Tài liệu Báo cáo khoa học: Interleukin-1-inducible MCPIP protein has structural and functional properties of RNase and participates in degradation of IL-1b mRNA doc

Interleukin-1-inducible MCPIP protein has structural and
functional properties of RNase and participates in
degradation of IL-1b mRNA
Danuta Mizgalska
1
, Paulina We˛ grzyn
1
, Krzysztof Murzyn
2
, Aneta Kasza
1
, Aleksander Koj
1
,
Jacek Jura
3
, Barbara Jarza˛b
4
and Jolanta Jura
1
1 Department of Cell Biochemistry, Jagiellonian University, Krakow, Poland
2 Department of Biophysics, Jagiellonian University, Krakow, Poland
3 National Research Institute of Animal Production, Balice, Poland
4 M. Sklodowska-Curie Memorial Institute, Gliwice, Poland
Introduction
Macrophages and hepatic cells are important players
in the inflammatory processes initiated in response to a
variety of agents, including viral or bacterial infections,
thermal and mechanical trauma, or malignant growth.
Recently, we have used human monocyte-derived mac-
rophages exposed to interleukins IL-1b and IL-6, and

studied changes of gene expression using microarrays
[1]. Among the cytokine-modulated genes, one was
highly activated by IL-1b, but not by IL-6. This tran-
script was found to correspond to the ZC3H12A gene
(also called the MCPIP gene), originally described as a
gene activated by monocyte chemoattractant protein-1
(MCP-1) and encoding MCPIP [2]. This protein con-
sists of 599 amino acids, and has a putative nuclear
localization signal and two proline-rich potential acti-
vation domains, one between residues 100 and 126 and
the other between residues 458 and 536 [2]. Moreover,
amino acid sequence analysis revealed a single zinc-
Keywords
IL-1b transcript degradation; inflammation;
MCPIP; PIN domain; RNase
Correspondence
J. Jura, Department of Cell Biochemistry,
Faculty of Biochemistry, Biophysics and
Biotechnology, 7 Gronostajowa Street,
30-387 Krakow, Poland
Fax: +48 12 664 6902
Tel: +48 12 664 6359
E-mail:
(Received 28 April 2009, revised
23 September 2009, accepted 20 October
2009)
doi:10.1111/j.1742-4658.2009.07452.x
In human monocyte-derived macrophages, the MCPIP gene (monocyte
chemoattractant protein-induced protein) is strongly activated by inter-
leukin-1b (IL-1b). Using bioinformatics, a PIN domain was identified, span-

ning amino acids 130-280; such domains are known to possess structural
features of RNases. Recently, RNase properties of MCPIP were confirmed
on transcripts coding for interleukins IL-6 and IL-12p40. Here we present
evidence that siRNA-mediated inhibition of the MCPIP gene expression
increases the level of the IL-1b transcript in cells stimulated with LPS,
whereas overexpression of MCPIP exerts opposite effects. Cells with an
increased level of wild-type MCPIP showed lower levels of IL-1b mRNA.
However, this was not observed when mutant forms of MCPIP, either
entirely lacking the PIN domain or with point mutations in this domain,
were used. The results of experiments with actinomycin D indicate that lower
levels of IL-1b mRNA are due to shortening of the IL-1b transcript half-life,
and are not related to the presence of AU-rich elements in the 3¢ UTR. The
interaction of the MCPIP with transcripts of both IL-1b and MCPIP
observed in an RNA immunoprecipitation assay suggests that this novel
RNase may be involved in the regulation of expression of several genes.
Abbreviations
ARE, AU-rich element; dsRBD, double-stranded RNA-binding domain; GM-CSF, granulocyte ⁄ macrophage colony-stimulating factor; iNOS,
inducible nitric oxide synthase; IL-1b, interleukin-1b; IL-6, interleukin-6; KH domain, K homology domain; LPS, lipopolysaccharide; MCP-1,
monocyte chemoattractant protein-1; MCPIP, MCP-1 induced protein; PAZ, Piwi Argonaut and Zwille; PDB, protein data bank; PMA, phorbol
12-myristate-13-acetate; RRM, RNA-recognition motif; TNF, tumor necrosis factor.
7386 FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS
finger motif (CCCH) in MCPIP, which prompted
Zhou et al. to propose a hypothetical function for this
molecule as a novel transcription factor [2]. They also
showed that MCPIP is a negative regulator of macro-
phage activation, affecting LPS-induced tumor necrosis
factor (TNF) and inducible nitric oxide synthase
(iNOS) promoters [3].
Recently, Matsushita et al. [4] reported that MCPIP
has an essential role in preventing immune disorders.

Mice with MCPIP deficiency (Zc3h12a) ⁄ )) showed
growth retardation, severe splenomegaly and lymphoa-
denopathy. Matsushita et al. [4] also observed infiltra-
tion of plasma cells in the lung, the para-epithelium of
the bile duct, the pancreas, lymph nodes and spleen.
Moreover, Zc3h12a-deficient mice suffered from severe
anemia, had increased serum immunoglobulin levels and
showed auto-antibody production. Transcriptome com-
parison of LPS-treated macrophages from wild-type and
Zc3h12a-deficient mice revealed that a particular set of
genes, namely IL-6, Ifng, Calcr and Sprr2d, was highly
expressed in knockout mice [4]. Further experiments
showed that MCPIP has RNase properties and regulates
IL-6 and IL12p40 mRNA stability [4]. Regulation of the
half-life of the IL-6 transcript involved its 3¢ UTR but
was independent of AU-rich elements (AREs) [4].
In our studies, we have observed that, in addition to
IL-1b, the MCPIP gene is also strongly activated by
LPS, TNFa and phorbol 12-myristate-13-acetate
(PMA). Moreover, we found that MCPIP has RNase
properties and regulates the half-life of IL-1b mRNA
and its own transcript. Overexpression of MCPIP in cells
activated with LPS results in downregulation of the
IL-1b transcript and silencing of MCPIP gene expression
has an opposite effect on the IL-1b mRNA level. RNase
properties of MCPIP are exerted by the PIN domain, as
shown by experiments with mutated MCPIP (either lack-
ing the PIN domain, or with two conservative amino
acid mutations within PIN domain). Interaction of
MCPIP with the IL-1b transcript and its own transcript

was confirmed by an RNA immunoprecipitation assay.
Our findings are in agreement with those of Mats-
ushita et al. [4]. The data show that MCPIP has more
inflammatory targets than described so far, and that
this protein is an important regulator of inflammatory
processes.
Results
Influence of proinflammatory cytokines on MCPIP
transcript level
Real-time PCR was used to determine the modulation
of expression of the MCPIP gene in two types of cells:
HepG2 cells stimulated with IL-1b or IL-6, and mono-
cyte lymphoma cells stimulated with IL-1b, TNFa,
PMA or LPS. As shown in Fig. 1A, the MCPIP tran-
script level in HepG2 cells stimulated with IL-1b for
A
B
C
Fig. 1. Regulation of MCPIP gene expression. Changes in MCPIP
gene expression in HepG2 and U937 cells measured by real-time
PCR. The results are means ± SD of three independent experi-
ments (*P < 0.05; **P < 0.01 versus control). (A) Changes in the
expression of the MCPIP gene after stimulation of HepG2 cells
with IL-1b (15 ngÆmL
)1
). Cells were stimulated for 0.25, 0.5, 1, 2,
4, 8, 12 and 24 h. Unstimulated cells collected at each time point
served as controls. (B) Expression of the MCPIP gene after stimula-
tion of HepG2 cells with IL-6 (15 ngÆmL
)1

