JBC Ideal method for primary cell transfection

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M407923200 on August 9, 2004

J. Biol. Chem., Vol. 279, Issue 41, 42774-42786, October 8, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/41/42774    most recent
M407923200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Nakajima, H.
Right arrow Articles by Hoshiko, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nakajima, H.
Right arrow Articles by Hoshiko, S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Critical Role of the Automodification of Poly(ADP-ribose) Polymerase-1 in Nuclear Factor-{kappa}B-dependent Gene Expression in Primary Cultured Mouse Glial Cells*

Hidemitsu Nakajima{ddagger}, Hiroshi Nagaso, Nobukazu Kakui, Midori Ishikawa, Toyokazu Hiranuma, and Shigeru Hoshiko

From the Pharmaceutical Research Center, Meiji Seika Kaisha Limited, 760 Moro-oka-cho, Kohoku-ku, Yokohama 222-8567, Japan

Received for publication, July 14, 2004


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Synthesis of ADP-ribose polymers catalyzed by poly-(ADP-ribose) polymerase-1 (PARP-1) has been implicated in transcriptional regulation. Recent studies with PARP-1 null mice and PARP-1 inhibitors have also demonstrated that PARP-1 has an essential role in nuclear factor-{kappa}B (NF-{kappa}B)-dependent gene expression induced by various inflammatory stimuli. In this study, we used primary cultured mouse glial cells to investigate the role of poly(ADP-ribosyl)ation by PARP-1 in NF-{kappa}B-dependent gene expression. PARP-1 inhibitors and the antisense RNA for PARP-1 mRNA suppressed lipopolysaccharide (LPS)-induced expression of tumor necrosis factor-{alpha} and inducible nitric-oxide synthase, suggesting that PARP-1 activity has a critical role in synthesis. Western blotting with anti-poly(ADP-ribose) antibody revealed that PARP-1 itself was mainly poly(ADP-ribosyl)ated in glial cells, i.e. automodified PARP-1 (AM-PARP). The amounts of AM-PARP were not affected by LPS treatment, but were decreased by PARP-1 inhibitors. Electrophoretic mobility shift assay revealed that PARP-1 inhibitors and the antisense RNA for PARP-1 mRNA reduced the LPS-induced DNA binding of NF-{kappa}B. Non-modified PARP-1 also reduced the DNA binding of NF-{kappa}B via its physical association with NF-{kappa}B, whereas AM-PARP had no effect. On the other hand, enhancement of the automodification of PARP-1 by the addition of NAD+, its substrate, promoted the DNA binding of NF-{kappa}B. Furthermore, in in vitro transcription assay, the addition of AM-PARP or NAD+ to nuclear extracts promoted NF-{kappa}B p50-dependent transcription. These results indicate that automodification of PARP-1 positively up-regulates formation of the NF-{kappa}B·DNA complex and enhances transcriptional activation. Therefore, AM-PARP may be critical for the NF-{kappa}B-dependent gene expression of some inflammatory mediators in glial cells.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Poly(ADP-ribose) polymerase-1 (PARP-11; EC 2.4.2.30 [EC] ) is an abundant nuclear protein that is activated by DNA strand breakage and that catalyzes the covalent attachment of poly-(ADP-ribose) (PAR) from NAD+ to numerous nuclear proteins and transcription factors, including histones; DNA polymerase {alpha} and {beta}; p53; and PARP-1, itself being the major target, via its automodification domain (1, 2). Besides PARP-1, another six PARPs have been identified: short PARP, PARP-2, PARP-3, tankylase-1/2, and vault PARP (2, 3). However, the physiological roles of poly(ADP-ribosyl)ation of nuclear proteins and transcription factors induced by PARPs are not completely understood. The initially identified subtype of the enzyme, PARP-1, has been thought to play a central role in the process of poly(ADP-ribosyl)ation because poly(ADP-ribosyl)ation is markedly reduced in most tissues of PARP-1 null mice (4). Transient poly(ADP-ribosyl)ation by PARP-1 can be induced by a wide variety of environmental stimuli, including reactive oxygen, ionizing radiation, and genotoxic stress (1, 2). Thus, PARP-1 has been suggested to regulate DNA repair (5). On the other hand, overactivation of PARP-1 by massively damaged DNA consumes NAD+ and consequently ATP, resulting in necrotic cell death by energy failure (3, 6).

