJ
2
-suppression of luciferase reporter gene
expression yielding between 1.5- and 1.7-fold reduc-
tions of luciferase expression (Fig. 7A). Consistent
with the latter, complete deletion of the PPARc (a)
and RXR III half sites while retaining the PPARc (b)
and RXR IV half sites generated the )170 subfrag-
ment that was fully responsive to 15d-PGJ
2
-suppres-
sion of luciferase expression. On the other hand, either
deletion of the latter PPARc (b) and RXR IV half
sites, such as within the )154 subfragment, or dis-
ruption of the RXR IV half site centred at )148 by
site-directed mutagenesis (sequence AGTTCA to
ATTTTA) within the )320 bp subfragment abolished
15d-PGJ
2
-suppression of Prm3 activity (Fig. 7A).
Thereafter, EMSAs were carried out to investigate
direct nuclear factor binding to the latter cis-acting
PPARc ⁄ RXR response elements within the )161 to
)148 region of Prm3 using a radiolabelled ds DNA
probe spanning )168 to )141 (PPARc ⁄ RXR probe B;

Kin281) and nuclear extracts prepared from either
vehicle- and 15d-PGJ
2
-treated HEL 92.1.7 cells. A dif-
fuse protein:DNA complex was observed following
incubation of the c
32
P-radiolabelled PPARc ⁄ RXR
probe B with either nuclear extracts prepared from
vehicle- (Fig. 7B; lane 2) or 15d-PGJ
2
-treated (Fig. 7B;
lane 6) HEL cells. Nuclear factor-DNA complex
formation using nuclear extracts prepared from both
vehicle-treated (Fig. 7B; lane 3) or 15d-PGJ
2
treated-
(Fig. 7B; lane 7) HEL cells was competed by a 50-fold
excess of the corresponding nonlabelled ds PPARc ⁄
RXR oligonucleotide, respectively, and by a ds oligo-
nucleotide containing a consensus PPARc response
element derived from the acyl-CoA oxidase gene
(Fig. 7B; lanes 4 and 8). On the other hand, nuclear
factor binding to the PPARc ⁄ RXR probe B (spanning
nucleotides )168 to )141) was not competed by an
excess of ds oligonucleotides in which both the
PPARc(b) plus RXR IV sites were mutated, using
nuclear extract prepared from either vehicle- (Fig. 7B;
lane 5) or 15d-PGJ
2

-treated (Fig. 7B; lane 9) HEL
cells, respectively. Taken together these data demon-
strate that either mutation or deletion of the PPARc
(a) ⁄ RXR IV half sites centred at )159 and )148,
respectively, abolishes 15d-PGJ
2
–suppression of Prm3
directed gene expression. Moreover, data from EMSAs
have confirmed the specific binding of nuclear factors
from HEL cells to the )168 to )141 region of Prm3
and, consistent with previous data (Fig. 6B), confirm
that nuclear factor-DNA complex formation is largely
independent of 15d-PGJ
2
stimulation.
Examination of PPARc ⁄ RXRa interactions within
the PPRE
To further investigate the identity and specificity of the
DNA ⁄ protein interactions within the PPRE centred at
)159 and )148 of Prm3, we examined the ability of
recombinant human (h) PPARc2 and ⁄ or mouse (m)
RXRa to directly bind to the latter cis-acting
PPARc ⁄ RXR response elements. Hence, the PPARc2
and ⁄ or RXRa transcription factors were transcribed
and translated in vitro in a rabbit reticulocyte lysate
cell free system and proteins of approximately
53–54 kDa corresponding to PPARc2 and RXRa were
readily detectable by SDS ⁄ PAGE following translation
in the presence of [
35

S]methionine (Fig. 8A, lanes 2
and 3). Thereafter, the ability of the translated
PPARc2 and ⁄ or RXRa factors to bind to the c
32
P-
radiolabelled ds DNA probe spanning )168 to )141
(PPARc ⁄ RXR probe B; Kin281) was investigated.
The recombinant PPARc2 or RXRa transcription
factors exhibited weak, though detectable, binding to
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4762 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
the radiolabelled PPARc ⁄ RXR probe B (Fig. 8B,
lanes 2 and 3, respectively). However, coincubation of
PPARc2 with RXRa significantly augmented trans-
cription factor binding to the PPARc ⁄ RXR probe B
(Fig. 8B, lane 4) indicating that both PPARc2 and
RXRa are required for efficient transcription factor
binding. Moreover, PPARc:RXR complex binding to
the latter probe was efficiently competed by an excess
of the corresponding nonlabelled ds PPARc ⁄ RXR
oligonucleotide (Fig. 8B, lane 5) but was not competed
by an excess of ds oligonucleotides in which the
PPARc(b) site, the RXR IV site or the PPARc(b) plus
A
B
Fig. 7. Identification of the site of action 15d-PGJ
2
site of action within Prm3 by site-directed and deletion mutagenesis. (A) Recombinant
pGL3Basic plasmids encoding Prm3ab ()320 to +1), Prm3abc ()192 to +1), Prm3abe ()186 to +1), Prm3abf ()175 to +1), Prm3abd ()170
to +1), Prm3aa ()154 to +1) or the site-directed variants Prm3ab

PPARc(a)
* or Prm3ab
RXRIV
*, where the PPARc(a) and RXR IV half sites within
Prm3ab were mutated, were transiently cotransfected along with pRL-TK into HEL 92.1.7. Thirty-six hours post-transfection, cells were incu-
bated with either 15d-PGJ
2
(10 lM) or the vehicle [0.1% (v ⁄ v) dimethylsulfoxide] for 16 h. Mean firefly relative to renilla luciferase activity is
expressed in arbitrary relative luciferase units (RLU ± SEM; n ¼ 5). The asterisks (*) indicate that luciferase expression in HEL cells was sig-
nificantly reduced in 15d-PGJ
2
treated cells relative to vehicle treated cells; **P £ 0.02. (B) EMSAs were carried out using a c
32
P-radio-
labelled ds DNA probe (Kin281 and its complement corresponding to nucleotides )168 to )141 of Prm3)and nuclear extract (4 lg) prepared
from vehicle- (lanes 2–5) or 15d-PGJ
2
– (10 lM; lanes 6–9) preincubated HEL 92.1.7 cells as described in the Experimental procedures sec-
tion. The c
32
P-radiolabelled probe was incubated: without nuclear extract (lane 1); with nuclear extract (lanes 2 and 6); with nuclear extract
in the presence of a 50-fold excess of: nonlabelled ds specific competitor oligonucleotide (Kin281 and its complement, lanes 3 and 7); non-
labelled ds oligonucleotide containing a consensus acyl coA oxidase PPARc response element (Kin342 and its complement, lanes 4 and 8);
nonlabelled ds noncompetitor oligonucleotide (Kin289 and its complement, corresponding to nucleotides )168 to )141 of Prm3 in which
both the PPRE
PPARc(b)
* and RXR
IV
half-site half-sites were disrupted by site-directed mutagenesis (lanes 5 and 9). DNA ⁄ nuclear factor com-
plexes were subject to polyacrylamide gel electroporesis followed by autoradiography, as outlined in Experimental procedures section.