). Cells were stimulated
for 4, 8, 12 and 24 h. Unstimulated cells served as a control. (C)
Changes in expression of the MCPIP gene after stimulation of
U937 cells with IL-1b (15 ngÆmL
)1
), TNFa (10 ngÆmL
)1
), LPS
(100 ngÆmL
)1
) or PMA (100 nM). Cells were stimulated for 1, 2 and
4 h with the above factors. Unstimulated cells served as a control.
D. Mizgalska et al. MCPIP protein as an RNase
FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS 7387
periods ranging from 0.25–4 h increases up to approxi-
mately 35 times compared to unstimulated controls.
The maximum expression is observed at 2 h after stim-
ulation, followed by a slow decrease. In contrast to
IL-1b, another cytokine important for hepatic cells,
IL-6, had no effect on MCPIP gene expression
(Fig. 1B). In U937 cells, TNFa appears to be the stron-
gest stimulant, increasing the MCPIP transcript level
more than 7-fold in comparison with unstimulated
cells, but IL-1b, LPS and PMA also activated the
MCPIP gene (Fig. 1C). These experiments showed that
MCPIP belongs to the group of early-response genes.
Bioinformatic analysis of MCPIP amino acid
sequence
Application of fold recognition methods allowed to
identify potential structural templates for the 130-290

region of the MCPIP1 sequence [5]. Ten top scoring
Protein Data Bank (PDB) records included 1O4W
(the PIN domain from Archaeoglobus fulgidus AF0591
protein; six hits from the various fold recognition serv-
ers), 1A76 (FLAP endonuclease-1 from Methanococ-
cus jannaschii) and 1W8I (AF1683 protein of unknown
function from Archeoglobus Fulgidus). For greater con-
fidence that the structural alignments created by the
GeneSilico.pl server [6] are meaningful, we compared
the reported AF0591 PIN ⁄ MCPIP PIN alignment with
previously published alignments of human SMG6
(PDB ID: 2HWW) PIN and AF0591 PIN sequences
[7,8]. The resulting alignment is shown in Fig. 2B.
Four acidic residues (D141, E185, D226, D244) in PIN
domains involved in binding Mg
2+
ions are conserved
in all three sequences. Other conserved residues are
R263, V139 and N144. Determination of the MCPIP
domain architecture was completed using the DISO-
PRED [9] and SPRITZ servers [10], which consistently
indicated that regions 1-50, 90-130 and 290-540 of
MCPIP are disordered, with as yet unidentified func-
tion (Fig. 2A).
Distribution of MCPIP transcript in human tissues
and cellular localization of MCPIP protein
A blot loaded with mRNAs from several human tis-
sues was subjected to Northern blot analysis using a
molecular probe spanning the entire coding sequence
of the MCPIP gene or the b-actin gene (control). In

comparison with the b-actin transcript, the level of
MCPIP mRNA was very low. From the analyzed tis-
sues, highest expression of the MCPIP gene was
observed in leukocytes, with intermediate levels in
heart and placenta, and the lowest levels in spleen, kid-
ney, liver and lung. The transcript was not detectable
in brain, thymus, muscles, small intestine and colon
(Fig. 3A). The high expression of the MCPIP gene in
leucocytes suggests the functional importance of the
protein encoded by this transcript for the immuno-
logical system.
Originally, Zhou et al. [2] described localization of
MCPIP in the nucleus. Recently Matsushita et al. [4]
showed cytoplasmic localization of MCPIP. We ana-
lyzed the cellular localization of MCPIP using two
approaches. First, subcellular fractions were isolated
A
B
Fig. 2. The domain architecture of MCPIP. (A) The disordered regions (D.R.) are shown as grey rectangles (positions 1-50, 90-130 and 340-
540), the PIN domain is shown as a striped rectangle (positions 130-280), and the C3H type zinc finger (Z.F.) as a white rectangle (positions
306-322). Unannotated protein regions are shown as black bars. (B) Alignment of the SMG6 PIN domain (PDB: 2hww), the bacterial hypo-
thetical protein PIN domain (PDB: 1o4w) and the MCPIP PIN domain. Identical residues are indicated by asterisks, conserved substitutions
as semi-colons and semi-conserved substitution as dots. The PIN domain of MCPIP has conserved four acidic residues (D141, E185, D226
and D244) involved in binding Mg
2+
ions during RNA degradation.
MCPIP protein as an RNase D. Mizgalska et al.
7388 FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS
from HepG2 cells overexpressing MCPIP. The results
of western blot analysis indicate that MCPIP is local-

ized in the cytoskeleton fraction (Fig. 3B). In the
second approach, HepG2 cells were transfected either
with an empty vector, or with a vector expressing a
fusion of MCPIP with red fluorescent protein (RFP).
Using various concentrations of the plasmid vector
overexpressing the fusion protein MCPIP–RFP, we
observed that the protein was localized in the cyto-
plasm, in the form of granules (Fig. 3C). It is possible
that formation of granules is the result of stress (e.g.
related to overexpression). Such accumulation of pro-
tein has been described previously for other factors
engaged in mRNA degradation: for example, triste-
traproline and T-cell-restricted intracellular antigen 1
both accumulate in stress granules during environmen-
tal stress, which are regarded as dynamic cytoplasmic
foci that contain untranslated mRNAs [11,12].
Furthermore, the results of western blots show that
MCPIP is detected in cytoskeleton. These data are in
agreement with results obtained by Henics et al. [13],
who showed that cytoskeleton proteins have binding
capacity for proteins involved in mRNA metabolism.
Moreover, cytoskeleton proteins, such as actin or
tubulin, are thought to be crucial for compartmentali-
zation and translation of various mRNAs, including
those containing AU-rich instability motifs in the 3 ¢
UTR [14,15]. We performed co-immunoprecipitation
and mass spectrometry analyses to identify putative
proteins interacting with MCPIP (data not shown). Of
the identified proteins, 73% are proteins involved in
mRNA stability, but there are also proteins involved

in other cellular processes, such as actin. Further stud-
ies characterizing proteins interacting with MCPIP will
clarify the dynamics and features of this protein.
Role of MCPIP in regulation of the endogenous
IL-1b transcript level
After identification of a PIN domain in the MCPIP by
use of bioinformatics, we decided to check whether
MCPIP has any effect on IL-1b transcript level. We
hypothesized that if IL-1b strongly regulates MCPIP
gene expression, MCPIP may be involved in regulation
of the IL-1b mRNA level. To elucidate the effect of
MCPIP on the endogenous IL-1b transcript level, we
used primary cultures of human fibroblasts. These cells
are a good model to study inflammatory processes,
because, in contrast to HepG2 cells, expression of
IL-1b in fibroblasts is significant and these cells are
also important in inflammatory defence. Human fibro-
blasts were transfected with either MCPIP-specific siR-
NA or a vector overexpressing MCPIP and then
A
B
C
Brain
Heart
Skeletal muscle
Colon
Thyroid
Spleen
Kidney
Cytosol

Membrane
Nucleus
Cytoskeleton
Liver
Small intestine
Placenta
Lungs
Leukocytes
Fig. 3. Tissue distribution of transcript and cellular localization of
MCPIP. (A) A human tissue Northern blot (Clontech) was hybridized
with a radioactively labelled MCPIP cDNA probe and visualized by
autoradiography. A b-actin probe served as a control for equal loading.
A single 2.0 kb band was seen in all lanes. A 1.8 kb actin isoform is vis-
ible for heart and skeletal muscle. The duration of exposure of the
Northern blot with MCPIP probe was 20 times longer compared to the
blot with the b-actin probe. (B) A Qproteome kit (Qiagen) was used to
isolate cytosolic, membrane, nuclear and cytoskeletal proteins. Pro-
tein concentrations were determined using a BCA protein assay
(Sigma). Aliquots of 10 lg from each fraction were used in western
blot analysis. Polyclonal antibodies against MCPIP were used at
concentration of 1 lgÆmL
)1
. As markers specific for cytoskeleton,
nucleus, membrane and cytosolic fraction, respectively, antibodies
against actin (mouse monoclonal, 1:4000, Sigma), lamin C (rabbit poly-
clonal, 1:1000, Abcam), MnSOD (rabbit polyclonal, 1:2000, Sigma)
and GAPDH (rabbit polyclonal, 1:1000, Abcam) were used. (C) A
genetic construct overexpressing recombinant MCPIP with fluores-
cent red protein in the pmaxFP-Red-N vector (Amaxa) was generated
for determination of the cellular localization of MCPIP. HepG2 cells

