Tài liệu Báo cáo khoa học: The plasminogen activator inhibitor 2 transcript is destabilized via a multi-component 3¢ UTR localized adenylate and uridylate-rich instability element in an analogous manner to cytokines and oncogenes pdf

The plasminogen activator inhibitor 2 transcript is
destabilized via a multi-component 3¢ UTR localized
adenylate and uridylate-rich instability element in an
analogous manner to cytokines and oncogenes
Stan Stasinopoulos
1
, Mythily Mariasegaram
1
, Chris Gafforini
1
, Yoshikuni Nagamine
2
and Robert
L. Medcalf
1
1 Monash University, Australian Centre for Blood Diseases, Melbourne, Victoria, Australia
2 Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland
Introduction
The generation of the serine protease plasmin by the
plasminogen activator system is a critical event in a
variety of physiological processes, including fibrino-
lysis, development, wound healing and cell migration
[1–4]. Plasmin generation is regulated by two plasmin-
ogen activators: urokinase-type plasminogen activator
in the extracellular environment and tissue-type plas-
minogen activator in the circulation. The proteolytic
activities of both tissue-type plasminogen activator and
urokinase-type plasminogen activator are controlled by
plasminogen activator inhibitor types 1 and 2 (PAI-1
and PAI-2, respectively). One of the enigmatic features
of PAI-2 is that, although it can inhibit extracellular

and receptor-bound urokinase-type plasminogen
Keywords
3¢ untranslated region; adenylate and
uridylate-rich element; mRNA decay;
plasminogen activator inhibitor type 2
Correspondence
R. Medcalf, Australian Centre for Blood
Diseases, Monash University, 6th Floor
Burnet Building, AMREP, 89 Commercial
Road, Melbourne 3004, Australia
Fax: +61 3 9903 0228
Tel: +61 3 9903 0133
E-mail:
(Received 21 August 2009, revised 23
December 2009, accepted 28 December
2009)
doi:10.1111/j.1742-4658.2010.07563.x
Plasminogen activator inhibitor type 2 (PAI-2; SERPINB2) is a highly-
regulated gene that is subject to both transcriptional and post-transcrip-
tional control. For the latter case, inherent PAI-2 mRNA instability was
previously shown to require a nonameric adenylate-uridylate element in the
3¢ UTR. However, mutation of this site was only partially effective at
restoring complete mRNA stabilization. In the present study, we have
identified additional regulatory motifs within the 3¢ UTR that cooperate
with the nonameric adenylate-uridylate element to promote mRNA destabi-
lization. These elements are located within a 74 nucleotide U-rich stretch
(58%) of the 3¢ UTR that flanks the nonameric motif; deletion or substitu-
tion of this entire region results in complete mRNA stabilization. These
new elements are conserved between species and optimize the destabilizing
capacity with the nonameric element to ensure complete mRNA instability

in a manner analogous to some class I and II adenylate-uridylate elements
present in transcripts encoding oncogenes and cytokines. Hence, post-tran-
scriptional regulation of the PAI-2 mRNA transcript involves an interaction
between closely spaced adenylate-uridylate elements in a manner analogous
to the post-transcriptional regulation of oncogenes and cytokines.
Abbreviations
ARE, adenylate and uridylate rich element; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GM-CSF, granulocyte macrophage-
colony-stimulating factor; IL, interleukin; PAI-2, plasminogen activator inhibitor type 2; REMSA, RNA electrophoretic mobility shift assays;
RPA, RNase protection analysis.
FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS 1331
activator [5,6], it exists primarily as a nonglycosylated
intracellular protein. Over the past decade, evidence
has accumulated to suggest a role for PAI-2 in intra-
cellular events associated with apoptosis [7–11], prolif-
eration and differentiation [4,12], and the innate
immune response [7,13–15]. PAI-2 has also generated a
substantial level of interest because of its impressive
regulatory profile. It is one of the most responsive
genes known (i.e. it can be induced over 1000-fold),
and is regulated in a cell type-dependent manner by
phorbol esters [16,17], the phosphatase inhibitor, oka-
daic acid [18], tumour necrosis factor a [19,20], lipo-
polysaccharide [21,22] and elevated levels of serum
lipoprotein (a) [23]. Although there is a significant
transcriptional component to the regulation of PAI-2
expression by these agents, in recent years, the role of
post-transcriptional regulation has come to the fore
because a number of studies have shown that the half-
life of PAI-2 mRNA can also be altered in a treatment
and cell type-dependent manner [19,22,24–26].

Post-transcriptional control of gene expression is
particularly important for controlling the levels of tran-
siently induced transcripts. Many of these transcripts
have extremely short half-lives, and this is usually attrib-
uted to the presence of adenylate and uridylate-rich
instability elements (AREs) located with the 3¢ UTR
[27]. Instability regions in the 3¢ UTR can comprise
single- or multiple-ARE elements that either interact
with each other or act independently to define the fate
of a transcript in response to a specific physiological
state [28–33]. AREs are usually 50–100 nucleotides in
length and contain single or multiple copies of the
consensus motif AUUUA, UUAUUUA(U ⁄ A)(U⁄ A) or
UUAUUUAUU embedded within a U-rich sequence
[34,35]. AREs have been classed into three groups
(groups I, II and III), depending on their particular
AU-rich sequence content [35].
Functional studies have indicated that AREs initially
accelerate mRNA deadenylation, which is then followed
by the degradation of the mRNA body [28,36,37].
A number of in vitro studies have also reported that
both AREs and ARE-binding proteins can interact with
the exosome, which then degrades the body of the
transcript with 3¢-to5¢ polarity [38–40]. Recent in vivo
studies, however, have elucidated a mammalian 5¢-to3¢
ARE decay pathway that is localized to P-bodies via an
ARE interaction with tristetraprolin and BRF1 [41–44].
However, both 5¢-to3¢ and 3¢-to5¢ pathways can be
simultaneously engaged in mRNA decay in an ARE-
mediated manner [45], suggesting that the pathway

of mammalian ARE-mediated mRNA decay can be
flexible. Recently, an excellent database compiling ARE
containing transcripts was established [46] and it has
been predicted that approximately 8% of human genes
code for transcript that contain AREs [47].
In a previous study, we defined the functional destabi-
lizing ARE element in the 3¢ UTR PAI-2 as a single
nonameric AU-rich sequence (UUAUUUAUU) located
304 nucleotides upstream of the poly(A) tail [24,48] and
suggested that tristetraprolin was a candidate PAI-2-
nonameric element binding protein involved in desta-
bilizing the PAI-2 mRNA transcript [49]. However,
subsequent work from our group demonstrated that
mutagenesis of the nonameric element only partially sta-
bilized the b-globin-PAI-2 3¢ UTR transcript [48], sug-
gesting the presence of additional functional
destabilizing regions within the PAI-2 3¢ UTR. In the
present study, we reveal that the nonameric ARE resides
within a 108 nucleotide U-rich (54%) region consisting
of three pentameric AU elements (one of which is a no-
nameric motif) and one atypical AU-rich region, and
that this extended region fully accounts for the complete
destabilizing activity of the PAI-2 3¢ UTR. Further-
more, functional mapping within the 108 AU-rich region
revealed that the essential destabilizing sequences, con-
sisting of the first two pentameric motifs and the atypical
AU-rich region, resided within a continuous 74 nucleo-
tide region, which we now define as the functional PAI-2
mRNA ARE element. The nonameric motif indeed com-
prises the core sequence that is essential for constitutive