There are many reports suggesting that PARP-1 is also involved in regulation of gene expression at the transcriptional step (2, 3). PARP-1 seems to play dual roles in transcription. Poly(ADP-ribosyl)ation of transcription factors such as Yin-Yang 1 (7), RNA polymerase II-associated factors (8), and p53 (9) results in reversible silencing of transcription by impairing the DNA binding of these proteins. In other instances, PARP-1 was found to have only one function, stimulating the DNA binding activity of transcription factors such as Oct-1 (10) and B-Myb (11). Recent reports have also shown that PARP-1 is required for specific nuclear factor-{kappa}B (NF-{kappa}B)-dependent gene expression and acts as a coactivator for NF-{kappa}B in vitro (13, 14). Indeed, the NF-{kappa}B-dependent transcription of some inflammatory mediators in response to endotoxin (13) or pro-inflammatory cytokines such as tumor necrosis factor-{alpha} (TNF-{alpha}) and interleukin-1{beta} (IL-1{beta}) (12–14) is almost completely abrogated in PARP-1 null mice. Thus, anti-inflammatory effects of PARP-1 inhibitors have been extensively discussed in relation to various inflammation-related diseases (15, 16). However, the exact biochemical mechanism by which PARP-1 regulates NF-{kappa}B-dependent transcription is obscure. To date, some groups have reported that the enzyme activity of PARP-1 might directly influence NF-{kappa}B-dependent transcription. Kameoka et al. (17) showed that poly(ADP-ribosyl)ation markedly suppresses the DNA binding activity of NF-{kappa}B via direct modification in vitro. Chang and Alvarez-Gonzalez (18) demonstrated that the DNA binding activity of NF-{kappa}B p50 is NAD+-dependent and reversibly regulated by the automodification of PARP-1 under cell-free conditions. In contrast, Hassa et al. (14) concluded that neither the enzyme activity nor the DNA binding activity of PARP-1 is required for NF-{kappa}B-dependent transcription. Thus, there are contradictory results (15).

To further investigate the role of poly(ADP-ribosyl)ation by PARP-1 in NF-{kappa}B-dependent transcription in inflammation, we examined the effects of several PARP-1 inhibitors and the antisense RNA for PARP-1 mRNA on lipopolysaccharide (LPS)-induced expression of some inflammatory mediators in primary cultured mouse glial cells. Cultured glial cells, composed of macroglia (astrocytes and oligodendrocytes) and microglia, were found to be a good in vitro model for neuro-inflammatory diseases such as stroke (19) and Parkinson's disease (20), which are implicated in the activation of PARP-1. Although these previous reports have focused on the activation of neuronal PARP-1, the pathophysiological significance of poly(ADP-ribosyl)ation catalyzed by PARP-1 in glial cells remains unclear. Here, we report that poly(ADP-ribosyl)ated PARP-1 itself, i.e. automodified PARP-1, has a critical role in NF-{kappa}B-dependent transcription and gene expression in glial cells.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Isotopes—[adenine-2,8-3H]NAD+, [adenylate-32P]NAD+, [{alpha}-32P]GTP, and [{gamma}-32P]ATP were purchased from PerkinElmer Life Sciences.

Antibodies—The following antibodies were obtained from the indicated commercial sources: mouse anti-GAPDH monoclonal antibody (mAb) (MAB374, Chemicon International, Inc., Temecula, CA); mouse anti-PARP-1 mAb (SA-250) and rabbit anti-PAR polyclonal antibody (pAb) (SA-276) (BIOMOL Research Labs Inc., Plymouth Meeting, PA); rabbit anti-PARP-1 pAb (Roche Applied Science, Mannheim, Germany); mouse anti-PAR mAb (4335-MC) and guinea pig anti-PAR pAb (4336-PC) (Trevigen, Gaithersburg, MD); mouse anti-PAR mAb (10H, Alexis Biochemicals, Lausen, Switzerland); mouse anti-NF-{kappa}B p50 mAb (sc-8414), rabbit anti-p50 pAb (sc-7178), and rabbit anti-NF-{kappa}B p65 pAb (sc-109) (Santa Cruz Biotechnology, Santa Cruz, CA); and mouse anti-inducible nitric-oxide synthetase (iNOS) mAb (Transduction Laboratories, Lexington, KY).