were transfected with a vector overexpressing MCPIP and with an
empty vector (control) using Lipofectamine 2000, according to
the manufacturer’s protocol. 4¢-6-diamidino-2-phenylindole (DAPI)
staining was performed to visualize nuclei. Localization of MCPIP in
cytoplasm was determined by use of a fluorescent microscope.
D. Mizgalska et al. MCPIP protein as an RNase
FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS 7389
stimulated with LPS (Fig. 4A,B). Inhibition of MCPIP
gene expression by siRNA (by approximately 40%) led
to an increase in the endogenous IL-1b transcript level
(by approximately 60%) in comparison to control cells
transfected with scrambled siRNA. In contrast, cells
overexpressing MCPIP had lower IL-1b transcript level
(by approximately 80%) in comparison to control cells
transfected with an empty vector. Despite the moder-
ate transfection efficiency (30% for overexpression and
25% for siRNA), the overall conclusion from the
experiment with fibroblasts is that MCPIP regulates
the amount of IL-1b mRNA.
Involvement of PIN domain of MCPIP in the
stability of IL-1 b mRNA
In order to find out whether the PIN domain in
MCPIP is responsible for IL-1b mRNA regulation, we
performed an experiment with HepG2 cells (Fig. 5A).
These cells were transfected with a genetic construct
encoding IL-1b cDNA and either wild-type MCPIP, a
mutant form of MCPIP (D141A and D226A; D ⁄ A-
MCPIP), or an empty vector. A drastic decrease in the
IL-1b transcript level was observed when cells were
co-transfected with a vector coding for wild-type

MCPIP in comparison to cells transfected with an
empty vector or one expressing D ⁄ A-MCPIP (Fig. 5,
lanes 2,3 and 4, lower band). The levels of mutant and
wild-type MCPIP mRNAs were higher when these
vectors were co-transfected with a vector overexpress-
ing IL-1b (Fig. 5, lanes 3 and 4, upper bands). The
absence of a plasmid encoding IL-1b prevented the
increase in the MCPIP mRNA level for both forms
(Fig. 5A, lanes 5 and 6, upper bands). Moreover, in a
co-transfection study with the IL-1b-expressing vector,
the level of mRNA for the mutant form of MCPIP was
higher (Fig. 5A, lane 4, upper band) than when wild-
type MCPIP was co-transfected (Fig. 5A, lane 3, upper
band). At this stage of the study, the following hypo-
thesis may be presented: some overexpressed IL-1b is
secreted and stimulates the endogenous form of MCPIP,
resulting in the observed higher level of MCPIP when
A
B
Fig. 4. MCPIP regulates IL-1b transcript in LPS-treated human
fibroblasts. (A) Human fibroblasts were transfected with siRNA
specific either for MCPIP (50 n
M)orGAPDH (50 nM), or with non-
specific siRNA (negative control, 50 n
M). The transfection efficiency
observed for carboxyfluorescein (FAM)-labeled scrambled siRNA
was estimated as 25%. Cells treated only with the transfection
reagent siPORT NeoFX were used as a control. At 48 h post-trans-
fection, cells were treated with LPS (100 ngÆmL
)1

) for 6 h. RNA
was isolated and real-time PCR was performed using primers spe-
cific for MCPIP cDNA (control of silencing, right panel) or IL-1b (left
panel), using GAPDH as a positive control (data not shown). For
GAPDH, 50% inhibition of gene expression was obtained (data not
shown) and for MCPIP approximately 40% inhibition of gene
expression was obtained. The IL-1b transcript level after LPS treat-
ment was expressed as a percentage of the basal transcript level
in untreated cells. The IL-1b transcript level was compared to that
in cells transfected with scrambled siRNA (assumed to be 100%).
The values represent the means of three independent experiments
(**P < 0.01, versus control). (B) Human fibroblasts were transfect-
ed with an empty vector (pcDNA3) or a vector overexpressing wild-
type MCPIP. The obtained transfection efficiency was 30%, as esti-
mated by use of GFP coding vector. Twenty-four hours post-trans-
fection, cells were treated with LPS (100 ngÆmL
)1
) for 6 h. RNA
was isolated and real-time PCR was performed using primers spe-
cific for MCPIP cDNA (control of overexpression, right panel) or for
IL-1b cDNA (left panel). In each case, the IL-1b transcript level after
LPS treatment was expressed as a percentage of the basal
transcript level in untreated cells. The IL-1b transcript level was
compared to that in cells transfected with pcDNA3 (assumed to be
100%). The values represent the means of three independent
experiments (**P < 0.01, versus control).
MCPIP protein as an RNase D. Mizgalska et al.
7390 FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS
plasmids expressing IL-1b are co-transfected. Evidence
supporting this idea includes the presence of MCPIP in

HepG2 cells overexpressing IL-1b and its absence in
cells transfected with an empty vector (data not shown).
Wild-type MCPIP (endogenous and exogenous) starts
degradation of IL-1b mRNA and also of its own tran-
script. For this reason, the level of mRNA for the wild-
type form of MCPIP and the level of mRNA for IL-1b
(Fig. 5A, lane 3) were lower in comparison to the level
of mRNA for MCPIP with a mutation within the PIN
domain (Fig. 5A, lane 4). The same tendency is seen at
the protein level (data not shown).
In order to determine whether the observed IL-1b
mRNA changes are due to a decrease in transcription
or an increased degradation rate, we performed experi-
ments with actinomycin D. HepG2 cells stably
overexpressing IL-1b mRNA were co-transfected with
an empty vector or a vector overexpressing wild-type
MCPIP or its mutant forms [mutMCPIP (without the
PIN domain) and D ⁄ A-MCPIP (with two mutated
residues within PIN domain)]. The cells were harvested
0, 3, 6 and 9 h after addition of actinomycin D, and
the level of exogenous IL-1b mRNA was analyzed by
Northern blot. A significant reduction in the IL-1b
transcript level was observed only when the wild-type
MCPIP was overexpressed (Fig. 5B).
Importance of AREs in regulation of IL-1b mRNA
by MCPIP
In order to determine whether AREs are important in
mRNA degradation triggered by MCPIP, we used
A
B

Fig. 5. Involvement of MCPIP in IL-1b mRNA stability. (A) For determination of the stability of exogenous IL-1b mRNA, HepG2 cells were
transfected with an empty vector pcDNA3, a vector overexpressing IL-1b mRNA, and a vector overexpressing either wild-type MCPIP or its
mutant form containing the D141A and D226A mutations (D ⁄ A-MCPIP). The level of transcripts coding for MCPIP and IL-1b was determined
by Northern blot. Lane 1, pcDNA3; lane 2, pcDNA3 ⁄ IL-1b; lane 3, wild-type MCPIP ⁄ IL-1b; lane 4, D ⁄ A-MCPIP ⁄ IL1b; lane 5, pcDNA3 ⁄ MCPIP;
lane 6, pcDNA3 ⁄ D ⁄ A-MCPIP. The figure is representative of three independent experiments, and the level of analyzed transcripts was com-
pared to that for 28S rRNA. (B) To determine the half life of the IL-1b transcript, HepG2 cells stably overexpressing IL-1b mRNA were trans-
fected either with an empty vector or with vector overexpressing wild-type MCPIP or its mutant forms: mutMCPIP without the PIN domain
or D ⁄ A-MCPIP with two mutated amino acids located within the PIN domain (D141A and D226A). Transcription was inhibited by addition of
actinomycin D, and RNA was collected after 3, 6 and 9 h. The mRNA level for IL-1b (top panel) and MCPIP (bottom panel) was determined
by Northern blot. The middle panel shows inverted images of ethidium bromide-stained 18S and 28S rRNA on the membrane. The graphs
below the Northern blot present data from densitometry of the IL-1b transcript level normalized to the rRNA amount, calculated from four
independent experiments (*P < 0.05 compared with control).
D. Mizgalska et al. MCPIP protein as an RNase
FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS 7391
A
B
C
Fig. 6. Importance of AREs in MCPIP-dependent mRNA degradation. (A) The 3¢ UTR sequence of the transcript coding for IL-1b (NCBI
accession number NM_000576). The genetic construct used in the transfection experiment encoding IL-1b cDNA without AREs (IL-1)ARE)
contained 343 bp of the 3¢ UTR region (nucleotides 898–1227 of the 3¢ UTR region). The construct encoding IL-1b cDNA with AREs (IL-
1+ARE) contained 423 bp of the 3¢ UTR (nucleotides 898–1318 of the 3¢ UTR region). The stop codon (positions 895-897) and AU-rich region
(positions 1242-1258) are indicated by open boxes. Capital letters indicate the sequence of reverse primers used in generation of the PCR
products used for construction of vectors expressing transcripts for IL-1b with ⁄ without AREs (IL-1)ARE and IL-1+ARE). Primers used for
amplification of AREs used in the generation of a genetic construct containing a minimal promoter and cDNA for luciferase with AREs (p-luc-
ARE) are indicated by shaded boxes. (B) Three genetic constructs were generated in order to determine whether MCPIP is involved in IL-1b
mRNA degradation: IL-1b cDNA without AREs [IL-1)ARE; detailed description in (A)], IL-1b cDNA with AREs [IL-1+ARE; detailed description
in (A)] and MCPIP cDNA (MCPIP). HepG2 cells were transfected with the following combination of vectors: pcDNA3, pcDNA3 ⁄ IL-1+ARE,
pcDNA3 ⁄ IL-1)ARE, pcDNA3 ⁄ MCPIP, MCPIP ⁄ IL-1+ARE, MCPIP ⁄ IL-1)ARE and pcDNA3 ⁄ MCPIP. The pcDNA3 vector served as a control for
the experiment, and to equalize the amount of introduced genetic material. The level of IL-1b and MCPIP transcripts was measured by
Northern blot 24 h after transfection. In lanes 1 and 6, the endogenous IL-1b transcript is visible. The left panel is representative of three