mRNA decay; however, its optimal destabilizing activity
is only achieved in a cooperative manner with either one
of two auxiliary AREs. The results obtained support the
concept that AU-rich instability elements can be
composed of multiple AREs that act in a synergistic
manner to destabilize or stabilize transcripts depending
on the physiological status of the cell. Finally, our
studies show that PAI-2 mRNA harbours a spatial and
functional class I ARE profile that is more analogous to
that of highly-regulated cytokines and oncogenes,
including granulocyte macrophage-colony-stimulating
factor (GM-CSF), interleukin (IL)-8 and c-fos.
This may explain why the regulation of the PAI-2 gene
differs so vastly from the broader family of serine
proteases.
Results
Mutation of the PAI-2 3¢ UTR nonameric
sequence results in only partial mRNA
stabilization
To re-assess the mRNA destabilizing characteristics
of the PAI-2 3¢ UTR, we established a tetracycline-
regulated system to accurately determine mRNA decay
rates. Accordingly, we used a HT-1080 fibrosarcoma
PAI-2 mRNA decay requires a multicomponent ARE S. Stasinopoulos et al.
1332 FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS
TET-OFF system (Clontech, Mountain View, CA, USA)
in combination with a plasmid pTETBBB (referred to as
pTETGLO), which contains the gene for b-globin under
the control of a tetracycline-regulated promoter that
allows transcription only in the absence of tetracycline or

a derivative (e.g. doxycycline) [50]. We cloned the full-
length wild-type PAI-2 3¢ UTR, and a mutant PAI-2
3¢ UTR containing a four nucleotide substitution within
the nonameric ARE (UUAUUUAUU to UUA
AAG
CUU) sequence into the unique BglII site in the b-globin
3¢ UTR of plasmid pTETGLO to create plasmids pTET
GLO
PAI)2
and pTETGLO
ARE II-MUT
, respectively.
These plasmids, including the empty vector, pTETGLO,
were transiently transfected into HT-1080 fibrosarcoma
TET-OFF cells and the decay characteristics (t
1 ⁄ 2
min)
of the various transcripts were determined after the addi-
tion of doxycycline by RNase protection analysis (RPA).
As shown in Fig. 1, the half-life of the wild-type b-globin
transcript was greater than 480 min, beyond the end
point of the experiment (based on the composite curve of
three separate experiments presented in Fig. 1), demon-
strating the high stability of this transcript. The half-life
of the b-globin
PAI)2
transcript was reduced to 158 min,
whereas the half-life of the b-globin
ARE II-MUT
transcript

only increased to 301 min (Fig. 1). This demonstrates
that mutation of this element only partially stabilized the
b-globin
ARE II-MUT
transcript, which is in agreement
with previous studies from our laboratory using a differ-
ent mRNA decay system [48] and also supports the
hypothesis that the PAI-2 3¢ UTR contains uncharacter-
ized functional instability elements.
The PAI-2 3¢ UTR mRNA destabilizing elements
are localized to a 108 nucleotide U-rich (54%)
sequence
Analysis of the PAI-2 3¢ UTR sequence (Fig. 2)
revealed that the nonameric element resided within a
108 nucleotide U-rich (54%) sequence and was flanked
at the 5¢ and 3¢ ends by two classical pentameric ARE
(AUUUA) motifs and an atypical AU-rich region
(AUUUUAUAUAAU) immediately abutting 3¢ to the
nonamer. This 108 nucleotide ‘extended ARE’ can, by
structure and sequence homology, be categorized as a
class I ARE element [35]. Furthermore, these classical
pentameric elements could be the source of the addi-
tional destabilizing sequences within the ‘extended
ARE’ (Fig. 2), which could act independently or in a
cooperative manner with the nonameric ARE.
To determine whether the ‘extended ARE’ possessed
all of the destabilizing elements within the PAI-2
3¢ UTR, the entire 108 nucleotide sequence was deleted
from the 3¢ UTR to create plasmid pTET-
GLO

3¢ UTRDARE
. This plasmid was transiently trans-
fected into HT1080 TET-OFF cells and the half-life of
the b-globin
ARED
transcript was shown to be
> 480 min (Fig. 3). Hence, this deletion resulted in
significant mRNA stabilization, with mRNA decay
kinetics reminiscent of the wild-type b-globin transcript
(Fig. 1). In addition, replacement of the 108 nucleotide
‘extended ARE’ with an equivalent length of an irrele-
vant sequence also substantially stabilized the tran-
script (data not shown) to an extent similar to that
seen previously with the b-globin and the b-globin
ARED
transcripts. Moreover, cloning the 108 nucleotide
‘extended ARE’ into the BglII site in the b-globin
3¢ UTR, creating plasmid pTETGLO
EXT.ARE
, resulted
A
B
C
Fig. 1. The PAI-2 3¢ UTR localized nonameric sequence only par-
tially contributes to PAI-2 mRNA instability. (A) Rabbit-b-globin-PAI-
23¢ UTR constructs prepared for the transient transfection of
HT1080-TET OFF cells. (B) HT1080-TET OFF cells were transfected
with the TET-responsive b-globin reporter plasmids described in (A).
After 16 h of incubation, doxycycline was added and total RNA was
isolated at the indicated times and analysed by RPA. The graph in

(C) corresponds to the experiments shown in (B) (n = 3–6). Each
point represents the mean ± SE.
S. Stasinopoulos et al. PAI-2 mRNA decay requires a multicomponent ARE
FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS 1333
in decay characteristics similar to the b-globin
PAI)2
wild-type transcript (t
1 ⁄ 2
178 min and t
1 ⁄ 2
198 min,
respectively) (Fig. 4). Collectively, these experiments
demonstrate that the 108 nucleotide ‘extended ARE’
contains all the essential destabilizing elements in the
PAI-2 3¢ UTR.
ARE I and III are not independent functional
destabilizing elements
To assess the relative contribution of these additional
ARE elements, the essential residues in ARE I and III
were mutated either alone or in combination to create
plasmids pTETGLO
ARE I-MUT
, pTETGLO
ARE III-MUT
and pTETGLO
ARE I+III-MUT
(Fig. 2) within the context
of the full-length PAI-2 3¢ UTR, and the influence of
these mutations on the mRNA decay characteristics was
determined. As shown in Fig. 5, the estimated half-lives

of these transcripts were 231 min for the b-globin
PAI)2
wild-type transcript, 204 min f or the b-globin
ARE I-MUT
,
193 min for the b-globin
ARE III-MUT
and 224 min
for the b-globin
ARE I+III-MUT
. These experiments dem-
onstrate that ARE I and III, both of which are
composed of classical pentameric sequence AUUUA,
do not independently contribute to the instability of the
PAI-2 transcript.
ARE I acts as a functional auxiliary element to
the core destabilizing ARE II site
To assess the possibility that the destabilizing activity
exhibited by the ‘extended ARE’ was the result of
A
B
C
Fig. 2. The PAI-2 3¢ UTR contains a 108 nucleotide functional ‘extended ARE’. A diagrammatic representation of the PAI-2 3¢ UTR showing
the location and sequence of the AU-rich regions of interest within the ‘extended ARE’, and the sequences of the various ‘extended ARE’
mutants that were generated.
Fig. 3. Deletion of the ‘extended’ ARE from the PAI-2 3¢ UTR
results in a stabilization reminiscent to the wild-type b-globin tran-
script. (A) Rabbit-b-globin-PAI-2 3¢ UTR constructs prepared for the
transient transfection of HT1080-TET OFF cells. Plasmids pTET-
GLO

