Negative regulation of TLR4 via targeting of the proinflammatory tumor suppressor PDCD4 by the MicroRNA miR-21

Abstract

The tumor suppressor PDCD4 is a proinflammatory protein that promotes activation of the transcription factor NF-kappaB and suppresses interleukin 10 (IL-10). Here we found that mice deficient in PDCD4 were protected from lipopolysaccharide (LPS)-induced death. The induction of NF-kappaB and IL-6 by LPS required PDCD4, whereas LPS enhanced IL-10 induction in cells lacking PDCD4. Treatment of human peripheral blood mononuclear cells with LPS resulted in lower PDCD4 expression, which was due to induction of the microRNA miR-21 via the adaptor MyD88 and NF-kappaB. Transfection of cells with a miR-21 precursor blocked NF-kappaB activity and promoted IL-10 production in response to LPS, whereas transfection with antisense oligonucleotides to miR-21 or targeted protection of the miR-21 site in Pdcd4 mRNA had the opposite effect. Thus, miR-21 regulates PDCD4 expression after LPS stimulation.

Full text

nature immunology VOLUME 11 NUMBER 2 FEBRUARY 2010 141
Many negative regulatory control mechanisms exist to limit the toxic
effects of lipopolysaccharide (LPS)1. These include soluble decoy
receptors, such as soluble Toll-like receptor 4 (TLR4)2, and splice
variants of signal-transduction proteins, including MyD88-s3, IRAK-
M4 and TAG5, which interfere with signal-transduction pathways. The
inhibitor of transcription factor NF-κB α-subunit (IκBα) is promptly
resynthesized by NF-κB to block excessive transcription factor activ-
ity after treatment with LPS6. The production of anti-inflammatory
cytokines is also induced by LPS signaling, including interleukin 10
(IL-10), which has paracrine effects on neighboring cells to negatively
regulate the action of NF-κB, and proinflammatory cytokines such as
IL-6 and IL-12 (ref. 7). Noncoding RNA products known as micro-
RNAs (miRNAs) have also been described, such as miR-146a, which
is induced by LPS and negatively targets signaling proteins such as
IRAK1 and TRAF6 at the post-transcriptional level8.
PDCD4 was first described as a protein induced by apoptotic
stimuli9 that acts as a tumor suppressor10. It is induced by cytokine
treatment, consistent with a predicted NF-κB site in its promoter11.
PDCD4 has been shown to positively influence tumor necrosis
factor–induced activation of NF-κB12. It has also been shown to be a
translational inhibitor through interaction with members of the eIF4
family of eukaryotic translation-initiation factors13,14. Target mRNAs
include those encoding IL-10 and IL-4 (ref. 15), whose production is
therefore suppressed by PDCD4. PDCD4-deficient mice are resistant
to models of inflammatory disease, such as experimental autoimmune
encephalomyelitis and streptozotocin-induced type II diabetes. It is
likely that the proinflammatory effect of PDCD4 is due to its role in
NF-κB function and its ability to suppress IL-10 translation.
PDCD4 is targeted for proteasomal degradation by β-TRCP
ubiquitin ligases activated by growth factors during tumor
promotion16,17. The miRNA miR-21 targets Pdcd4 mRNA post-
transcriptionally, blocking production of PDCD4 protein18–20. This
miRNA is upregulated in many cancers, including lymphoma, leukemia
and solid tumors and therefore has been called an ‘oncomiR’, which
may explain the loss of PDCD4 during neoplastic transformation21.
Here we examine the role of PDCD4 in the inflammatory response
to LPS. Similar to experimental autoimmune encephalomyelitis and
type II diabetes15, we found PDCD4-deficient mice were protected
from the lethality of LPS, and we examined its role in LPS signaling.
We found that LPS modulated the expression of PDCD4 through the
induction of miR-21 and obtained evidence that this modulation reg-
ulated NF-κB activity while promoting IL-10 production. Our study
identifies miR-21 as a negative regulator of TLR4 signaling through
the targeting of PDCD4.
RESULTS
PDCD4 is required for the lethality of LPS
To examine the role of PDCD4 in TLR signaling and inflammation, we
injected PDCD4-deficient and wild-type control mice with LPS and
monitored their survival. PDCD4-deficient mice were less susceptible
to LPS, with lower mortality than wild-type mice (Fig. 1a). Analysis of
circulating cytokine concentrations 24 h after LPS injection showed that
1School of Biochemistry & Immunology, Trinity College, Dublin, Ireland. 2Department of Pathology, Coombe Women’s Hospital, Dublin, Ireland. 3Department
of Histopathology, School of Medicine, Trinity College, Dublin, Ireland. 4Department of Pathology and Laboratory Medicine, School of Medicine, University of
Pennsylvania, Philadelphia, USA. Correspondence should be addressed to L.A.J.O. (laoneill@tcd.ie).
Received 11 September; accepted 21 October; published online 29 November 2009; corrected online 6 December 2009; doi:10.1038/ni.1828
Negative regulation of TLR4 via targeting of the
proinflammatory tumor suppressor PDCD4 by the
microRNA miR-21
Frederick J Sheedy1, Eva Palsson-McDermott1, Elizabeth J Hennessy1, Cara Martin2,3, John J O’Leary2,3,
Qingguo Ruan4, Derek S Johnson4, Youhai Chen4 & Luke A J O’Neill1
The tumor suppressor PDCD4 is a proinflammatory protein that promotes activation of the transcription factor NF-B and
suppresses interleukin 10 (IL-10). Here we found that mice deficient in PDCD4 were protected from lipopolysaccharide
(LPS)-induced death. The induction of NF-B and IL-6 by LPS required PDCD4, whereas LPS enhanced IL-10 induction in
cells lacking PDCD4. Treatment of human peripheral blood mononuclear cells with LPS resulted in lower PDCD4 expression,
which was due to induction of the microRNA miR-21 via the adaptor MyD88 and NF-B. Transfection of cells with a
miR-21 precursor blocked NF-B activity and promoted IL-10 production in response to LPS, whereas transfection with
antisense oligonucleotides to miR-21 or targeted protection of the miR-21 site in Pdcd4 mRNA had the opposite effect.
Thus, miR-21 regulates PDCD4 expression after LPS stimulation.
ARTICLES
© 2010 Nature America, Inc. All rights reserved.

14 2 VOLUME 11 NUMBER 2 FEBRUARY 2010 nature immunology
ARTICLES
IL-6 concentrations were lower in PDCD4-deficient mice treated with
LPS (Fig. 1b), consistent with lower susceptibility. Analysis of cytokine
production at earlier times showed striking differences between PDCD4-
deficient mice and wild-type mice in terms of production of the anti-
inflammatory cytokine IL-10 (Fig. 1c). At 1 h after LPS injection, IL-10
serum concentrations in PDCD4-deficient mice were greater than those
in wild-type mice, an effect also evident at 3 h after injection. These data
indicate that PDCD4 has a proinflammatory role in LPS signaling.
PDCD4 expression is regulated by TLR4
We investigated whether PDCD4 is a target of TLR signaling. LPS
resulted in higher expression of PDCD4 protein in RAW264.7 mouse
macrophages, evident at 1 h (Fig. 2a); this decreased and was abolished
at 24 h. The effect at 24 h was concentration dependent and was evident
at a range of LPS concentrations from 0.1 ng/ml to 100 ng/ml (Fig. 2b).
A profound decrease in Pdcd4 mRNA in response to LPS was also evi-
dent from 4 h (Fig. 2c). We also observed the effect on PDCD4 protein in
mouse primary bone marrow–derived macrophages (BMDMs; Fig. 2d).
