NF-κB independent signaling pathway is responsible for LPS-induced GDNF gene expression in primary rat glial cultures

(1)

Title

NF-κB independent signaling pathway is responsible for LPS-

induced GDNF gene expression in primary rat glial cultures( 本文

(Fulltext) )

Author(s)

TANAKA, Tatsuhide; OH-HASHI, Kentaro; SHITARA,

_{Hironobu; HIRATA, Yoko; KIUCHI, Kazutoshi}

Citation

[Neuroscience Letters] vol.[431] no.[3] p.[262]-[267]

Issue Date

2008-02-06

Rights

Elsevier Ireland Ltd

Version

著者最終稿 (author final version) postprint

URL

http://hdl.handle.net/20.500.12099/33327

(2)

NF-κκκκB independent signaling pathway is responsible for LPS-induced GDNF gene expression

in primary rat glial cultures

Tatsuhide Tanaka, Kentaro Oh-hashi, Hironobu Shitara, Yoko Hirata, Kazutoshi Kiuchi＊

Department of Biomolecular Science, Faculty of Engineering, Gifu University, 1-1 Yanagido, Gifu

501-1193, Japan

The number of text pages of the whole manuscript : 17

The number of figures : 4

＊_{Corresponding author. Tel: +81 58 293 2651; fax: +81 58 230 1893}

E-mail address: kiuchi@biomol.gifu-u.ac.jp (K. Kiuchi). * Manuscript

(3)

Abstract

Glial cell line-derived neurotrophic factor (GDNF), a distant member of the transforming growth

factor-β superfamily, was originally purified and cloned as a potent survival factor for midbrain

dopaminergic neurons. Some studies have characterized the transcriptional regulation of the GDNF

gene, but its regulatory mechanisms have yet to be well defined, especially under

pathophysiological conditions. In this study, we used a pharmacological approach to study the

expression of the rat GDNF gene induced by lipopolysaccharide (LPS) in primary cultures of glial

cells. MG132, a blocker of nuclear factor κB (NF-κB) activation, did not apparently affect

LPS-induced GDNF gene expression, whereas it attenuated the up-regulation of iNOS genes via

Toll-like receptor (TLR) 4. In primary glial cultures, LPS increased the phosphorylation levels of

c-Jun amino-terminal kinase 1 (JNK1) and p38 mitogen-activated protein kinase (MAPK); in

primary microglial cultures, it enhanced phosphorylation of extracellular signal-regulated kinase

(Erk). Of the several MAP kinase inhibitors tested, a JNK-specific inhibitor blocked LPS-induced

GDNF transcription in primary cultures of microglia, but not of astrocytes. These results suggest

that LPS up-regulates GDNF transcription through an NF-κB independent pathway, and that JNK is

responsible for LPS-stimulated GDNF transcription in primary cultures of microglia.

(4)

Glial cell line-derived neurotrophic factor (GDNF), a distant member of the transforming

growth factor-β superfamily, was originally purified and cloned as a potent survival factor for

midbrain dopaminergic neurons [1]. GDNF mRNA is widespread in the central and peripheral

nervous systems as well as outside the nervous system. GDNF acts as a morphogen in kidney

development and regulates the differentiation of spermatogonia [2]. Thus, GDNF is a

multifunctional molecule that regulates development and the differentiation of a variety of cells, and

that functions as a neurotrophic factor for specific groups of neurons. Moreover, GDNF is

up-regulated in glial cells and macrophages during pathophysiological conditions such as spinal

cord injury and cerebral ischemia [3-6].Some studies have characterized the mouse and human

GDNF genes [7-10], but to date, its transcriptional regulation remains poorly understood, especially

under pathophysiological conditions.

Lipopolysaccharide (LPS), a component of Gram-negative cell walls recognized by Toll-like

receptor (TLR) 4 in hosts, is an adjuvant for the adaptive immune response. The LPS-TLR4

interaction up-regulates co-stimulatory molecules on antigen-presenting cells [11, 12], and is

reported to be a potent inducer of GDNF transcription in astrocytes and microglia [13, 14]. Two

pathways have been reported for LPS-TLR4 signaling [15, 16]. One is dependent on both MyD88

and on Toll-interleukin-1 receptor (TIR) domain-containing adaptor protein. The other is dependent

on TIR domain-containing adaptor inducing interferon-β and its related adaptor molecule. The

former pathway activates nuclear factor κB (NF-κB) and MAPK cascades, while the latter activates

(5)

Since the putative binding site of NF-κB exists in intron 1 of the mouse GDNF gene [8], we

wondered whether NF-κB was responsible for induction of GDNF transcription induced by LPS in

primary cultures of glial cells. In fact, we found in the present study that LPS-induced GDNF

transcription in primary microglial cultures depends not on NF-κB, but on the c-Jun N-terminal

kinase (JNK) cascade.