A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4763
RXR IV sites were mutated (Fig. 8B; lanes 6–8,
respectively).
The fidelity of PPARc:RXRa binding to the PPRE
was further demonstrated by examining their binding
to a c
32
P-labelled ds oligonucleotide spanning the
)168 ⁄ )141 region of Prm3 in which both the
PPARc(b) (sequence GGTTGT to TTCTGT) and
RXR IV (sequence AGTTCA to ATTTTA) sites are
disrupted by mutagenesis (PPARc ⁄ RXR probe C;
Kin289). No DNA ⁄ protein binding or complex forma-
tion was observed between either PPARc, RXRa alone
or PPARc plus RXRa to the radiolabelled
PPARc ⁄ RXR probe C (Fig. 8C). These data confirm
that PPAR c binds to the PPRE within Prm3 as a het-
erodimer with RXR and that both the PPARc(b) and
RXR IV half sites centred at )159 and )148, respect-
ively, are required for binding.
Discussion
TXA
2
is a potent mediator of platelet activation and
aggregation and a constrictor of vascular and bron-
chial smooth muscle and together with prostacyclin,
for example, it plays a key role in the maintenance of
haemostasis [23]. TXA
2

also induces a diversity of
other actions and is widely implicated as a mediator in
a range of vascular, thrombotic and inflammatory dis-
eases [24–29]. In humans, the actions of TXA
2
are
mediated by two receptor isoforms termed TP a and
TPb [30,31,41]. Whilst the significance of two receptors
for TXA
2
in humans but not in other species is cur-
rently unknown, there is increasing evidence that they
are physiologically distinct displaying differences in
their basic mechanisms of intracellular signalling,
modes of desensitization and patterns of expression
[32–39]. Moreover, recent studies established that
expression of TPa and TPb are under the transcrip-
tional control of two distinct promoters within the sin-
gle TP gene whereby the originally described promoter
(Prm)1 regulates TPa expression and the recently iden-
tified Prm3 regulates TPb expression [40,41].
In the present study we have demonstrated that the
cyclopentanone prostaglandin 15d-PGJ
2
inhibits the
activity of Prm3 in the platelet-like megakaryocytic
human erythroleukemic (HEL) 92.1.7 cell line in a
concentration and time dependent manner but had no
effect on Prm1- or Prm2-directed reporter gene expres-
sion in HEL cells. Moreover, the effect of 15d-PGJ

2
on Prm1 and Prm3 correlated with its effect on TPa
and TPb mRNA expression in HEL 92.1.7 cells yield-
ing 1.5-fold reductions in both Prm3-directed reporter
gene expression and TPb mRNA expression while
AB C
Fig. 8. Investigation of PPARc ⁄ RXRa interactions within the putative PPAR
PPARc(b)
⁄ RAR
IV
site within Prm3. (A) in vitro transcripts of
hPPARc2 (lane 2) and mRXRa (lane 3) were translated in vitro in the presence of [
35
S]methionine, where translations carried out in the
absence of exogenous RNA acted as a control (lane 1). The arrow to the right of (A) indicates the position of the hPPARc2 (lane 2) and
mRXRa (lane 3) in vitro translated products. (B and C) for EMSAs, parallel in vitro translations of either hPPARc2, mRXRa, hPPARc2 plus
mRXRa transcripts or, as controls, without added exogenous RNA were carried out where [
35
S]methionine was replaced with an equivalent
concentration of methionine, as outlined in Experimental procedures. Thereafter, c
32
P-radiolabelled ds DNA probes corresponding to
PPARc ⁄ RXR probe B (Kin281 and its complement; B) or, as a control, PPARc ⁄ RXR probe C (Kin289 and its complement in which the both
the PPARc(b) and RXR IV half sites were mutated; (C) were used in EMSAs using in vitro translations carried out without added RNA (B and
C; lane 1); with hPPARc2 RNA (B and C, lane 2); with mRXRa RNA (B and C, lane 3); with hPPARc2 plus mRXRa RNA (B and C, lane 4);
with hPPARc2 plus mRXRa RNA in the presence of excess nonlabelled specific competitor ds Kin281 and its complement (B, lane 5) or
Kin289 and its complement (C, lane 5); with hPPARc2 plus mRXRa RNA in the presence of excess nonlabelled ds Kin285 and its comple-
ment (corresponding to nucleotides )168 to )141 of Prm3 in which the PPARc(b) half-site was disrupted; B and C, lane 6; with hPPARc2
plus mRXRa RNA in the presence of excess nonlabelled ds Kin283 and its complement (corresponding to nucleotides )168 to )141 of Prm3
in which the RXR IV half-site was disrupted; Panels B and C, lane 7); with hPPARc2 plus mRXRa RNA in the presence of excess nonlabelled

ds Kin289 and its complement (corresponding to nucleotides )168 to )141 of Prm3 in which the PPAR c (b) and RXR IV half-sites were dis-
rupted; B and C, lane (8). DNAÆprotein complexes were subject to polyacrylamide gel electrophoresis followed by autoradiography, as out-
lined in Experimental procedures.
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4764 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
showing no reduction in Prm1 activity or TPa mRNA
expression. The reduction in TPb mRNA expression in
response to 15d-PGJ
2
in HEL cells was reflected in a
corresponding reduction in the overall level of TP
expression, as assessed by radioligand binding studies,
and in reductions in TP-mediated [Ca
2+
]
i
mobilization,
in response to the TXA
2
mimetic U46619. On the
other hand, 15d-PGJ
2
having no significant effect on
signalling by the control EP
1
receptor in response to
its agonist 17 phenyl trinor PGE
2
. Furthermore, in
human embryonic kidney (HEK) 293 cells 15d-PGJ

2
had no effect on Prm1, Prm2 or indeed Prm3-activity
or TPa or TPb mRNA expression (data not shown).
The latter observation is consistent with previous
reports that HEK 293 cells do not express significant
levels of functional PPARc [48] and thereby implies
that PPARc isoforms may have a role in mediating
the 15d-PGJ
2
–inhibition of Prm3 activity. The
involvement of both PPARc and the retinoic acid X
receptor (RXR) in the regulation of Prm3 was
explored through coexpression studies whereby trans-
fection of either PPARc2 or PPARc3 along with the
RXRa isoform significantly augmented 15d-PGJ
2
-inhi-
bition of Prm3 activity while expression of either
PPARc2, PPARc3 or RXRa alone had no effect.
Hence, these data established that TPb, but not TPa,
expression may be down-regulated by the endogenous
PPARc agonist 15d-PGJ
2
through suppression of
Prm3-directed gene expression. The specific inhibition
of TPb mRNA expression mediated by PPARc is fur-
ther evidence that TPa and TPb are under the control
of distinct transcriptional regulatory mechanisms.
It was noteworthy that 10 lm 15d-PGJ
2

was
required for maximum inhibition of Prm3 activity, a
concentration that is likely to be considerably higher
that its circulating plasma concentration. However, it
is indeed likely that local concentration of the autocoid
15d-PGJ
2
, such as in inflammatory exudates, may be
substantially higher than circulatory levels and may
vary significantly depending on the (patho)physiologic
state. In addition, a number of PG transport systems
exist that may raise intracellular concentrations of
15d-PGJ
2
under certain conditions [49–51]. Interest-
ingly the highest in vivo concentrations of 15d-PGJ
2
is
present during the resolution phase of inflammation
[1], suggesting that TPb expression may be down-regu-
lated in inflammation or following vascular injury.
As stated, the majority of nuclear hormone receptors
up-regulate gene expression through heterodimeriza-
tion with the common retinoic acid X receptor (RXR)
and binding to specific response elements within target
gene promoters. The PPARc–RXR heterodimer is
reported to preferentially bind to the PPRE consisting
of a direct repeat of the sequence AGGTCA spaced by
one nucleotide (DR1) [52,53]. For example, the rat
acyl Co oxidase gene contains the first characterized

PPRE and has the sequence AGGACA a AGGTCA
[4]. However, in general PPREs are poorly conserved
exhibiting considerable sequence variation such as in
the case of previously characterized PPREs within
the lipoprotein lipase, apolipoprotein AII, enoyl-CoA
hydratase and 3-ketoacyl-CoA thiolase genes, as illus-
trated in Table 1.
In this study to locate the PPARc responsive
region(s) within Prm3 ()1394 to +1), progressive 5¢
deletions of sequences 5¢ of )320 had no significant
effect while further deletion to the )154 bp subfrag-
ment abolished 15d-PGJ
2
-suppression of Prm3. matin-
spector
tm
analysis [47] identified 4 putative retinoic
acid X receptor (RXR I–IV) and two putative PPARc
[PPARc (a) and PPARc(b)] half sites between )320
and )154. Additional 5¢deletion analysis established
that the RXR half sites centred at )300 (RXR I) and
)268 (RXR II) do not play a role in 15d-PGJ
2
-inhibi-
tion of Prm3 and implied that the functional PPARc
response element(s) (PPREs) may be located between
nucleotides )192 and )154 within Prm3 or the sur-
rounding sequences. Specific nuclear factor binding to
the latter region was confirmed by EMSA using the
radiolabelled PPARc ⁄ RXR probe A spanning nucleo-