LPS caused a slight increase 4 h after stimulation, whereas at 8 h and 24 h
after LPS, a substantial decrease in PDCD4 was evident. Similarly, treat-
ment of primary BMDMs with other TLR ligands such as Pam3CSK4
(a TLR2 ligand) and poly(I:C) (a TLR3 ligand) induced PDCD4 protein
at earlier times. However, at 24 h after treatment, there was much less
PDCD4 (Fig. 2d). Notably, in human peripheral blood mononuclear
cells (PBMCs), LPS treatment increased the expression of PDCD4
protein at 4 h and 8 h, with a decrease occurring at 24 h (Fig. 2e).
PDCD4 regulates NF-B and IL-10
To explain the proinflammatory role of PDCD4 in LPS signaling,
we examined its ability to affect IL-10 production. Transfection of
RAW264.7 cells with small interfering RNA (siRNA) specific for
PDCD4 at a concentration of 50 nM and 100 nM decreased endo-
genous PDCD4 expression, causing a knockdown of 50%, as measured
by densitometry scanning (Fig. 3a, top). LPS induced a decrease in
PDCD4 protein both in cells transfected with control siRNA and in
cells treated with PDCD4-specific siRNA. We observed more LPS-
induced production of IL-10 in cells transfected with increasing
amounts of siRNA specific for PDCD4 (Fig. 3a, bottom). A similar
increase in IL-10 production was also evident in PDCD4-deficient
BMDMs after 4 h of treatment with LPS at concentrations of 10 ng/ml
and 100 ng/ml (Fig. 3b). PDCD4 is known to inhibit cap-depend-
ent translation of mRNAs with complex 5′ untranslated regions
(UTRs), of which Il10 is an example. To determine if this is the case,
we analyzed Il10 mRNA abundance in PDCD4-deficient BMDMs.
Unexpectedly, we found lower Il10 mRNA expression in PDCD4-
deficient BMDMs that produced more IL-10 protein than did wild-
type cells (Fig. 3c). At 4 h after LPS treatment, PDCD4-deficient
BMDMs expressed one-fourth less Il10 mRNA than did wild-type
cells. Other genes that are regulated by eIF4E, which are considered to
be sensitive to PDCD4 activity, also show differences in mRNA abun-
dance when translation is modulated22,23. To rule out the possibility
of transcriptional effects of PDCD4 deficiency on IL-10 expression,
we transfected RAW264.7 cells with an Il10 promoter–luciferase con-
struct alongside siRNA specific for PDCD4. LPS induced the activity
of this promoter at a concentration of 100 ng/ml (Fig. 3d). However,
when PDCD4 expression was knocked down, we detected no substan-
tial difference in Il10 promoter activity.
We also examined the role of PDCD4 in NF-κB activation by LPS
with TLR4-expressing human embryonic kidney cells (HEK293-
TLR4 cells) transfected with an NF-κB–luciferase reporter plasmid.
Transfection with siRNA specific for PDCD4 at a concentration of
200 nM decreased endogenous PDCD4 in HEK293-TLR4 cells (62%
knockdown, as measured by densitometry scanning; Fig. 3e, top).
The knockdown of endogenous PDCD4 resulted in less activation of
NF-κB at all concentrations of LPS tested, with 60–70% inhibition
occurring at an siRNA concentration of 200 nM, equivalent to the
degree of PDCD4 knockdown (Fig. 3e, bottom). We also tested the
induction of IL-6, which is NF-κB dependent, in PDCD4-deficient
BMDMs (Fig. 3f). This response was impaired in PDCD4-deficient
cells after 24 h of treatment with LPS at a concentration of 10 ng/ml
and 100 ng/ml. We detected significantly less Il6 mRNA in PDCD4-
deficient BMDMs than in wild-type BMDMs after 24 h of LPS
treatment (Fig. 3g), which indicated this impairment occurs as a
result of less NF-κB-induced transcription of Il6.
To explain this function of PDCD4, we examined early signal-
ing events in wild-type and PDCD4-deficient BMDMs. Analysis
of IκBα protein by immunoblot showed that there was more IκBα
degradation at 0.5 h in the PDCD4-deficient cells (Fig. 3h, top).
Consistent with a positive effect of PDCD4 on NF-κB activation, the
LPS (d)
0
20
40
60
80
100
Survival (%)
Mouse lL-6 (ng/ml)
Mouse lL-10 (ng/ml)
012345678910
WT Pdcd4–/–
Pdcd4–/–
a
b c
0
20
40
60
80
WT
0
1
2
3
4
1 h 3 h
WT
Pdcd4
–/
–
Figure 1 PDCD4-deficient mice are protected from the lethality of
LPS. (a) Survival of wild-type (WT) control mice and PDCD4-deficient
(Pdcd4−/−) mice 8–12 weeks of age injected intravenously with LPS,
monitored over a period of 10 d; results are plotted as a percentage
of total numbers (n = 10 mice per group). (b,c) Enzyme-linked
immunosorbent assay (ELISA) of mouse IL-6 (b) and mouse IL-10 (c) in
blood samples from wild-type and PDCD4-deficient mice (n = 10 (b) or
5 (c) mice per group), 24 h (b) and 1 h and 3 h (c) after LPS injection.
Data are representative of three (a) or two (b,c) independent experiments.
a
PDCD4
β-actin
β-actin
β-actin
LPS (h) 0 1 2 4 8 24
PDCD4
LPS (ng/ml) 0 0.1 1 10 100
c
b
0 4 8 24
LPS (h)
e
0
0.5
1.0
0 4 8 16 24
PDCD4
LPS (h)
β-actin
PDCD4
LPS (h)
d
0 4 8 24
LPS
0 4 24 0 4 24
Poly (I:C)Pam
3
CSK
4
Pdcd4 mRNA
(relative)
Figure 2 PDCD4 protein expression is regulated by LPS. (a) Immunoblot
analysis of PDCD4 in RAW264.7 cells treated for 0–24 h (above lanes) with
LPS (100 ng/ml). (b) Immunoblot analysis of PDCD4 in RAW264.7 cells
treated for 24 h with 0–100 ng/ml (above lanes) of LPS. (c) Quantitative
RT-PCR analysis of Pdcd4 mRNA in RAW264.7 cells treated for 0–24 h
(horizontal axis) with LPS, presented relative to Pdcd4 mRNA in untreated
cells. (d) Immunoblot analysis of PDCD4 in mouse BMDMs treated for 0–24 h
(above lanes) with LPS (1 ng/ml), Pam3CSK4 (100 ng/ml) or poly(I:C)
(12.5 µg/ml). (e) Immunoblot analysis of PDCD4 in human PBMCs treated
for 0–24 h (above lanes) with LPS. β-actin expression serves as a loading
control. Data are representative of three independent experiments (c; mean
± s.d.) or at least three independent experiments (a,b,d,e).
© 2010 Nature America, Inc. All rights reserved.

nature immunology VOLUME 11 NUMBER 2 FEBRUARY 2010 143
ARTICLES
NF-κB-dependent resynthesis of IκBα at later times was markedly
impaired in PDCD4-deficient macrophages. These results indicate
that the effect of PDCD4 on NF-κB activation is not due to modula-
tion of the IKK complex and is probably indirect and secondary to its
primary function as a repressor of translation. PDCD4 is known to
affect the activity of the transcription factor AP-1 in tumor progres-
sion24. Here we analyzed activation of the kinase Jnk by assessing its
phosphorylation with an antibody specific for Jnk phosphorylation.
Jnk activation was impaired in response to LPS in PDCD4-deficient
BMDMs compared with its activation in wild-type cells (Fig. 3h,
middle). These data confirm that PDCD4 has a proinflammatory role
in LPS signaling events.