PAI)2
, containing the full-length PAI-2 3¢ UTR, and pTET-
GLO
3¢ UTRDARE
in which the ‘extended ARE’ is deleted. (B, C)
HT1080-TET OFF cells were transfected with the TET-responsive
b-globin reporter plasmids described in (A) and the b-globin mRNA
decay curves were quantified by RPA as described in Fig. 1 and
the Experimental procedures. The experiments shown in (C) were
repeated three times and each point represents the mean ± SE.
PAI-2 mRNA decay requires a multicomponent ARE S. Stasinopoulos et al.
1334 FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS
cooperation between the classical AU-rich elements,
double-ARE mutants (ARE II and ARE I; or ARE II
and ARE III) were created within the context of
the full-length PAI-2 3¢ UTR to create constructs
pTETGLO
ARE I+II-MUT
and pTETGLO
ARE II+III-MUT
and their decay characteristics were determined. The
b-globin
ARE I+II-MUT
transcript was significantly
stabilized (t
1 ⁄ 2
> 480 min; Fig. 6), to a level reminis-
cent to that seen for the b-globin (Figs 1 and 4) and
b-globin
ARED

(Fig. 3) transcripts, compared to the
wild-type b-globin
PAI)2
( 192 min) transcript in
this series of experiments (Fig. 6). This result suggests
that ARE I is an essential functional auxiliary element
to the core destabilizing ARE II sequence and that this
combination of AU-rich elements (ARE I⁄ ARE II)
plays a central role in determining the half-life of
the PAI-2 mRNA transcript under physiological
conditions.
Curiously, the half-life of the b-globin
ARE II+III
double mutant transcript was only partially stabilized
to 347 min (Fig. 6), which is also reminiscent of the
half-life of 333 min for the b-globin
ARE II-MUT
transcript (Fig. 1). This implies that ARE III is
unlikely to cooperate with the AREII ⁄ nonameric
element to contribute to the destabilizing activity
of the ‘extended ARE’ in the presence of an active
ARE I.
A
B
C
Fig. 4. The 108 nucleotide ‘extended ARE’ independently confers
mRNA instability in an analogous manner to the PAI-2 full-length
3¢ UTR. (A) Rabbit-b-globin-PAI-2 3¢ UTR constructs prepared for
the transient transfection of HT1080-TET OFF cells. Plasmids pTET-
GLO, Plasmids pTETGLO

PAI)2
, containing the full-length PAI-2
3¢ UTR, and pTETGLO
EXT.ARE
containing the 108 nucleotide
‘extended ARE’. (B, C) HT1080-TET OFF cells were transfected
with the TET-responsive b-globin reporter plasmids described in (A)
and the b-globin mRNA decay curves were quantified as described
in Fig. 1 and according the northern hybridization protocol (see
Experimental procedures). The experiments shown in (C) were
repeated three times and each point represents the mean ± SE.
The dotted line represents 50% mRNA remaining.
A
B
Fig. 5. The PAI-2 ‘extended ARE’ contains two classical pentamer-
ic sequences (AUUUA), designated as ARE I and III, that do not
independently function as instability elements. (A) Rabbit-b-globin-
PAI-2 3¢ UTR constructs prepared for the transient transfections of
HT1080-TET OFF cells. The full-length PAI-2 3¢ UTR was cloned
into the 3¢ UTR of b-globin creating plasmid pTETGLO
PAI)2
. A five
nucleotide substitution (as shown in Fig. 2) was introduced into the
ARE I and the ARE III pentameric sequences, individually or in com-
bination, to create pTETGLO
ARE I-MUT
, pTETGLO
ARE III-MUT
and
pTETGLO

ARE I+III-MUT
. (B) HT1080-TET OFF cells were transfected
with the TET-responsive b-globin reporter plasmids described in (A)
and the b-globin mRNA decay curves were quantified by RPA as
described in Fig. 1 and the Experimental procedures. The experi-
ments shown in (B) were repeated two or three times and each
point represents the mean ± SE.
S. Stasinopoulos et al. PAI-2 mRNA decay requires a multicomponent ARE
FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS 1335
The ‘extended ARE’ contains an alternate atypical
AU-rich auxiliary element that interacts with the
core ARE II sequence
Comparison of the human PAI-2 3¢ UTR with
those from a number of mammalian species (Fig. 7)
using clustalw [50a] analyses revealed a high degree
of conservation between the ARE II (nonameric)
sequences and a 12 nucleotide atypical AU-rich
sequence (labelled ARE IV) immediately 3¢ to ARE II.
To determine the extent to which this sequence
contributed to the decay rate, the same seven nucleo-
tide substitution (gUUAUUUAUUau
gcauuccuau) was
introduced into the abutting atypical ARE IV site
within the context of the full-length 3¢ UTR (Fig. 2)
to create the plasmid pTETGLO
ARE IV-MUT
.As
determined by our TET-regulated globin mRNA decay
system, disruption of this element resulted in a half-life
of the b-globin

ARE IV-MUT
transcript of  223 min
(Fig. 8) compared to the half-life of the b-globin
PAI)2
wild-type transcript ( 182 min), which is unlikely to
be a significant difference. Hence, ARE IV is unlikely
to function as an independent PAI-2 mRNA destabi-
lizing element.
To determine whether the adjacent elements (ARE
II and IV) could destabilize the transcript in an
additive or cooperative manner in an analogous way
to the AREI⁄ AREII region, both the ARE II and
the abutting ARE IV sequence were mutated
(gUUA
AAGCUUaugcauuccuau) within the context of
the full-length 3¢ UTR to create the plasmid pTET-
GLO
ARE II+IV-MUT
. This plasmid was transiently
transfected into HT1080 TET-OFF cells and the half-
life of the b-globin
ARE II+IV-MUT
transcript was sub-
stantially increased (t
1 ⁄ 2
> 480 min) (Fig. 8), which is
equivalent to the high level of stability of the b-globin,
the b-globin
ARED
and the b-globin

ARE I+II-MUT
tran-
scripts (Figs 1, 3 and 6, respectively).
RNA electrophoretic mobility shift assays (REMSA)
were next performed to determine whether these adja-
cent ARE sites played a role in protein binding activ-
ity. Initial experiments confirmed that the extended
wild-type ARE sequence provided specific protein
binding sites for cytoplasmic proteins extracted from
HT1080 TET-OFF cells (Fig. S1A). Subsequent analy-
ses further indicate that mutations introduced into
ARE II substantially reduced protein binding activity,
which is consistent with our previous results using
shorter RNA probes [48]. However, mutations intro-
duced into the adjacent ARE IV had only a minimal
effect on binding activity. When both the ARE II and
IV sites were mutated simultaneously, binding activity
was reduced to the level seen with mutations in ARE
II alone (Fig. S1B). Hence, ARE IV does not appear
to modulate protein binding activity to the ‘extended
ARE’, despite the fact that it contributes to mRNA
stability. Whether this is a consequence of the limita-
tion of the REMSA approach or the influence of
alternative functional AREs (e.g. ARE I) remains
unknown.
Taken together, the results obtained in the present
study suggest that the functional PAI-2 3¢ UTR insta-
bility sequence consists of an essential core nonameric
sequence, for which the optimal destabilizing activity
A

