Báo cáo khoa học: Rapamycin inhibits lipopolysaccharide induction of granulocyte-colony stimulating factor and inducible nitric oxide synthase expression in macrophages by reducing the levels of octamer-binding factor-2 doc

As both iNOS and G-CSF are potential Oct-2 target genes, we tested the effect of rapamycin on their expression and found that it reduced the LPS-induced increase in iNOS and G-CSF mRNA l

granulocyte-colony stimulating factor and inducible nitric oxide synthase expression in macrophages by reducing the levels of octamer-binding factor-2

Yuan-Yi Chou1, Jhen-I Gao1, Shwu-Fen Chang2, Po-Yuan Chang3and Shao-Chun Lu1

1 Department of Biochemistry and Molecular Biology, College of Medicine, National Taiwan University, Taipei, Taiwan

2 Graduate Institute of Medical Sciences, Taipei Medical University, Taiwan

3 Department of Internal Medicine, National Taiwan University Hospital, National Taiwan University, Taipei, Taiwan

Introduction

Macrophages play a critical role in the host defense

against bacterial pathogens The toll-like receptor 4

(TLR4) on the surface of macrophages recognizes

lipopolysaccharide (LPS), a Gram-negative bacterial endotoxin, and induces the production of proinflam-matory cytokines [1,2] In LPS-stimulated macrophages,

Keywords

granulocyte-colony stimulating factor

(G-CSF); inducible nitric oxide synthase

(iNOS); lipopolysaccharide (LPS);

macrophage; mammalian target of

rapamycin (mTOR); octamer-binding factor-2

(Oct-2); rapamycin

Correspondence

S.-C Lu, Room 810, No.1, Jen Ai Road

Section 1, Department of Biochemistry and

Molecular Biology, College of Medicine,

National Taiwan University, Taipei 10051,

Taiwan

Tel: +886 2 2312 3456, ext 88224

Fax: +886 2 2391 5295

E-mail: lsc@ntu.edu.tw

Website: http://www.mc.ntu.edu.tw/

department/ibmb/

(Received 22 July 2010, revised 5 October

2010, accepted 21 October 2010)

doi:10.1111/j.1742-4658.2010.07929.x

This article reports an inhibitory effect of rapamycin on the lipopolysac-charide (LPS)-induced expression of both inducible nitric oxide synthase (iNOS) and granulocyte-colony stimulating factor (G-CSF) in macrophages and its underlying mechanism The study arose from an observation that rapamycin inhibited the LPS-induced increase in octamer-binding factor-2 (Oct-2) protein levels through a mammalian target of rapamycin (mTOR)-dependent pathway in mouse RAW264.7 macrophages As both iNOS and G-CSF are potential Oct-2 target genes, we tested the effect of rapamycin

on their expression and found that it reduced the LPS-induced increase in iNOS and G-CSF mRNA levels and iNOS and G-CSF protein levels Blocking of mTOR-signaling using a dominant-negative mTOR expression plasmid resulted in inhibition of the LPS-induced increase in iNOS and G-CSF protein levels, supporting the essential role of mTOR Forced expression of Oct-2 using the pCG–Oct-2 plasmid overcame the inhibitory effect of rapamycin on the LPS-induced increase in iNOS and G-CSF mRNA levels Chromatin immunoprecipitation assays showed that LPS enhanced the binding of Oct-2 to the iNOS and G-CSF promoters and that this effect was inhibited by pretreatment with rapamycin Moreover, RNA interference knockdown of Oct-2 reduced iNOS and G-CSF expression in LPS-treated cells The inhibitory effect of rapamycin on the LPS-induced increase in Oct-2 protein levels and on the iNOS and G-CSF mRNA levels was also detected in human THP-1 monocyte-derived macrophages This study demonstrates that rapamycin reduces iNOS and G-CSF expression at the transcription level in LPS-treated macrophages by inhibiting Oct-2 expression

Abbreviations

Akt-in, Akt inhibitor; ChIP, chromatin immunoprecipitation; DN, dominant-negative; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; G-CSF, granulocyte-colony stimulating factor; iNOS, inducible nitric oxide synthase; LPS, lipopolysaccharide; mTOR, mammalian target of rapamycin; NO, nitric oxide; Oct-1, octamer-binding factor-1; Oct-2, octamer-binding factor-2; PI3K, phosphoinositide 3-kinase; PMA, 4b-phorbol 12-myristate 13-acetate; RR, rapamycin-resistant; TLR4, toll-like receptor 4; TSA, trichostatin A.