Table 1. PPARc responsive gene. 5¢ PPARc and 3¢ RXR direct repeat hexameric sequences are in uppercase separated by 1–5 nucleotides
(DR1–DR5) given in lowercase.
PPARc regulated gene PPARc response element Reference
h.TPb receptor (Prm3) GGTTGT gtagg AGTTCA –
Acyl-CoA oxidase A AGGACA a AGGTCA [54]
Acyl-CoA oxidase B AGGTAG a AGGTCA [54]
Lipoprotein Lipase TGCCCT t TCCCCC [58]
Apolipoprotein AII CAACCT t TACCCT [68]
Enoyl-CoA hydratase GACCTA tt GAACTA t TACCTA [69]
3-Ketoacyl-CoA thiolase AGACCT t TGAACC [70]
Perilipin TCACCT t TCACCC [71]
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4765
tides )198 to )150 and nuclear extracts from vehicle-
and 15d-PGJ
2
-treated HEL 92.1.7 cells and this
binding was shown to be independent of 15d-PGJ
2
stimulation.
Thereafter, the precise site of action of 15d-PGJ
2
was identified by further 5¢ deletion analysis dissecting
the )192 to )154 region of Prm3. Deletion of the
PPARc (a) and RXR III half sites to generate the
)170 subfragment retained 15d-PGJ
2
-suppression of
Prm3 activity. On the other hand, either deletion of
the PPARc (b) and RXR IV half sites, such as in the

)154 subfragment, or site-directed mutagenesis of the
RXR IV within the )320 bp subfragment abolished
15d-PGJ
2
-suppression of Prm3 activity. Sequence ana-
lysis of the latter PPRE within Prm3 indicates that it
consists of two imperfect hexameric half sites including
the 5¢ PPARc(b) (sequence GGTTGT) and 3¢ RXR IV
(sequence AGTTCA) half site centred at )159 and at
)148, respectively, separated by five (DR5) nucleotides
(Table 1). Specific nuclear factor binding to the latter
PPRE was verified by EMSAs using the radiolabelled
ds PPARc ⁄ RXR probe B spanning )168 to )141 and
nuclear extracts prepared from either vehicle and 15d-
PGJ
2
treated HEL 92.1.7 cells. Furthermore, these
experiments showed that nuclear factor binding to
probe B was competed using excess unlabelled ds
oligonucleotide containing a conserved PPRE derived
from the acyl CoA gene [54]. Therefore, these data
suggest that despite the sequence divergence between
the PPREs derived from Prm3 and the acyl CoA gene,
both sequences are bound by PPARc ⁄ RXR heterodi-
mers in vitro. Similar to the EMSAs involving the
extended PPARc ⁄ RXR probe A, nuclear factor bind-
ing was independent of 15d-PGJ
2
stimulation. The lig-
and independent nature of the nuclear factor binding

to Prm3 is consistent with the general model of Dus-
sault and Froman, 2000 [55] whereby it is proposed
that PPAR c–RXR heterodimers bind to PPREs in the
absence or presence of ligand and that transcriptional
activation results from a ligand dependent conforma-
tional change in PPARc possibly leading to recruit-
ment of coactivator molecules [55]. The fact that in the
case of Prm3 that nuclear factor:DNA complex forma-
tion was independent of ligand and did not display a
substantially altered mobility shift in the pres-
ence ⁄ absence of ligand suggests that an altered confor-
mation, rather than recruitment of cofactors, may be
involved in mediating the ligand-dependent transcrip-
tional repression.
As previously stated, the interdependent nature of
PPARc:RXRa binding to the PPRE within Prm3 was
suggested by over-expression of PPARc2, PPARc3
and RXRa whereby the inhibitory effect of 15d-PGJ
2
on Prm3 activity was augmented by the coexpression
of either hPPARc2 or hPPARc3 along with RXRa in
HEL 92.1.7 cells but not by over-expression of the
PPARc2 ⁄ 3 or RXRa factors alone. However, over-
expression studies alone cannot definitively establish
the PPARc ⁄ RXRa heterdimers are actually required
to regulate Prm3 in response to 15d-PGJ
2
. Other
approaches such as functional knock-out through the
use of RNA

interference
(RNA
i
) to disrupt PPARc or
RXRa expression or chromosomal immunoprecipita-
tion (CHIP) analysis are also possible but have proven
technically difficult owing to the inability to transfect
HEL cells to high efficiency for RNA
i
and due to
the failure of CHIP analysis to discriminate between
the PPARc ⁄ RXR site located between )159–148 and
the adjacent site located between )182–175 (data not
shown). Hence, to extend these studies, the direct bind-
ing of and the specific requirement for both PPARc
and RXRa for efficient complex formation with the
PPRE centred at )159 and )148 within Prm3 was veri-
fied by examining the ability of in vitro translated
PPARc2 and ⁄ or RXRa factors to bind to the radiola-
belled PPARc ⁄ RXR probe B. Although some weak
binding was observed following incubation with either
the individual PPARc2 or RXRa proteins, binding
was substantially increased following coincubation
with both the PPARc and RXRa proteins with the
radiolabelled PPARc ⁄ RXR probe B. On the other
hand, neither the individual PPARc2 or RXRa pro-
teins nor the PPARc–RXRa heterodimer bound to
the radiolabelled PPAR c ⁄ RXR probe C spanning the
)168 ⁄ )141 region of Prm3 in which both the
PPARc(b) and RXR IV sites were disrupted by muta-

genesis. Collectively, these data confirm that PPARc
binds to the PPRE within Prm3 as a heterodimer with
RXR and that both the PPARc(b) and RXR IV half
sites centred at )159 and )148, respectively, are
required for binding and for 15d-PGJ
2
-inhibition of
Prm3 activity and TPb expression.
Although reports of PPARc-mediated inhibition, as
opposed to activation, of gene expression are becoming
increasingly common [16,56], transcriptional inhibition
is mainly documented to involve PPARc-mediated
trans-repression of a host of transcription factors inclu-
ding NF-jB, AP-1, STAT1, SP1 or sequestration of
coactivators CBP and ⁄ or p300 rather than involving
direct binding of the PPARc–RXR heterodimer to the
PPRE [12–14,19,22]. For example, PPARc activation
suppresses expression of the rat TXA
2
receptor (TP)
through direct interaction between PPARc and SP1
[22]. Hence, the study herein involving Prm3 of the
human TP gene represents the first reported study in
which the inhibitory effects of PPARc are mediated by
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4766 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
direct binding of the PPARc–RXRa heterodimer to a
PPRE. An extensive investigation of the interactions of
retinoic acid receptor (RAR)–RXR and PPAR–RXR
heterodimers with RAR response elements (RREs) and