Other Materials—The following reagents were obtained from the indicated commercial sources: benzamide and nicotinamide (Nacalai Tesque, Kyoto, Japan); 3-aminobenzamide (Sigma); 1,5-dihydroxyisoquinoline (RBI, Natick, MA); pcDNA3, Dulbecco's modified Eagle's medium, and SuperScript II RNase H– reverse transcriptase (Invitrogen); cytokine enzyme-linked immunosorbent assay kits, protein A and G-Sepharose, and the enhanced chemiluminescence ECL detection system (Amersham Biosciences); the gel shift assay kit, the HeLaScribe® nuclear extract in vitro transcription system, the {phi}X174 DNA/HinfI dephosphorylation marker, and the human recombinant p50 subunit of NF-{kappa}B (Promega, Madison, WI); LPS (Escherichia coli 0111:B4) (Difco); and Effectene transfection reagent and the RNeasy mini-kit (QIAGEN Japan, Tokyo, Japan). All other materials we used were of analytical grade.

Cell Culture and Transfection—Primary mixed glial cultures were prepared from neonatal BALB/c mice (1–2 days old) as described previously (21) with some modifications. Cells harvested by centrifugation at 800 x g for 5 min were resuspended; grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (JRH Biosciences, Lenexa, KS), 100 units/ml penicillin, and 100 µg/ml streptomycin; and seeded in 6- or 24-well culture plates. The cells were incubated at 37 °C in a humidified atmosphere of 5% CO2. The medium was changed twice each week. The cultures were used at 14–18 days after plating. Immunohistochemical analysis indicated that the cultures consisted of 72% astrocytes and 22% microglia (data not shown). For transfection of the antisense RNA expression vector for PARP-1 mRNA, microglial cells were mechanically isolated from the primary mixed glial culture as described previously (22) and then seeded in a 24-well plate at 1–2 x 105 cells/well. The purity of the cultures was >95% as estimated by morphological criteria and by their reactivity toward MRC-OX42 (CD11b, a maker of microglia). A 3.0-kb full-length mouse PARP-1 cDNA was cloned in an antisense orientation in the expression vector pcDNA3 as described previously (23). The resulting antisense or mock vector (pcDNA3) was transfected into microglia (1 h after plating) using the Effectene transfection reagent for 24 h according to the manufacturer's protocol. The efficacy of transfection was assessed by PARP catalytic activity (24). Treatment of the cells with the antisense vector down-regulated the resulting PARP activity to 62% versus the control; efficacy was calculated as 38% of the total cells (n = 4) (data not shown).

Assessment of Cell Viability—The viability of glial cells was assessed by Alamar BlueTM (BIOSOURCE) according to the manufacturer's protocol.

Measurement of Cytokines and Nitrite—The culture supernatants were collected at the indicated times, and the levels of TNF-{alpha}, IL-1{beta}, and IL-6 were measured using the respective mouse enzyme-linked immunosorbent assay kits according to the manufacturer's protocol. The detection limit for assays was 50 pg/ml. The nitrite concentration was measured with a standard Griess reaction adapted to microplates as described previously (25). The sensitivity of this assay was ~0.5 µM.

Western Blotting—For sample preparation, the attached cells were washed twice with ice-cold phosphate-buffered saline, solubilized in lysis buffer (50 mM Tris-HCl, pH 7.2, 1 mM EDTA, 150 mM NaCl, 0.1% SDS, 0.5% sodium deoxycholate, 1% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml aprotinin, and 10 µg/ml leupeptin), and sonicated on ice for 10 s. After centrifugation at 15,000 x g for 10 min at 4 °C, the cleared supernatants were obtained; an equal volume of SDS sample buffer (0.25 M Tris-HCl, pH 6.8, 2% SDS, 10% 2-mercaptoethanol, 30% glycerol, and 0.01% bromphenol blue) was added; and the sample was heated at 100 °C for 5 min. Protein concentrations of the samples were determined by the Bradford assay (Bio-Rad). Protein samples were separated by 5–20% SDS-PAGE (DRC, Tokyo) and transferred to Immobilon P membranes (Millipore, Tokyo). The membranes were incubated for 1 h with 5% bovine serum albumin in phosphate-buffered saline containing 0.05% Tween 20 and 0.02% NaN3 to block nonspecific binding and then incubated overnight at 4 °C with anti-PAR pAb (1:500), anti-PAR mAb (1:1000), anti-PARP-1 mAb (1:5000), or anti-iNOS mAb (1:5000). Subsequently, the membranes were incubated with an affinity-purified peroxidase-conjugated secondary antibody (Zymed Laboratories Inc.). PAR-bound proteins, PARP-1, and iNOS were detected using the ECL detection system according to the manufacturer's protocol. The membranes were also reprobed with anti-GAPDH mAb (1:500). These band intensities were quantified with NIH Image Version 1.61.