gene expression is controlled by the activation of

various protein kinases, such as protein kinase A,

pro-tein kinase C, Src-related kinases, mitogen-activated

protein kinases and phosphoinositide 3-kinase (PI3K),

downstream of the TLR4-signaling pathway [3–5] Of

the LPS-induced genes, those coding for inducible

nitric oxide synthase (iNOS) and granulocyte-colony

stimulating factor (G-CSF) attracted our interest,

because both proteins are important in the host’s

defense against microbial infection iNOS catalyzes the

production of nitric oxide (NO) to combat invading

pathogens [6] and G-CSF stimulates the production,

growth and function of neutrophils [7,8] Both iNOS

and G-CSF genes are expressed in macrophages and

their expression is strongly induced by LPS at the

transcriptional level [9–11] In addition, a nuclear

factor-kappa B-binding element, a nuclear factor-interleukin-6

(also named C⁄ EBPb)-binding element and an

oct-amer element in the promoter of the iNOS and G-CSF

genes have been reported to be essential for full

pro-moter activities of these two genes following

stimula-tion with LPS [9,10,12,13] However, there is clinical

evidence that excessive concentrations of NO or

G-CSF exacerbates inflammatory responses and causes

tissue damage [14,15] It is therefore necessary to

main-tain appropriate levels of both iNOS and G-CSF

during inflammation, and the expression of these two

genes should be tightly controlled and regulated by

similar mechanisms

Activation of PI3K leads to the production of

phos-phatidylinositol 3,4-bisphosphate and

phosphatidylino-sitol 3,4,5-trisphosphate [16,17], which subsequently

activate downstream signaling molecules, such as Akt

[18] and mammalian target of rapamycin (mTOR),

reg-ulating various biological processes, including cell

cycling, cell survival and protein synthesis [19] There

is growing evidence that mTOR (activated by TLR4

via PI3K⁄ Akt) is crucial in monocytes and

macrophag-es for coordinating the innate immune rmacrophag-esponse, but

how mTOR exerts its effect is poorly understood

[20,21] A well-known function of mTOR is to regulate

protein synthesis by activating p70 S6 kinase and by

inhibiting eukaryotic translation initiation factor

4E-binding protein (eIF4E-BP1) [19] Rapamycin, a

potent immunosuppressor, exerts its function by binding

to the FK506-binding protein (FKBP12) and inhibits

the activity of mTOR complex 1 (mTORC1) and

subse-quently inhibits the translation of target mRNA, but

has no effect on mTOR complex 2 (mTORC2) [22] In

contrast, in LPS-stimulated innate immune cells,

rapa-mycin alters the expression of cytokine genes at the

transcriptional level [21], and LPS-induced expression of

iNOSin macrophages has been reported to be partially

inhibited by rapamycin at the mRNA level [23] These findings suggest that mTOR may be involved in gene expression through mechanisms other than trans-lational control Octamer-binding factor-2 (Oct-2) is a transcription factor that binds to the octamer element (ATGCAAAT) in the promoter of its target gene We have previously demonstrated that expression of Oct-2 can be induced by LPS in macrophages and that it is involved in the LPS-induced upregulation of resistin and iNOS expression [24,25] In addition, we reported that treatment of macrophages with LY294002 decre-ased LPS-induced Oct-2 expression at the protein level, subsequently reducing the expression of resistin [24]

As LY294002 inhibits the activity of both PI3K and mTOR [26], we speculated that the induction of Oct-2 expression by LPS might occur through an mTOR-dependent pathway and be inhibited by the mTOR inhibitor, rapamycin In addition, Oct-2 is involved in the expression of LPS-inducible genes that contain octamer in their promoters [24,25] Thus, rapamycin may inhibit the expressions of iNOS and G-CSF at the transcriptional level in LPS-treated macrophages In this study, the effects of rapamycin on the LPS-induced increase in Oct-2 protein and on the expres-sion of iNOS and G-CSF protein and mRNA were evaluated in macrophages, and the involvement of Oct-2 in LPS⁄ mTOR-induced iNOS and G-CSF expression was investigated further

Results

Rapamycin inhibits the LPS-induced increase in Oct-2 protein levels in RAW264.7 macrophages Oct-2 mRNA and Oct-2 protein levels increased in a time-dependent manner when RAW264.7 cells were exposed to LPS for 1 to 24 h Oct-2 mRNA levels increased by 60% at 1 h and reached a maximum, of about threefold higher than the basal level, after 4–8 h

of stimulation with LPS, then showed a subsequent decrease at 24 h (Fig 1A, upper panel) Oct-2 protein levels showed an 80% increase at 1 h, reached a maxi-mum of about 10-fold higher than the basal level at

8 h and were maintained at this level for at least 24 h (Fig 1A, bottom panel) To confirm that the LPS-induced increase in Oct-2 protein was inhibited by the PI3K inhibitor LY294002 [24], the cells were treated with 12.5, 25 or 50 lm LY294002 for 30 min before, and during, treatment with LPS for 4 h Fig 1B shows that LY294002 inhibited the LPS-induced increase in Oct-2 protein levels in a dose-dependent manner, reaching 70% inhibition at 50 lm LY294002 A sim-ilar, but lower, inhibitory effect on the LPS-induced

Oct-2 protein increase was also seen in cells treated

with 0.5 lm wortmannin, another PI3K inhibitor (data

not shown) As Akt is a well-characterized downstream

effecter of the PI3K signaling pathway, we examined

whether it is involved in the LPS-induced increase in

Oct-2 protein Pretreatment with 25, 50 or 75 lm Akt

inhibitor (Akt-in) before treatment with LPS for 4 h

resulted in a dose-dependent decrease in Oct-2 protein,

with maximal inhibition (approximately 70%) being

achieved using 50 or 75 lm Akt-in (Fig 1C)

Inhibi-tors (LY294002 and Akt-in) had no effect on Oct-2

expression in cells not stimulated with LPS (Fig 1D)

To further evaluate the involvement of Akt in the

LPS-induced increase of Oct-2 protein, RAW264.7

cells were transfected with the dominant-negative

(DN)-Akt expression plasmid for 24 h before 4 h of

stimulation with LPS, and this resulted in a reduction

of about 60% in Oct-2 protein expression compared

with cells not treated with the expression plasmid

(Fig 1E) However, the levels of Oct-2 mRNA were

not altered by either LY294002 or Akt-in in LPS-trea-ted cells (Fig 1F) These results show that LPS upre-gulates Oct-2 protein in macrophages through a PI3K⁄ Akt-dependent pathway As mTOR is one of several downstream targets of PI3K and because LY294002 is also an mTOR inhibitor [26], it is possi-ble that mTOR is also involved in LPS-induced Oct-2 expression To test this possibility, cells were pretreated with increasing amounts of rapamycin before treat-ment with LPS and then the expression of Oct-2 was analyzed Fig 2A shows that rapamycin inhibited the increase in Oct-2 protein levels in LPS-treated cells in

a dose-dependent manner, whereas the Oct-2 mRNA levels in LPS-treated cells were not affected (Fig 2C) Rapamycin had no effect on Oct-2 expression in cells not stimulated with LPS (Fig 2B) To verify the involvement of mTOR in LPS-induced Oct-2 expres-sion, RAW264.7 cells were transfected with either the DN-mTOR expression plasmid or the rapamycin-resistant (RR)-mTOR expression plasmid 24 h before