PPREs, respectively, was carried out by DiRenzo et al.
[5]. They found that the ability of the RAR–RXR het-
erodimer to either activate or suppress transcription
was dependent on the spacing (so-called ‘DR spacing’)
between the hexameric half sites within the RRE itself.
Briefly, binding of RAR–RXR heterodimers to DR5
elements resulted in transcriptional activation, whereas
binding to DR1 elements resulted in transcriptional
repression [5,57]. In contrast, it is reported that tran-
scriptional activation mediated by the PPARc–RXR
heterodimer occurs following binding to DR1 contain-
ing PPREs [58]. Although the binding of PPARc–
RXR heterodimers to DR5 containing elements has
yet to be characterized, it is evident that the spacing
between the hexanucleotide motifs within nuclear hor-
mone response elements, such as the RREs, dramatic-
ally influence the transcriptional response mediated by
these elements and their specific transcription factor
complex [5]. Therefore, we propose that the ligand
activation of PPARc–RXR heterodimer bound to the
DR5 element centred at )159 and )148 within Prm3
results in transcriptional repression rather than the
transcriptional activation typically observed within the
classical DR1 containing PPREs. Whether transcrip-
tional repression of Prm3 by the PPARc–RXR also
involves recruitment of corepressors, such as such as
N-CoR [59], requires further investigation.
Epidemiological studies have shown that atheroscler-
osis accounts for some 80% of all deaths associated
with type 2 diabetes mellitus, of which roughly 75%

are attributable to coronary artery disease and the
remainder to cerebrovascular or peripheral vascular
complications [60]. Consistent with this, the enhanced
risk of cardiovascular disease in patients with diabetes
mellitus and in animal models of diabetes is associated
with both the increased synthesis and action of TXA
2
,
most likely due to increased TP expression [42–45].
The thiazolidinedione (TZD) class of insulin-sensitizing
drugs have proven to be a major therapeutic advance
in the treatment of type II diabetes. These drugs act as
potent agonists of PPARc and activation of this recep-
tor is central to the insulin sensitizing actions of TZDs
[61]. Recently an anti atherosclerotic role for PPARc
has been suggested due to the inhibitory effects
of PPARc activation on the expression of a number of
pro-atherosclerotic, pro-thrombotic and a range of
other critical vascular-related genes [16,17]. This is sup-
ported by clinical data whereby PPARc agonists have
been shown to lower blood pressure [62], prevent
cardiac mass increase and cardiac functional impair-
ment in diabetic patients [63]. Moreover, PPARc
ligands suppress gene expression of the human
cyclooxygenase (COX) II in epithelial cells [20], the rat
TXA
2
synthase in macrophages [21] and the rat TP
gene in VSMC [22]. In addition, troglitazone reduced
TXA

2
production and action in human platelets and
HEL cells [22,64]. Taken together, these findings have
led to the proposal that clinically PPARc ligands may
attenuate the synthesis and action of TXA
2
and, in
turn, may exert a beneficial effect on cardiovascular
complications in patients with diabetes mellitus. Hence,
the observed inhibitory effect of the endogenous
PPARc agonist 15d-PGJ
2
on TPb expression reported
herein may account for some of the observed beneficial
effects of PPARc agonists in the treatment of cardio-
vascular disease associated with diabetes mellitus and
is entirely consistent with the growing recognition of
the importance of PPARc and its agonists in the regu-
lation of those events. In addition, these data point to
further differences in the modes of transcriptional
regulation of the individual TPa and TPb isoforms in
humans and imply potentially important physiologic
differences between TPa and TPb such as during
inflammation, diabetes mellitus and associated cardio-
vascular disease.
Experimental procedures
Materials
pGL3Basic, pGL3Enhancer, pRL-Thymidine Kinase (pRL-
TK) and Dual LuciferaseÒ Reporter Assay System were
obtained from Promega Corporation (Madison, WI, USA).

[c
32
P]ATP (6000 CiÆ mmol
)1
at 10 mCiÆmL
)1
) was from
Valeant Pharmaceuticals (ICN) (Costa Mesa, CA, USA).
Anti-PPARc (sc-7273x) antibody was from Santa Cruz Bio-
technology (Santa Cruz, CA, USA). 15-deoxy-D
12,14
-PGJ
2
was from Calbiochem-Novabiochem (Nottingham, UK).
All other reagents were molecular biology grade. Mamma-
lian plasmids encoding human PPARc2 (pSG5-hPPARc2),
human PPARc3 (pcDNA3-hPPARc3) and mouse retinoic
acid X receptor a (pSG5-mRXRa) were a kind gift from
C. Haby (CNRS Inserm, France).
Construction of luciferase-based genetic reporter
plasmids
The plasmids pGL3b:Prm1, pGL3b:Prm2 and pGL3b:Prm3
containing promoter (Prm)1, Prm2 and Prm3 sequences
from the human TXA
2
receptor (TP) gene in pGL3Basic
reporter vector have been previously described [40]. In addi-
tion, pGL3b:Prm3a, pGL3b:Prm3ab and pGL3b:Prm3aa
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4767

containing subdeletions of Prm3 have been described [46].
To identify the PPARc responsive site within Prm3, a range
of 5¢ deletions were generated. Specifically, for all 5¢ dele-
tions, PCR fragments were generated using pGL3b:Prm3
as template and the antisense primer Kin113 (5¢-dAGAG
ACGCGTGGCTCCGGAGCCCTGAGGGATC-3¢, com-
plementary to nucleotides )19 to +3 where the underlined
sequence corresponds to an Mlu1 cloning site) in combina-
tion with specific sense primers. The following lists the
identities of the Prm3 gene fragments and the correspond-
ing plasmids generated in the vector pGL3Basic (annotated
pGL3b) and the identity of the specific sense primer, its
sequence and corresponding nucleotides where – designa-
tions indicate nucleotides 5¢ of the translational ATG start
codon which is designated +1 and the Kpn1 cloning site is
underlined.
1. Prm3aba; pGL3b:Prm3aba. (Primer Kin211, 5¢-dGAG
A
GGTACCGCTGCAGTGAGCCTTGA TTG-3¢, nucleo-
tides )276 to )257).
2 Prm3abb; pGL3b:Prm3abb. (Primer Kin212, 5¢-dGAG
A
GGTACCGAGCAAGACTCTGTCTC AAA-3¢, nucleo-
tides )229 to )209).
3. Prm3abc; pGL3b:Prm3abc. (Primer Kin236, 5¢-dGAG
AGGTACCCCGGAGAGGATATTTGAGCTG-3¢, nucleo-
tides )192 to )171).
4. Prm3abe; pGL3b:Prm3abe. (Primer Kin240, 5¢-dGA
GA
GGTACCAGGATATTTGAGCTGGGGCATTG-3¢,

nucleotides )186 to )163).
5. Prm3abf; pGL3b:Prm3abf. (Primer Kin241, 5¢-dGA
GA
GGTACCGCTGGGGCATTGAAGGTTGTGT-3¢,
nucleotides )175 to )153).
6. Prm3abd; pGL3b:Prm3abd. (Primer Kin237, 5¢-dGA
GA
GGTACCGGCATTGAAGGTTGTGTAGG-3¢, nucleo-
tides )170 to )150).
Each of the latter recombinant plasmids was verified by
double stranded DNA sequencing.
Site-directed mutagenesis
Site-directed mutagenesis was carried out using the Quik-
Change
TM
(Stratagene) method. Mutation of the consensus
PPARc-responsive elements (PPRE) half site, designated
PPRE
PPARc(a)
with the sequence TGAGCT to TAGGCT
centred at )182, within Prm3 was performed using the plas-
mid pGL3b:Prm3ab as template and mutator primers
Kin250 (5¢-dACCGGAGAGGATATTTAGGCTGGGGCA
TTGAAGGTTG-3¢; sense primer) vs. Kin251 (5¢-dCAA
CCTTCAATGCCCCAGCCTAAATATCCTCTCCGGT-3¢;
antisense primer) to generate pGL3b:Prm3ab
PPARc(a)
*.
Mutation of the consensus retinoic acid X responsive
element (RXR) half site with the sequence AGTTCA

to ATTTTA centred at )148 within Prm3 was per-
formed using the template pGL3b:Prm3ab in combination
with the primers Kin275 (5¢-dGAAGGTTGTGTAGG
ATTTTACCAGAGCTACTTACACTG-3¢; sense primer)
vs. Kin276 (5¢-dCAGTGTAAGTAGCTCTGGTAAA
ATCCT
ACACAACCTTC-3¢; antisense primer to generate
pGL3b:Prm3ab
RXRIV
*. Sequences corresponding to the
mutated bases are in bold. Each of the latter plasmids was
verified by double stranded DNA sequencing.
Cell culture
All mammalian cells were grown at 37 °C in a humid envi-
ronment with 5% (v ⁄ v) CO
2
. Human erythroleukemic
92.1.7 (HEL) cells and human embryonic kidney (HEK)
293 cells were cultured in RPMI 1640, 10% (v ⁄ v) foetal
bovine serum and in Eagle’s minimal essential medium
(MEM), 10% (v ⁄ v) foetal bovine serum, respectively.
Throughout the various assays, HEL cells were treated with
either the drug vehicle ([0.1% (v ⁄ v) dimethylsulfoxide] or
with 15d-PGJ
2
(0–40 lm; 0–24 h) and it was determined
that 80% of cells remained viable following treament under
maximum conditions. For all assays, only viable cells were
used.
Assay of luciferase activity