independent experiments. The panel on the right shows densitometric measurements of the IL-1b transcript level normalized to the 18S
rRNA band. Student’s t test was used to determine significant differences from the control (*P < 0.05, **P < 0.01). (C) Two genetic con-
structs consisting of a minimal promoter and cDNA for luciferase with (p-luc-ARE) or without AREs (p-luc) from the IL-1b transcript were
generated. HepG2 cells were co-transfected with combination of p-luc-ARE and pcDNA3 empty vector (control), or wild-type MCPIP or a
mutant form of MCPIP lacking the domain with RNase activity. The same combinations were used for the plasmid expressing luciferase
transcript without ARE sequences in the 3¢ UTR (p-luc). As an internal transfection control, a pEF1 ⁄ Myc-His ⁄ Gal vector encoding b-galactosi-
dase was used. The bars represent the luciferase activity calculated from three independent experiments, normalized to that for b-galactosi-
dase. Overexpression of MCPIP was confirmed by western blotting (right upper panel). Semi-quantitative RT-PCR was performed for the
luciferase transcript to show the reporter transcript level (lower panel). The GAPDH
transcript served as a housekeeping gene.
MCPIP protein as an RNase D. Mizgalska et al.
7392 FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS
three genetic constructs: (a) a construct overexpressing
IL-1b transcript with AREs (long 3¢ UTR, Fig. 6A),
(b) a construct overexpressing IL-1b transcript without
AREs (short 3¢ UTR, Fig. 6A), and (c) a construct
overexpressing the MCPIP transcript. HepG2 cells
were co-transfected with vector containing IL-1b
cDNA with or without AREs or with the vector con-
taining MCPIP cDNA. The levels of IL-1b transcript
with ⁄ without AREs and the level of MCPIP transcript
were determined using Northern blotting 24 h after
transfection. The endogenous MCPIP transcript is not
detectable due to the short exposure time.
The level of IL-1b transcripts containing AREs was
lower than the level of IL-1b mRNA without AREs
(Fig. 6B, lanes 2–5). These results are in agreement with
data indicating that transcripts with AREs are more sus-
ceptible to degradation [16]. In both cases, overexpres-
sion of MCPIP (IL-1b + ARE and IL - 1b )ARE)

triggers a significant decrease in the level of IL-1b
mRNAs. Thus, the MCPIP level correlates with the fast
degradation of IL-1b transcript independently of the
presence of AREs (Fig. 6B, lanes 2 and 5). It is possible
that in the region located before the AREs, consisting of
343 bp of the 3¢ UTR of the IL-1b transcript, there is a
putative RNA stem–loop structure sequence similar to
that present in the 3¢ UTR of IL-6 mRNA [4,17]. The
presence of such a sequence may account for the degra-
dation of both types of IL-1b transcripts, with and with-
out AREs. The influence of MCPIP on the IL-1b
transcript level was also noted for the endogenous IL-1b
transcript (Fig.6B, lane 1 versus lane 6).
To confirm that AREs are not involved in the
observed IL-1b transcript turnover, we prepared two
genetic constructs consisting of a minimal promoter
and the cDNA for luciferase with (p-luc)ARE) or
without AREs (p-luc) from the IL-1b transcript
(Fig. 6A). The plasmid p-luc)ARE was transfected
into HepG2 cells in combination with a plasmid cod-
ing for wild-type MCPIP or its mutant form lacking
the PIN domain. The same combination was used for
the plasmid expressing the luciferase transcript without
ARE sequences in the 3¢ UTR (p-luc). The luciferase
activity was measured after 22 h. As shown in Fig. 6C,
the presence of ARE sequence has no influence on the
MCPIP-dependent turnover of luciferase transcript
by MCPIP. In both cases, when the constructs
p-luc)ARE or p-luc were used, differences in luciferase
transcript level ⁄ activity were not detectable between

the control and samples overexpressing wild-type
MCPIP or its mutant form lacking the PIN domain.
Nevertheless, the levels of luciferase transcript
(Fig. 6C, right panel) and luciferase activity (Fig. 6C,
left panel) were significantly lower in all samples
(including control) when AREs were present. This
A
B
Fig. 7. RNA immunoprecipitation. (A) Control of the experimental model. HepG2 cells were transfected with a vector overexpressing wild-
type MCPIP. RNA was isolated and cDNA synthesis was performed. Expression of genes encoding exogenous MCPIP (cells with a vector
overexpressing MCPIP) or endogenous transcripts encoding IL-1b and GADPH was determined by semi-quantitative PCR (with EF-2 as a
reference transcript). A total RNA sample (0.1 lg RNA) served as a negative control (cDNA-). (B) To bind RNA templates to proteins,
HepG2 cells were subjected to crosslinking with formaldehyde, lysed and incubated with antibodies specific for MCPIP or with IgG as a
negative control. Then the samples were digested with DNase to remove traces of genomic DNA. Reverse transcription was performed
using RNA isolated from immunoprecipitates, followed by PCR with MCPIP-, IL-1 b- and GADPH-specific primers (35 cycles). The same RNA
but not reverse-transcribed served as a negative control for the PCR reaction. Specific transcripts coding for MCPIP and IL-1b were detected
only in immunoprecipitates obtained with antibodies specific for MCPIP, but not with IgG antibodies. Specificity of the PCR product was
determined by restriction analysis (data not shown). PCR product was not observed for the negative control, i.e. with primers specific for
the GAPDH gene.
D. Mizgalska et al. MCPIP protein as an RNase
FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS 7393
means that IL-1b AREs are important in luciferase
mRNA degradation but this process is performed in
an MCPIP-independent manner.
Interaction of MCPIP protein with the IL-1b
transcript
An RT-PCR reaction was performed to confirm that
HepG2 cells synthesize endogenous IL-1b mRNA. As
shown in Fig. 7A, there is expression of the gene
encoding IL-1b, although the transcript level is low in