GAPDH

Oct

-2

β-actin

LPS 0 1 2 4 8 24

Time (h)

Oct-2

*

1 2 3 4 5 6

LY (μM )

LPS

Oct-2

β-actin

0 0 12.5 25 50

1 2 3 4 5

Oct-2

β-actin

LPS

Akt-in (μM )

0 0 25 50 75

1 2 3 4 5

LY LPS

Oct-2

β-actin

Akt-in

LPS

Oct-2

β-actin

1 2 3 4

1 2 3 4

Oct-2

LPS

β-actin

Vector DN-Akt

1 2 3 4

GAPDH Oct-2

LPS

LY Akt-in

1 2 3 4

Fig 1 LPS induces the expression of Oct-2 mRNA and Oct-2 protein in RAW264.7 macrophages, and the induction of Oct-2 protein by LPS

is blocked by a PI3K or an Akt inhibitor (A) RAW264.7 cells were treated with 100 ngÆmL)1of LPS for 0, 1, 2, 4, 8 or 24 h, then the levels

of Oct-2 mRNA (upper panels) or Oct-2 protein (lower panels) were determined by RT-PCR or western blot analysis, respectively GAPDH or b-actin was used as the respective internal control The asterisk indicates a nonspecific signal (B and C) RAW264.7 cells were untreated (lane 1), were treated with LPS for 4 h (lane 2), or were pretreated for 30 min with increasing concentrations of LY294002 (B) or Akt-in (C) before treatment with LPS (100 ngÆmL)1) for 4 h (lanes 3–5); the levels of Oct-2 and b-actin were then measured by western blotting (D) RAW264.7 cells were untreated (lanes 1 and 2) or were pretreated with 50 l M LY294002 (upper panels) or 50 l M Akt-in (lower panels) for

30 min (lanes 3 and 4), then incubated with (lanes 2 and 4) or without (lanes 1 and 3) LPS for 4 h The Oct-2 and b-actin levels were measured by western blotting (E) Cells were transiently transfected with 7.5 lg of empty vector or DN-Akt expression plasmid, then, after

24 h, were incubated for 4 h in the presence or absence of LPS and the levels of Oct-2 and b-actin were analyzed by western blotting (F) Cells were untreated (lane 1) or were treated for 4 h with LPS in the absence (lane 2) or presence of 25 l M LY294002 (lane 3) or 50 l M

Akt-in (lane 4), then the Oct-2 and GAPDH mRNA levels were analyzed by RT-PCR The results are representative of three independent experiments LY, LY294002.

treatment with LPS for 4 h Transfection with the

DN-mTOR expression plasmid inhibited the increase

in Oct-2 protein in LPS-treated RAW264.7

macro-phages by  60% (Fig 2D), while transfection with

the RR-mTOR expression plasmid overcame the

inhibitory effect of rapamycin on the increase in Oct-2

protein levels in LPS-treated cells (Fig 2E) The

phos-phorylation levels of p70 S6 kinase were assayed to

verify the activation of mTOR Fig 2F shows that

phospho-p70 S6 kinase was detectable after 30 min of

LPS treatment and that this effect was blocked by

rapamycin treatment or by transfection with the

DN-mTOR expression plasmid None of these

treat-ments affected the total amount of p70 S6 kinase protein

These results strongly suggest that mTOR activation is

essential for LPS-induced Oct-2 protein expression

Rapamycin decreases LPS-induced iNOS

expression in RAW264.7 macrophages

The octamer element plays an essential role in the

LPS-induced activation of the promoters of the iNOS,

G-CSF and resistin [12,13,24,25] and we previously demonstrated that inhibition of Oct-2 expression by trichostatin A (TSA), a potent histone deacetylase inhibitor, or by LY294002 leads to reduced iNOS

or resistin expression, respectively, in LPS-treated mac-rophages [24,25] It was possible that rapamycin might also reduce the expression of these genes We therefore tested the effect of rapamycin on the LPS-induced expression of iNOS and G-CSF Fig 3A shows that detectable levels of iNOS and G-CSF mRNAs were induced in RAW264.7 cells by 4 h of incubation with LPS and that both effects were inhibited by pretreat-ment with rapamycin LPS-induced iNOS protein expression (4 h of incubation, Fig 3B) and nitrite pro-duction (24 h of incubation, Fig 3C) were also inhib-ited in cells pretreated with rapamycin Furthermore, DN-mTOR-transfected cells expressed only 46% as much iNOS protein as control cells in response to LPS (Fig 3D) and produced only 43% as much nitrite (Fig 3E) These results show that rapamycin inhibited LPS-induced iNOS expression in RAW264.7 macro-phages through an mTOR-dependent pathway

Oct-2

β-actin

LPS

Rapa (ng·mL –1 )

0 0 100 200 400

1 2 3 4 5

Rapa LPS

Oct-2

β-actin

1 2 3 4

0.0 0.5 1.0 1.5 2.0 2.5 3.0 GAPDH Oct-2

1 2 3

Rapa LPS

Oct-2

β-actin LPS Vector DN-mTOR

1 2 3 4

RR-mTOR Rapa

LPS

Oct-2 β-actin

1 2 3 4

p-p70S6K p70S6K

Rapa

LPS

DN-mTOR Vector Control

1 2 3 4 5 6

Fig 2 Rapamycin reduces the LPS-induced increase in Oct-2 protein levels, but not in Oct-2 mRNA levels, in RAW264.7 macrophages through an mTOR-dependent pathway (A) RAW264.7 cells were untreated (lane 1), were treated with 100 ngÆmL)1of LPS for 4 h (lane 2)

or were pretreated for 30 min with increasing concentrations of rapamycin before LPS treatment for 4 h (lanes 3–5); the concentrations of Oct-2 and b-actin were then measured after western blotting (B) RAW264.7 cells were untreated (lanes 1 and 2), or were pretreated with