Induction of miR-21 by LPS
To explain the lower PDCD4 protein abundance after LPS treatment,
we monitored expression of the PDCD4-targeting miRNA miR-
21 after LPS treatment. We used the induction of a known LPS-
responsive miRNA, miR-146a8, as a positive control. LPS treatment
induced miR-21 expression in RAW264.7 macrophages, which was
apparent from 4 h (Fig. 4a, left). LPS treatment led to strong induc-
tion of both miR-21 and miR-146a at 24 h with similar kinetics
in this cell type. The effect of LPS on miR-21 induction after 24 h
in RAW264.7 was dose dependent, with 100 ng/ml of LPS being
the optimal dose, and again we observed a similar pattern for
miR-146a (Fig. 4b).
Analysis of the induction of miRNA in primary mouse BMDMs
gave a result similar to that obtained with the RAW264.7 cell
line (Fig. 4c). We found that miR-21 was induced by LPS to an
extent similar to that of miR-146a. We also examined the human
monocytic cell line THP-1 after 18 h of treatment with LPS
(Fig. 4d). We found that miR-146a was induced by LPS, consist-
ent with earlier reports8; however, we observed no upregulation
of miR-21 in this cell line. We also observed upregulation of
miR-21 by LPS in human PBMCs (Fig. 4e). We found fourfold more
miR-21, with a similar effect for miR-146a.
Induction of miR-21 by LPS requires MyD88 and NF-B
We examined the induction of miR-21 in immortalized BMDMs
deficient in the TLR adaptor proteins MyD88 and TRIF (Fig. 4f).
LPS caused a tenfold induction of miR-21 in wild-type cells after
18 h. This effect was abolished in the absence of MyD88 and was
only slightly impaired in TRIF-deficient BMDMs (Fig. 4f, left). We
noted a similar dependency on MyD88 for miR-146a; however, in
contrast to miR-21 induction, miR-146a induction by LPS also
required TRIF (Fig. 4f, right). We next examined the role of NF-κB
in miR-21 induction by LPS, as the promoter region has a putative
NF-κB site located at position −248 (5′-GTGGGAGGTGCCT-3′),
as predicted by the Genomatix MatInspector software package. We
found induction of miR-21 by LPS in wild-type mouse embryonic
fibroblasts, but this was completely abolished in mouse embryonic
fibroblasts deficient in the NF-κB subunit p65 and was actually
lower than basal expression (Fig. 4g, left). We observed a similar
dependency on p65 for miR-146a (Fig. 4g, right), but to a lesser
extent than for miR-21.
LPS decreases PDCD4 protein via miR-21 induction
Having examined the induction of miR-21 by LPS, we then examined the
ability of miR-21 to regulate PDCD4 abundance after TLR4 signaling. We
assessed this effect by transfecting RAW264.7 cells with antisense oligo-
nucleotides specific to miR-21 (anti-miR-21) or control antisense RNA
(Fig. 5a). Treatment with LPS for 24 h resulted in less PDCD4 in cells
treated with control antisense RNA. Notably, pretreatment with anti-
miR-21 blocked the LPS-induced decrease in PDCD4, particularly at
anti-miR-21 concentrations of 12.5 nM and 25 nM, resulting in 1.3-fold
and 1.57-fold more PDCD4 protein, respectively, than in control LPS-
treated cells, as measured by densitometry scanning. Notably, to verify
a e
PDCD4
PDCD4
siRNA (nM) Ctrl 50 100 200
PDCD4
β
-actin
β
-actin
β
-actin
PDCD4
siRNA (nM) Ctrl 50 100
LPS
– + – + – +
6
2
4
8
10
12
14
16 200 nM Ctrl
50 nM siRNA
200 nM siRNA
*
f
0
0.2
0.4
0.6
0.8
1.0
1.2
WT
Pdcd4
–/–
**
b
siRNA (nM)
*
*
0
0.5
1.0
1.5
2.0
2.5
3.0
IL-10 production
(relative)
IL-10 production
(relative)
Il10 mRNA (relative)
Il10 promoter
luciferase (relative)
NF-
κ
B luciferase
(relative)
IL-6 production
(relative)
Il6 mRNA (relative)
Ctrl 50 100
c d
LPS (h)
I
κ
B
α
p-Jnk
WT
Pdcd4
–/–
h
100 94 33 38Band intensity (%)
100
32
51
20
52
10
Band intensity (%)
0 0.5 1 4 8 18 0 0.5 1 4 8 18
g
0
0.2
0.4
0.6
0.8
1.0
1.2
WT
Pdcd4–/–
***
0
0.2
0.4
0.6
0.8
1.0
1.2
WT
Pdcd4–/–
**
0.5
1.0
1.5
2.0
2.5
0 1 10 100LPS (ng/ml) 0 1 10 100LPS (ng/ml) 10 100LPS (ng/ml)
Ctrl
siRNA
Ctrl
siRNA
PDCD4
siRNA
Total Jnk
0.2
0.4
0.6
1.0
1.2
1.4
1.6
1.8
2.0
10 100
0.8
WT
Pdcd4
–/–
LPS (ng/ml)
*** ***
RAW264.7 cells transfected with an Il10 promoter–luciferase reporter and 100 nM PDCD4-specific siRNA or 100 nM control nontargeting siRNA
and then stimulated for 24 h with 0–100 ng/ml of LPS (horizontal axis). Above, immunoblot analysis of PDCD4 in unstimulated RAW264.7 cells
transfected with control or PDCD4-specific siRNA. (e) Luciferase activity in HEK293-TLR4 transfected with an NF-κB-luciferase reporter and 50 or
200 nM PDCD4-specific siRNA (key) and then stimulated for 24 h with 0–100 ng/ml of LPS (horizontal axis). Top, immunoblot analysis of PDCD4
protein in HEK293-TLR4 transfected with control siRNA or 50 or 200 nM PDCD4-specific siRNA (above lanes); below lanes, densitometry of band
intensity relative to that of cells transfected with control siRNA. (f) ELISA of IL-6 production in wild-type and PDCD4-deficient BMDMs stimulated for
24 h with 10 or 100 ng/ml of LPS. (g) Quantitative RT-PCR analysis of Il6 mRNA in wild-type and PDCD4-deficient BMDMs stimulated for 24 h with
100 ng/ml of LPS. (h) Immunoblot analysis of IκBα protein (top) and Jnk phosphorylation (p-Jnk; middle) in wild-type and PDCD4-deficient BMDMs
stimulated for 0–18 h (above lanes) with LPS (10 ng/ml). β-actin serves as a loading control. In siRNA experiments, all cells were transfected with an
equal amount of total RNA normalized with negative control siRNA. Results for LPS-treated cells are presented relative to those for wild-type cells or
cells transfected with negative control RNA only. *P < 0.05, **P < 0.01 and ***P < 0.001 (two-tailed unpaired t-test). Data are representative of three
independent experiments (mean and s.d.).
Figure 3 Regulation of TLR signaling by PDCD4. (a) Immunoblot analysis of PDCD4 (above) and
ELISA of mouse IL-10 production (below) in RAW264.7 cells transfected with 50 or 100 nM
PDCD4-specific siRNA and then stimulated for 24 h with LPS (100 ng/ml). Below lanes,
densitometry of band intensity relative to that of cells transfected with control siRNA (Ctrl).
(b) ELISA of IL-10 production in wild-type and PDCD4-deficient BMDMs stimulated for 4 h
with 10 or 100 ng/ml of LPS. (c) Quantitative RT-PCR analysis of Il10 mRNA in wild-type and
PDCD4-deficient BMDMs stimulated for 4 h with 10 ng/ml of LPS. (d) Luciferase activity (below) in
© 2010 Nature America, Inc. All rights reserved.