Genetic reporter assays were carried out using the Dual
Luciferase Assay System
TM
. HEK 293 cells were plated in
MEM, 10% (v ⁄ v) foetal bovine serum in six well dishes at
1 · 10
5
cells per well. At 70–80% confluence, cells were
cotransfected with pGL3Basic control vector, encoding fire-
fly luciferase, or its recombinant derivatives (0.4 lgÆwell
)1
)
along with pRL-TK (50 ngÆwell
)1
), encoding renilla luci-
ferase, using Effectene (Qiagen, Valencia, CA, USA) as
recommended by the supplier. Approximately 36 h post
transfection, the medium was replaced with fresh MEM,
10% foetal bovine serum and supplemented with 15d-PGJ
2
(10 lm) or vehicle [0.1% (v ⁄ v) dimethylsulfoxide]. After
16 h, cells were washed in phosphate buffered saline
(NaCl ⁄ P
i
), were lysed and harvested by scraping in 350 lL
Reporter Lysis Buffer (Promega) and centrifuged at
14 000 g for 1 min at R.T.
HEL cells were transfected using the DMRIE-C transfec-
tion reagent (Invitrogen Life Technologies, Carlsbad, CA,
USA). Briefly, 0.5 mL of serum free RPMI 1640 medium

was dispensed into a six well dish and 6 lL of DMRIE-C
reagent was added. Thereafter, 0.5 mL of serum free RPMI
1640 medium containing 2 lg of recombinant pGL3 Basic
vector and 200 ng of pRL-TK was added and DNA ⁄
DMRIE-C reagent was complexed by incubation at room
temperature for 30 min. Thereafter, 0.2 mL of serum free
RPMI 1640 medium containing 2 · 10
6
HEL cells were
added and were incubated for 4 h (37 °CinaCO
2
incuba-
tor) after which 2 mL of RPMI 1640 medium containing
15% (v ⁄ v) foetal bovine serum was added. Approximately
36 h post transfection, the medium was replaced with fresh
RPMI, 10% foetal bovine serum and supplemented with
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4768 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
15d-PGJ
2
(10 lm) or, for concentration response studies,
with 0–40 lm 15d-PGJ
2
or, as controls, with vehicle [0.1%
(v ⁄ v) dimethylsulfoxide]. After 16 h, cells were washed in
ice-cold NaCl ⁄ P
i
and harvested by centrifugation at 1200 g
for 5 min at 4 °C. Cell pellets were resuspended in Reporter
Lysis Buffer (100 lL), were lysed by repeated trituration.

Cell lysates were prepared by centrifugation at 14 000 g for
1 min at R.T.
To investigate the effect of over-expression of PPARc2,
PPARc3 and ⁄ or RXRa on 15d-PGJ
2
-induced inhibition of
Prm3 activity, HEL cells were cotransfected with the speci-
fic pGL3b:Prm3 reporter (1 lg) plus 200 ng of pRL-TK
along with 1 lg of either pSG5-hPPARc2, pcDNA3-
hPPARc3 and ⁄ or pSG5-mRXRa coding for PPARc2,
PPARc3 and RXRa, respectively. For the latter transfec-
tions, the total amount of transfected DNA (3.2 lg) was
kept constant by using a corresponding amount of empty
vector (pSG5 or pcDNA3) DNA. Approximately 36 h
post-transfection, the medium was replaced with fresh
RPMI, 10% foetal bovine serum and was supplemented
with the vehicle [0.1% (v ⁄ v) dimethylsulfoxide] or with
10 lm 15d-PGJ
2
. After 16 h, cell lysates were prepared as
previously described above.
HEK 293 and HEL cell supernatants were assayed for
both firefly and renilla luciferase activity using the Dual
Luciferase Assay System
TM
, essentially as previously des-
cribed [46]. Relative firefly to renilla luciferase activities
were calculated as a ratio and were expressed in relative
luciferase units (RLU).
In vitro transcription ⁄ translation

Linearized pSG5-hPPARc2 and pSG5-mRXRa plasmids
(4 lg) were transcribed in vitro at 38 °C for 1 h in the pres-
ence of I X in vitro transcription buffer (40 m m Tris ⁄ HCl,
pH 7.5, 6 mm MgCl
2
,2mm spermidine, 10 mm NaCl),
10 mm dithiothreitol, 100 l RNAsin, 0.5 mm rNTPs, T7
RNA polymerase (15–20 UÆmL
)1
;2lL) in a reaction vol-
ume of 100 lL. Thereafter, the reaction products were trea-
ted with DNase 1 (RQ RNase-free DNase I, 1 lÆlL
)1
;
2 lL) for 10 min at 37 °C followed phenol ⁄ chloroform [65].
In vitro translation reactions (25 lL) contained rabbit
reticulocyte lysate (nuclease treated; 14 lL), 1 mm amino
acids mix (–methionine; 4 lL), [
35
S]methionine (1175 CiÆ
mmol
)1
,10mCÆmL
)1
; 1.6 lL), RNasin (16 l) and primed
with the in vitro transcribed RNA template (2 lg) or, as
controls with an equivalent volume of H
2
O. Parallel in vitro
translations were carried out where the radiolabelled

[
35
S]methionine was replaced with an equivalent concentra-
tion of cold methionine (8.5 lm; 1.6 lL). In vitro transla-
tion reactions were incubated for 1 h at 30 °C followed by
1 h at 37 °C. The identity and size of the [
35
S]methionine
labelled translation products was confirmed by SDS ⁄ poly-
acrylamide gel electrophoresis (12% gels) followed by auto-
radiography.
Electrophoretic mobility shift assay
Oligonucleotides corresponding to the sense and antisense
strands of each probe (0.35 lm of each) were annealed and
32
P end-labelled as previously described [46].
Nuclear extracts were prepared from either vehicle [0.1%
(v ⁄ v) dimethylsulfoxide]- or 15d-PGJ
2
(10 lm for 16 h)-
treated HEL cells essentially as previously described [46].
Nuclear extract (4 lg total protein) or in vitro translated
PPARc2 and ⁄ or RXRa protein lysates (2 lL) or in vitro
translation reactions from the unprogrammed reticulocyte
lysates (2 lL), acting as a negative control, were incubated
for 15 min at R.T with ⁄ without a 50-fold molar excess of
unlabelled d ⁄ s competitor or noncompetitor oligonucleotide
(2 lm)in1· binding buffer [20 mm Hepes, pH 7.9., 50 mm
KCl, 0.1 mm EDTA, 0.1 mm EGTA, 0.5 mm dithiothreitol,
4% (w ⁄ v) Ficoll, 0.5 lg poly(dI-dC) (Sigma, St Louis, MO,