Measurement of Cellular PARP-1 Activity—Measurement of cellular PARP-1 activity was performed as described previously (26). Briefly, cells were washed twice with ice-cold phosphate-buffered saline; resuspended in 0.5 ml of assay buffer containing 56 mM Hepes-KOH, pH 7.5, 28 mM KCl, 28 mM NaCl, 2 mM MgCl2, 30 µM digitonin, 125 µM NAD+, and 18 kBq/ml [adenine-2,8-3H]; and incubated for 10 min at 37 °C. After incubation, ice-cold 20% trichloroacetic acid was added, and the samples were further incubated for 30 min at 4 °C. The samples were then washed twice with ice-cold 10% trichloroacetic acid and solubilized overnight in 2% SDS and 0.1 N NaOH at 37 °C. The radioactivity of each sample was measured in a liquid scintillation counter.

Reverse Transcription-PCR—Total RNA from glial cells was prepared using RNeasy. cDNA was synthesized by SuperScript II RNase H– reverse transcriptase according to the manufacturer's protocol. The sequences of the oligonucleotide primers used as a control for RNA isolation and reverse-transcription were as follows: for mouse TNF-{alpha} (354 bp), 5'-TTCTGTCTACTGAACTTCGGGGTAATCGGTCC-3' (upstream) and 5'-GTATGAGATAGCAAATCGGCTGACGGTGTGGG-3' (downstream); and for mouse GAPDH (558 bp), 5'-TGCTGAGTATGTCGTGGAGTCT-3' (upstream) and 5'-AATGGGAGTTGCTGTTGAAGTC-3' (downstream). One PCR cycle was run under the following conditions: DNA denaturation at 94 °C for 30 s, primer annealing at 60 °C for 1 min, and DNA extension at 72 °C for 1 min (28 cycles for TNF-{alpha} and 22 cycles for GAPDH). Semiquantification of mRNA levels of TNF-{alpha} and 18 S ribosomal RNA (used as a control) was carried out using a TaqMan cytokine gene expression plate and ABI PRISM 7700 (PerkinElmer Life Sciences) according to the manufacturer's protocol.

Preparation of Automodified PARP-1 (AM-PARP) and Non-modified PARP-1 (NM-PARP)—Construction of recombinant baculovirus containing cDNA of human His-tagged PARP-1, its expression in insect cells, and protein purification are described elsewhere (27). AM-PARP was prepared by the following procedure. Purified His-PARP-1 was incubated in buffer containing 100 mM Tris-HCl, pH 8.0, 10 mM MgCl2, 1mM dithiothreitol, 1 µM NAD+, and 20 µg/ml synthetic octameric DNA (5'-GGAATTCC-3') at 37 °C. After 5 min, the reactions were terminated by passing the mixture through a spin column (prepared with a 1-ml cylinder filled with Sephadex G-25 and equilibrated with 10 mM Tris-HCl, pH 7.2, and 1 mM EDTA) by centrifugation at 800 x g for 3 min. The concentration of the resulting proteins was measured, and the proteins were stored at –30 °C until used. NM-PARP was obtained by the same protocol as used for AM-PARP, except for incubation with 1 µM NAD+.

Preparation of Nuclear Extracts—Nuclear extracts from unstimulated or LPS-stimulated glial cells (1–5 x 105 cells) were prepared by the method of Lukasiuk et al. (28) with slight modifications. Cells were washed twice with ice-cold phosphate-buffered saline; lysed in 400 µlof buffer containing 10 mM Hepes-KOH, pH 7.8, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 1 mM sodium vanadate, 0.5 mM phenylmethylsulfonyl fluoride, 2.5 µg/ml aprotinin, 1 µg/ml pepstatin A, and 0.5 µg/ml leupeptin containing 0.1% Nonidet P-40 for 10 min on ice; vortexed vigorously for 15 s; and centrifuged at 12,000 x g for 3 min. The pelleted nuclei were resuspended in 100 µl of buffer containing 20 mM Hepes-KOH, pH 7.8, 0.4 M KCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 1 mM sodium vanadate, 0.5 mM phenylmethylsulfonyl fluoride, 2.5 µg/ml aprotinin, 1 µg/ml pepstatin A, and 0.5 µg/ml leupeptin; mixed vigorously for 15 min at 4 °C; and centrifuged at 15,000 x g for 30 min. Supernatants containing the nuclear proteins were stored at –80 °C.