200 ngÆmL)1of rapamycin for 30 min (lanes 3 and 4), followed by incubation with (lanes 2 and 4) or without (lanes 1 and 3) LPS for 4 h The levels of Oct-2 and b-actin were measured after western blotting (C) RAW264.7 cells were untreated (lane 1), were treated with LPS for

4 h (lane 2), or were pretreated for 30 min with 200 ngÆmL)1of rapamycin before LPS treatment for 4 h, then the Oct-2 and GAPDH mRNA levels were measured by RT-PCR The levels of Oct-2 mRNA were normalized to those for GAPDH and the results were expressed relative

to those in the untreated control (relative value = 1) The values are expressed as the mean ± sd of three independent experiments (D and E) RAW264.7 cells were transiently transfected with 7.5 lg of empty vector or plasmid encoding DN-mTOR (D) or RR-mTOR (E), then, 24 h later, were incubated for 4 h with or without LPS in the absence or presence of 200 ngÆmL)1of rapamycin; the Oct-2 and b-actin levels were analyzed after western blotting (F) RAW264.7 cells were untreated (lanes 1 and 2) or were preincubated with 200 ngÆmL)1of rapamy-cin for 30 min (lane 3) or transiently transfected for 24 h with 7.5 lg of empty vector (lanes 4 and 5) or plasmid encoding DN-mTOR (lane 6), then incubated in the presence or absence of LPS for 30 min and the levels of total and phosphorylated p70 S6 kinase were determined

by western blot analysis Similar results were obtained in three separate experiments Rapa, rapamycin.

Rapamycin inhibits G-CSF expression in

LPS-treated RAW264.7 macrophages

To examine whether LPS-induced G-CSF expression

was also sensitive to rapamycin treatment, RAW264.7

cells were pretreated with rapamycin for 30 min before

treatment with LPS for different periods of time The

levels of G-CSF mRNA (Fig 4A) and of G-CSF

pro-tein (Fig 4B) were below the detection limit in

untreated cells and increased in a time-dependent

man-ner in response to LPS stimulation Pretreatment with

rapamycin resulted in approximately 50% less G-CSF

protein in the medium (Fig 4B), which was probably

caused by attenuation of G-CSF promoter activity

(Fig 4C) and mRNA levels (Fig 3A) by rapamycin

Transfection of RAW264.7 cells with the DN-mTOR

expression plasmid also resulted in a reduction of

about 50% in the LPS-induced increase in G-CSF

pro-tein in the medium (6 h of LPS treatment, Fig 4D),

suggesting that activation of mTOR is essential for

induction of G-CSF expression by LPS In a previous

study, we showed that forced expression of Oct-2

restores iNOS protein levels in TSA- and LPS-treated cells [25] In the present study, ectopic expression of Oct-2 by transfection with increasing amounts of pCG-Oct-2 overcame the rapamycin-mediated inhibi-tion of iNOS and G-CSF mRNA expression in LPS-treated RAW264.7 cells in a dose-dependent manner (4 h of incubation; Fig 4E) These results confirm that Oct-2 plays a critical role in the upregulation of iNOS and G-CSF expression by LPS in RAW264.7 macro-phages; moreover, these results also show that the LPS-induced increase in G-CSF expression is sensitive

to rapamycin and suggest that mTOR activation is required for G-CSF expression in response to LPS

Oct-2 is directly involved in LPS-induced iNOS and G-CSF expression in RAW264.7 macrophages

To further examine the involvement of Oct-2 in LPS-induced iNOS and G-CSF gene expression in RAW264.7 cells, a chromatin immunoprecipitation (ChIP) assay was performed to examine the binding of Oct-2 to the iNOS and G-CSF promoters in vivo In

GAPDH

iNOS G-CSF

LPS Rapa

1 2 3

β-actin

iNOS

LPS LY Rapa

1 2 3 4

Time (h)

0 4 8 12 16 20 24

0 10 20

30

LPS LPS + Rapa

*

*

Vector

β-actin

iNOS

LPS

DN-mTOR

1 2 3 4

0 20 40 60 80 100

*

Vector DN-mTOR LPS Nitrite induction (% of control)

Fig 3 Rapamycin reduces LPS-induced NO production and the expression of iNOS mRNA and iNOS protein in RAW264.7 macrophages (A and B) RAW264.7 cells were either untreated or pretreated with LY294002 (50 l M ) or rapamycin (200 ngÆmL)1) for 30 min and then treated with LPS (100 ngÆmL)1) for 4 h The levels of iNOS and G-CSF mRNAs (A) and the level of iNOS protein (B) were analyzed by RT-PCR and western blotting, respectively GAPDH mRNA or b-actin was used as the internal control Similar results were obtained in at least three independent experiments (C) Cells were treated with LPS in the absence or presence of 200 ngÆmL)1of rapamycin for 0, 16 or

24 h, then the NO levels in the culture medium were determined using a Griess reagent system kit (D and E) RAW264.7 cells were transiently transfected with 7.5 lg of empty vector or DN-mTOR expression plasmid, then, after 24 h, were incubated for 4 h (D) or 24 h (E)

in the presence or absence of LPS; the levels of iNOS and b-actin were analyzed by western blotting (D) and the NO levels in the medium were determined using a Griess reagent system kit (E) The values for the LPS-treated cells were divided by those for the non-LPS-treated cells and are shown as a percentage of the values for the cells transfected with control vector (relative value = 100) All results are expressed as the mean ± SD of three independent experiments *P < 0.01 compared with the untreated control or the LPS-treated control,

as appropriate LY, LY294002; Rapa, rapamycin.