14 4 VOLUME 11 NUMBER 2 FEBRUARY 2010 nature immunology
ARTICLES
this effect of miR-21 was brought about through specific and direct
targeting of Pdcd4 mRNA, we designed a morpholino oligonucleotide
specific to the miR-21 site of PDCD4 (morpho-21). This morpho-
21 oligonucleotide should target specifically the Pdcd4 3′ UTR at the
miR-21 site and block the activity of any miRNA at that position.
Treatment with LPS again resulted in less PDCD4 protein expression
in control cells transfected with the morpholino than in unstimulated
cells (Fig. 5b). However, transfection of morpho-21 followed by LPS
stimulation protected Pdcd4 protein, particularly at a morpholino
oligonucleotide concentration of 5 µM, which resulted in twofold
more PDCD4 protein than in control LPS-treated cells, as measured
by densitometry scanning. LPS therefore regulates the translation of
Pdcd4 mRNA through the induction of miR-21.
PDCD4 is regulated by the proteasome
PDCD4 protein has been shown to be targeted for degradation in
tumor promotion through the action of ubiquitin ligases and the 26S
proteasome16,17. We therefore determined if LPS could also target
the amount of PDCD4 protein via the proteasome. Pretreatment of
RAW264.7 cells with vehicle control and subsequent stimulation with
LPS resulted in a slight induction of LPS 1 h after stimulation and a
decrease at 6 h after stimulation (Fig. 5c). Pretreatment with the protea-
some inhibitor MG132 blocked the decrease in PDCD4 at 6 h. However,
as prolonged MG132 treatment is toxic to cells, the effect of MG132
on the decrease at 24 h noted before could not be analyzed. However,
it was obvious the proteasome did affect the amount of PDCD4
protein at earlier time points after LPS stimulation. To explain this
phenomenon, we examined the ability of PDCD4 protein to become
ubiquitinated in response to LPS by transfection of HEK293-TLR4
cells with hemagglutinin (HA)-tagged ubiquitin and subsequent LPS
stimulation. We immunoprecipitated PDCD4 and subsequently ana-
lyzed HA-tagged ubiquitin by immunoblot (Fig. 5d). We detected
ubiquitination of endogenous PDCD4 at 4 h after LPS treatment in
the presence of MG132. Thus, proteasomal degradation activated by
LPS can degrade ubiquitinated PDCD4 protein in the cell.
Regulation of IL-10 production and NF-κB activation by miR-21
To determine if miR-21, through its regulation of PDCD4, affects the
production of IL-10, we transfected RAW264.7 cells with the miR-21
precursor pro-miR-21 (Fig. 6a). Transfection of pro-miR-21, which
will generate mature miR-21, resulted in less PDCD4, particularly
at a concentration of 100 nM (Fig. 6b). LPS again resulted in less
PDCD4 protein (Fig. 6b). At the same time, transfection with pro-
miR-21 also resulted in more production of IL-10 induced by LPS,
with significantly more IL-10 production evident at a pro-miR-21
concentration of 100 nM (Fig. 6a, left). Transfection of cells with
anti-miR-21, shown above to increase PDCD4 during LPS signaling
(Fig. 5b), blunted the production of IL-10 by LPS (Fig. 6a, middle).
To confirm that the effect of miR-21 was brought about specifically
through the targeting of PDCD4, we transfected RAW264.7 cells with
increasing amounts of morpho-21, shown above to protect PDCD4
from the LPS-induced decrease (Fig. 5c). Again we examined the
effect of morpho-21 on LPS-induced production of IL-10 and, as with
anti-miR-21, found that it significantly inhibited IL-10 production
(Fig. 6a, right). To confirm that the specificity of the effect of miR-21
on IL-10 production was due to targeting of PDCD4, we transfected
BMDMs from wild-type and PDCD4-deficient mice with pro-miR-21
or control RNA and monitored the effect on IL-10 production at 24 h
a b
0
2
4
6
8
miRNA induction (fold)
miRNA induction (fold)
miRNA induction (fold)
miRNA induction (fold)
miR-21 induction (fold)
miR-21 induction (fold)
miR-146a induction (fold)
miR-146a induction (fold)
miRNA induction (fold)
10
12
14
16
4 8 12 16 20 24
miR-21 miR-21
miR-146a miR-146a
Time (h)
f g
0
2
4
6
8
10
12
1 10 100
LPS (ng/ml)
*
***
0
1
2
3
4
5
6
7
8
miR-21
miR-146a
miR-21
miR-146a
miR-21
miR146a
c
BMDM
0
4
8
12
16
20
d
THP-1
0
1
2
3
4
5
6
e
PBMC
*
***
** **
*** ***
*
*
**
0
4
8
12
WT
Myd88
–/–
Trif
–/–
WT
WT
Myd88
–/–
Rela
–/–
WT
Rela
–/–
Trif
–/–
0
4
8
12
16
0
1
2
3
4
0
1
2
3
4
5
6
Figure 4 Induction of miR-21 by LPS treatment in macrophages. (a) Time-course analysis of the induction of miR-21 and miR-146a in RAW264.7 cells
stimulated for 4–24 h (horizontal axis) with LPS (100 ng/ml). (b) Dose-response of the induction of miR-21 and miR-146a in RAW264.7 after treatment
for 24 h with 1–100 ng/ml (horizontal axis) of LPS. (c–e) Induction of miR-21 and miR-146a in mouse BMDMs (c), THP-1 cells (d) and human PBMCs (e)
treated for 24 h with LPS (100 ng/ml). (f) Induction of miR-21 (left) and miR-146a (right) in immortalized Myd88−/−, Trif−/− or wild-type (C57BL/6) BMDMs
stimulated for 18 h with LPS (100 ng/ml). (g) Induction of miR-21 (left) and miR-146a (right) in wild-type and p65-deficient (Rela−/−) mouse embryonic
fibroblasts stimulated for 8 h with LPS (100 ng/ml). Results for LPS-treated cells are presented relative to those for untreated cells. *P < 0.05, **P < 0.01
and ***P < 0.001 (two-tailed unpaired t-test). Data are from at least three independent experiments (mean and s.d.).
a
PDCD4
β-actin
Anti-21 (nM) Ctrl
Ctrl
Ctrl6.25 12.5 25 6.25 12.5 25
Unstim LPS
b
PDCD4
β-actin β-actin
Morpho-21 (µM)
1.0 5.0 10
Ctrl 1.0 5.0 10
Unstim LPS
1.00
1.12
2.01
1.63
Band intensity
(relative):
Band intensity
(relative):
Band intensity
(relative):
c
PDCD4
LPS (h) 0 1 6 0 1 6
Vehicle MG132
1.00
1.18
0.54
1.00
1.21
1.10
d
0 0 4LPS (h)
IP: PDCD4
IB: α-HA
HA-UbEV
Lysates
IB: PDCD4
1.00
1.24
1.30
1.57
Figure 5 Regulation of PDCD4 expression in LPS signaling. (a) Immunoblot analysis of PDCD4 in RAW264.7 cells transfected with various
concentrations of anti-miR-21 (above lanes) and then left untreated (Unstim) or treated for 24 h with LPS (100 ng/ml). (b) Immunoblot analysis of
PDCD4 in RAW264.7 cells transfected with various concentrations of morpho-21 (above lanes) and then left untreated or treated for 24 h with LPS
(100 ng/ml). (c) Immunoblot analysis of PDCD4 in RAW264.7 cells pretreated with 1 µM MG132 or vehicle control and then treated for 0–6 h (above
lanes) with LPS (100 ng/ml). Below lanes (a–c), densitometry of band intensity relative to that of cells transfected with control RNA (Ctrl; a,b) or no
LPS (c), set as 1. (d) Immunoprecipitation (IP; with anti-PDCD4) of endogenous PDCD4 together with overexpressed HA-tagged ubiquitin (HA-Ub)
from transfected HEK293-TLR4 cells stimulated with LPS (100 ng/ml) in the presence of MG132 (100 µg/ml), followed by immunoblot analysis (IB)
with anti-HA (α-HA). EV, empty vector (control). Bottom, immunoblot analysis of PDCD4 in lysates without immunoprecipitation (loading control).