Primary cultures of astroglial and microglial cells were obtained from Sprague-Dawley rats

on postnatal day 1 (P1) according to previously reported methods, with a slight modification [17,

18]. Briefly, rat cerebral cortex was digested with 0.125% trypsin for 15 min. After incubation,

0.004% DNase and 0.03% trypsin inhibitor were added to the dissociated tissue. Cells were passed

through a 40-µm nylon mesh. The resultant cell suspension was diluted with DMEM supplemented

with 10% fetal bovine serum and 50 µg/ml kanamycin, and seeded into poly-D-lysine-coated dishes.

Microglial cells from the astrocyte-monolayer sheet were removed by appropriate shaking. The

cells were treated with 1 µg/ml LPS for the indicated times. The concentrations of other agents used

in this study were as follows: 5 µM MG132 (a blocker of NF-κB activation), 5 µM actinomycin D

(a transcription inhibitor), 10 µg/ml cycloheximide (a translation inhibitor), 10 µM U0126 [a

mitogen-activated protein kinase (MAPK) kinase 1 (MKK-1) inhibitor], 30 µM SP600125 (a JNK

inhibitor), 10 µM SB202190 (a p38 MAPK inhibitor), 0.2 µM Gö6976 (a cPKC-specific inhibitor),

5 µM rottlerin (a PKCδ-specific inhibitor), 1 µM Ro-31-8220 (a pan-PKC inhibitor). Cells were

pretreated with each inhibitor for 30 min (MG132, PKC inhibitors) or 60 min (actinomycin D,

(6)

To estimate the mRNA levels of each gene by reverse transcription-polymerase chain reaction

(RT-PCR), total RNA was extracted from cultured cells with TRIzol (Invitrogen) and converted to

cDNA by reverse transcription using an oligo(dT)12-18 primer (Invitrogen) to prime superscript III

RNase-free reverse transcriptase (Invitrogen) according to the manufacturer’s instructions. The

intended DNA was amplified with 0.2 µM of its primer pairs and 0.5 µg of resultant cDNA. PCR

primer pairs used in this study were as follows: GDNF sense primer,

5’-CGGGACTCTAAGATGAAGTTATGGGATGTCGTG-3’; GDNF anti-sense primer,

5’-GGGTCAGATACATCCACACCGTTTAGCGGAATGC-3’. iNOS sense primer,

5’-TTGGTGTTTGGGTGCCGGC-3’; iNOS anti-sense primer,

5’-CCATAGGAAAAGACTGCACCGAAG-3’. β-actin sense primer,

5’-TGTATGCCTCTGGTCGTACC-3’; β-actin anti-sense primer,

5’-CAACGTCACACTTCATGATGG-3’. After 20-36 cycles of amplification, cDNAs were

separated by electrophoresis on 1.5% agarose gels and visualized using ethidium bromide. The

fluorescence intensity of each band was scanned and quantified using NIH-image software.

For Western blot analysis, cells were lysed with 10 mM Tris buffer, pH 7.4, containing 150

mM NaCl, 5 mM EDTA, 1% Triton X-100, 1% deoxycholic acid, 0.1% SDS and 1 mM sodium

vanadate. The protein concentration was determined by DC Protein Assay (Bio-Rad). Equal

amounts of cell lysates were subjected to SDS-PAGE, transferred to nitrocellulose membranes

(Amersham) and identified by an enhanced chemiluminescence kit (Amersham) using antibodies

(7)

p38, and phospo-p38 (Cell Signaling).

Primary cultures of glial cells from P1 rats were treated with 1 µg/ml LPS for up to 12 h,

and then levels of GDNF and iNOS transcripts were determined. The induction of iNOS

transcription by LPS served as a positive control. Transcriptional profiles of GDNF and iNOS were

similar, with both genes strongly induced in astrocytes 3 h after LPS stimulation. Elevated levels of

GDNF and iNOS transcripts persisted for at least 12 h (Fig. 1A). In microglia, GDNF transcription

was induced within 1 h, and it decreased slightly at 12 h after LPS stimulation (Fig. 1B). As shown

in Fig. 1C and D, pretreatment with actinomycin D, a transcription inhibitor, blocked the

LPS-induced increase in GDNF mRNA, suggesting that this induction is due to increased

transcription, not mRNA stabilization. Cycloheximide, a translation inhibitor, enhanced the

background level of GDNF transcription, but it did not significantly affect the increased

transcription induced by LPS stimulation (Fig. 1C). Similar results were observed in the case of

primary microglial cultures (Fig. 1D).