cells treated with LPS for 6 h, but not in control cells,

an iNOS promoter region from nucleotides)90 to 154

and a G-CSF promoter region from nucleotides)70 to

)248, encompassing the octamer, were pulled down by

anti-Oct-2 IgG (Fig 5), but not by a control IgG, and

this effect was blocked by pretreatment with rapamycin

This shows that Oct-2 binds to the iNOS and G-CSF

promoters and that the binding is sensitive to

rapamycin Moreover, specific knockdown of Oct-2 by

transfection with an Oct-2 RNA interference (RNAi)

plasmid (pLL3.7–Oct-2) resulted in decreased levels of

iNOS protein in the cell lysate (4 h of incubation,

Fig 6A) and G-CSF protein in the culture medium

(6 h of incubation, Fig 6B) of LPS-induced cells

These results support the critical role of Oct-2 in

LPS-induced iNOS and G-CSF expression

LPS-induced expression of Oct-2, G-CSF and iNOS in the THP-1 human monocyte/macrophage cell line is sensitive to rapamycin treatment

To determine whether rapamycin also inhibited Oct-2, iNOS, and G-CSF expression in human macrophages, THP-1, a human monocyte⁄ macrophage cell line, was induced to differentiate by incubation with 160 nm 4b-phorbol 12-myristate 13-acetate (PMA) for 24 h, then the cells were treated with rapamycin and LPS (100 ngÆmL)1, 6 h) As in mouse macrophages, Oct-2 protein expression was induced by LPS, and rapamy-cin inhibited this effect by 34% (Fig 7A) Moreover, rapamycin also inhibited the LPS-induced expression

of iNOS and G-CSF mRNAs by 37% and 45%, respectively (Fig 7B)

G-CSF

GADPH LPS 0 2 4 6 12 24

Time (h)

1 2 3 4 5 6

Time (h)

–1 )

0 100 200 300 400

500 LPS LPS + Rapa

*

0 20 40 60 80 100

*

LPS Rapa

Vector DN-mTOR

G-CSF induction (%

0 20 40 60 80 100

*

LPS

GAPDH G-CSF

0 0 0 0.2 0.4 0.6 pCG-Oct-2 (μg)

LPS Rapa

0 1.0 0.5 0.7 0.9 1.1 Fold

Oct-2

β-actin 0.2 1.0 0.4 0.7 0.9 1.3 Fold

1 2 3 4 5 6

0.2 1.0 0.5 1.0 1.4 1.7 Fold iNOS

Fig 4 Rapamycin reduces LPS-induced G-CSF expression in RAW264.7 macrophages (A and B) RAW264.7 cells were untreated or were pretreated with rapamycin (200 ngÆmL)1) for 30 min, followed by treatment with 100 ngÆmL)1of LPS for 0, 2, 4, 6, 8, 12 or 24 h (A) Total RNA was then isolated and the levels of G-CSF and GAPDH mRNAs were determined by RT-PCR The result is representative of those obtained in three similar experiments (B) The levels of G-CSF protein in the medium were determined by ELISA The values are the mean ± sd of an experiment performed in triplicate, and similar results were observed in three separate experiments (C) RAW264.7 cells were transfected with pG-CSF( )283 ⁄ +35)-Luc and phRLTK, then, 24 h later, were pretreated with rapamycin and with LPS for 6 h and the Photinus and Renilla luciferase activities were measured The levels of Photinus luciferase activity were normalized to Renilla luciferase activity and expressed relative to those in the LPS-treated controls (relative value = 100) (D) RAW264.7 cells were transfected and treated

as in Fig 3E, except that LPS treatment was for 6 h, then the G-CSF protein levels in the medium were determined by ELISA The values for the LPS-treated cells were divided by those for the non-LPS-treated cells and are shown as a percentage of the values for the cells trans-fected with control vector (relative value = 100) The values are the mean ± SD of three independent experiments *P < 0.01 compared with the untreated control or with the LPS-treated control, as appropriate (E) RAW264.7 cells were transfected with 2 lg of control pCG vector

or with increasing amounts of pCG–Oct-2 (made up to 2 lg with pCG) and cultured for 24 h, then untreated (lane 1), treated with LPS for

4 h (lane 2) or pretreated for 30 min with 200 ngÆmL)1of rapamycin, then treated with LPS for 4 h (lanes 3–6); iNOS and G-CSF mRNA were detected by RT-PCR (upper panels) and Oct-2 protein in cell lysates was measured by western blotting (lower panels) GAPDH or b-actin was used as the respective internal control The results are representative of three independent experiments Rapa, rapamycin.

We evaluated the effects of rapamycin on the expres-sion of Oct-2 and its potential target genes, iNOS and G-CSF, in RAW264.7 cells and in THP-1 cells The results demonstrated that rapamycin reduced Oct-2 protein levels and attenuated iNOS and G-CSF expres-sion at the level of transcription in LPS-stimulated macrophages Similar results were obtained by trans-fection of cells with DN-mTOR, indicating that the LPS-induced increase in Oct-2, iNOS and G-CSF expression occurs through an mTOR-dependent path-way It is very likely that the increase in Oct-2, iNOS,

or G-CSF expression occurs through the PI3K⁄ Akt ⁄ m-TOR signaling pathway, as mm-TOR is downstream of PI3K-Akt, and treatment of cells with LY294002 or an Akt inhibitor, or transfection of cells with DN-Akt, also resulted in a decrease in the LPS-induced increase

in Oct-2, iNOS and G-CSF expression (Figs 1B,C,E, and 3B and 7A, and data not shown)