For miRNA-morpholino transfections, all cells were transfected with an equal amount of total RNA normalized with negative control anti-miRNA or
a morpholino targeted to a redundant site in the mouse Pdcd4 3′ UTR. Data are representative of three independent experiments.
© 2010 Nature America, Inc. All rights reserved.

nature immunology VOLUME 11 NUMBER 2 FEBRUARY 2010 145
ARTICLES
after LPS (Fig. 6c). There was slightly more IL-10 production in
BMDMs transfected with pro-miR-21 than in control cells. This effect
was not present, however, in BMDMs from PDCD4-deficient mice,
which indicated that the effect of pro-miR-21 was specifically due to
targeting of Pdcd4 mRNA.
Again we examined the effect of miR-21 on NF-κB activity
with HEK293-TLR4 cells transfected with anti-miR-21, as well as
those transfected with pro-miR-21 or morpho-21, and measured
NF-κB-linked luciferase activity in response to LPS. We assessed the
activity of anti-miR-21 and pro-miR-21 in these cells with a miR-21-
dependent luciferase construct (which contains a synthetic miR-21
binding site in the 3′ UTR of the luciferase gene). Transfection of
cells with pro-miR-21 resulted in less luciferase activity (Fig. 6d, left),
whereas anti-miR-21 resulted in more miR-21-dependent luciferase
activity (Fig. 6d, right). Transfection of pro-miR-21 also negatively
regulated the activation of NF-κB, with 50% inhibition occurring
at a pro-miR-21 concentration of 100 nM (Fig. 6e). Notably, we
observed the opposite trend when we transfected anti-miR-21 into
cells. We observed a doubling in LPS-induced NF-κB activity with
transfection of 50 nM anti-miR-21 relative to the activity in cells
transfected with a control antisense RNA (Fig. 6e, middle). Similarly,
transfection of cells with 10 µM morpho-21 resulted in more LPS-
induced NF-κB activity than that in cells transfected with control
morpholino. Together these results demonstrate that miR-21 affects
IL-10 production and also NF-κB activation through its effect on
PDCD4 expression.
DISCUSSION
Here we have found that the control of PDCD4 expression is a key
step in the negative regulation of the inflammatory response to LPS,
acting as a molecular switch between the proinflammatory (NF-κB)
and anti-inflammatory (IL-10) response. This switch is a decrease
in PDCD4 protein abundance, which is brought about through the
induction of miR-21. This process positively influences IL-10 produc-
tion while negatively regulating NF-κB activity, presumably to control
the LPS response that can be lethal.
Our study has demonstrated upregulation of miR-21 by LPS in
many cell types, including macrophages, mouse embryonic fibrob-
lasts and PBMCs. We found that the induction of miR-21 by LPS
was dependent on MyD88 and also dependent on NF-κB, consist-
ent with the presence of an NF-κB-binding site in the miR-21 pro-
moter25,26. There is upregulation of miR-21 in a variety of disease
states. Overexpression of miR-21 has been reported in many types
of cancer27,28. Also, miR-21 has been reported to be upregulated in
many inflamed states, including the inflamed lung in LPS-treated
mice29, allergic airway inflammation30, osteoarthritis31, psoriasis and
atopic eczema32, disease-active ulcerative colitis tissue33, Helicobacter
pylori–associated gastric cancer34, cardiac muscle injury35 and cardiac
hypertrophy36. Therefore, miR-21 may be an indicator of inflamma-
tion and, given its role in tumorigenesis, might be an important link
between cancer and inflammation.
Evidence is now emerging indicating that TLR activation affects
the expression of many miRNAs. The first LPS-responsive miRNA
reported was miR-146a8, and here we have shown that its induction
was similar to that of miR-21 in macrophages. Studies have shown
considerable upregulation of miR-155 in macrophages37 as well as
dendritic cells38, to an extent higher than reported here for miR-21
and miR-146a. Experiments with animals genetically deficient in
miR-155 have confirmed its importance in the generation of adap-
tive immunity39,40. Additional TLR-responsive miRNAs include miR-
132 (ref. 8), miR-9 (ref. 41), miR-147 (ref. 42) and miR-346 (ref. 43);
these are upregulated in various cell types after stimulation with
TLR ligands. Some miRNAs have been reported to be downregu-
lated after LPS treatment, including let-7i, which is thought to target
TLR4 itself44, and miR-125b45. As many of those miRNAs regulated
by TLR signaling are also dysregulated in cancer27, it is possible that
miRNAs form a key link between inflammation and cancer and that
the induction of specific miRNAs, including miR-21, by TLRs may
be a key step in tumor progression.
PDCD4 has been demonstrated to function as an inhibitor of
cap-dependent translation of complex mRNAs through its interaction
with the eukaryotic initiation factors eIF4a and eiF4G13,14. Gene prod-
ucts inhibited by this mechanism include growth factors and cytokines,
including IL-10. Furthermore, PDCD4 has been linked to NF-κB acti-
vation by an unknown mechanism12. We have demonstrated here
that PDCD4 was required for NF-κB activation and attenuated IL-10
production in LPS signaling and was involved in the lethality of LPS.
Here we have shown that the suppressive effect of PDCD4 on IL-10
occurred at the translational level, as PDCD4-deficient macrophages
had more IL-10 protein yet less Il10 mRNA, consistent with published
reports that IL-10 translation is highly regulated46–48 and that more
eIF4E-directed translation can lead to more mRNA turnover22,23. In
addition, we have shown that PDCD4 affected NF-κB activation by
an undefined mechanism that promotes Il6 transcription, which may
involve a positive effect on Jnk. It is likely that other targets for PDCD4
exist in TLR signaling that are regulated at the translational level.
a d
Band
intensity (%)
– + – + – +
LPS
Pro-miR-21
(nM)
PDCD4
b
100
43
99
28
62
02
Pro-miR-21 (nM) Anti-miR-21 (nM) Morpho 21 (µM)
*
**
** **
***
2.5
IL-10 production
(relative)
IL-10 production
(relative)
miR-21 luciferase
(relative)
2.0
1.5
1.0
0.5
1.6
1.8
0.6
0.2
1.4
1.2
0Ctrl
Ctrl
25 50
50
100
100
Ctrl 25 50 100 Ctr l 1 5 10
c
Ctrl
Pro-
miR-21
WT
Pdcd4
–/–
0 2550 0 25 50
Pro-miR-21
(nM)
Anti-miR-21
(nM)
β-actin
1.0
0.8
0.6
0.4
0.2
0
e
NF-κB luciferase
activity (relative)
*
*
0
5
10
15
20
25
30
Ctrl Ctrl 50 100
Pro-miR-21 (nM) Anti-miR-21 (nM)
– + + + Ctrl Ctrl 25 50
– + + + Ctrl Ctrl 10
– + +LPS
Morpho-21 (µM)
Figure 6 Regulation of PDCD4 function by miR-21 in TLR signaling (a) ELISA of mouse IL-10 in RAW264.7 cells transfected with various doses
(horizontal axes) of pro-miR-21 (left), anti-miR-21 (middle) or morpho-21 (right) and then stimulated for 24 h with LPS (100 ng/ml). (b) Immunoblot
analysis of PDCD4 in RAW264.7 cells transfected with 50 or 100 nM pro-miR-21 (above lanes) and then stimulated for 24 h with LPS (100 ng/ml).