comparison to that of the GAPDH transcript. To
investigate interactions of MCPIP with the IL-1b tran-
script, immunoprecipitation of RNA-binding protein
was carried out. In this experiment, HepG2 cells were
transfected with vector overexpressing MCPIP. Then
HepG2 cells were subjected to crosslinking with form-
aldehyde in order to bind RNA to protein complexes,
lysed and incubated with antibodies specific for
MCPIP, or with IgG as a negative control. After
extraction of RNA from immunoprecipitates, reverse
transcription was performed, followed by PCR with
primers specific for MCPIP, IL-1b and GADPH
(Fig. 7B). The amount of IL-1b product was signifi-
cantly lower than that of MCPIP because endogenous
expression of the IL-1b gene in HepG2 cells is quite
low (Fig. 7A). Amplification of the MCPIP transcript
in the immunoprecipitates indicates that this protein
interacts with its own transcript and probably regulates
its stability. We did not observe any product in PCR
reactions with primers specific for GAPDH. This
indicates that IL-1b and MCPIP mRNAs, but not
GAPDH mRNA, interact with MCPIP. The obtained
result confirms the involvement of MCPIP in mRNA
processing.
Discussion
It is well known that gene expression of proinflam-
matory mediators is tightly controlled, both at the
transcriptional and post-transcriptional levels. The
majority of transcripts coding for cytokines and
chemokines (IL-1, IL-2, IL-6, TNFa, GM-CSF, inter-

ferons b and c) contain regulatory elements in the 3¢
UTR that are important for the half life of cytokine
mRNAs. For this reason, transcription of genes encod-
ing cytokines is not sufficient to guarantee the appear-
ance of functional proteins. As reported by Kaspar
and Gehrke [18] and Dinarello [19], stimulation of
macrophages by mild stimulants (such as C5a comple-
ment or some modified proteins) leads to accumulation
of specific mRNAs without their translation. This is
due to the formation of translationally inactive
mRNA–protein complexes. These complexes may
either result in mRNA degradation or its conversion
into translationally active forms by the action of
strong stimulants, such as endotoxin (LPS).
One of the regulatory sequences involved in deter-
mining mRNA half life are the AREs, which can stim-
ulate deadenylation of mRNA [20] and its degradation
by the exosome in the 3¢fi5¢ direction [21]. In addi-
tion to mRNA degradation mediated by the exosome,
RNA can be degraded through removal of the
7-methyl guanosine cap, followed by 5¢fi3¢ mRNA
degradation in cytoplasmic processing bodies
(P-bodies) [22]. One of the best known examples of
AREs binding proteins involved in inflammatory
processes is tristetraprolin [23]. It was shown that
tristetraprolin enhances the deadenylation and
degradation of mRNA for granulocyte ⁄ macrophage
colony-stimulating factor, TNFa and IL-2.
Using human monocyte-derived macrophages
exposed to IL-1b, we found that the MCPIP gene is

strongly regulated by this cytokine [1]. Our study on
the function of the novel protein started with the use
of bioinformatic tools to analyse MCPIP structure.
It was already known that MCPIP has a single
CCCH zinc-binding domain [2], which, similarly to
K homology (KH), RNA-recognition motif (RRM),
double-stranded RNA-binding domain (dsRBD) and
Piwi Argonaut and Zwille (PAZ), represents an
RNA-binding domain. These domains are usually
linked to larger globular enzymatic domains, such as
nucleases and nucleotide modifying enzymes that are
important in RNA metabolism [24,25]. Our bioinfor-
matic analyses of the MCPIP sequence showed that
MCPIP contains a centrally located PIN-like domain,
presumably with endo ⁄ exonuclease activity, together
with relatively large fragments that show disordered
structure.
The PIN domain spans amino acids 130-280.
Despite only marginal sequence similarity, the various
PIN domains that are found in all kingdoms of life
have a highly conserved structure with 5¢fi3¢ exonu-
clease activity [26]. We found that the PIN domain of
MCPIP, similar to another protein with confirmed
RNase activity – SMG6 (2hww) [7] – has the canonical
four conserved acidic residues (D141, E185, D226 and
D244; Fig. 2B) that are involved in binding Mg
2+
ions
during RNA degradation. By analogy to SMG6, we
hypothesized that MCPIP may be directly involved in

mRNA degradation, with IL-1b mRNA being one of
the targets. To confirm our hypothesis, we used genetic
constructs overexpressing full-length cDNA for IL-1b,
wild-type MCPIP and its mutant forms. The level of
IL-1b mRNA was decreased in cells overexpressing
MCPIP protein as an RNase D. Mizgalska et al.
7394 FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS
wild-type MCPIP in comparison with a mutant form.
Inhibition of MCPIP expression by siRNA leads to
the increase in IL-1b transcript level. These data are in
agreement with the results obtained by Liang et al. [3].
MCPIP-mediated downregulation of the half life of the
IL-1b transcript was confirmed in the presence of acti-
nomycin D in HepG2 cells stably overexpressing IL-1b
mRNA. We found that the stability of IL-1b transcript
is dependent on the type of MCPIP expressed by the
vector introduced into these cells. In control cells
(empty vector) and cells overexpressing a mutant form
of MCPIP (without the PIN domain, or MCPIP with
two mutated amino acid residues), the transcript cod-
ing for IL-1b was stable. However, when wild-type
MCPIP was present, a decrease in the transcript level
for IL-1b was observed. We have also shown that
IL-1b transcript degradation is ARE-independent. We
used genetic constructs with cDNA coding for interleu-
kin-1 but either containing or lacking the region with
ARE sequences. Co-transfection of constructs with or
without ARE sequences with the vector overexpressing
MCPIP resulted in efficient disappearance of the IL-1b
transcript. The degradation was independent of the

presence of AREs, as confirmed in the experiment with
luciferase vector.
Recently, Matsushita et al. showed that MCPIP is
engaged in the regulation of mRNA stability in inflam-
matory processes. These authors found that MCPIP
has RNase properties and regulates the stability of IL-6
and IL-12p40 mRNA [4]. MCPIP-dependent degrada-
tion of IL-6 mRNA is performed via its 3¢ UTR, but
the mechanism is independent of ARE sequences [4].
Data obtained in knockout mice for the MCPIP gene
showed that MCPIP is essential for inhibition of the
development of severe autoimmune responses that cul-
minate in lethality [4]. Knockout mice showed increased
production of IL-6 and IL-12p40 but not TNF alpha in
macrophages, due to failure of mRNA degradation.
Our data showing that MCPIP is involved in
mRNA degradation are in agreement with the results
of Matsushita et al. [4]. As for IL-6 mRNA [4], cis-act-
ing elements that are important in MCPIP-mediated
IL-1b mRNA degradation are localized in the 3¢ UTR
of this transcript. Moreover, as for the IL-6 transcript
[4], IL-1b ARE sequences do not participate in degra-
dation performed by MCPIP.
In addition to regulation of transcripts encoding
proinflammatory cytokines (IL-1b and IL-6), it
appears that MCPIP regulates its own transcript. This
was demonstrated by the RNA immunoprecipitation
experiment. Self-regulation has already been described
for proteins engaged in the control of RNA stability,
such as tristetraprolin [27].

Our results confirm the data of Matsushita et al. [4]
showing that MCPIP has RNase properties performed
by the PIN domain. We found that MCPIP is acti-
vated by many stimulants, and regulates the level of
IL-1b mRNA and of its own transcript. Further stud-
ies are necessary to clarify the biological role of
MCPIP as a nuclease and to identify the cis-acting
sequences in the 3¢ UTR of the IL-1b and MCPIP
transcripts that directly interact with MCPIP. How-
ever, as MCPIP regulates the stability of transcripts
encoding the two major proinflammatory cytokines
IL-1b and IL-6, as well as of its own transcript, it may
be regarded as a very important potential regulator of
inflammatory processes.
Experimental procedures
Cell culture and stimulation with cytokines
All cell lines were grown under standard conditions ()37 °C
and 5% CO
2
). HepG2 cells (American Type Culture Collec-
tion, Manassas, VA, USA) were cultured in Dulbecco’s mod-
ified Eagle’s medium with 1000 mgÆL
)1
d-glucose (Gibco,
Carlsbad, CA, USA) supplemented with 5% fetal bovine
serum. Cells were grown in plastic Petri dishes (6 cm in diam-
eter, BD Biosciences, Franklin Lakes, NJ, USA) to approxi-
mately 70% confluence. At 15 h before stimulation with
cytokine, cells were washed twice in NaCl ⁄ P
i