B
C
Fig. 6. The ARE I pentameric sequence can optimize the mRNA
destabilizing activity of the ARE II nonameric sequence. (A) Rabbit-
b-globin-PAI-2 3¢UTR constructs prepared for the transient transfec-
tion of HT1080-TET OFF cells. Double ARE mutants were con-
structed by combining ARE I-MUT and ARE II-MUT to create
plasmid pTETGLO
ARE I+II-MUT
and by combining ARE II and ARE III
to create plasmid pTETGLO
ARE II+III-MUT
(Fig. 2). (B, C) HT1080-TET
OFF cells were transfected with the TET-responsive b-globin repor-
ter plasmids described in (A) and the b-globin mRNA decay curves
were quantified by RPA as described in Fig. 1 and the Experimental
procedures. The graph in (C) corresponds to the experiments
shown in (B) (n = 3–5). Each point represents the mean ± SE.
PAI-2 mRNA decay requires a multicomponent ARE S. Stasinopoulos et al.
1336 FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS
depends on the cooperative activity of two auxiliary
elements. One is a pentameric motif (ARE I) located
55 nucleotides upstream of the core ARE II element,
and the second is an atypical AU-rich sequence (ARE
IV) abutting 3¢ to the core ARE II sequence (Fig. 7).
This functional multidomain ARE structure has been
observed in a variety of class I and II ARE elements,
including those of c-fos, GM-CSF, IL-8 [28,29,33].
Discussion
PAI-2 is a serine protease inhibitor and is a highly-reg-

ulated member of the plasminogen activator system,
and is one of the most highly inducible genes known.
Its expression can be dramatically increased in
response to cytokines, growth factors, hormones,
lipopolysaccharides and tumour promoters
[16,18,20,21,51]. Although the impressive induction of
PAI-2 has been attributed to transcriptional events,
work from the early to mid-1990s demonstrated that
PAI-2 gene expression could be regulated post-trans-
criptionally via the modulation of mRNA stability
[19,24].
We previously demonstrated that human PAI-2
mRNA was inherently unstable, with a half-life of
 1 h and that most of the destabilizing activity was
attributed to the 3¢ UTR [24] and, to a lesser extent,
an instability element within exon 4 of the coding
region [52]. It was originally predicted the nonameric
ARE (UUAUUUAUU) located 304 nucleotides
upstream of the poly(A) tail was largely responsible
for the 3 ¢ UTR driven-instability of the PAI-2 tran-
script. However, mutagenesis of this nonameric ARE
only partially stabilized both a HGH-PAI2-3¢ UTR
chimeric transcript [48] and a b-globin-PAI-2 3¢ UTR
chimeric transcript (present study). Work from other
groups has demonstrated that the presence of a single
nonameric element [UUAUUUA(U ⁄ A)(U ⁄ A)] within a
3¢ UTR has a modest effect on the stability of a repor-
ter transcript [34,53]; as such, we predicted that the
PAI-2 3¢ UTR contained additional functional instabil-
ity elements, AU-rich or otherwise [37,54,55], that

could contribute to the overall decay rate of the tran-
script.
Our analysis of the PAI-2 mRNA 3¢ UTR sequence
revealed that the nonameric element was present in
the centre of a 108 nucleotide class I type of ARE
element consisting of three copies of the AUUUA
motif that were evenly distributed within a U-rich
(54%) region and an atypical ARE (AREIV) immedi-
ately adjacent to the nonameric element. Moreover,
this region did not contain three to six clustered
AUUUA motifs, which is indicative of class II ARE
elements [35,56]. On the basis of this sequence analy-
sis, we hypothesized that this 108 nucleotide AU-rich
sequence contained all the essential destabilizing ele-
ments in the PAI-2 3¢ UTR, and we confirmed this
by demonstrating that either deleting the 108 nucleo-
tide ARE (Fig. 3) or replacing it with an irrelevant
sequence of equivalent length (data not shown) stabi-
lized the transcript to a level equivalent to that seen
for the wild-type b-globin transcript. Furthermore,
we also demonstrated that the 108 nucleotide ARE
was sufficient to destabilize the b-globin transcript
with kinetics similar to those seen with the PAI-2
full-length 3¢ UTR.
The pentameric motif (AUUUA) is the minimal
active destabilizing sequence element when present
within an appropriate AU-rich or U-rich environment.
We therefore tested the hypothesis that each of these
pentameric AREs (ARE I, II and III) contributed
equally to the overall transcript instability and, as

such, were functionally equivalent in an analogous
manner to the three pentameric motifs located in the
c-fos transcript [33]. However, this set of experiments
(Fig. 5) demonstrated that ARE I and III did not con-
tribute to transcript instability, either individually or in
combination (Fig. 5). We then sought an alternative
model to explain the destabilizing characteristics of the
PAI-2 ‘extended ARE’ element.
Fig. 7. CLUSTALW analyses of the PAI-2 ‘extended ARE’ from different mammalian species reveals a high degree conservation in the ARE II
and ARE IV regions; human (accession no. J02685), Pan troglodytes (accession no. XM_001148307), mouse (accession no. X16490), and rat
(accession no. X64563). The open boxes indicate the relative positions of the human ARE I, II, III and IV elements.
S. Stasinopoulos et al. PAI-2 mRNA decay requires a multicomponent ARE
FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS 1337
We next investigated the possibility that the struc-
ture of the PAI-2 ‘extended ARE’ was based on a
multidomain model consisting of an essential, func-
tional destabilizing core domain (e.g. the ARE II
nonameric sequence), for which the destabilizing activ-
ity was optimized by the presence of nearby auxiliary
AU-rich sequences. Figures 6 and 8 demonstrate that
the b-globin-PAI-2 3¢ UTR chimeric transcript was
only stabilized in an manner comparable to the b-glo-
bin and the b-globin
DARE
transcripts, upon the intro-
duction of two different sets of double mutations (e.g.
ARE II + ARE IV mutant and ARE II + ARE I
mutant). Taking into consideration the fact that muta-
genesis of either ARE I or ARE IV in isolation (Figs 5
and 8, respectively) did not influence the transcript’s

decay rate, we propose that the ARE II nonameric
sequence forms the core destabilizing domain of the
PAI-2 ARE and that its optimal destabilizing activity
requires the contribution of either the 3¢ abutting ARE
IV (AUUUUAUAUAAU) sequence or the 5¢ ARE I
pentameric motif. Apart from optimizing the destabi-
lizing activity of the core ARE II sequence, the ARE
IV and ARE I elements also appear to buffer effects
of mutations in the core ARE II nonameric sequence,
thereby retaining the AREs destabilizing activity, albeit
less efficiently (Fig. 1). Subsequently, we suggest that
the PAI-2 ARE IV and ARE I elements can act as
auxiliary elements to the PAI-2 core ARE II sequence.
To investigate the means by which these ARE elements
cooperate in modulating PAI-2 mRNA stability,
REMSA analyses were performed to determine the
role of the ARE II and ARE IV sites in the binding of
proteins to the ‘extended ARE’. Binding of cytoplas-
mic proteins to the ‘extended ARE’ probe was first
shown to be specific as determined by competition
titration experiments. ARE II was shown to play a sig-
nificant role in this binding activity because a four
nucleotide substitution introduced into the ARE II
caused substantial decrease in binding activity. By con-
trast, mutagenesis of ARE IV had no noticeable effect
on the binding of proteins to the ‘extended ARE’ and
had no additional suppression of protein binding activ-
ity in the presence of the mutated ARE II. Hence,
ARE IV does not appear to modulate protein binding
activity to the ‘extended ARE’. The means by which

ARE IV cooperates with ARE II to destabilize mRNA
still remains unknown. The role of the ARE 1 site was
not investigated in the present study and will be the
subject of future research.
Functional multidomain ARE structures have been
observed in a variety of class I and II ARE elements,
including those of c-fos, GM-CSF and IL-8, amongst
others (Fig. 9) [28,29,33], and appear to function via
similar mechanisms. Of greatest relevance to the PAI-2
ARE is the c-fos multidomain class I ARE, for which
the structure and function has been characterized in
detail; this ARE is composed of two structurally dis-
tinct but functionally interdependent domains [33]
(Fig. 9). The c-fos ARE core sequence consists of three
pentameric motifs embedded within a U-rich region
and is independently capable of destabilizing a tran-
script. The c-fos ARE auxiliary domain II is a
20 nucleotide U-rich sequence that cannot indepen-
A
B
C
Fig. 8. An atypical AU-rich sequence (ARE IV) abutting 3¢ to the
ARE II pentameric sequence can optimize the mRNA destabilizing
activity of the ARE II nonameric sequence. (A) Rabbit-b-globin-PAI-2
3¢ UTR constructs prepared for the transient transfection of
HT1080-TET OFF cells. The AU-rich sequence (ARE IV) abutting 3¢
to the ARE II pentameric motif was mutated to create plasmid
pTETGLO
ARE IV-MUT
; a double ARE mutant that combined ARE