USA)]. The appropriate
32
P-radiolabelled d ⁄ s oligonucleo-
tide was then added to each reaction and incubated for
20 min at room temperature. Following incubation, binding
reactions were subjected to electrophoresis through a 4%
polyacrylamide gel (20 cm · 20 cm) in 89 mm Tris ⁄ borate
and 2 mm EDTA for 3 h at RT. Thereafter, gels were dried
and analysed by autoradiography.
The sequences of the competitor ⁄ noncompetitor oligonu-
cleotides used were as follows: (a) PPARc ⁄ RXR probe A
(Kin242; 5¢-dCCGGAGAGGATATTTGAGCTGGGGCA
TTGAAGGTTGTGTAGGAGTTC-3¢; corresponding to
nucleotides )198 to )150 of Prm3). (b) PPARc ⁄ RXR
probe B (Kin281; 5¢-dCATTGAAGGTTGTGTAGGA
GTTCACCA-3¢; corresponding to nucleotides )168 to
)141 of Prm3). (c) PPRE
PPARc (b)*
mutant, (Kin285; 5¢-d
CATTGAATTC TGTGTAGGAGTTCACCA-3 ¢; corres-
ponding to nucleotides )168 to )141 of Prm3 where bases
mutated from the wild type Prm3 sequence are in bold face
italics). (d) PPRE
RXRIV*
mutant (Kin283; 5 ¢-dCATT
GAAGGTTGTGTAGGAT TTTACCA-3¢; corresponding
to nucleotides )168 to )141 of Prm3 where bases mutated
from the wild type Prm3 sequence are in bold face italics).
(e) PPRE
PPARc (b)*,RXRIV*

double mutant or PPARc ⁄ RXR
probe C (Kin289; 5¢-dCATTGAATTC TGTGTAGGAT TT
TACCA-3¢; corresponding to nucleotides )168 to )141 of
Prm3 where bases mutated from the wild type Prm3
sequence are in bold face italics). (f) Consensus PPARc
response element derived from the acyl-CoA oxidase gene
(Kin342; 5¢-dAGCTGGACCAGGACAAAGGTCACGTT-3¢.
(g) AP-1 (Kin189; 5¢-dGGTGGTGACTGATCCCTCAGG
GC-3¢; corresponding to nucleotides )32 to ) 10 of Prm3).
(h) October 1 ⁄ 2 (Kin195; 5¢-dTAATCACAAGCAAA
TCTTCTCTC-3¢ corresponding to nucleotides )115 to )92
of Prm3). (i) SP-1 consensus element (Promega) with
the sequence, 5¢-dATTCGATCGGGGCGGGGCGAG-3¢.
Note, only sequences of the forward oligonucleotides are
given and sequences of the corresponding complementary
strands are omitted.
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4769
Reverse transcriptase-polymerase chain reaction
HEL 92.1.7 cells (5 · 10
6
cells approximately) were preincu-
bated for 16 h with 10 lm 15d-PGJ
2
or, as a control, with
the vehicle [0.1% (v ⁄ v) dimethylsulfoxide]. Total RNA was
isolated using TRIzol reagent (Invitrogen Life Technol-
ogies). DNase 1-treatment and RT-PCR was carried out
using oligonucleotide primers to specifically amplify TPa,
TPb and glyceraldehyde-3-phosphate dehydrogenase

(GAPDH) mRNA sequences were previously described
[40]. Southern blot analysis of the RT-PCR products were
carried out using
32
P-radiolabelled oligonucleotides probes
specific for TPa ⁄ TPb and GAPDH mRNA sequences as
previously described [40]. Images were captured and quanti-
fied by using a Typhoon PhosphorImage Analyser (Amer-
sham).
Radioligand binding studies
HEL cells were preincubated with 10 lm 15d-PGJ
2
or, as
controls, in the presence of an equivalent volume of the
vehicle [0.1% (v ⁄ v) dimethylsulfoxide] for 24 or 48 h; cells
were harvested at 500 g at 4 °C for 5 min and washed three
times with ice-cold Ca
2+
⁄ Mg
2+
-free NaCl ⁄ P
i
. TP radiolig-
and binding assays were carried out at 30 °C for 30 min in
100 lL reactions in the presence of 20 nm [
3
H]SQ29,548, as
previously described [66]. Protein determinations were car-
ried out using the Bradford assay [67].
Measurement of intracellular calcium

mobilization
HEL cells were preincubated with 10 lm 15d-PGJ
2
or, as
controls, in the presence of an equivalent volume of the vehi-
cle [0.1% (v ⁄ v) dimethylsulfoxide] for 24 or 48 h. Measure-
ments of intracellular calcium ([Ca
2+
]
i
) mobilization in
FURA2 ⁄ AM preloaded cells in response to the TP agonist
U46619 (1 lm) or the EP
1
agonist 17 phenyl trinor PGE
2
(1 lm) was carried as previously described [66]. The results
presented in the figures are representative profiles from at
least four independent experiments and are plotted as chan-
ges [Ca
2+
]
i
mobilized [D[Ca
2+
]
i
(nm)] as a function of time
(s) upon ligand stimulation. Changes in [Ca
2+

]
i
mobilization
were determined by measuring peak rises in intracellular
[Ca
2+
]
i
mobilized (D[Ca
2+
]
i
) and were calculated as mean
changes in D[Ca
2+
]
i
± SEM (nm) and values are reported
at the end of each figure legend, where appropriate.
Statistical analysis
Statistical analyses were analysed using the one-way analy-
sis of variance (anova; Fig. 4) or using two-tailed Students’
unpaired t-test using the graphprism3 Software (Graphpad
Software, San Diego, CA, USA) analysis package. All
values are expressed as mean ± standard error of the mean
(SEM). P-values of less than or equal to (£) 0.05 were con-
sidered to indicate statistically significance differences using
both t-tests and anova one-way comparisons.
Acknowledgements
This work was supported by grants from The Well-

come Trust and The Health Research Board.
References
1 Gilroy DW, Colville-Nash PR, Willis D, Chivers J,
Paul-Clark MJ & Willoughby DA (1999) Inducible
cyclooxygenase may have anti-inflammatory properties.
Nat Med 5, 698–701.
2 Colville-Nash PR & Gilroy DW (2000) COX-2 and the
cyclopentenone prostaglandins – a new. the book of
inflammation? Prostaglandins Other Lipid Mediat. 62,
33–43.
3 Perez-Sala D, Cernuda-Morollon E & Canada FJ
(2003) Molecular basis for the direct inhibition of AP-1
DNA binding by 15-deoxy-Delta 12,14-prostaglandin
J2. J Biol Chem 278, 51251–51260.
4 Tugwood JD, Issemann I, Anderson RG, Bundell KR,
McPheat WL & Green S (1992) The mouse peroxisome
proliferator activated receptor recognizes a response ele-
ment in the 5¢ flanking sequence of the rat acyl CoA
oxidase gene. EMBO J 11, 433–439.
5 DiRenzo J, Soderstrom M, Kurokawa R, Ogliastro
MH, Ricote M, Ingrey S, Horlein A, Rosenfeld MG &
Glass CK (1997) Peroxisome proliferator-activated
receptors and retinoic acid receptors differentially con-
trol the interactions of retinoid X receptor heterodimers
with ligands, coactivators, and corepressors. Mol Cell
Biol 17, 2166–2176.
6 Gelman L, Zhou G, Fajas L, Raspe E, Fruchart JC &
Auwerx J (1999) p300 interacts with the N- and C-term-
inal part of PPARgamma2 in a ligand-independent and
-dependent manner, respectively. J Biol Chem 274,

7681–7688.
7 Zhu Y, Qi C, Calandra C, Rao MS & Reddy JK (1996)
Cloning and identification of mouse steroid receptor
coactivator-1 (mSRC-1), as a coactivator of peroxisome
proliferator-activated receptor gamma. Gene Expr 6,
185–195.
8 Puigserver P, Wu Z, Park CW, Graves R, Wright M &
Spiegelman BM (1998) A cold-inducible coactivator of
nuclear receptors linked to adaptive thermogenesis. Cell
92, 829–839.
9 Castillo G, Brun RP, Rosenfield JK, Hauser S, Park
CW, Troy AE, Wright ME & Spiegelman BM (1999)
An adipogenic cofactor bound by the differentiation
domain of PPARgamma. EMBO J 18, 3676–3687.
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4770 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS
10 Heinlein CA, Ting HJ, Yeh S & Chang C (1999) Identi-
fication of ARA70 as a ligand-enhanced coactivator for
the peroxisome proliferator-activated receptor gamma.
J Biol Chem 274, 16147–16152.
11 Yang W, Rachez C & Freedman LP (2000) Discrete
roles for peroxisome proliferator-activated receptor
gamma and retinoid X receptor in recruiting nuclear
receptor coactivators, Mol Cell Biol 20, 8008–8017.
12 Rossi A, Kapahi P, Natoli G, Takahashi T, Chen Y,
Karin M & Santoro MG (2000) Anti-inflammatory
cyclopentenone prostaglandins are direct inhibitors of
IkappaB kinase. Nature 403, 103–108.
13 Delerive P, De Bosscher K, Besnard S, Vanden Berghe
W, Peters JM, Gonzalez FJ, Fruchart JC, Tedgui A,