After establishing that LPS up-regulates GDNF transcription in these neuronal cultures, we

examined whether NF-κB was responsible. MG132, a blocker of NF-κB activation, did not affect

the up-regulation of GDNF transcription induced by LPS, whereas it attenuated this up-regulation in

the case of the iNOS gene (Fig. 2A and B). Similar phenomena were observed using a distinctive

NF-κB inhibitor, BAY 11-7082 (data not shown). That LPS-induced iNOS transcription in primary

glial cells requires activation of NF-κB is consistent with previous reports [19, 20]. These results

(8)

glial cultures.

MG132 was found to enhance GDNF transcription in the absence of LPS. We sought to

determine whether this might reflect a connection between GDNF transcription and NF-κB activity.

We found by Western blotting that the nuclear accumulation of p65, one of the NF-κB subunits, was

not increased in the presence of MG132 (data not shown). Given recent reports that MG132 induces

COX-2 and IL-8 transcription via MAPK cascades [21, 22], we tested the effect of a MAPK

inhibitor on MG132-induced GDNF transcription. We found that a p38 MAPK inhibitor blocked

MG132’s ability to stimulate GDNF transcription in primary astrocytes (data not shown). As a final

test for a link between LPS-induced GDNF transcription and NF-κB, we took advantage of our

previous studies identifying a putative NF-κB binding site in intron 1 of GDNF [8]. Primary

cultures of astrocytes were transfected with pLG3-luciferase GDNF promoter 1 construct including

this intron as an enhancer. The luciferase activity did not change by the stimulation of LPS, despite

the presence of the presumably functional NF-κB binding site (data not shown). These results

support the notion that LPS stimulates GDNF transcription in primary glial cultures through a

pathway or pathways independent of NF-κB.

To investigate which of the MAPKs is activated by LPS treatment in primary cultures of astrocytes

and microglia, the phosphorylation levels of p44/42 MAPK (Erk1/2), JNK1, and p38 MAPK were

estimated by Western blot analysis using antibodies against the nonphosphorylated and

phosphorylated proteins. We found that JNK1 and p38 MAPK phosphorylation in astrocytes was

(9)

phospho-JNK1 and phospho-p38 MAPK increased 30 min after LPS treatment (Fig. 3B). On the

contrary, p44/42 MAPK phosphorylation in microglia was activated after 5 min of exposure to LPS,

whereas in astrocytes, the amount of phospho-p44/42 MAPK was unchanged within 60 min after

LPS treatment. We next examined whether MAPK cascades contribute to LPS-induced GDNF

transcription in primary cultures of astrocytes and microglia. Specific pharmacological MAPK

inhibitors (U0126 for MEK1/2, SP600125 for JNK, and SB202190 for p38 MAPK) were employed

to address this question. We found that none of these inhibitors influenced LPS-induced GDNF

transcription in astrocytes (Fig. 4A), whereas the JNK inhibitor significantly attenuated this

induction in microglia (Fig. 4B). Western blot analysis indicated that SP600125 (30 µM) did indeed

block the LPS-induced phosphorylation of c-Jun in astrocytes (data not shown). In contrast, none of

these inhibitors affected LPS-induced iNOS transcription in either astrocytes or microglia (Fig. 4C

and D). Finally, we sought to test whether PKC signaling pathways might be involved in

LPS-induced GDNF transcription. We found that Gö6976 (a cPKC inhibitor), rottlerin (a PKCδ

inhibitor), and Ro-31-8220 (a pan-specific PKC inhibitor) all failed to attenuate LPS-induced

GDNF transcription in primary glial cultures (data not shown). These results indicate that

LPS-induced GDNF transcription in primary cultures of rat microglia may be related to JNK

cascades, but not to NF-κB pathways. These results with GDNF contrast with LPS-induced iNOS

transcription, which is known to depend on NF-κB activation in primary glial cultures [19, 20].

LPS can be recognized by TLR-4, which is a type I transmembrane protein whose

(10)

cytoplasmic adaptor molecule, MyD88, through a homophilic interaction between their TIR

domains. MyD88 also possesses a death domain, which mediates an association with interleukin-1

receptor-associated kinase (IRAK). Subsequently, IRAK phosphorylates transforming growth factor

receptor-associated factor 6 (TRAF-6), which in turn activates the IKK complex as well as MKK3/6

and MKK7 [24]. In other words, LPS is able to activate NF-κB, as well as p38 MAPK and JNK.