(c) ELISA of IL-10 production in wild-type C57BL/6 and PDCD4-deficient BMDMs transfected with 100 nM pro-miR-21 and then stimulated
for 24 h with LPS (10 ng/ml). (d) Luciferase activity in HEK293-TLR4 cells transfected with a miR-21–luciferase reporter and 0, 25 or 50 nM
pro-miR-21 (left) or anti-miR-21 (right). (e) Luciferase activity in HEK293-TLR4 transfected with an NF-κB luciferase reporter and 50 or 100 nM
pro-miR-21 oligonucleotide (left), anti-miR-21 oligonucleotide (middle) or morpho-21 (right) and then stimulated for 24 h with LPS (1 ng/ml).
In all RNA-morpholino transfections, cells were transfected with an equal amount of total RNA normalized with negative control pro-miRNA, negative
control anti-miRNA or a morpholino targeted to a redundant site in the mouse Pdcd4 3′ UTR. In a,d,e, results with LPS are presented relative to those
obtained with negative control RNA. *P < 0.05, **P < 0.01 and ***P < 0.001 (two-tailed unpaired t-test). Data are representative of three independent
experiments (a,d,e; mean and s.d.), three experiments (b) or two independent experiments (c; mean and s.d. of triplicate samples).
© 2010 Nature America, Inc. All rights reserved.

14 6 VOLUME 11 NUMBER 2 FEBRUARY 2010 nature immunology
ARTICLES
We have also demonstrated here that modulation of miR-21 had
effects opposite to those of PDCD4 in LPS signaling. It attenuated
NF-κB activation and promoted IL-10 production in response to LPS.
Notably, through the use of target protection of the PDCD4 miR-21
site and through the transfection of PDCD4-deficient BMDMs with
miR-21, we have demonstrated that miR-21 has an important role in
negatively regulating these processes specifically through the targeting
of PDCD4, thereby limiting excessive inflammation.
Highlighting the importance of the removal of PDCD4 for the
appropriate development of inflammatory responses was the find-
ing that PDCD4 can also be targeted for proteasomal degradation
in LPS signaling. We have demonstrated here that inhibition of the
26S proteasome by MG132 stabilized PDCD4 protein at early time
points after LPS stimulation. In addition, we have identified PDCD4
as a key target protein for ubiquitination downstream of TLR
signaling. The existence of multiple mechanisms to control PDCD4
protein expression highlights the importance of this molecule in the
immune response. Degradation of the pre-existing pool of PDCD4
protein occurred through ubiquitination and the action of the 26S
proteasome at early times after LPS treatment (6 h). This occurred
in addition to the post-transcriptional silencing of Pdcd4 mRNA
through the action of miR-21 to block the translation of new PDCD4
protein, which, because of the delay in miR-21 induction, had its
effect at later times (24 h). As for the induction of PDCD4 protein
at earlier time points after LPS treatment (1 h), this was not accom-
panied by an increase in Pdcd4 mRNA, which began to decrease
4 h after LPS administration. It is likely that at earlier time points,
because of lower basal amounts of miR-21, translation of PDCD4 can
occur without halting.
PDCD4 abundance was high initially and facilitated NF-κB acti-
vation while suppressing IL-10. LPS decreased PDCD4 via miR-21
to limit NF-κB activity while promoting IL-10 production, thereby
providing a negative regulatory loop for TLR4 signaling. The link
between miR-21 and PDCD4 and between NF-κB and IL-10 is direct,
as shown by use of a morpholino oligonucleotide directed specifically
to the miR-21 site of mouse Pdcd4 mRNA. This stabilized PDCD4
protein and had the opposite effect in terms of enhanced NF-κB and
decreased IL-10 relative to that in cells treated with siRNA specific
for PDCD4 or PDCD4-deficient cells. It is therefore reasonable to
conclude that miR-21 limits PDCD4 in LPS signaling, leading to a
decrease in NF-κB and an increase in IL-10, which in turn regulates
inflammatory processes induced by LPS. In conclusion, our study
has identified an axis involving miR-21 and PDCD4 that provides
important new insight into the negative regulation of TLR4 signal-
ing. In terms of the importance of TLR4 in vaccine adjuvancy49 or for
inflammation in processes such as sepsis, rheumatoid arthritis and
allergic asthma50, this finding might present new opportunities for
boosting or limiting TLR4 activation therapeutically.
METHODS
Methods and any associated references are available in the online
version of the paper at http://www.nature.com/natureimmunology/.
ACKNOWLEDGMENTS
We thank D. Golenbock (University of Massachusetts) for wild-type, MyD88-
deficient and TRIF-deficient BMDMs immortalized by retrovirus; R. Hay
(University of St. Andrews) for p65-deficient and matched wild-type control
mouse embryonic fibroblasts and for HA-tagged ubiquitin; R. Hofmeister
(Universitaet Regensburg) for 5× NF-κB reporter luciferase plasmid;
A.M. Cheng (Ambion) for the pMIR-REPORT miR-21 reporter luciferase plasmid;
and A. Bowie (Trinity College, Dublin) for the Il10 promoter–luciferase plasmid.
Supported by Science Foundation Ireland and the Irish Research Council for
Science, Engineering and Technology (RS/2005/190).
AUTHOR CONTRIBUTIONS
F.J.S. did the functional experiments on PDCD4 and miR-21 and cowrote the
manuscript; E.P.-M. did experiments on PDCD4 degradation by the proteasome
and experiments on signals in PDCD4-deficient cells; E.J.H. helped with
experiments on miR-21; C.M. and J.J.O. provided advice on miRNA profiling
experiments; Q.R., D.S.J. and J.Y.C. did the experiments on the lethality of LPS in
PDCD4–deficient mice and supplied BMDMs from the mice; and L.A.J.O. directed
the work and cowrote the manuscript.
Published online at http://www.nature.com/natureimmunology/.
Reprints and permissions information is available online at http://npg.nature.com/
reprintsandpermissions/.
1. Liew, F.Y., Xu, D., Brint, E.K. & O’Neill, L.A. Negative regulation of toll-like receptor-
mediated immune responses. Nat. Rev. Immunol. 5, 446–458 (2005).
2. Iwami, K.I. et al. Cutting edge: naturally occurring soluble form of mouse Toll-like receptor
4 inhibits lipopolysaccharide signaling. J. Immunol. 165, 6682–6686 (2000).
3. Janssens, S., Burns, K., Vercammen, E., Tschopp, J. & Beyaert, R. MyD88S, a
splice variant of MyD88, differentially modulates NF-κB- and AP-1-dependent gene
expression. FEBS Lett. 548, 103–107 (2003).
4. Kobayashi, K. et al. IRAK-M is a negative regulator of Toll-like receptor signaling.
Cell 110, 191–202 (2002).
5. Palsson-McDermott, E.M. et al. TAG, a splice variant of the adaptor TRAM,
negatively regulates the adaptor MyD88-independent TLR4 pathway. Nat. Immunol.
10, 579–586 (2009).
6. Sun, S.C., Ganchi, P.A., Ballard, D.W. & Greene, W.C. NF-κB controls expression
of inhibitor IκBα: evidence for an inducible autoregulatory pathway. Science 259,
1912–1915 (1993).
7. Murray, P.J. STAT3-mediated anti-inflammatory signalling. Biochem. Soc. Trans. 34,
1028–1031 (2006).
8. Taganov, K.D., Boldin, M.P., Chang, K.J. & Baltimore, D. NF-κB-dependent induction
of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune
responses. Proc. Natl. Acad. Sci. USA 103, 12481–12486 (2006).
9. Shibahara, K. et al. Isolation of a novel mouse gene MA-3 that is induced upon
programmed cell death. Gene 166, 297–301 (1995).
10. Yang, H.S., Knies, J.L., Stark, C. & Colburn, N.H. Pdcd4 suppresses tumor
phenotype in JB6 cells by inhibiting AP-1 transactivation. Oncogene 22,
3712–3720 (2003).