Electrophoretic Mobility Shift Assay (EMSA)—The nuclear extracts (10 µg) were preincubated in a total volume of 19 µl of gel shift binding buffer (10 mM Tris-HCl, pH 7.5, 1 mM MgCl2, 0.5 mM dithiothreitol, 0.5 mM EDTA, 50 mM NaCl, 5% glycerol, and 0.05 mg/ml poly(dI-dC)) for 10 min on ice. The sample was mixed with 1 µlof 32P-labeled NF-{kappa}B probe (5'-AGTTGAGGGGACTTTCCCAGGC-3'), which was end-labeled with [{gamma}-32P]ATP using T4 DNA polynucleotide kinase; and the mixtures were then further incubated for 20 min at room temperature. For EMSA with purified proteins, purified recombinant NF-{kappa}B p50 (350 ng) was preincubated with purified His-PARP-1 (20, 100, or 200 ng; non-modified or automodified) for 20 min at room temperature in buffer containing 20 mM Tris-HCl, pH 8.0, 60 mM KCl, 5 mM MgCl2, 1 mM dithiothreitol, 0.05% Nonidet P-40, 10% glycerol, and 50 µg/ml bovine serum albumin (18) and mixed with the 32P-labeled probe, and the mixtures were further incubated for 30 min at room temperature. Additionally, LPS-stimulated nuclear extract (10 µg) and purified His-PARP-1 (200 ng; non-modified or automodified) were incubated for 20 min at room temperature in gel shift binding buffer and mixed with the 32P-labeled probe, and the mixtures were then further incubated for 20 min at room temperature. After the reactions, the protein·DNA complexes were separated on a 6% nondenaturing polyacrylamide gel (DRC) using the gel shift assay kit according to the manufacturer's protocol. In the supershift assay, 2 µg of anti-p65 pAb were added, and the samples were incubated for 30 min at 4 °C before the addition of 32P-labeled probes. The competition experiments were performed by adding a 100-fold molar excess of unlabeled probes with the 32P-labeled probes.

Poly(ADP-ribosyl)ation in Nuclear Extracts—Assays were performed as described by Chang and Alvarez-Gonzalez (18). Each 10-µg sample in assay buffer containing 100 mM Tris-HCl, pH 8.0, 20 mM MgCl2, and 1 mM dithiothreitol was preincubated with 10 mM 3-aminobenzamide for 5 min at 37 °C before the addition of 10 µM NAD+ to start a reaction. Control assays were performed in the same manner, except that no 3-aminobenzamide was added. After 10 min, an equal volume of 2x gel shift binding buffer was added to each mixture, which was then incubated for 20 min at room temperature before EMSA. Alternatively, the poly(ADP-ribosyl)ated proteins were resolved by Western blotting with rabbit anti-PAR pAb as described above.

Co-immunoprecipitation Assay—In a cell-free system, the procedure reported previously (18) was used. NM- or AM-PARP (200 ng) and the p50 subunit (350 ng) were preincubated for 30 min at 4 °C in buffer containing 10 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 0.1% Nonidet P-40. Anti-p50 pAb (2 µg) was added, and the mixture was incubated for 60 min at 4 °C. Protein G-Sepharose beads (1:1 slurry) were added next, and the samples were further incubated for 60 min at 4 °C with rocking. After centrifugation at 5000 x g for 1 min, both the first supernatants and the pellet of beads were collected. The beads were further washed five times, and then the bead-bound proteins were eluted by adding SDS sample buffer and heating at 100 °C for 5 min. The eluted proteins and the supernatants were submitted to Western blotting with anti-p50 mAb, anti-PARP-1 mAb, or anti-PAR mAb (10H) as described above. To assess the association of PARP-1 with the p50 subunit in vivo, glial nuclear extracts (1 mg) prepared as described above were initially incubated with both rabbit control IgG (5 µg) and protein G-Sepharose beads for 30 min at 4 °C with rocking to block nonspecific binding. After centrifugation, the supernatants were obtained, and the following procedure, which was similar to the procedure used in the cell-free system, was performed.