Our results suggest that the JNK cascade, most likely activated by IRAK and/or TRAF-6, plays an

important role in LPS-induced GDNF transcription in primary cultured microglia, but not astrocytes.

It is interesting that these two cell types may use different pathways for LPS-induced GDNF

transcription. The signaling mechanisms whereby LPS stimulates GDNF transcription may vary

from cell to cell, and a clearer understanding of this will require more detailed study of the precise

mechanisms involved. Further investigation into the cause of GDNF induction under

pathophysiological conditions may provide new insights into the gene’s crucial roles in

neurodegenerative diseases.

Acknowledgments: This study was supported by the research grant from the Ministry of Education,

(11)

References

[1] L.F. Lin, D.H. Doherty, J.D. Lile, S. Bektesh, F. Collins, GDNF: a glial cell line-derived

neurotrophic factor for midbrain dopaminergic neurons, Science 260 (1993) 1130-1132.

[2] M.S. Airaksinen, M. Saarma, The GDNF family: signaling, biological functions and therapeutic

value, Nat. Rev. Neurosci. 3 (2002) 383-394.

[3] K. Satake, Y. Matsuyama, M. Kamiya, H. Kawakami, H. Iwata, K. Adachi, K. Kiuchi,

Up-regulation of glial cell line-derived neurotrophic factor (GDNF) following traumatic spinal

cord injury, Neuroreport 11 (2000) 3877-3881.

[4] T. Ikeda, H. Koo, Y.X. Xia, T. Ikenoue, B.H. Choi, Bimodal upregulation of glial cell

line-derived neurotrophic factor (GDNF) in the neonatal rat brain following ischemic/hypoxic

injury, Int. J. Dev. Neurosci. 20 (2002) 555-562.

[5] K. Yamagata, M. Tagami, K. Ikeda, S. Tsumagari, Y. Yamori, Y. Nara, Differential regulation of

glial cell line-derived neurotrophic factor (GDNF) mRNA expression during hypoxia and

reoxygenation in astrocytes isolated from stroke-prone spontaneously hypertensive rats, Glia 37

(12)

[6] M. Hashimoto, A. Nitta, H. Fukumitsu, H. Nomoto, L. Shen, S. Furukawa, Involvement of glial

cell line-derived neurotrophic factor in activation processes of rodent macrophages, J. Neurosci.

Res. 79 (2005) 476-487.

[7] N. Matsushita, Y. Fujita, M. Tanaka, T. Nagatsu, K. Kiuchi, Cloning and structural organization

of the gene encoding the mouse glial cell line-derived neurotrophic factor, GDNF, Gene 203

(1997) 149-157.

[8] M. Tanaka, S. Ito, K. Kiuchi, Novel alternative promoters of mouse glial cell line-derived

neurotrophic factor gene, Biochem. Biophys. Acta 1494 (2000) 63-74.

[9] L. Grim, E. Holinski-Feder, J. Teodoridis, B. Scheffer, D. Schindelhauer, T. Meitinger, M.

Ueffing, Analysis of the human GDNF gene reveals an inducible promoter, three exons, triplet

repeat within the 3’-UTR and alternative splice products, Hum. Mol. Genet. 7 (1998) 1873-1886.

[10] P.A. Baecker, W.H. Lee, A.N. Verity, R.M. Eglen, R.M. Johnson, Characterization of a

promoter for the human glial cell line-derived neurotrophic factor gene, Mol. Brain Res. 69

(13)

[11] M. Yamamoto, S. Sato, H. Hemmi, K, Hoshino, T. Kaisho, H. Sanjo, O. Takeuchi, M.

Sugiyama, M. Okabe, K. Takeda, S. Akira, Role of adaptor TRIF in the MyD88-independent

toll-like receptor signaling pathway, Science 301 (2003) 640-643.

[12] K. Hoebe, X. Du, P. Georgel, E. Janssen, K. Tabeta, S.O. Kim, J. Goode, P. Lin, N. Mann, S.

Mudd, K. Crozat, S. Sovath, J. Han, B. Beutler, Identification of Lps2 as a key transducer of

MyD88-independent TIR signaling, Nature 424 (2003) 743-748.

[13] R. Kuno, Y. Yoshida, A. Nitta, T. Nabeshima, J. Wang, Y. Sonobe, J. Kawanokuchi, H.

Takeuchi, T. Mizuno, A. Suzumura, The role of TNF-α and its receptors in the production of

NGF and GDNF by astrocytes, Brain Res. 1116 (2006) 12-18.

[14] M. Hashimoto, A. Nitta, H. Fukumitsu, H. Nomoto, L. Shen, S. Furukawa,

Inflammation-induced GDNF improves locomotor function after spinal cord injury,

Neuroreport 16 (2005) 99-102.