11. Onishi, Y., Hashimoto, S. & Kizaki, H. Cloning of the TIS gene suppressed by
topoisomerase inhibitors. Gene 215, 453–459 (1998).
12. Yang, H.S. et al. A novel transformation suppressor, Pdcd4, inhibits AP-1 transactivation
but not NF-κB or ODC transactivation. Oncogene 20, 669–676 (2001).
13. Yang, H.S. et al. The transformation suppressor Pdcd4 is a novel eukaryotic
translation initiation factor 4A binding protein that inhibits translation. Mol. Cell.
Biol. 23, 26–37 (2003).
14. Loh, P.G. et al. Structural basis for translational inhibition by the tumour suppressor
Pdcd4. EMBO J. 28, 274–285 (2009).
15. Hilliard, A. et al. Translational regulation of autoimmune inflammation and lymphoma
genesis by programmed cell death 4. J. Immunol. 177, 8095–8102 (2006).
16. Dorrello, N.V. et al. S6K1- and βTRCP-mediated degradation of PDCD4 promotes
protein translation and cell growth. Science 314, 467–471 (2006).
17. Schmid, T. et al. Translation inhibitor Pdcd4 is targeted for degradation during
tumor promotion. Cancer Res. 68, 1254–1260 (2008).
18. Asangani, I.A. et al. MicroRNA-21 (miR-21) post-transcriptionally downregulates
tumor suppressor Pdcd4 and stimulates invasion, intravasation and metastasis in
colorectal cancer. Oncogene 27, 2128–2136 (2008).
19. Frankel, L.B. et al. Programmed cell death 4 (PDCD4) is an important functional
target of the microRNA miR-21 in breast cancer cells. J. Biol. Chem. 283,
1026–1033 (2008).
20. Lu, Z. et al. MicroRNA-21 promotes cell transformation by targeting the programmed
cell death 4 gene. Oncogene 27, 4373–4379 (2008).
21. Cmarik, J.L. et al. Differentially expressed protein Pdcd4 inhibits tumor promoter-induced
neoplastic transformation. Proc. Natl. Acad. Sci. USA 96, 14037–14042 (1999).
22. Rousseau, D., Kaspar, R., Rosenwald, I., Gehrke, L. & Sonenberg, N. Translation
initiation of ornithine decarboxylase and nucleocytoplasmic transport of cyclin D1
mRNA are increased in cells overexpressing eukaryotic initiation factor 4E. Proc.
Natl. Acad. Sci. USA 93, 1065–1070 (1996).
23. Graff, J.R. et al. Translation of ODC mRNA and polyamine transport are suppressed
in ras-transformed CREF cells by depleting translation initiation factor 4E. Biochem.
Biophys. Res. Commun. 240, 15–20 (1997).
24. Talotta, F. et al. An autoregulatory loop mediated by miR-21 and PDCD4 controls
the AP-1 activity in RAS transformation. Oncogene 28, 73–84 (2009).
25. Fujita, S. et al. miR-21 Gene expression triggered by AP-1 is sustained through a
double-negative feedback mechanism. J. Mol. Biol. 378, 492–504 (2008).
26. Loffler, D. et al. Interleukin-6 dependent survival of multiple myeloma cells involves
the Stat3-mediated induction of microRNA-21 through a highly conserved enhancer.
Blood 110, 1330–1333 (2007).
27. Calin, G.A. & Croce, C.M. MicroRNA signatures in human cancers. Nat. Rev. Cancer
6, 857–866 (2006).
28. Cho, W.C. OncomiRs: the discovery and progress of microRNAs in cancers. Mol.
Cancer 6, 60 (2007).
© 2010 Nature America, Inc. All rights reserved.

nature immunology VOLUME 11 NUMBER 2 FEBRUARY 2010 147
ARTICLES
29. Moschos, S.A. et al. Expression profiling in vivo demonstrates rapid changes in lung
microRNA levels following lipopolysaccharide-induced inflammation but not in the
anti-inflammatory action of glucocorticoids. BMC Genomics 8, 240 (2007).
30. Lu, T.X., Munitz, A. & Rothenberg, M.E. MicroRNA-21 is up-regulated in allergic
airway inflammation and regulates IL-12p35 expression. J. Immunol. 182,
4994–5002 (2009).
31. Iliopoulos, D., Malizos, K.N., Oikonomou, P. & Tsezou, A. Integrative microRNA and
proteomic approaches identify novel osteoarthritis genes and their collaborative
metabolic and inflammatory networks. PLoS ONE 3, e3740 (2008).
32. Sonkoly, E. et al. MicroRNAs: novel regulators involved in the pathogenesis of
psoriasis? PLoS ONE 2, e610 (2007).
33. Wu, F. et al. MicroRNAs are differentially expressed in ulcerative colitis and alter
expression of macrophage inflammatory peptide-2α. Gastroenterology 135,
1624–1635 (2008).
34. Zhang, Z. et al. miR-21 plays a pivotal role in gastric cancer pathogenesis and
progression. Lab. Invest. 88, 1358–1366 (2008).
35. Ji, R. et al. MicroRNA expression signature and antisense-mediated depletion reveal
an essential role of MicroRNA in vascular neointimal lesion formation. Circ. Res.
100, 1579–1588 (2007).
36. Cheng, Y. et al. MicroRNAs are aberrantly expressed in hypertrophic heart:
do they play a role in cardiac hypertrophy? Am. J. Pathol. 170, 1831–1840
(2007).
37. O’Connell, R.M., Taganov, K.D., Boldin, M.P., Cheng, G. & Baltimore, D. MicroRNA-
155 is induced during the macrophage inflammatory response. Proc. Natl. Acad.
Sci. USA 104, 1604–1609 (2007).
38. Ceppi, M. et al. MicroRNA-155 modulates the interleukin-1 signaling pathway in
activated human monocyte-derived dendritic cells. Proc. Natl. Acad. Sci. USA 106,
2735–2740 (2009).
39. Thai, T.H. et al. Regulation of the germinal center response by microRNA-155.
Science 316, 604–608 (2007).
40. Rodriguez, A. et al. Requirement of bic/microRNA-155 for normal immune function.
Science 316, 608–611 (2007).
41. Bazzoni, F. et al. Induction and regulatory function of miR-9 in human monocytes
and neutrophils exposed to proinflammatory signals. Proc. Natl. Acad. Sci. USA
106, 5282–5287 (2009).
42. Liu, G. et al. miR-147, a microRNA that is induced upon Toll-like receptor
stimulation, regulates murine macrophage inflammatory responses. Proc. Natl. Acad.
Sci. USA (2009).
43. Alsaleh, G. et al. Bruton’s tyrosine kinase is involved in miR-346-related regulation
of IL-18 release by lipopolysaccharide-activated rheumatoid fibroblast-like
synoviocytes. J. Immunol. 182, 5088–5097 (2009).
44. Chen, X.M., Splinter, P.L., O’Hara, S.P. & LaRusso, N.F. A cellular micro-RNA,
let-7i, regulates Toll-like receptor 4 expression and contributes to cholangiocyte
immune responses against Cryptosporidium parvum infection. J. Biol. Chem. 282,
28929–28938 (2007).
45. Tili, E. et al. Modulation of miR-155 and miR-125b levels following lipopolysaccharide/
TNF-α stimulation and their possible roles in regulating the response to endotoxin
shock. J. Immunol. 179, 5082–5089 (2007).
46. Naora, H., Altin, J.G. & Young, I.G. TCR-dependent and -independent signaling
mechanisms differentially regulate lymphokine gene expression in the murine
T helper clone D10.G4.1. J. Immunol. 152, 5691–5702 (1994).