Rapamycin has been reported to downregulate NO production by inhibiting phosphorylation [27] or

Oct-2

iNOS promoter

IgG

Input

G-CSF promoter

LPS

Rapa

Fig 5 Rapamycin inhibits the binding of Oct-2 to the iNOS and

G-CSF promoters in LPS-treated RAW264.7 macrophages RAW264.7

cells were either untreated or pretreated with rapamycin for 30 min

and treated with or without LPS for 6 h, then ChIP assays were

performed using anti-Oct-2 IgG (top row) or control IgG (center

row), and a 179 bp G-CSF promoter fragment ( )248 to )70 bp) and

a 244 bp iNOS promoter fragment ( )90 to 154 bp) were amplified

by PCR Ten per cent of the chromatin DNA used for

immunopre-cipitation was subjected to PCR and is indicated as ‘input’ (bottom

row) The results are representative of three independent

experi-ments Rapa, rapamycin.

Control RNAi LPS

β-actin

iNOS

Oct -2

Oct -2 RNAi

A

0 20 40 60 80 100

*

Control RNAi Oct -2

B

Fig 6 Knockdown of Oct-2 reduces the LPS-induced increase in

iNOS and G-CSF expression in RAW264.7 macrophages.

RAW264.7 cells were transiently transfected with either pLL

3.7-scrambled (Sc) or pLL 3.7-Oct-2 RNA interference (RNAi), then,

24 h later, were untreated or were treated with LPS for 4 h (A) or

6 h (B); the amounts of Oct-2, iNOS and b-actin proteins were then

detected by western blotting (A) and the concentration of G-CSF

protein in the medium was measured by ELISA (B) The values for

the LPS-treated cells were divided by those for the non-LPS-treated

cells and are shown as a percentage of the values for the cells

transfected with control vector (relative value = 100) The values

are expressed as the mean ± sd of three independent

experi-ments *P < 0.01 compared with the LPS-treated control.

Oct-2 β-actin LY LPS Akt-in Rapa

1 2 3 4 5 6 7 8

GAPDH G-CSF

LPS Rapa

iNOS

1 2 3 4

A

B

Fig 7 LPS induces Oct-2 protein and iNOS and G-CSF mRNA expression in PMA-differentiated human THP-1 monocyte ⁄ macro-phages, and this effect is blocked by a PI3K, Akt or mTOR inhibitor (A) THP-1 cells were induced to differentiate by exposure to

160 n M PMA for 24 h and then incubated with or without LPS for

6 h in the absence or presence of LY294002 (50 l M ), Akt-in (50 l M )

or rapamycin (200 ngÆmL)1); the levels of Oct-2 and b-actin protein were then determined by western blotting (B) Differentiated THP-1 cells were untreated (lane 1), treated with 200 ngÆmL)1of rapamy-cin for 30 min (lane 3), or treated for 6 h with LPS in the absence (lane 2) or presence (lane 4) of 200 ngÆmL)1of rapamycin; iNOS, G-CSF and GAPDH mRNA levels were then determined by RT-PCR The results are representative of three independent experiments.

LY, LY294002; Rapa, rapamycin.

inducing proteasomal degradation [28] of iNOS

pro-tein, or by decreasing the secretion of interferon-b [29]

However, Attur et al [23] showed that rapamycin

inhibits the LPS-induced accumulation of iNOS

mRNA in macrophages Our data showed that

rapa-mycin inhibited the accumulation of iNOS mRNA at

the transcriptional level and that this effect was caused

by a lower amount of Oct-2 These results are in

agree-ment with our previous results showing that

downregu-lation of Oct-2 is responsible for the decrease in

LPS-induced iNOS expression in TSA-pretreated

macro-phages [25] Moreover, we observed that LPS-induced

expression of G-CSF was decreased by pretreating

macrophages with rapamycin (this study) or TSA (data

not shown), which may also be attributed to the

decrease in Oct-2 expression caused by these reagents

The involvement of Oct-2 in iNOS and G-CSF

expres-sion was further supported by the finding that

knock-down of Oct-2 expression also attenuated iNOS and

G-CSF protein expression (Fig 6) Mutation of the

octamer in the iNOS and G-CSF promoters results in

a reduction of more than 90% in the LPS-induced

pro-moter activities of these genes [13,30], suggesting the

essential role of the octamer and the factors binding to

it in LPS-induced gene expression Although

involve-ment of octamer-binding factor-1 (Oct-1), another

oct-amer-binding protein, cannot be excluded in the

LPS-induced expression of iNOS and G-CSF, it is very

likely that Oct-2 plays a more important role, as the

level of expression of iNOS and G-CSF changed in

parallel with that of Oct-2, whereas Oct-1 is

constitu-tively expressed and its expression is not changed by

treatment with LPS (data not shown) Interestingly, we

observed that LPS induced an increase of about

1.6-fold in Oct-2 mRNA levels, which was not changed by

treatment with rapamycin This result suggests that

rapamycin may regulate the production of Oct-2

protein at the post-transcriptional level Treatment

with lactacystin, a specific inhibitor of the 26S

protea-some, did not change the levels of Oct-2 protein in the

presence or absence of LY294002, Akt-in or rapamycin

in LPS-treated cells (data not shown) As mTOR

regu-lates cell growth and protein synthesis by increasing

the phosphorylation of two major downstream targets

– 4E-BP and ribosomal p70 S6 kinase – it is very

pos-sible that rapamycin downregulates the Oct-2 protein

level by inhibiting Oct-2 protein synthesis

Although recombinant G-CSF is routinely used to

treat neutropenia and to mobilize hematopoietic stem

cells from the bone marrow into the peripheral blood

before transplantation [31], it is also expressed

endoge-nously in a variety of cell types in response to

treat-ment with LPS, tumor necrosis factor-a, interleukin-1b,

PMA or interferon-c [32] Expression of G-CSF can

be regulated at both the transcriptional and post-transcriptional levels [33]; however, the regulatory mechanisms are poorly understood To our knowledge, this is the first study showing that rapamycin reduces the LPS-induced expression of G-CSF at the transcrip-tional level in macrophages by blocking mTOR activ-ity G-CSF functions as an anti-inflammatory cytokine