[15] K. Miyake, Innate recognition of lipopolysaccharide by Toll-like receptor 4-MD-2, Trends

Microbiol. 12 (2004) 186-192.

(14)

783-801.

[17] J. Hu, C. Francis, L. José, Guevara, L.J. Van Eldik, S100β stimulates inducible nitric oxide

synthase activity and mRNA levels in rat cortical astrocytes, J. Biol. Chem. 271 (1996)

2543–2547.

[18] K.H. Lee, S.J. Yun, K.N. Nam, Y.S. Gho, E.H. Lee, Activation of microglial cells by

ceruloplasmin, Brain Res. 1171 (2007) 1-8.

[19] S.J. Suh, T.W. Chung, M.J. Son, S.H. Kim, T.C. Moon, K.H. Son, H.P. Kim, H.W. Chang, C.H.

Kim, The naturally occurring biflavonoid, ochnaflavone, inhibits LPS-induced iNOS expression,

which is mediated by ERK1/2 via NF-κB regulation, in RAW264.7 cells, Arch. Biochem.

Biophys. 447 (2006) 136-146.

[20] N.S. Chandel, W.C. Trzyna, D.S. McClintock, P.T. Schumacker, Role of oxidants in NF-κB

activation and TNF-α gene transcription induced by hypoxia and endotoxin, J. Immunol. 165

(2000) 1013-1021.

[21] K.J. Woo, J.W. Park, T.K. Kwon, Proteasome inhibitor-induced cyclooxygenase-2 expression

(15)

Biophys. Res. Commun. 342 (2006) 1334-1340.

[22] A. Gerber, A. Heimburg, A. Reisenauer, A. Wille, T. Welte, F. Buhling, Proteasome inhibitors

modulate chemokine production in lung epithelial and monocytic cells, Eur. Respir. J. 24 (2004)

40-48.

[23] L.A. O’Neill, C.A. Dinarello, The IL-1 receptor/toll-like receptor superfamily: crucial

receptors for inflammation and host defense, Immunol. Today 21 (2000) 206-209.

[24] S. Akira, K. Takeda, T. Kaisho, Toll-like receptors: critical proteins linking innate and acquired

(16)

Figure legends

Fig. 1. LPS-stimulated transcription of GDNF in primary cultures of rat astrocytes and microglia.

Astrocytes (A) or microglia (B) from P1 rats were treated with 1 µg/ml LPS for up to 12 h. After

preparation of total RNA from these cells, RT-PCR was performed as described in section 2.

LPS-induced GDNF transcription levels were determined in the presence of the transcription

inhibitor actinomycin D (5 µM; Act-D), or the translation inhibitor cycloheximide (10 µg/ml; CHX).

Astrocytes (C) or microglia (D) from P1 rats were pre-incubated with each inhibitor for 1 h,

followed by 1 µg/ml LPS for 3 h. After preparation of total RNA from the cells, RT-PCR was

performed as described in section 2. The fluorescence intensity of each band was scanned and

quantified using NIH-image software. Levels of GDNF mRNA are normalized against those of

β-actin mRNA, respectively. Data are obtained from two independent experiments.

Fig. 2. Effects of MG132 on the stimulation of GDNF and iNOS transcription by LPS. Astrocytes

(A, C) or microglia (B, D) from P1 rats were incubated with 1 µg/ml LPS or 5 µM MG132. After

preparation of total RNA from the cells, RT-PCR was performed as described in section 2. The

fluorescence intensity of each band was scanned and quantified using NIH-image software. Levels

of GDNF (A, B) and iNOS (C, D) mRNA are normalized against those of β-actin mRNA,

respectively, and the relative mRNA levels of GDNF and iNOS are expressed as ratio of that of

(17)

mean ± SEM. * p<0.05. N.S., not significantly different.

Fig. 3. Western blot analysis of MAPK phosphorylation. Astrocytes (A) or microglia (B) were

incubated with 1 µg/ml LPS for the indicated times. Cell lysates were analyzed by immunoblot

using antibodies against phosphorylated and non-phosphorylated p44/42 MAPK, JNK1 and p38

MAPK. The intensity of each band was scanned and quantified using NIH-image software. Levels

of phosphorylated MAPKs (p44/42 MAPK, JNK1 and p38 MAPK) are normalized against those of

non-phosphorylated MAPKs, respectively. In the case of p44/42 MAPK, open column represents

the intensity of normalized phospho-p42 while close column does that of normalized phospho-p44.

Arrowhead indicates the band of phospho-JNK1. This experiment was repeated three times and