Immunodepletion Assay—To remove AM-PARP in glial cells by immunodepletion (29), total cell extracts (100 µg/50 µl) or LPS-treated nuclear extracts (50 µg/25 µl) were preincubated for 60 min at 4 °C with one of the following antibodies (2 µg/assay): rabbit control IgG, anti-PARP-1 pAb, or rabbit anti-PAR pAb. Subsequently, protein A-Sepharose beads (1:1 slurry) were added, and the samples were further incubated for 45 min at 4 °C with rocking. After centrifugation at 5000 x g for 1 min, the supernatants were obtained, and free proteins were submitted to Western blotting with anti-PARP-1 mAb or rabbit anti-PAR pAb as described above.

In Vitro Transcription Assay (Runoff Assay)—In vitro transcription reactions were performed as described by Dignam et al. (30) following the HeLaScribe® nuclear extract in vitro transcription system protocol provided by the manufacturer. Briefly, 25-µl reaction mixtures were prepared by combining HeLa nuclear extract (8 units/reaction) with ATP, UTP, and CTP (0.4 mM each) and GTP (0.012 mM) in buffer containing 3 mM MgCl2. [{alpha}-32P]GTP (3000 Ci/mmol, 10 mCi/ml) was used to label runoff transcripts. The standard reactions were performed in the presence of 50 ng of human recombinant NF-{kappa}B p50 or bovine serum albumin, and 30 ng of linearized pNF-{kappa}B-Luc plasmid (Stratagene) digested with BsrGI were used as a template. The commercially available plasmid has five NF-{kappa}B-binding sites (TGGGGACTTTCCGC) in the promoter region and a unique restriction site by BsrGI in the open reading frame of the luciferase gene (see Fig. 9A). The reactions were started by the addition of HeLa nuclear extract and incubated at 30 °C for 45 min. The reactions were terminated by the addition of 175 µl of stop solution containing of 0.3 M Tris-HCl, pH 7.4, 0.3 M sodium acetate, 0.5% SDS, 2 mM EDTA, and 3 µg/ml yeast tRNA. RNA was isolated using the RNeasy mini-kit and resuspended in nuclease-free water. The samples were heated at 90 °C for 10 min. RNA was separated on a 7 M urea and 5% acrylamide denaturing gel run at 300 V. Gels were dried, and the radioactivities of runoff transcripts (489 nucleotides) were measured using the Fuji BAS-3000 system. To assess the effects of automodification of PARP-1 on p50-dependent transcription, preincubation of purified His-PARP-1 (118 ng; non-modified or automodified) or NAD+ (10 µM) with HeLa nuclear extracts was performed as described above.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 9.
Effects of AM-PARP and NAD+ on p50-dependent transcription in HeLa nuclear extracts. A, shown is a schematic of putative NF-{kappa}B-binding sites in linearized pNF-{kappa}B-Luc template used in a runoff assay. The hatched regions indicate the five NF-{kappa}B-binding sites (nucleotides –131 to –67). ORF indicates the start of the open reading frame, and the cross-hatched region indicates the TATA box. The unique restriction site by BsrGI is indicated at +489 in the open reading frame. B, runoff assay was performed using 30 ng of template, 8 units of HeLa nuclear extract, and 1–50 ng of p50 subunit (lanes 2–5) or 50 ng of bovine serum albumin (BSA; lane 1) as described. {alpha}-Amanitin ({alpha}-ama; 1 µg/ml), a specific RNA polymerase II inhibitor, was co-incubated with the mixture in the presence of the p50 subunit during the standard reaction (lane 6). The 32P-labeled size markers of {phi}X174 DNA/HinfI are indicated in nucleotides (nt; lane M). The autoradiogram is representative of three independent experiments. C, the radioactivities of runoff transcripts (489 nucleotides) shown in B were quantified using the BAS-3000 system as described under "Experimental Procedures." Data are indicated as the means ± S.E. of six samples from three experiments. D, shown is the effect of the addition of NM- or AM-PARP to HeLa nuclear extract on p50-dependent transcription in vitro. p50-dependent transcription was assessed by runoff assay (lane 2) in the presence of NM-PARP (200 ng; lane 3) or AM-PARP (200 ng; lane 4). The autoradiogram is representative of three independent experiments. Data are indicated as the means ± S.E. of the percent of the control group (lane 1 and open bar).*, p < 0.05 (Student's t test). E, shown is the effect of the addition of NAD+ to HeLa nuclear extract on basal (open bars) or p50-dependent (gray bars) transcription in vitro. HeLa nuclear extract (8 units) was preincubated for 10 min at 37 °C with 10 µM NAD+ (lanes 2, 3, 5, and 6) in the presence (lanes 3 and 6) or absence (lanes 2 and 5) of 10 mM 3-aminobenzamide (3AB). The autoradiogram is representative of three independent experiments. Data are indicated as the means ± S.E. of the percent of the control group (lane 1). *, p < 0.05 (Student's t test) compared with the control group and the NAD+-treated group; #, p < 0.01 (Student's t test) compared with the NAD+-treated group and the NAD+/3-aminobenzamide-treated group.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Effects of PARP-1 Inhibitors on the LPS-induced Expression of TNF-{alpha} in Primary Cultured Mouse Glial Cells—Treatment of mixed glial cells with LPS (1 µg/ml) increases the release of TNF-{alpha} and nitrite in a time-dependent manner (21, 25). The release of TNF-{alpha} and nitrite reached maximal levels at 6 and 24 h, respectively (data not shown). In the following experiments, the levels of TNF-{alpha} and nitrite released into the culture medium were determined at these time points after treatment of the cells with LPS (1 µg/ml).