and placed in
Dulbecco’s modified Eagle’s medium containing 0.5% fetal
bovine serum. U937 cells (American Type Culture Collec-
tion) were cultured and stimulated in RPMI-1640 medium
(Gibco ⁄ BRL) supplemented with fetal bovine serum to a
final concentration of 10%. Human skin fibroblasts were
grown in modified Eagle’s medium (MEM) with 1000 mgÆL
)1
d-glucose supplemented with 10% fetal bovine serum.
The final concentrations of stimulants were optimized
experimentally, and were as follows: human IL-1b,15
ngÆmL
)1
(BiolMol, Plymouth Meeting, PA, USA); human
recombinant IL-6, 15 ngÆmL
)1
(Strathmann Biotec AG,
Hamburg, Germany); human TNFa,10ngÆmL
)1
(Suntory
Pharmaceuticals, Cambridge, MA, USA); LPS, 100 ngÆmL
)1
(Sigma, St Louis, MO, USA); PMA, 100 nm (Calbiochem,
Darmstadt, Germany).
Real-time PCR
Total RNA was isolated from unstimulated cells (control)
and cytokine-stimulated cells, as described previously [28].
The first strand of cDNA was synthesized from 2 lg of total
RNA using MMLV reverse transcriptase (Promega, Madi-
son, WI, USA) and oligo(dT) primer. Real-time PCR was

performed using SYBR Green PCR master mix (DyNAmoÔ
HS SYBR Green qPCR; Finnzymes, Espoo, Finland), 1 ll
cDNA and 20 ng suitable forward and reverse primers. For
D. Mizgalska et al. MCPIP protein as an RNase
FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS 7395
amplification, the following primers were used: IL-1b cDNA,
5¢-GATGTCTGGTCCATATGAACTG-3¢ (forward) and
5¢-TGGGATCTACACTCTCCAGC-3¢ (reverse); MCPIP
cDNA, 5¢-GGAAGCAGCCGTGTCCCTATG-3¢ (forward)
and 5¢-TCCAGGCTGCACTGCTCACTC-3¢ (reverse). Each
sample was normalized to reference genes: elongation
factor 2 (EF-2), 5¢-GACATCACCAAGGGTGTGCA-3¢
(forward) and 5¢-TTCAGCACACTGGCATAGAGGC-3¢
(reverse); GAPDH,5¢-CCGAGCCACATCGCTCAGAC-3¢
(forward) and 5¢-GTTGAGGTCAATGAAGGGGTC-3¢
(reverse). The relative level of transcripts was quantified by
the DDC
T
method.
Northern blot analysis
For determination of MCPIP and IL-1b transcript levels,
10 lg of total RNA were separated in a 1% formaldehyde
agarose gel and blotted to a nitrocellulose membrane. Pre-
hybridization and hybridization were performed at 65 °Cin
1% SDS, 1 m NaCl, 10% dextran sulfate solution. Probes
were generated by PCR amplification of the entire coding
region of studied genes. A random primer labelling kit
(Promega) was used to label 30 ng of PCR product with
[a-
32

P]dCTP. After washing (20 min at room temperature
in 2· SSC, 20 min at 65 °Cin2· SSC, and twice for
20 min at 65 °Cin1· SSC ⁄ 1% SDS), RNA blots were
scanned using a Molecular Imager FX (Bio-Rad, Hercules,
CA, USA). All densitometry measurements were performed
using Quantity One software, and transcripts amounts were
normalized to the rRNA level.
In order to analyse the tissue distribution of the MCPIP
transcript, a blot loaded with 2 lg of each of the poly(A)
+
RNAs from several human tissues (Clontech, Mountain
View, CA, USA) was used. The blot was later stripped and
reprobed using the b-actin probe provided with the kit to
ensure equal loading. The probes for MCPIP and actin
transcript were labelled with the same specific activity.
Hybridization was performed as described above.
Bioinformatic analysis of the MCPIP amino acid
sequence
The Genesilico.pl server [5] was used to predict the domain
architecture of MCPIP. The disordered regions of MCPIP
were predicted using the DISOPRED [9] and SPRITZ [10]
servers. The PCONS method [6] was used for consensus fold
recognition prediction for the MCPIP PIN domain. Multiple
sequence alignment of target ⁄ template PIN domain
sequences was performed using clustalx [8].
Generation of genetic constructs
The coding sequences of the wild-type and mutant forms of
MCPIP lacking the PIN domain were obtained by two-step
PCR. For the MCPIP gene, a first round of PCR was car-
ried out using forward primer 5¢-CCGCTGGCGCATG

GCGGGTAGG-3¢ and reverse primer 5¢-GGGGTGGGCC
TCAGGGCTGGG-3¢. Then nested primers 5¢-GTCTGA
GCTATGAGTGGCCC-3¢ (forward) with a restriction site
for KpnI and 5¢-CAGCTTACTCACTGGGGTGC-3¢
(reverse) with a restriction site for EcoRI were used to
obtain a region of 1813 bp, corresponding to positions -9
to 1804 relative to the start codon (ATG) of the sequence
NM_025079. The obtained fragment was cloned into a
pcDNA3 vector (Invitrogen, Carlsbad, CA).
The same sequence was used to design primers for gener-
ation of a mutant form of MCPIP, without a PIN domain
(mutMCPIP). The fragment from 412-888 bp, correspond-
ing to the PIN domain, was removed from the entire cod-
ing sequence of the MCPIP gene. The coding sequence
lacking the PIN domain was obtained in two steps. First,
the fragment of 423 bp situated in front of the PIN domain
was amplified by PCR using forward primer 5¢-
CGGGATCCCGGAGTCTGAGCTATGAGTG-3¢ with a
restriction site for BamHI and reverse primer 5¢-CACT
GGTCTCAGGTCGCTG-3¢ with a restriction site for BsaI,
and cloned into pcDNA3 at the same restriction sites. Then
a second fragment situated after the PIN domain was gen-
erated by PCR using primers 5¢-AGCGACCTGAGAC
CACTCACTTTGGAGCAC-3¢ (BsaI) and 5¢-GGAATTC
CCTCACTGGGGTGCTGG-3¢ (EcoRI). This fragment
was cloned into pcDNA3 containing the first fragment at
the BsaI ⁄ EcoRI restriction sites. A genetic construct with
mutation of two conserved amino acids residues within the
PIN domain (D141A and D226A) was generated by site-
directed mutagenesis (Stratagene, Cedar Creek, TX, USA).

A genetic construct overexpressing recombinant MCPIP
with fluorescent red protein (pmaxFN-MCPIP-RFP) was
generated by PCR using forward primer 5¢-CGCTAG
CTATGGGCCCCTGTGGAGA-3¢ with a restriction site
for NheI and reverse primer 5¢-GAGATCTAACTCACTG
GGGTGCTGGG-3¢ with a restriction site for BglII. The
PCR product was cloned into the pmaxFP-Red-N vector
(Amaxa, Cologne, Germany) using the same restriction sites
to produce pmaxFP-MCPIP-RFP. The coding sequences of
IL-1b with AREs (IL-1+ARE) and IL-1b without AREs
(IL-1)ARE) were obtained by two-step PCR. For the IL-
1
b transcript with ⁄ without AREs, the first round of PCR
was performed using forward primer 5¢-CAAGGCACAA
CAGGCTGCTC-3¢ and reverse primer 5¢-GAGAGCA
CACCAGTCCAAATTG-3¢. Then a second round of PCR
was performed with nested primers 5¢-TGAAGCA
GCCATGGCAG-3¢ (forward) with a restriction site for
KpnI and 5¢-GGAAGCGGTTGCTCATCAG-3¢ (reverse)
with a restriction site for XbaI for IL-1)ARE or 5¢-CAGA
CACTGCTACTTCTTG-3¢ (reverse), also with an XbaI
restriction site, for IL-1+ARE (Fig. 6A). Primers were
designed using the sequence accession number NM_000576.
MCPIP protein as an RNase D. Mizgalska et al.
7396 FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS
All PCR products were cloned into the pcDNA3 vector
using appropriate restriction sites. All genetic constructs
were sequenced before the transfection experiment.
To prepare a reporter vector consisting of a minimal pro-
moter and cDNA for luciferase with or without AREs from