II-MUT and ARE IV was constructed to create plasmid
pTETGLO
ARE II+IV-MUT
(Fig. 2). (B, C) HT1080-TET OFF cells were
transfected with the TET-responsive b-globin reporter plasmids
described in (A) and the b-globin mRNA decay curves were quanti-
fied by RPA as described in Fig. 1. The experiments shown in
(C) were repeated three times and each point represents the
mean ± SE.
PAI-2 mRNA decay requires a multicomponent ARE S. Stasinopoulos et al.
1338 FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS
dently destabilize the transcript; however, when pres-
ent in the appropriate context (i.e. immediately
up- or downstream of the core domain I) [33], it
can stimulate the deadenylation rate and thereby
increase the decay rate of the transcript. Moreover,
domain II of c-fos also serves the essential function
of buffering the effects of mutations occurring within
domain I [33].
The relative location of a functional auxiliary
domain, with respect to the core domain is flexible
because auxiliary domains have been identified either
5¢ or 3¢ to the core domains in class I and II AREs
[28,29,33] (Fig. 9); moreover, placing the c-fos auxil-
iary domain either 5¢ or 3¢ to the core domain resulted
in a similar deadenylation and overall mRNA decay
rate [33]. The PAI-2 ARE is unusual in that in
addition to an auxiliary domain (Fig. 7, the atypical
AU-rich ARE IV; Fig. 8) immediately 3¢ to the core,
the ARE I (AUUUA) element located 5¢ to the core

element (Figs 7 and 9) also behaves as a functional
auxiliary domain in the presence of a mutated ARE
IV (Fig. 8). Whether the two PAI-2 auxiliary domains
are simultaneously active cannot be determined from
the data obtained in the present study, although it
does remain a plausible hypothesis. However, we
suggest that, under normal physiological conditions,
the destabilizing activity of the core domain is prefer-
entially optimized by auxiliary domain I (Fig. 7, the
atypical AU-rich ARE IV; Fig. 9) based on the high
degree of homology in the equivalent sequences of
other species (Fig. 7). Moreover, the addition of the
second auxiliary sequence, domain I (Fig. 9), can sup-
port the destabilizing activity of the core domain in
the absence of domain IV (Fig. 8).
In summary, under normal physiological conditions,
the PAI-2 mRNA transcript is unstable, which we now
attribute to the presence of a multidomain AU-rich
element within the 3 ¢ UTR (Fig. 9). ARE-mediated
PAI-2 mRNA instability significantly contributes to
the low constitutive levels of PAI-2 protein; however,
the ARE can also modulate PAI-2 mRNA stability
during physiological conditions that require high levels
of PAI-2 gene expression and, subsequently, the contri-
bution of post-transcriptional regulation to PAI-2 gene
expression cannot be underestimated. The present
study has focused on the characterization and fine
mapping of the functional destabilizing AU-rich region
within the PAI-2 3¢ UTR under physiological condi-
tions that result in an unstable PAI-2 mRNA tran-

script. We have shown that the PAI-2 ARE is a
74 nucleotide multidomain (Fig. 9) class I element con-
sisting of a destabilizing core nonameric element with
activity that is supported by one of two auxiliary ele-
ments. Hence, in an attempt to severely inhibit constit-
utive PAI-2 gene expression, nature has evolved a
functional, mutation insensitive, multidomain ARE
element. We are currently determining the contribution
and mechanism of this ARE, and the individual ARE
domains, to PAI-2 mRNA stabilization and, subse-
quently, PAI-2 gene expression.
Experimental procedures
Plasmids and mutant construction
The vector pTETBBB was provided by A. B. Shyu (Univer-
sity of Texas Medical School, Houston, TX, USA). This
plasmid contains the gene for b-globin under the control of
a tetracycline-regulated promoter that allows the transcrip-
tion of this gene in the absence of tetracycline within an
appropriate mammalian cell line (e.g. HT1080-TET OFF).
pTETBBB is referred to as pTETGLO throughout the
present study.
The PAI-2 3¢ UTR was amplified from plasmids pCMV-
glo-PAI-2 3¢ UTR and pCMV-glo-PAI-2 3¢ UTR-ARE
MUT [48] with primers SJS133 and SJS134, and cloned into
the BglII site in the b-globin 3¢ UTR in pTETGLO, to gener-
ate pTETGLO
PAI)2
and pTETGLO
ARE II-MUT
,

respectively.
Fig. 9. Multidomain structure of PAI-2 (accession no. J02685), c-fos (accession no. NM_005252), GM-CSF (accession no. M11220) and IL-8
(accession no. Y00787) AREs. Sequences of the AREs are shown with the AUUUAs underlined, and the relative positions of the core and
auxiliary domains are overlined.
S. Stasinopoulos et al. PAI-2 mRNA decay requires a multicomponent ARE
FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS 1339
Mutant variants of the PAI-2 3¢ UTR were generated via
overlap extension PCR mutagenesis [57] using SJS133 and
SJS134 as the external primers and the constructs pTET-
GLO
PAI)2
and pTETGLO
ARE II-MUT
as the templates.
Mutation of the ARE I-AUUUA and ARE III-AUUUA
sequences used primers SJS172 and SJS173, and SJS174 and
SJS175, respectively. Mutation of the atypical AU-rich
sequence used primers SJS259 and SJS260, and the creation
of the ARE II ⁄ ARE IV double mutant used primers SJS261
and SJS262. The mutagenesis of ARE I and III introduced
HindIII restriction sites and so the creation of the pTET-
GLO
3¢ UTRDARE
involved digesting construct pTET-
GLO
ARE I+III-MUT
mutant with HindIII to remove the
108 bp ARE, gel purifying the larger fragment and self-liga-
tion. The PAI-2 ‘extended ARE’ was amplified from plasmid
pTETGLO

PAI)2
using primers SJS137 and SJS138, and
cloned into the BglII site in the b-globin 3¢ UTR in pTET-
GLO, to generate pTETGLO
EXT.ARE
. The sequences of the
primers used in the present study are listed in Table 1.
Cell culture and transfection
HT1080-TET OFF cells (Clontech) were maintained in
DMEM supplemented with 10% fetal bovine serum and
100 lgÆmL
)1
G418 (Life Technologies, Inc. Carlsbad, CA,
USA). Cells were maintained at 37 °C in the presence of
5% CO
2
. Transient transfections were performed via the
Fugene (Roche, Basel, Switzerland) method according to
the manufacturer’s instructions. A typical mRNA decay
experiment involved seeding five 35 mm plates with
5.0 · 10
5
cells and incubating overnight. The next day, each
plate was transfected with a total of 1 lg of plasmid DNA
and incubated at 37 °C for 5 h. These cells were then
washed once with NaCl ⁄ P
i
, trypsinized, combined and
equally seeded into five 35 mm to ensure equal transfection
efficiency within samples and the plates were returned to