47. Nemeth, Z.H. et al. Adenosine augments IL-10 production by macrophages through an
A2B receptor-mediated posttranscriptional mechanism. J. Immunol. 175, 8260–8270
(2005).
48. Sharma, A. et al. Posttranscriptional regulation of interleukin-10 expression by
hsa-miR-106a. Proc. Natl. Acad. Sci. USA 106, 5761–5766 (2009).
49. van Duin, D., Medzhitov, R. & Shaw, A.C. Triggering TLR signaling in vaccination.
Trends Immunol. 27, 49–55 (2006).
50. Cook, D.N., Pisetsky, D.S. & Schwartz, D.A. Toll-like receptors in the pathogenesis
of human disease. Nat. Immunol. 5, 975–979 (2004).
© 2010 Nature America, Inc. All rights reserved.

nature immunology doi:10.1038/ni.1828
ONLINE METHODS
Reagents. Ultrapure TLRgrade LPS was from Alexis. Pam3CSK4 was from
Calbiochem and poly(I:C) was from Amersham Biosciences. MG132 was
from Calbiochem. SMARTpool siRNAs specific for human and mouse
PDCD4 and negative control SMARTpool siRNA were from Dharmacon.
Pro-miR-21, anti-miR-21, specific to human and mouse miRNA, and the
respective negative control RNAs were from Ambion. Morpholino oligo-
nucleotides with the following sequences were designed in association with
GeneTools: PDCD4-miR-21 (morpho-21), 5′-AAGTAGCTTATCAGAACAC
CCACAC-3′, and PDCD4-control (morpho-ctrl), 5′-GATCAGGTCCTAAA
CATGGCACTTA-3′.
Cell culture and animal handling. RAW264.7 and THP-1 cells were from the
European Collection of Cell Cultures. HEK293-TLR4 cells (HEK293-TLR4-
MD2-CD14) were from Invivogen. Wild-ty pe C57BL/6 mice were housed
and maintained at Trinity College, Dublin. Wild-type, MyD88-deficient
and TRIF-deficient BMDMs immortalized by retrovirus were provided by
D. Golenbock. PDCD4-deficient mice were housed and maintained at the
University of Pennsylvania. The p65-deficient and matched wild-type control
mouse embryonic fibroblasts were provided by R. Hay. All cells were main-
tained in either RPMI media or DMEM (Invivogen) supplemented with 10%
(vol/vol) FCS and penicillin and streptomycin. For BMDM extraction, animals
were killed humanely according to the regulations of the European Union and
the Irish Department of Health. Bone marrow was extracted from femurs and
tibias and red blood cells were lysed with Red Blood Cell Lysing Buffer (Sigma),
and the resulting cells were grown in medium conditioned with macrophage
colony-stimulating factor. For analysis of human PBMCs, mononuclear cells
were isolated from whole blood with a Ficoll gradient (Sigma) and were grown
in RPMI media as described above. For analysis of the lethality of LPS, mice
were injected intraperitoneally with LPS at a dose of 30 mg per kg body weight.
The survival of mice was monitored for 10 d, after which all surviving mice
were killed humanely in accordance with the Animal Research Committee of
the University of Pennsylvania.
RNA extraction and PCR. Cells were grown to 5 × 106 to 10 × 106 cells and
RNA was extracted with the RNeasy Kit (Qiagen), modified to obtain small
RNA species. For miRNA analysis, individual miRNA TaqMan assays for the
endogenous reference RNA RNU6B, miR-21 and miR-146a were done accord-
ing to the manufacturer’s instructions (Applied Biosystems). For analysis
of gene expression, cDNA was prepared with the High-Capacity cDNA
Archive kit according to manufacturers’ instructions (Applied Biosystems),
and individual mRNAs were monitored with the following inventoried
TaqMan assays (Applied Biosystems): mouse Gapdh (glyceraldehyde phos-
phate dehydrogenase) assay, mouse Pdcd4 assay (Mm01266062_m1), mouse
Il6 assay (Mm99999064_m1) and mouse Il10 assay (Mm99999062_m1). The
AB7900FAST platform (Applied Biosystems) was used for all PCR, done in
triplicate. Changes in expression were calculated by the change in threshold
(∆∆CT) method with RNU6B as the endogenous control for miRNA analysis
and Gapdh as an endogenous control for gene-expression analysis and were
normalized to results obtained with untreated cells.
Transient transfection. For transfection of miRNA and morpholino oligo-
nucleotides, 5 × 106 RAW264.7 cells were transfected with the appropriate
RNA oligonucleotides with 2% Lipofectamine 2000 (Invitrogen). For siRNA
transfection, 5 × 106 RAW264.7 cells were transfected with 2% Lipofectamine
RNAiMax (Invitrogen). Cells were allowed to recover for 24 h before treatment
with LPS for various times. For all RNA transfection, equal total concentra-
tions of RNA were used for each reaction, with negative control RNA mole-
cules used for normalization. For luciferase assays, 5 × 106 HEK293-TLR4 cells
were transfected with endotoxin-free 5× NF-κB reporter luciferase plasmid
(a gift from R. Hofmeister) and pRL-TK, the renilla luciferase reporter, with
6% GeneJuice (Novagen). Cells were allowed to recover for 24 h before being
transfected with RNA oligonucleotides as described above. Reporter gene
activity was measured with the Dual-Luciferase kit (Promega) 18 h after LPS
treatment. The pMIR-REPORT miR-21 reporter luciferase plasmid was from
A.M. Cheng. The Il10 promoter luciferase plasmid was a gift from A. Bowie.
HA-tagged ubiquitin was a gift from R. Hay.
ELISA. Cytokine concentration in supernatants were measured with ELISA
DuoSet Development systems according to the manufacturer’s instructions
(R&D Systems).
Coimmunoprecipitation of ubiquitinated PDCD4. HEK293-TLR4 cells
(2.5 × 106) were grown to 70% confluency and were transfected with 4 µg
plasmid encoding HA-tagged ubiquitin with 6% GeneJuice (Novagen). Cells
were allowed to recover for 24 h before stimulation for various times w ith
100 ng/ml of LPS, then were lysed as described51. For assistance in the detection
of ubiquitinated proteins, proteasomal degradation was prevented by treat-
ment of the cells with MG132 (1 µg/ml) for 4 h before lysis. Anti-PDCD4 (2 µg;
600-401-965; Rockland) was preincubated for 16 h with protein A/G PLUS-
Agarose beads (sc_2003; Santa Cruz Biotechnology) and subsequently washed
in lysis buffer. PDCD4 was isolated by incubation of beads with cell lysates
for 2 h at 4 °C. After beads were washed, immune complexes were eluted with
50 µl Laemmli sample buffer and separated by SDS-PAGE and proteins were
detected by immunoblot analysis with anti-HA (HA-101R; Sigma).
Immunoblot. Cells (5 × 105 to 10 × 105) were lysed with low-stringency lysis
buffer complete with protease inhibitors. Samples loaded with 5× denaturing
sample buffer were separated by 12% SDS-PAGE. Proteins were transferred to
polyvinylidene difluoride membrane by standard techniques and were subse-
quently analyzed by immunoblot with the relevant antibodies. Protein abundance
was calculated by densitometry scanning of the blot with IMAGE-J software;
PDCD4 abundance was normalized to β-actin abundance and is presented rela-
tive to results obtained with the control sample (percentage or fraction), set as 1.0.
The monoclonal antibody used to detect β-actin was from Sigma (AC-15).
Statistical tests. All statistical analyses used the Student’s t-test, unpaired for
normal distributions of at least three independent experiments.
51. Bowie, A. et al. A46R and A52R from vaccinia virus are antagonists of host IL-1 and
toll-like receptor signaling. Proc. Natl. Acad. Sci. USA 97, 10162–10167 (2000).
© 2010 Nature America, Inc. All rights reserved.