Haegeman G & Staels B (1999) Peroxisome proliferator-
activated receptor alpha negatively regulates the vascu-
lar inflammatory gene response by negative cross-talk
with transcription factors NF-kappaB and AP-1. J Biol
Chem 274, 32048–32054.
14 Li M, Pascual G & Glass CK (2000) Peroxisome prolif-
erator-activated receptor gamma-dependent repression
of the inducible nitric oxide synthase gene. Mol Cell
Biol 20, 4699–4707.
15 Murphy GJ & Holder JC (2000) PPAR-gamma ago-
nists: therapeutic role in diabetes, inflammation and
cancer. Trends Pharmacol Sci 21, 469–474.
16 Bishop-Bailey D (2000) Peroxisome proliferator-activa-
ted receptors in the cardiovascular system. Br J Pharma-
col 129, 823–834.
17 Thiemermann C (2004) Ligands of the peroxisome pro-
liferator-activated receptor-gamma and heart failure.
Br J Pharmacol 141, 1–3.
18 Marx N, Schonbeck U, Lazar MA, Libby P & Plutzky
J (1998) Peroxisome proliferator-activated receptor
gamma activators inhibit gene expression and migration
in human vascular smooth muscle cells. Circ Res 83,
1097–1103.
19 Delerive P, Martin-Nizard F, Chinetti G, Trottein F,
Fruchart JC, Najib J, Duriez P & Staels B (1999) Per-
oxisome proliferator-activated receptor activators inhibit
thrombin-induced endothelin-1 production in human
vascular endothelial cells by inhibiting the activator pro-
tein-1 signaling pathway. Circ Res 85, 394–402.
20 Subbaramaiah K, Lin DT, Hart JC & Dannenberg AJ

(2001) Peroxisome proliferator-activated receptor
gamma ligands suppress the transcriptional activation of
cyclooxygenase-2. Evidence for involvement of activator
protein-1 and CREB-binding protein ⁄ p300. J Biol Chem
276, 12440–12448.
21 Ikeda Y, Sugawara A, Taniyama Y, Uruno A, Igarashi
K, Arima S, Ito S & Takeuchi K (2000) Suppression of
rat thromboxane synthase gene transcription by peroxi-
some proliferator-activated receptor gamma in macro-
phages via an interaction with NRF2. J Biol Chem 275,
33142–33150.
22 Sugawara A, Uruno A, Kudo M, Ikeda Y, Sato K,
Taniyama Y, Ito S & Takeuchi K (2002) Transcription
suppression of thromboxane receptor gene by peroxi-
some proliferator-activated receptor–gamma via an
interaction with Sp1 in vascular smooth muscle cells.
J Biol Chem 277, 9676–9683.
23 Narumiya S, Sugimoto Y & Ushikubi F (1999) Prosta-
noid receptors: structures, properties, and functions.
Physiol Rev 79, 1193–1226.
24 Oates JA, FitzGerald GA, Branch RA, Jackson EK,
Knapp HR, Roberts LJ & 2nd. (1988) Clinical implica-
tions of prostaglandin and thromboxane A2 forma-
tion(1). N Engl J Med 319, 689–698.
25 Fitzgerald DJ, Rocki W, Murray R, Mayo G & Fitz-
Gerald GA (1990) Thromboxane A2 synthesis in preg-
nancy-induced hypertension. Lancet 335, 751–754.
26 Neri Serneri GG, Gensini GF, Masotti G, Abbate R,
Morettini A, Poggesi L & Fortini A (1984) Role of
prostacyclin and thromboxane A2 in ischaemic heart

disease. Adv Exp Medical Biol 164, 175–185.
27 Mehta JL, Lawson D, Mehta P & Saldeen T (1988)
Increased prostacyclin and thromboxane A2 biosynth-
esis in atherosclerosis. Proc Natl Acad Sci USA 85,
4511–4515.
28 Remuzzi G, Imberti L, Rossini M, Morelli C, Carminati
C, Cattaneo GM & Bertani T (1985) Increased glomeru-
lar thromboxane synthesis as a possible cause of protei-
nuria in experimental nephrosis. J Clin Invest 75,
94–101.
29 Matsuo Y, Takagawa I, Koshida H, Kawabata T,
Nakamura M, Ida T, Zhou L & Marumo F (1995)
Antiproteinuric effect of a thromboxane receptor
antagonist, S-1452, on rat diabetic nephropathy and
murine lupus nephritis. Pharmacology 50, 1–8.
30 Hirata M, Hayashi Y, Ushikubi F, Yokota Y,
Kageyama R, Nakanishi S & Narumiya S (1991) Cloning
and expression of cDNA for a human thromboxane A2
receptor. Nature 349, 617–620.
31 Raychowdhury MK, Yukawa M, Collins LJ, McGrail
SH, Kent KC & Ware JA (1994) Alternative splicing
produces a divergent cytoplasmic tail in the human
endothelial thromboxane A2 receptor. J Biol Chem 269,
19256–19261.
32 Kinsella BT (2001) Thromboxane A2 signalling in
humans: a ‘Tail’ of two receptors. Biochem Soc Trans
29, 641–654.
33 Hirata T, Ushikubi F, Kakizuka A, Okuma M &
Narumiya S (1996) Two thromboxane A2 receptor
isoforms in human platelets. Opposite coupling to

adenylyl cyclase with different sensitivity to Arg60 to
Leu mutation. J Clin Invest 97, 949–956.
34 Vezza R, Habib A & FitzGerald GA (1999) Differential
signaling by the thromboxane receptor isoforms via the
novel GTP-binding protein, Gh. J Biol Chem 274,
12774–12779.
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4771
35 Parent JL, Labrecque P, Orsini MJ & Benovic JL
(1999) Internalization of the TXA2 receptor alpha and
beta isoforms. Role of the differentially spliced cooh
terminus in agonist-promoted receptor internalization.
J Biol Chem 274, 8941–8948.
36 Parent JL, Labrecque P, Driss Rochdi M & Benovic JL
(2001) Role of the differentially spliced carboxyl termi-
nus in thromboxane A2 receptor trafficking: identifica-
tion of a distinct motif for tonic internalization. J Biol
Chem 276, 7079–7085.
37 Walsh MT, Foley JF & Kinsella BT (2000) The alpha,
but not the beta, isoform of the human thromboxane
A2 receptor is a target for prostacyclin-mediated desen-
sitization. J Biol Chem 275, 20412–20423.
38 Reid HM & Kinsella BT (2003) The alpha, but not the
beta, isoform of the human thromboxane A2 receptor is
a target for nitric oxide-mediated desensitization. Inde-
pendent modualtion of Tp alpha signaling by nitric
oxide and prostacyckin. J Biol Chem 278 , 51190–51202.
39 Miggin SM & Kinsella BT (1998) Expression and tissue
distribution of the mRNAs encoding the human throm-
boxane A2 receptor (TP) alpha and beta isoforms.