the incubator for further incubation.
In vitro transcription and RNase protection assay
and northern hybridization
A cDNA library prepared with 1 lg of total RNA
extracted from an HT1080 TET-off cell line transiently
transfected with pTETBBB was used to generate the rabbit
b-globin and glyceraldehyde 3-phosphate dehydrogenase
(GAPDH) riboprobes. Accordingly, a 295 bp b-globin frag-
ment that spans the first intron was amplified using primers
SJS167 and SJS170 and a 155 bp GAPDH fragment was
amplified using primers ALS030 and SJS209; these frag-
ments were then cloned into the pGEM-T Easy vector
(Promega, Madison, WI, USA) generating pTEasy-Globin
and pTEasy-GAPDH. For in vitro transcription, 500 ng of
SpeI linearized pTEasy-Globin and SacII linearized
pTEasy-GAPDH were incubated for 1 h in the presence of
50 lCi [a-
32
P]UTP (PerkinElmer Life and Analytical Sci-
ences, Inc., Waltham, MA, USA), 10 lm UTP, 0.5 mm
ATP, 0.5 mm CTP, 0.5 mm GTP, 40 U of RNase Inhibitor
(Promega Corporation, Madison, WI, USA) and either
Table 1. PCR and overlap PCR mutagenesis primers. The name, nucleotide sequence, orientation and GenBank nucleotide reference (where
available) are provided. The introduced mutations are underlined,the restriction enzyme sites are italicized, and lower case indicates the T7
promoter sequence. PAI-2 cDNA (accession no. M18082), GADPH cDNA (accession no. M33197), pTETBBB plasmid sequence from Profes-
sor A. B. Shyu (University of Texas Medical School, Houston, TX, USA). nt, nucleotide.
Primer Nucleotide sequence (5¢-to3¢) Orientation
SJS133 CGGA
AGATCT
AACTAAGCGTGCTGCTTC Forward (nt 1281–1298 PAI-2)

SJS134 TACG
AGATCT
GTTGTTTGGAAGCAGGTT Reverse (nt 1860–1843 PAI-2)
SJS137 CGGA
AGATCT
GGGATCATGCCCATTTAG Forward (nt 1491–1508 PAI-2)
SJS138 TACG
AGATCT
TAGCTACATTAAATAGGC Reverse (nt 1620–1603 PAI-2)
SJS172 GGGATCATGCCCA
AGCTTATTTTCCTTACT Forward (nt 1491–1520 PAI-2)
SJS173 AGTAAGGAAAATA
AGCTTGGGCATGATCCC Reverse (nt 1520–1491 PAI-2)
SJS174 GCTCACTGCCTA
AGCTTTGTAGCTAATAAAG Forward (nt 1596–1625 PAI-2)
SJS175 CTTTATTAGCTACA
AAGCTTAGGCAGTGAGC Reverse (nt 1625–1596 PAI-2)
SJS259 CTTTGTTATTTATTAT
GCATTCCTATGGTGAGTT Forward (nt 1552–1585 PAI-2)
SJS260 AACTCACCAT
AGGAATGCATAATAAATAACAAAG Reverse (nt 1585–1552 PAI-2)
SJS261 CTTTGTTA
AAGCTTATGCATTCCTATGGTGAGTT Forward (nt 1552–1585 PAI-2)
SJS262 AACTCACCAT
AGGAATGCATAAGCTTTAACAAAG Reverse (nt 1585–1552 PAI-2)
SJS167 CCTCTTACACTTGCTTTTGAC Forward (nt 455–474 pTETBBB)
SJS170 GCAAAGGTGCCTTTGAGGTTG Reverse (nt 897–878 pTETBBB)
ALS030 GACCCCTTCATTGACCTCAACTA Forward (nt 163–185 GAPDH)
SJS209 CTTGATTTTGGAGGGATCTC Reverse (nt 318–299 GAPDH)
SJS275 TTAGCTACATTAAATAGGCAG Reverse (nt 1620–1601 PAI-2)

SJS276 GtaatacgactcactataGGGATCATGCCCATTTAG T7Forward (nt 1491–1508 PAI-2)
PAI-2 mRNA decay requires a multicomponent ARE S. Stasinopoulos et al.
1340 FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS
50 U of T7 Polymerase for pTEasy-Globin or 50 U of SP6
polymerase for pTEasy-GAPDH. The radioactive products
were purified according to the RPA III kit instructions
(Ambion Inc., Austin, TX, USA).
The RNase protection assay was carried out using the
RPA III kit (Ambion) according to the manufacturers
instructions with 7.5 lg of total RNA isolated from the
transiently transfected HT1080 TET-OFF cells. The prod-
ucts were resolved on a denaturing gel (7 m urea ⁄ 5%
PAGE) and visualized by autoradiography. Signals were
quantified using the ImageQuant, version 5 (Amersham
Biosciences, Piscataway, NJ, USA) and the results were
presented in graphical form after correcting for variations
in GAPDH levels between time point samples.
Total RNA (10 lg) was resolved on a 1% formaldehyde-
agarose gel and transferred to a Hybond-N nylon mem-
brane (GE Healthcare, Piscataway, NJ, USA). RNA blots
were stained with methylene blue to confirm for equal load-
ing and transfer. Hybridization was performed by the
Rapid-Hyb hybridization protocol (GE Healthcare) using
random primed [a-
32
P]dATP-labelled cDNA probes corre-
sponding to rabbit b-globin and human GAPDH isolated
from plasmids pTEasy-Globin and pTEasy-GAPDH.
Hybridization signals were visualized and quantified with
a PhosphorImager (Molecular Dynamics, Sunnyvale, CA,

USA).
Tet-off b-globin mRNA decay assay
HT1080-TET OFF cells were transiently transfected with
the pTETGLO and various pTETGLO-PAI-2 3¢ UTR
plasmids and seeded into 35 mm plates as described above.
After 16 h of incubation, doxycycline was added to stop
transcription from the tetracycline responsive promoter and
total RNA was harvested in Trizol at the indicated times.
Total RNA (7.5 lg) from each sample was analysed by
RNase protection assay or by northern analysis (10 lgof
total RNA) using the intron spanning
32
P-labelled globin
and GAPDH probes described above. The b-globin and
GAPDH mRNA band intensities were visualized and
quantified with a Phosphorimager Storm (Molecular
Dynamics). For each sample, the hybridization intensity
value of b-globin was normalized to GAPDH mRNA
levels. The normalized b-globin mRNA levels at time 0 h
was set to 100% and all other normalized b-globin mRNA
values are plotted over time relative to the value at 0 h and
the half-lives were calculated for each transfection group.
REMSAs
Radiolabelled RNAs were prepared by in vitro transcription
as described above, using DNA templates generated by
PCR that had an incorporated core T7 promoter region.
This DNA was used to transcribe a 128 nucleotide sequence
that included the 108 nucleotide ARE region and ten
nucleotides of the 5¢ and 3¢ flanking sequence (wild-type or
various mutants). Primers SJS276 and SJS275 were used to

generate the various DNA fragments from plasmids pTET-
GLO
PAI)2
, pTETGLO
ARE II-MUT
, pTETGLO
ARE IV-MUT
and pTETGLO
ARE II+IV-MUT
. The unrelated 114 nucleo-
tide RNA probe was derived from T7 transcribed KpnI
digested pBluescript KS+. The RNA probes were purified
on a 6% polyacrylamide-urea gel, eluted in a 500 mm
NH
4
CH
3
COO, 1 mm EDTA solution overnight at room
temperature, ethanol precipitated at –80 °C and resus-
pended in water (30 000 c.p.s.ÆlL
)1
). To prepare extracts
for REMSAs, cells (80% confluence) were collected by
trypsinization, washed with NaCl ⁄ P
i
, then lysed for 10 min
on ice in 300 lL of cytoplasmic extraction buffer (10 mm
Hepes, pH 7.1, 3 mm MgCl
2
,14mm KCl, 0.2% NP-40,

1mm dithiothreitol, supplemented with protease inhibitors
(complete EDTA free; Roche) and phosphatase inhibitors
(Phospho-Stop; Roche). The nuclei were pelleted for 1 min
at 1000 g at 4 °C, and the supernatant containing the cyto-
plasmic fraction was aliquoted, snap-frozen in liquid nitro-
gen, and stored at –80 °C. Cytoplasmic protein extracts
and unlabelled RNA transcripts (for ‘extended ARE’ and
unrelated RNA) for competition titration experiments were
prepared as described previously [48].
For the binding reactions, 0.5–5 l g of protein extract
was preincubated with 150 lg of heparin in a total
volume of 20 lL for 10 min at room temperature in
cytoplasmic extraction buffer before the addition of the
RNA probe (60 000 c.p.s.). After 30 min of incubation at
room temperature, samples were treated with 0.6–3.0 U
of RNase T1 (Roche Molecular Biochemicals, Indianapo-
lis, IN, USA) for 10 min at room temperature and then
subjected to electrophoresis through a 5% native PAGE,
and protein–RNA complexes were visualized by auto-
radiography.
References
1 Hibino T, Matsuda Y, Takahashi T & Goetinck PF
(1999) Suppression of keratinocyte proliferation by
plasminogen activator inhibitor-2. J Invest Dermatol
112, 85–90.
2 Jensen PJ, Wu Q, Janowitz P, Ando Y & Schechter
NM (1995) Plasminogen activator inhibitor type 2: an
intracellular keratinocyte differentiation product that is
incorporated into the cornified envelope. Exp Cell Res
217, 65–71.