IL-1b cDNA (p-luc)ARE and p-luc), the region flanking
the AREs in IL-1b cDNA was amplified using forward
primer 5¢-ATTCGCTCCCACATTCTGATG-3¢ and reverse
primer 5¢-ACACTGCTACTTCTTGCC-3¢. Both primers
had restriction sites for XbaI. The PCR product was cloned
into pGL3-Promoter (Promega) at the XbaI site, and orien-
tation of the insert was confirmed by PCR.
Transient transfection
For transient transfection experiments, HepG2 cells and
human fibroblasts were cultured in six-well plates to
approximately 80% confluency. The next day, cells were
transfected using Lipofectamine 2000 (Invitrogen) in serum-
free Opti-MEM (Invitrogen), according to the manu-
facturer’s protocol. Empty pcDNA3 vector was used as a
control for the experiment or to equalize the amount of
genetic material introduced.
siRNA transfection
The pre-designed siRNAs obtained from Ambion (Austin,
TX, USA) included GAPDH siRNA as a positive control
and siRNA with a scrambled sequence as a negative control.
We tested six different siRNA specific for the MCPIP gene.
The most potent inhibitor appeared to be the sequence 5¢-
CCCUGUUGAUACACAUUGUTT-3¢. Human fibroblasts
were cultured in 12-well plates. The siRNA concentration
and amount of transfection reagent required were optimized
experimentally. Before transfection, cells were trypsinized,
centrifuged at 1000 g at 4 °C for 5 min, and resuspended in
fresh medium. The lipid-based transfection reagent siPORT
NeoFX (4 ll per well) from Ambion and siRNAs (50 nm
final concentration) were separately diluted in 50 lL

Opti-MEM and then mixed together. Transfection com-
plexes were allowed to form for 10 min, and overlaid with
9 · 10
4
cells per well. As an additional control, some wells
were treated with transfection reagent only. Silencing of
MCPIP was confirmed by real-time PCR in three indepen-
dent experiments.
Stability assay for IL-b1 mRNA
HepG2 cells stably transfected with a vector overexpressing
IL-1b mRNA were transfected with an empty vector or one
overexpressing either wild-type MCPIP or its mutant forms
– mutMCPIP without the PIN domain and D ⁄ A-MCPIP
with two conservative amino acid mutations within the PIN
domain. Transcription was stopped 24 h after transfection
by adding 5 lgÆmL
)1
actinomycin D. The cells were har-
vested, and RNA was isolated at various time points as
indicated. Northern blot analyses were performed to deter-
mine the level of exogenous IL-1b mRNA.
Reporter gene assay
Transient transfection experiments were performed using
Lipofectamine 2000 reagent (Invitrogen) in a 12-well plate.
In each well, the pGL3 reporter gene with ARE from IL-1b
3¢ UTR (p-luc)ARE) or without AREs (p-luc) was
co-transfected with an empty vector (pcDNA3), and vector
expressing either wild-type or mutant MCPIP lacking the
PIN domain (mutMCPIP). Luciferase assays were per-
formed 22 h after transfection using the Dual-Light System

(Tropix, Foster City, CA, USA) reporter gene as described
by the manufacturer. The luciferase activity of each con-
struct was normalized to b-galactosidase activity as an
integral transfection control. Total RNA was isolated as
described above. Semi-quantitative RT-PCR was performed
using primers specific for luciferase (forward 5¢-AGA
GATACGCCCTGGTTCCT-3¢; reverse 5¢-AATCTGACG
CAGGCAGTTCT-3¢) and GAPDH transcripts (sequences
given above).
Generation of rabbit antibodies specific for
MCPIP, and western blot analysis
E. coli BL21 cells were transfected with pQE-31 vector
(Qiagen, Du
¨
sseldorf, Germany) containing the entire coding
sequence of the MCPIP transcript. The recombinant pro-
tein was purified from bacterial culture using BD Talon
metal affinity resins with cobalt ions (BD Biosciences) and
used for immunization of a New Zealand White rabbit
(three times 300 lg of antigen). Antibodies were purified
from rabbit sera by protein A–Sepharose chromatography
and then checked for specificity. Generated antibodies were
used for western blot analysis of protein extracts from
HepG2 cells transfected with a vector overexpressing
MCPIP. Several independent experiments showed that there
was no cross-reactivity, and only a specific product of the
expected size was observed on membranes.
For subcellular fractionation of HepG2 cells overex-
pressing MCPIP, the Qproteome kit (Qiagen) was used,
and all steps were performed as described by the manufac-

turer. Consecutive centrifugation and extraction with suit-
able buffers allowed isolation of cytosolic, membrane,
nuclear and cytoskeletal proteins. Protein concentrations
were determined using a BCA protein assay (Sigma).
Aliquots of 10 lg from each fraction were used in western
blot analysis. To detect markers of particular fractions,
the following antibodies were used: polyclonal antibodies
against lamin C (Abcam, Cambridge, MA, USA) for the
nuclear fraction, polyclonal antibodies against GAPDH
D. Mizgalska et al. MCPIP protein as an RNase
FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS 7397
(Abcam) for the cytosolic fraction, polyclonal antibodies
against MnSOD (Sigma) for the membrane fraction and
monoclonal antibodies against b-actin (Sigma) for the
cytoskeleton fraction. Secondary antibodies used: goat
anti-mouse (BD Biosciences), goat anti-rabbit (Sigma).
RNA immunoprecipitation
HepG2 cells were transfected with plasmid coding for
MCPIP. After 24 h, cells were crosslinked with 1% formal-
dehyde, and subsequently washed with NaCl ⁄ P
i
containing
1mm phenylmethanesulfonyl fluoride and 125 mm glycine.
RNAsin (100 UÆmL
)1
, Promega) was added to all buffers.
Cells were lysed with buffer containing 1% SDS, 10 mm
EDTA, 50 mm Tris ⁄ HCl, pH 8. Lysates were centrifuged at
14 000 g for 10 min at 4 °C, and the supernatant obtained
was diluted 10-fold using IP buffer [0.01% SDS, 1.1%

Triton X-100, 1.2 mm EDTA, 16.7 mm NaCl, protease
inhibitor complex (Roche, Basel, Switzerland), 16.7 mm
Tris ⁄ HCl, pH 8]. Protein A–agarose (Roche) used for
pre-cleaning of the lysate, was washed with IP buffer,
blocked with 1 mgÆmL
)1
BSA, and 20 lgÆmL
)1
salmon
DNA (Sigma). Then the sample was divided into two parts
and antibodies were added: anti-MCPIP to the first and
rabbit polyIgG as a negative control to the second. Samples
were incubated overnight at 4 °C with constant mixing.
The next day, blocked protein A–agarose beads were
added. After 1 h of incubation at 4 °C, the immunoprecipi-
tates were centrifuged (at 14 000 g, 10 min, 4 °C) and
washed twice each with four buffers (B-I: 0.1% SDS, 1%
Triton X-100, 2 mm EDTA, 150 mm NaCl, 20 mm
Tris ⁄ HCl, pH 8; B-II: 0.1% SDS, 1% Triton X-100, 2 mm
EDTA, 500 mm NaCl, 20 mm Tris ⁄ HCl, pH 8; B-III:
0.25 m LiCl, 1% Nonidet P-40, 1% sodium deoxycholate,
1mm EDTA, 10 mm Tris ⁄ HCl, pH 8; B-IV: 1 mm
EDTA,10 mm Tris ⁄ HCl, pH 8). RNA was eluted from
beads with buffer containing 1% SDS and 0.1 m NaHCO
3
.
After addition of NaCl to 0.2 m concentration, samples
were reverse crosslinked by incubation at 65 °C for 4 h,
with subsequent protein digestion by proteinase K. RNA
was isolated from the immunoprecipitates using the phe-

nol:chlorophorm method. To facilitate precipitation with
isopropanol, yeast tRNA (Sigma) was added to the samples
to final concentration of 1 mg ÆmL
)1
. Residual DNA
removal was performed by DNase digestion (Ambion). The
RNA obtained was reverse-transcribed, and PCR reactions
were performed using IL-1b -, MCPIP- and GAPDH-spe-
cific primers (sequences given above).
Statistical analysis
The data from real-time PCR, Northern blotting and
measurements of luciferase activity were examined for
statistical significance using Student’s t test. All values are
means ± SD (n = 3).
Acknowledgments
The study was supported by the European Union’s
FP6 project Savebeta (LSHM-CT-2006-036903,
MTKD-CT-2006-042586 and COST Action BM0602)
and by the Polish Ministry of Scientific Research and
Information Technology (2 P05A01127, 63⁄ 6PR-
UE ⁄ 2007 ⁄ 7 and 339 ⁄ 6PR-UE ⁄ 2007 ⁄ 7). We would like
to thank Professor Sigurd Lenzen (Institute of Clinical
Biochemistry, Hannover Medical School, Hannover ⁄
Germany) and Dr Lindsay Ramage (Center of
Inflammation, University of Edinburgh, UK) for their
constructive criticism.
References
1 Jura J, Wegrzyn P, Korostyn
´
ski M, Guzik K, Oczko-