When the glial cells were pretreated for 30 min with two classical PARP-1 inhibitors, benzamide and nicotinamide, both inhibitors reduced LPS-induced release of TNF-{alpha} in a concentration-dependent manner (Fig. 1, A and B), and the IC50 values for benzamide and nicotinamide were 0.24 and 0.95 mM, respectively. Treatment with their inactive analogs, benzoic acid and nicotinic acid, had no effect (Fig. 1, A and B). Pretreatment of the cells with 10 mM 3-aminobenzamide and 0.1 mM 1,5-dihydroxyisoquinoline, which are more potent inhibitors of PARP-1 than either benzamide or nicotinamide (31), also significantly reduced TNF-{alpha} release to 3.8 ± 0.5 and 23.2 ± 5.8%, respectively (Fig. 1C). There was no significant difference in cell viability between the vehicle- and PARP-1 inhibitor-treated groups, even 24 h after treatment, indicating that these inhibitors at the concentrations used had no toxic effect on the cells (data not shown).



View larger version (42K):
[in this window]
[in a new window]
  		FIG. 1.
Effects of PARP-1 inhibitors and the antisense RNA for PARP-1 mRNA on LPS-induced release of TNF-{alpha} in primary cultured mouse glial cells. A–C, mixed glial cells were pretreated with either vehicle or various concentrations of PARP-1 inhibitors for 30 min, followed by additional incubation with LPS (1 µg/ml). Culture supernatants were collected after 6 h, and the concentrations of TNF-{alpha} (percent) were measured by enzyme-linked immunosorbent assay. Data are indicated as the means ± S.E. of 4–10 samples. **, p < 0.01 (analysis of variance, followed by Dunnett's multiple range test) compared with the LPS-treated control group (open bars). Note that the PARP-1 inhibitors benzamide (BZM), nicotinamide (NIC), 3-aminobenzamide (3AB), and 1,5-dihydroxyisoquinoline (DHIQ), but not the analogs benzoic acid (BA) and nicotinic acid (NA) reduced LPS-induced TNF-{alpha} release. D, microglial cells (1 x 105/well) were transfected with pcDNA3 (0.05 or 0.2 µg/well) as a mock vector (MOCK; lanes 2 and 3) or the antisense RNA vector for PARP-1 mRNA (AS; lanes 4 and 5) and then incubated for 24 h, followed by treatment with LPS (1 µg/ml) for 6 h. The expression of PARP-1 (114 kDa; upper panel) in the cells was assessed by Western blotting with anti-PARP-1 mAb. Anti-GAPDH mAb (36 kDa; middle panel) was used as a control. NT, no treatment (lane 1). E, the levels of TNF-{alpha} (pg/ml) were measured by enzyme-linked immunosorbent assay. Microglial cells were pretreated with benzamide (1 mM) for 30 min before LPS challenge. Data are indicated as the means ± S.E. of four to seven samples. **, p < 0.01 (Student's t test) compared with each control-treated group.