3 Shafren DR, Gardner J, Mann VH, Antalis TM &
Suhrbier A (1999) Picornavirus receptor down-
regulation by plasminogen activator inhibitor type 2.
J Virol 73, 7193–7198.
4 Yu H, Maurer F & Medcalf RL (2002) Plasminogen
activator inhibitor type 2: a regulator of monocyte
proliferation and differentiation. Blood 99, 2810–
2818.
S. Stasinopoulos et al. PAI-2 mRNA decay requires a multicomponent ARE
FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS 1341
5 Yebra M, Parry GC, Stromblad S, Mackman N,
Rosenberg S, Mueller BM & Cheresh DA (1996)
Requirement of receptor-bound urokinase-type plasmin-
ogen activator for integrin alphavbeta5-directed cell
migration. J Biol Chem 271, 29393–29399.
6 Kruithof EK, Baker MS & Bunn CL (1995) Biological
and clinical aspects of plasminogen activator inhibitor
type 2. Blood 86, 4007–4024.
7 Gan H, Lee J, Ren F, Chen M, Kornfeld H & Remold
HG (2008) Mycobacterium tuberculosis blocks crosslink-
ing of annexin-1 and apoptotic envelope formation on
infected macrophages to maintain virulence. Nat Immu-
nol 9, 1189–1197.
8 Park JM, Greten FR, Wong A, Westrick RJ, Arthur
JS, Otsu K, Hoffmann A, Montminy M & Karin M
(2005) Signaling pathways and genes that inhibit
pathogen-induced macrophage apoptosis–CREB
and NF-kappaB as key regulators. Immunity 23, 319–
329.
9 Antalis TM, La Linn M, Donnan K, Mateo L, Gardner

J, Dickinson JL, Buttigieg K & Suhrbier A (1998) The
serine proteinase inhibitor (serpin) plasminogen activa-
tion inhibitor type 2 protects against viral cytopathic
effects by constitutive interferon alpha ⁄ beta priming.
J Exp Med 187, 1799–1811.
10 Dickinson JL, Bates EJ, Ferrante A & Antalis TM
(1995) Plasminogen activator inhibitor type 2 inhibits
tumor necrosis factor alpha-induced apoptosis.
Evidence for an alternate biological function. J Biol
Chem 270, 27894–27904.
11 Kumar S & Baglioni C (1991) Protection from tumor
necrosis factor-mediated cytolysis by overexpression of
plasminogen activator inhibitor type-2. J Biol Chem
266, 20960–20964.
12 Lian X & Yang T (2004) Plasminogen activator inhibi-
tor 2: expression and role in differentiation of epidermal
keratinocyte. Biol Cell 96, 109–116.
13 Haile WB, Coleman JL & Benach JL (2006) Reciprocal
upregulation of urokinase plasminogen activator and its
inhibitor, PAI-2, by Borrelia burgdorferi affects bacterial
penetration and host-inflammatory response. Cell
Microbiol 8, 1349–1360.
14 Varro A, Noble PJ, Pritchard DM, Kennedy S, Hart
CA, Dimaline R & Dockray GJ (2004) Helicobacter
pylori induces plasminogen activator inhibitor 2 in gas-
tric epithelial cells through nuclear factor-kappaB and
RhoA: implications for invasion and apoptosis. Cancer
Res 64, 1695–1702.
15 Losick VP & Isberg RR (2006) NF-kappaB transloca-
tion prevents host cell death after low-dose challenge by

Legionella pneumophila. J Exp Med 203, 2177–2189.
16 Schleuning WD, Medcalf RL, Hession C, Rothenbuhler
R, Shaw A & Kruithof EK (1987) Plasminogen activa-
tor inhibitor 2: regulation of gene transcription during
phorbol ester-mediated differentiation of U-937 human
histiocytic lymphoma cells. Mol Cell Biol 7, 4564–4567.
17 Genton C, Kruithof EK & Schleuning WD (1987)
Phorbol ester induces the biosynthesis of glycosylated
and nonglycosylated plasminogen activator inhibitor 2
in high excess over urokinase-type plasminogen activa-
tor in human U-937 lymphoma cells. J Cell Biol 104,
705–712.
18 Medcalf RL (1992) Cell- and gene-specific interactions
between signal transduction pathways revealed by oka-
daic acid. Studies on the plasminogen activating system.
J Biol Chem 267, 12220–12226.
19 Pytel BA, Peppel K & Baglioni C (1990) Plasminogen
activator inhibitor type-2 is a major protein induced in
human fibroblasts and SK-MEL-109 melanoma cells by
tumor necrosis factor. J Cell Physiol 144
, 416–422.
20 Medcalf RL, Kruithof EK & Schleuning WD (1988)
Plasminogen activator inhibitor 1 and 2 are tumor
necrosis factor ⁄ cachectin-responsive genes. J Exp Med
168, 751–759.
21 Suzuki T, Hashimoto S, Toyoda N, Nagai S, Yamazaki
N, Dong HY, Sakai J, Yamashita T, Nukiwa T &
Matsushima K (2000) Comprehensive gene expression
profile of LPS-stimulated human monocytes by SAGE.
Blood 96, 2584–2591.

22 Schwartz BS & Bradshaw JD (1992) Regulation of plas-
minogen activator inhibitor mRNA levels in lipopolysac-
charide-stimulated human monocytes. Correlation with
production of the protein. J Biol Chem 267 , 7089–7094.
23 Buechler C, Ullrich H, Ritter M, Porsch-Oezcueruemez
M, Lackner KJ, Barlage S, Friedrich SO, Kostner GM
& Schmitz G (2001) Lipoprotein (a) up-regulates the
expression of the plasminogen activator inhibitor 2 in
human blood monocytes. Blood 97, 981–986.
24 Maurer F & Medcalf RL (1996) Plasminogen activator
inhibitor type 2 gene induction by tumor necrosis factor
and phorbol ester involves transcriptional and post-
transcriptional events. Identification of a functional
nonameric AU-rich motif in the 3¢-untranslated region.
J Biol Chem 271, 26074–26080.
25 Niiya K, Taniguchi T, Shinbo M, Ishikawa T, Tazawa
S, Hayakawa Y & Sakuragawa N (1994) Different regu-
lation of plasminogen activator inhibitor 2 gene expres-
sion by phorbol ester and cAMP in human myeloid
leukemia cell line PL-21. Thromb Haemost 72, 92–97.
26 Antalis TM & Dickinson JL (1992) Control of plasmin-
ogen-activator inhibitor type 2 gene expression in the
differentiation of monocytic cells. Eur J Biochem 205,
203–209.
27 Wilusz CJ & Wilusz J (2004) Bringing the role of
mRNA decay in the control of gene expression into
focus. Trends Genet 20, 491–497.
28 Xu N, Chen CY & Shyu AB (1997) Modulation of the
fate of cytoplasmic mRNA by AU-rich elements: key
PAI-2 mRNA decay requires a multicomponent ARE S. Stasinopoulos et al.