by activating antibacterial defense by neutrophils and

by reducing the release of proinflammatory mediators, thereby preventing the overactivation of monocytes and lymphocytes [34] Thus, a decrease in G-CSF might decrease the host-defense response during infec-tion This could explain, at least in part, the higher mortality rate in LPS-induced shock in rapamycin-treated mice compared with control mice [21] In con-trast, G-CSF has been demonstrated to induce chronic inflammation and autoimmunity and to exacerbate underlying inflammatory diseases in humans and mice G-CSF deficiency protects mice from collagen-induced arthritis [35], and this effect might be a result of G-CSF not only inducing neutrophil production, but also promoting neutrophil trafficking into inflamed joints [36] Further investigations are required to confirm the clinical usefulness of rapamycin in the treatment of inflammatory diseases

In summary, we demonstrated that rapamycin inhib-its the LPS-induced increase in iNOS and G-CSF expression through an Oct-2-dependent pathway Our results provide evidence for an interesting role of mTOR in the transcriptional control of LPS-induced gene expression Because of its potent immunosuppres-sive and antiproliferative properties, there is consider-able interest in the use of rapamycin for the treatment

of inflammatory diseases As G-CSF can function as either an anti-inflammatory or a proinflammatory cytokine, understanding the mechanism of the effect of rapamycin on G-CSF expression is important Whether rapamycin also inhibits G-CSF expression in other cells, such as rheumatoid synovial fibroblasts, deserves further investigation

Materials and Methods

Materials

LPS from Escherichia coli (serotype 0111:B4) and PMA were purchased from Sigma-Aldrich (St Louis, MO, USA) LY294002, a PI3K inhibitor, 1L6-hydroxymethyl-chiro-ino-sitol-2-(R)-2-O-methyl-3-O-octadecyl-sn-glycerocarbonate, an Akt inhibitor (Akt-in), and rapamycin, an mTOR inhibi-tor, were purchased from Calbiochem (San Diego, CA, USA) and were dissolved in dimethylsulfoxide Dulbecco’s

modified Eagle’s medium (DMEM), penicillin⁄ streptomycin

and fetal bovine serum were obtained from GibcoBrl⁄

Life-Technologies (Rockville, MD, USA) Rabbit polyclonal

anti-Oct-2 and anti-iNOS IgGs, and horseradish

peroxi-dase-conjugated anti-rabbit IgG were purchased from Santa

Cruz Biotechnology, Inc (Santa Cruz, CA, USA) Rabbit

polyclonal anti-p70 S6 kinase and phospho-p70 S6 kinase

(Thr389) IgGs and horseradish peroxidase-conjugated

anti-mouse IgG were purchased from Cell Signaling Technology

(Beverly, MA, USA) Mouse monoclonal anti-b-actin IgG

was purchased from Chemicon (Temecula, CA, USA)

Restriction endonucleases were obtained from New

Eng-land Biolabs (Beverly, MA, USA) The pGL3-Basic and

phRLTK reporter plasmids, the Dual-LuciferaseReporter

Assay System and the Griess reagent system were from

Pro-mega (Madison, WI, USA) The SuperFect Transfection

Reagent was purchased from Qiagen (Hilden, Germany),

the mouse G-CSF Quantikine ELISA kit from R&D

Sys-tems (Minneapolis, MN, USA), the ChIP assay kit from

Upstate Biotechnology (Lake Placid, NY, USA) and

pro-tein A–Sepharose beads from Amersham Biosciences

(Chandler, AZ, USA) Plasmids 1 and

pCG-Oct-2, and their parent plasmid, pCG, were gifts from Dr W

Herr (Cold Spring Harbor Laboratory, Cold Spring

Har-bor, NY, USA) [37] The expression plasmid for DN-Akt

(hemagglutinin-tagged Akt K179M) was kindly provided

by Dr Mien-Chie Hung (University of Texas M D

Anderson Cancer Center) [38], and those for DN-mTOR

(hemagglutinin-tagged mTOR N2343K) and RR-mTOR

(hemagglutinin-tagged mTOR S2035T) were kindly

pro-vided by Dr Kazuyoshi Yonezawa (Biosignal Research

Center, Kobe University) [39]

Cell culture and LPS treatment

RAW264.7, a murine macrophage cell line, was cultured in DMEM supplemented with 10% fetal bovine serum, 4 mm glutamine, 100 UÆmL)1 of penicillin and 100 lgÆmL)1 of streptomycin at 37C in 5% CO2, as described previously [24,25] The human acute monocytic leukemia cell line

THP-1 was obtained from the American Type Culture Collection, and was maintained and induced to differentiate using

160 nm PMA, as described previously [40] In these experi-ments, Oct-2, iNOS and G-CSF mRNA and protein levels,

NO production, the phosphorylation levels of p70 S6 kinase and the transactivation and activity of the G-CSF promoter were compared between untreated cells and cells treated with

100 ngÆmL)1 of LPS (RAW264.7 and THP-1 cells) When inhibitors were used, they were added 30 min before the LPS

Plasmid construction

A DNA fragment containing nucleotides )283 to +35 of the mouse G-CSF promoter was PCR-amplified from geno-mic DNA using the primers listed in Table 1; the under-lined sequences are MluI and BglII sites created to facilitate cloning The DNA fragment was cloned into the MluI⁄ BglII sites of the pGL3-basic luciferase reporter vector to obtain the pG-CSF()283 ⁄ +35)-Luc reporter plasmid The transcription start site (+1) was assigned according to Nag-ata et al [12,41] The Oct-2 short hairpin RNA plasmid (pLL 3.7-Oct-2) was constructed as described previously [24] All constructs were verified by restriction mapping and sequencing

Table 1 Primers used in this study.