Biochim Biophys Acta 1425, 543–559.
40 Coyle AT, Miggin SM & Kinsella BT (2002) Characteri-
zation of the 5¢ untranslated region of alpha and beta
isoforms of the human thromboxane A2 receptor (TP).
Differential promoter utilization by the TP isoforms.
Eur J Biochem 269, 4058–4073.
41 Nusing RM, Hirata M, Kakizuka A, Eki T, Ozawa K
& Narumiya S (1993) Characterization and chromoso-
mal mapping of the human thromboxane A2 receptor
gene. J Biol Chem 268, 25253–25259.
42 Carr ME (2001) Diabetes mellitus: a hypercoagulable
state. J Diabetes Complications 15, 44–54.
43 Hishinuma T, Koseki Y, Murai Y, Yamazaki T, Suzuki
K & Mizugaki M (1999) Urinary thromboxane
A2 ⁄ prostacyclin balance reflects the pathological state
of a diabetic. Prostaglandins Other Lipid Mediat 58,
263–271.
44 Okumura M, Imanishi M, Yamashita T, Yamamura Y,
Kim S, Iwao H, Tanaka S & Fujii S (2000) Renal pro-
duction of thromboxane and prostaglandins in a rat
model of type 2 diabetes. Life Sci 66, 371–377.
45 Lasserre B, Navarro-Delmasure C, Pham Huu Chanh
A, Catala J & Hollande E (2000) Modifications in the
TXA(2) and PGI(2) plasma levels and some other bio-
chemical parameters during the initiation and develop-
ment of non-insulin-dependent diabetes mellitus
(NIDDM) syndrome in the rabbit. Prostaglandins
Leukot Essent Fatty Acids 62, 285–291.
46 Coyle AT & Kinsella BT (2005) Characterization of
promoter 3 of the human thromboxane A receptor gene.

FEBS J 272, 1036–1053.
47 Quandt K, Frech K, Karas H, Wingender E & Werner T
(1995) MatInd and MatInspector: new fast and versatile
tools for detection of consensus matches in nucleotide
sequence data. Nucleic Acids Res 23, 4878–4884.
48 Guardiola-Diaz HM, Rehnmark S, Usuda N, Albrekt-
sen T, Feltkamp D, Gustafsson JA & Alexson SE
(1999) Rat peroxisome proliferator-activated receptors
and brown adipose tissue function during cold acclima-
tization. J Biol Chem 274, 23368–23377.
49 Lu R, Kanai N, Bao Y & Schuster VL (1996) Cloning,
in vitro expression, and tissue distribution of a human
prostaglandin transporter cDNA (hPGT). J Clin Invest
98, 1142–1149.
50 Nosjean O & Boutin JA (2002) Natural ligands of
PPARgamma: are prostaglandin J(2) derivatives really
playing the part? Cell Signal 14, 573–583.
51 Nishio T, Adachi H, Nakagomi R, Tokui T, Sato E,
Tanemoto M, Fujiwara K, Okabe M, Onogawa T,
Suzuki T, Nakai D, Shiiba K, Suzuki M, Ohtani H,
Kondo Y, Unno M, Ito S, Iinuma K, Nunoki K,
Matsuno S & Abe T (2000) Molecular identification of a
rat novel organic anion transporter moat1, which trans-
ports prostaglandin D(2), leukotriene C(4), and tauro-
cholate. Biochem Biophys Res Commun 275, 831–838.
52 Nakshatri H & Bhat-Nakshatri P (1998) Multiple para-
meters determine the specificity of transcriptional
response by nuclear receptors HNF-4, ARP-1, PPAR,
RAR and RXR through common response elements.
Nucleic Acids Res 26, 2491–2499.

53 Palmer CN, Hsu MH, Griffin HJ & Johnson EF (1995)
Novel sequence determinants in peroxisome proliferator
signalling. J Biol Chem 270, 16114–16121.
54 Kast-Woelbern HR, Dana SL, Cesario RM, Sun L, De
Grandpre LY, Brooks ME, Osburn DL, Reifel-Miller
A, Klausing K & Leibowitz MD (2004) Rosiglitazone
induction of INSIG-1 in white adipose tissue reveals a
novel interplay of PPARg and SREBP in the regulation
of adipogenesis. J Biol Chem 279, 23908–23915.
55 Dussault I & Forman BM (2000) Prostaglandins and
fatty acids regulate transcriptional signaling via the per-
oxisome proliferator activated receptor nuclear recep-
tors. Prostaglandins Other Lipid Mediat 62, 1–13.
56 Benson S, Wu J, Padmanabhan S, Kurtz TW &
Pershadsingh HA (2000) Peroxisome proliferator-acti-
vated receptor (PPAR)-gamma expression in human
vascular smooth muscle cells: inhibition of growth,
migration, and c-fos expression by the peroxisome pro-
liferator-activated receptor (PPAR)-gamma activator
troglitazone. Am J Hypertens 13, 74–82.
57 Vivat-Hannah V, Bourguet W, Gottardis M &
Gronemeyer H (2003) Separation of retinoid X receptor
homo- and heterodimerization functions. Mol Cell Biol
23, 7678–7688.
58 Li L, Beauchamp MC & Renier G (2002) Peroxisome
proliferator-activated receptor alpha and gamma ago-
nists upregulate human macrophage lipoprotein lipase
expression. Atherosclerosis 165, 101–110.
Effect of 15d-PGJ2 action on TP gene expression A. T. Coyle et al.
4772 FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS

59 Chen JD & Evans RM (1995) A transcriptional co-
repressor that interacts with nuclear hormone receptors.
Nature 377, 454–457.
60 Calles-Escandon J, Garcia-Rubi E, Mirza S & Morten-
sen A (1999) Type 2 diabetes: one disease, multiple car-
diovascular risk factors. Coron Artery Dis 10, 23–30.
61 Tan MH (2000) Current treatment of insulin resistance
in type 2 diabetes mellitus. Int J Clin Pract Suppl.
54–62.
62 Ogihara T, Rakugi H, Ikegami H, Mikami H & Masuo
K (1995) Enhancement of insulin sensitivity by troglita-
zone lowers blood pressure in diabetic hypertensives.
Am J Hypertens 8, 316–320.
63 Ghazzi MN, Perez JE, Antonucci TK, Driscoll JH,
Huang SM, Faja BW & Whitcomb RW (1997) Cardiac
and glycemic benefits of troglitazone treatment in
NIDDM. The Troglitazone Study Group. Diabetes 46,
433–439.
64 Hishinuma T, Yamazaki T & Mizugaki M (2000) Tro-
glitazone has a reducing effect on thromboxane produc-
tion. Prostaglandins Other Lipid Mediat 62, 135–143.
65 Sambrook J, Maniatis T & Fritsch FE (1989) Molecular
Cloning: A Laboratory Manual. 2nd edn. Cold Spring
Harbor Laboratory Press, Cold Spring Harbor, NY,
USA.
66 Kinsella BT, O’Mahony DJ & Fitzgerald GA (1997)
The human thromboxane A2 receptor alpha isoform
(TP alpha) functionally couples to the G proteins Gq
and G11 in vivo and is activated by the isoprostane
8-epi prostaglandin F2 alpha. J Pharmacol Exp Ther

281, 957–964.
67 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein util-
izing the principle of protein-dye binding. Anal Biochem
72, 248–254.
68 Vu-Dac N, Schoonjans K, Kosykh V, Dallongeville J,
Fruchart JC, Staels B & Auwerx J (1995) Fibrates
increase human apolipoprotein A-II expression through
activation of the peroxisome proliferator-activated
receptor. J Clin Invest 96, 741–750.
69 Marcus SL, Miyata KS, Zhang B, Subramani S, Rachu-
binski RA & Capone JP (1993) Diverse peroxisome pro-
liferator-activated receptors bind to the peroxisome
proliferator-responsive elements of the rat hydratase ⁄
dehydrogenase and fatty acyl-CoA oxidase genes but
differentially induce expression. Proc Natl Acad Sci
USA 90, 5723–5727.
70 Latruffe N, Nicolas-Frances V, Dasari VK & Osumi T
(1999) Studies on regulation of the peroxisomal beta-
oxidation at the 3-ketothiolase step. Dissection of the
rat liver thiolase B gene promoter. Adv Exp Medical
Biol 466, 253–259.
71 Arimura N, Horiba T, Imagawa M, Shimizu M & Sato
R (2004) The peroxisome proliferator-activated receptor
gamma regulates expression of the perilipin gene in adi-
pocytes. J Biol Chem 279, 10070–10076.
A. T. Coyle et al. Effect of 15d-PGJ2 action on TP gene expression
FEBS Journal 272 (2005) 4754–4773 ª 2005 FEBS 4773