Wojciechowska M, Jarza˛ b M, Kowalska M, Piechota
M, Przewocki P & Koj A (2008) Identification of inter-
leukin-1 and interleukin-6-responsive genes in human
monocyte-derived macrophages using microarrays. Bio-
chim Biophys Acta 1779, 383–389.
2 Zhou L, Azfer A, Niu J, Graham S, Choudhury M,
Adamski FM, Younce C, Binkley PF & Kolattukudy
PE (2006) Monocyte chemoattractant protein-1 induces
a novel transcription factor that causes cardiac myocyte
apoptosis and ventricular dysfunction. Circ Res 98,
1177–1185.
3 Liang J, Wang J, Azfer A, Song W, Tromp G,
Kolattukudy PE & Fu M (2008) A novel CCCH-zinc
finger protein family regulates proinflammatory activa-
tion of macrophages. J Biol Chem 283, 6337–6346.
4 Matsushita K, Takeuchi O, Standley DM, Kumagai Y,
Kawagoe T, Miyake T, Satoh T, Kato H, Tsujimura T,
Nakamura H et al. (2009) Zc3h12a is an RNase essen-
tial for controlling immune responses by regulating
mRNA decay. Nature 458, 1185–1190.
5 Wallner B & Elofsson A (2006) Identification of correct
regions in protein models using structural, alignment,
and consensus information. Protein Sci 15, 900–913.
6 Kurowski MA & Bujnicki JM (2003) GeneSilico protein
structure prediction meta-server. Nucleic Acids Res 31,
3305–3307.
7 Glavan F, Behm-Ansmant I, Izaurralde E & Conti E
(2006) Structures of the PIN domains of SMG6 and
SMG5 reveal a nuclease within the mRNA surveillance
complex. EMBO J 25, 5117–5125.

8 Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F
& Higgins DG (1997) The CLUSTAL_X windows
interface: flexible strategies for multiple sequence align-
ment aided by quality analysis tools. Nucleic Acids Res
25, 76–82.
9 Ward JJ, Sodhi JS, McGuffin LJ, Buxton BF & Jones
DT (2004) Prediction and functional analysis of native
MCPIP protein as an RNase D. Mizgalska et al.
7398 FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS
disorder in proteins from the three kingdoms of life.
J Mol Biol 337, 635–645.
10 Vullo A, Bortolami O, Pollastri G & Tosatto S (2006)
Spritz: a server for the prediction of intrinsically disor-
dered regions in protein sequences using kernel
machines. Nucleic Acids Res 34, W164–W168.
11 Stoecklin G, Stubbs T, Kedersha N, Wax S, Rigby WF,
Blackwell TK & Anderson P. (2004) MK2-induced triste-
traprolin:14-3-3 complexes prevent stress granule associa-
tion and ARE-mRNA decay. EMBO J 23, 1313–1324.
12 Kedersha NL, Gupta M, Li W, Miller I & Anderson P
(1999) RNA-binding proteins TIA-1 and TIAR link the
phosphorylation of eIF-2a to the assembly of mamma-
lian stress granules. J Cell Biol 147, 1431–1442.
13 Henics T (1999) Microfilament-dependent modulation
of cytoplasmic protein binding to TNFa mRNA
AU-rich instability element in human lymphoid cells.
Cell Biol Int 23, 561–570.
14 Sirenko O, Bo
¨
cker U, Morris JS, Haskill JS & Watson

JM (2002) IL-1b transcript stability in monocytes is
linked to cytoskeletal reorganization and the availability
of mRNA degradation factors. Immunol Cell Biol 80,
328–339.
15 Sweet TJ, Boyer B, Hu W, Baker KE & Coller J (2007)
Microtubule disruption stimulates P-body formation.
RNA 13, 493–502.
16 Roca FJ, Cayuela ML, Secombes CJ, Meseguer J &
Mulero V (2007) Post-transcriptional regulation of
cytokine genes in fish: a role for conserved AU-rich
elements located in the 3¢-untranslated region of their
mRNAs. Mol Immunol 44, 472–478.
17 Paschoud S, Dgar AM, Kuntz C, Grisoni-Neupert B,
Richman L & Kuhn LC (2006) Destabilization of inter-
leukin-6 mRNA requires a putative RNA stem–loop
structure, an AU-rich element, and the RNA-binding
protein AUF1. Mol Cell Biol 26, 8228–8241.
18 Kaspar RL & Gehrke L (1994) Peripheral blood
mononuclear cells stimulated with C5a or lipopolysac-
charide to synthesize equivalent levels of IL-1beta
mRNA show unequal IL-1beta protein accumulation
but similar polyribosome profile. J Immunol 153, 277–
286.
19 Dinarello C (1996) Biologic basis for interleukin-1 in
disease. Blood 87, 2095–2147.
20 Lai WS, Carballo E, Strum JR, Kennington EA,
Phillips RS & Blackshear PJ (1999) Evidence that
tristetraprolin binds to AU-rich elements and promotes
the deadenylation and destabilization of tumor necrosis
factor alpha mRNA. Mol Cell Biol 19, 4311–4323.

21 Haile S, Estevez AM & Clayton C (2003) A role for the
exosome in the in vivo degradation of unstable mRNAs.
RNA 9, 1491–1501.
22 Sheth U & Parker R (2003) Decapping and decay of
messenger RNA occur in cytoplasmic processing bodies.
Science, 300, 805–808.
23 Hau HH, Walsh RJ, Ogilvie RL, Williams DA, Reilly
CS & Bohjanen PR (2007) Tristetraprolin recruits func-
tional mRNA decay complexes to ARE sequences.
J Cell Biochem 100, 1477–1492.
24 Cerutti L, Mian N & Bateman A (2000) Domains in
gene silencing and cell differentiation proteins: the novel
PAZ domain and redefinition of the Piwi domain.
Trends Biochem Sci 25, 481–482.
25 Anantharaman V & Aravind L (2006) The NYN
domains: novel predicted RNAses with a PIN domain-
like fold. RNA Biol 3, 18–27.
26 Clissold PM & Ponting CP (2000) PIN domains in non-
sense-mediated mRNA decay and RNAi. Curr Biol 10,
R888–R890.
27 Brooks SA, Connolly JE & Rigby WF (2004) The role
of mRNA turnover in the regulation of tristetraprolin
expression: evidence for an extracellular signal-regulated
kinase-specific, AU-rich element-dependent, autoregula-
tory pathway. J Immunol 172, 7263–7271.
28 Jura J, Wegrzyn P, Zarebski A, Wladyka B & Koj A
(2004) Identification of changes in the transcriptome
profile of human hepatoma HepG2 cells stimulated with
interleukin-1b. Biochim Biophys Acta 1689, 120–133.
D. Mizgalska et al. MCPIP protein as an RNase

FEBS Journal 276 (2009) 7386–7399 ª 2009 The Authors Journal compilation ª 2009 FEBS 7399