Sequence (5¢ fi 3¢) Oligonucleotides used for plasmid construction

Reverse AGATCTGATTCTGGGTGATCTGGGCTGCA Oligonucleotides used for RT-PCR

Reverse AAATGGTCGTTTGGCTGAAG

Reverse ACTTGGGATGCTCCATGGTC

Reverse CTGGAAGGCAGAAGTGAAGG

Reverse CGACACCTCCAGGAAGCTCTG

Reverse GAATTCGTCATGGATGACCTTGGCCAG Oligonucleotides used for the ChIP assay

Reverse CTGGGGCAACTCAGGCTTA

Reverse CTACTCCGTGAAGTGAACAA

a Primers that can be used to amplify both human and mouse Oct-2, iNOS and GAPDH cDNA.

Transient transfection

Transient transfection was carried out using the SuperFect

Transfection Reagent, as described previously [24] Briefly,

RAW264.7 cells were plated and cultured overnight before

transfection Twenty-four hours after transfection, the cells

were either untreated or were pretreated with vehicle or

specific inhibitors for 30 min before incubation with

100 ngÆmL)1of LPS

RNA isolation and RT-PCR

Following treatment, total RNA in RAW264.7 and THP-1

cells was isolated by acid guanidinium thiocyanate⁄

phe-nol⁄ chloroform extraction, according to the method of

Chomczynski and Sacchi [42] The concentration and purity

of the RNA were measured by reading the absorbances at

260 nm and 280 nm, respectively, on a spectrophotometer

The levels of Oct-2, G-CSF, iNOS or

glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNAs were

deter-mined by semiquantitative RT-PCR using the primers listed

in Table 1 The amplified DNA fragments were separated

by electrophoresis through a 1% agarose gel and sequenced

to confirm their identity

Western blot analysis

Samples of cell lysates (40 lg of protein per lane) from

RAW264.7 and THP-1 cells were separated by SDS⁄ PAGE

on 8% gels and transferred onto a poly(vinylidene

difluo-ride) membrane, which was then blocked for 1 h at room

temperature with blocking buffer [TBST (150 mm NaCl,

10 mm Tris⁄ HCl and 0.05% Tween 20, pH 7.4) containing

5% (w⁄ v) nonfat dried milk] The blots were then

incu-bated overnight at 4C with primary antibodies in blocking

buffer, washed with TBST, and incubated for 40 min at

room temperature with horseradish peroxidase-conjugated

secondary antibodies in blocking buffer Immunoreactive

bands were detected using Western Lightning Western

Blot Chemiluminescence Reagent Plus (Perkin-Elmer,

Bos-ton, MA, USA), following the manufacturer’s instructions,

and by autoradiography using Kodak BioMax MR films

(Eastman Kodak, Rochester, NY, USA)

Quantification of G-CSF and nitrite in culture

medium

The concentration of G-CSF in the culture medium of

RAW264.7 cells was measured by an ELISA using a mouse

G-CSF Quantikine ELISA kit (R&D Systems) according to

the manufacturer’s instructions The levels of NO in the

culture medium of RAW264.7 cells were measured using a

Griess reagent system kit, according to the manufacturer’s

instructions [25] The limit of detection for G-CSF and

nitrite was 5 pgÆmL)1and 2.5 lm, respectively

Reporter gene activity assay

To measure G-CSF promoter activity, 0.9 lg of the G-CSF promoter-luciferase reporter plasmid pG-CSF ()283 ⁄ +35)-Luc was mixed with 0.1 lg of phRLTK plasmid and the mixture was used to transiently transfect RAW264.7 cells,

as described above Photinus and Renilla luciferase activities

in the cell lysates were assayed using the Dual-Luciferase Reporter Assay System, as described previously [24], and the light intensity produced by Photinus luciferase (test plasmid) was normalized to that produced by Renilla lucif-erase (control plasmid)

ChIP assay

The ChIP assay was performed as described previously [24] Briefly, after various treatments, the cells were fixed with 1% (w⁄ w) formaldehyde for 10 min at 37 C to crosslink proteins to DNA, collected by scraping and then sonicated

on ice by pulsing eight times for 10 s at a power setting of 30% using a Microson Ultrasonic Cell Disrupter KL

A portion of the sample was kept as input material and the rest of the fragmented chromatin was immunoprecipitated with anti-Oct-2 IgG or control rabbit IgG, then the cross-links were reversed by incubation at 65C for 4 h, the pro-teins were digested with proteinase K for 1 h at 45C and the DNA was recovered by phenol⁄ chloroform extraction and ethanol precipitation and used as the template for PCR with the primers listed in Table 1 The input material was similarly subjected to de-crosslinking, DNA recovery and PCR analysis Thirty PCR cycles were used for all ChIP experiments and 28 PCR cycles were used for the input samples

Statistical analysis

The results are shown as the mean ± SD Differences between means were evaluated using the Student’s t-test and were considered significant at P < 0.01

Acknowledgements

This study was supported by the National Science Council of Taiwan grant NSC97-2320-B-002-057-MY3 and NSC99-2320-B-038-009-MY3

References

1 Raetz CR (1990) Biochemistry of endotoxins Annu Rev Biochem 59, 129–170

2 Beutler B & Rietschel ET (2003) Innate immune sensing and its roots: the story of endotoxin Nat Rev Immunol

3, 169–176

3 Sweet MJ & Hume DA (1996) Endotoxin signal trans-duction in macrophages J Leukoc Biol 60, 8–26