1342 FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS
sequence features controlling mRNA deadenylation and
decay. Mol Cell Biol 17, 4611–4621.
29 Winzen R, Gowrishankar G, Bollig F, Redich N, Resch
K & Holtmann H (2004) Distinct domains of AU-rich
elements exert different functions in mRNA destabiliza-
tion and stabilization by p38 mitogen-activated protein
kinase or HuR. Mol Cell Biol 24, 4835–4847.
30 Bevilacqua A, Ceriani MC, Capaccioli S & Nicolin A
(2003) Post-transcriptional regulation of gene expression
by degradation of messenger RNAs. J Cell Physiol 195,
356–372.
31 Benjamin D & Moroni C (2007) mRNA stability and
cancer: an emerging link? Expert Opin Biol Ther 7,
1515–1529.
32 Akashi M, Shaw G, Hachiya M, Elstner E, Suzuki G &
Koeffler P (1994) Number and location of AUUUA
motifs: role in regulating transiently expressed RNAs.
Blood 83, 3182–3187.
33 Chen CY, Chen TM & Shyu AB (1994) Interplay of
two functionally and structurally distinct domains of
the c-fos AU-rich element specifies its mRNA-destabi-
lizing function. Mol Cell Biol 14, 416–426.
34 Lagnado CA, Brown CY & Goodall GJ (1994)
AUUUA is not sufficient to promote poly(A)
shortening and degradation of an mRNA: the
functional sequence within AU-rich elements may be
UUAUUUA(U ⁄ A)(U ⁄ A). Mol Cell Biol 14, 7984–7995.
35 Chen CY & Shyu AB (1995) AU-rich elements: charac-
terization and importance in mRNA degradation.

Trends Biochem Sci 20, 465–470.
36 Wilson T & Treisman R (1988) Removal of poly(A)
and consequent degradation of c-fos mRNA facilitated
by 3¢ AU-rich sequences. Nature 336, 396–399.
37 Chen CY & Shyu AB (1994) Selective degradation of
early-response-gene mRNAs: functional analyses of
sequence features of the AU-rich elements. Mol Cell
Biol 14, 8471–8482.
38 Gherzi R, Lee KY, Briata P, Wegmuller D, Moroni C,
Karin M & Chen CY (2004) A KH domain RNA bind-
ing protein, KSRP, promotes ARE-directed mRNA
turnover by recruiting the degradation machinery. Mol
Cell 14, 571–583.
39 Chen CY, Gherzi R, Ong SE, Chan EL, Raijmakers
R, Pruijn GJ, Stoecklin G, Moroni C, Mann M &
Karin M (2001) AU binding proteins recruit the exo-
some to degrade ARE-containing mRNAs. Cell 107,
451–464.
40 Mukherjee D, Gao M, O’Connor JP, Raijmakers R,
Pruijn G, Lutz CS & Wilusz J (2002) The mammalian
exosome mediates the efficient degradation of mRNAs
that contain AU-rich elements. EMBO J 21, 165–174.
41 Franks TM & Lykke-Andersen J (2007) TTP and BRF
proteins nucleate processing body formation to silence
mRNAs with AU-rich elements. Genes Dev 21, 719–
735.
42 Stoecklin G, Mayo T & Anderson P (2006) ARE-
mRNA degradation requires the 5¢-3¢ decay pathway.
EMBO Rep. 7, 72–77.
43 Lykke-Andersen J & Wagner E (2005) Recruitment and

activation of mRNA decay enzymes by two ARE-medi-
ated decay activation domains in the proteins TTP and
BRF-1.
Genes Dev 19, 351–361.
44 Fenger-Gron M, Fillman C, Norrild B & Lykke-Ander-
sen J (2005) Multiple processing body factors and the
ARE binding protein TTP activate mRNA decapping.
Mol Cell 20, 905–915.
45 Murray EL & Schoenberg DR (2007) A+U-rich insta-
bility elements differentially activate 5¢-3¢ and 3¢-5¢
mRNA decay. Mol Cell Biol 27, 2791–2799.
46 Bakheet T, Williams BR & Khabar KS (2006) ARED
3.0: the large and diverse AU-rich transcriptome.
Nucleic Acids Res 34, D111–D114.
47 Khabar KS (2005) The AU-rich transcriptome: more
than interferons and cytokines, and its role in disease.
J Interferon Cytokine Res 25, 1–10.
48 Maurer F, Tierney M & Medcalf RL (1999) An
AU-rich sequence in the 3¢-UTR of plasminogen
activator inhibitor type 2 (PAI-2) mRNA
promotes PAI-2 mRNA decay and provides a binding
site for nuclear HuR. Nucleic Acids Res 27,
1664–1673.
49 Yu H, Stasinopoulos S, Leedman P & Medcalf RL
(2003) Inherent instability of plasminogen activator
inhibitor type 2 mRNA is regulated by tristetraprolin.
J Biol Chem 278, 13912–13918.
50 Xu N, Loflin P, Chen CY & Shyu AB (1998) A broader
role for AU-rich element-mediated mRNA turnover
revealed by a new transcriptional pulse strategy. Nucleic

Acids Res 26, 558–565.
50a Thomson JD, Higgins DG & Gibson TJ (1994)
CLUSTAL W: improving the sensitivity of progressive
multiple sequence alignment through sequence weight-
ing, position-specific gap penalties and weight matrix
choice. Nucleic Acids Res 22, 4673–4680.
51 Bystrom J, Wynn TA, Domachowske JB & Rosenberg
HF (2004) Gene microarray analysis reveals
interleukin-5-dependent transcriptional targets in mouse
bone marrow. Blood 103, 868–877.
52 Tierney MJ & Medcalf RL (2001) Plasminogen
activator inhibitor type 2 contains mRNA instability
elements within exon 4 of the coding region.
Sequence homology to coding region instability
determinants in other mRNAs. J Biol Chem 276,
13675–13684.
53 Zubiaga AM, Belasco JG & Greenberg ME (1995) The
nonamer UUAUUUAUU is the key AU-rich sequence
motif that mediates mRNA degradation. Mol Cell Biol
15, 2219–2230.
54 Chen CY, Xu N & Shyu AB (1995) mRNA decay
mediated by two distinct AU-rich elements from c-fos
S. Stasinopoulos et al. PAI-2 mRNA decay requires a multicomponent ARE
FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS 1343
and granulocyte-macrophage colony-stimulating factor
transcripts: different deadenylation kinetics and
uncoupling from translation. Mol Cell Biol 15,
5777–5788.
55 Stoecklin G, Lu M, Rattenbacher B & Moroni C (2003)
A constitutive decay element promotes tumor necrosis

factor alpha mRNA degradation via an AU-rich
element-independent pathway. Mol Cell Biol 23, 3506–
3515.
56 Barreau C, Paillard L & Osborne HB (2005) AU-rich
elements and associated factors: are there unifying
principles? Nucleic Acids Res 33, 7138–7150.
57 Ho SN, Hunt HD, Horton RM, Pullen JK & Pease LR
(1989) Site-directed mutagenesis by overlap extension
using the polymerase chain reaction. Gene 77, 51–59.
Supporting information
The following supplementary material is available:
Fig. S1. The ‘extended ARE’ provides specific binding
sites for cytoplasmic proteins.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
PAI-2 mRNA decay requires a multicomponent ARE S. Stasinopoulos et al.
1344 FEBS Journal 277 (2010) 1331–1344 ª 2010 The Authors Journal compilation ª 2010 FEBS