 
To address the question of whether the inhibition of TNF-{alpha} release by PARP-1 inhibitors resulted from inhibition of TNF-{alpha} mRNA transcription, its expression in glial cells was examined by reverse transcription-PCR (Fig. 2). The expression of TNF-{alpha} mRNA in the cells was detectable at 15 min after LPS (1 µg/ml) stimulation and reached maximal levels at 1 h (data not shown). Pretreatment of the cells with benzamide and nicotinamide reduced the expression of TNF-{alpha} m RNA in a concentration-dependent manner (Fig. 2A). Maximal inhibition of benzamide at 10 mM and nicotinamide at 20 mM was 71 ± 16% (p < 0.05) and 87 ± 2.4% (p < 0.01), respectively (Fig. 2B). Their inactive analogs, benzoic acid and nicotinic acid, had no effect (Fig. 2, A and B). The mRNAs were also reduced by 10 mM 3-aminobenzamide and 0.1 mM 1,5-dihydroxyisoquinoline (data not shown).



View larger version (42K):
[in this window]
[in a new window]
  		FIG. 2.
Effects of PARP inhibitors (A and B) and the antisense RNA for PARP-1 mRNA (C and D) on the expression of TNF-{alpha} mRNA in primary cultured mouse glial cells. Shown are the results from 1.8% agarose electrophoresis (A and C) and semiquantification of TNF-{alpha} mRNA by TaqMan PCR (B and D). A and B, mixed glial cells were pretreated with 0.1–10 mM benzamide (BZM; lanes 3–5), 10 mM benzoic acid (BA; lane 6), 0.2–20 mM nicotinamide (NIC; lanes 8–10), or 20 mM nicotinic acid (NA; lane 11) for 30 min, followed by additional incubation with LPS (1 µg/ml) for 1 h. Lane 1 received no treatment (NT). Lanes 2 and 7 received LPS treatment alone. The size markers are indicated in base pairs (lane M). C and D, microglial cells (1 x 105/well) were transfected with pcDNA3 as a mock vector (MOCK; 0.2 µg/well; lane 2) or the antisense RNA vector for PARP-1 mRNA (AS; lane 3) and incubated for 24 h, followed by treatment with LPS (1 µg/ml) for 1 h. In B, the values are indicated as percent of each LPS-treated control group (closed bars) and normalized with the signals amplified from 18 S ribosomal RNA cDNA. * and **, p < 0.05 and 0.01 (analysis of variance, followed by Dunnett's multiple range test), respectively, compared with each LPS-treated control group. In D, the values are indicated as percent of LPS treatment alone (open bar) and also normalized with 18 S ribosomal RNA cDNA, as they were in B. *, p < 0.05 (Student's t test) compared with the mock-treated group.

 
Effects of the Antisense RNA for PARP-1 mRNA on TNF-{alpha} Levels in Primary Cultured Microglial Cells—All PARP-1 inhibitors used in this study bind to the well defined nicotinamide subsite of the NAD+-binding pocket (31), indicating that these compounds would also inhibit other PARPs containing this domain (32). Therefore, to assess the specific relevance of the activity of PARP-1 in LPS-induced synthesis of TNF-{alpha}, we examined isolated microglial cells in which endogenous PARP-1 could be reduced by transfection of the antisense RNA expression vector for PARP-1 mRNA (23). Western blotting revealed that antisense vector transfection decreased PARP-1 to 59% of that in cells treated with a mock vector for 24 h (Fig. 1D). In accordance with the results, the total activity of PARPs in the antisense vector-transfected cells was also reduced to 62% of that in the mock vector-treated cells (data not shown). Whereas LPS-induced release of TNF-{alpha} in the mock vector-treated cells remained unchanged, the release of TNF-{alpha} in the antisense vector-transfected cells (0.2 µg of DNA/well) was reduced to 63 ± 3.5%. It was also significantly reduced by 1 mM benzamide (Fig. 1E). The level of TNF-{alpha} mRNA in the antisense vector-transfected cells was also reduced to 53 ± 9.7% (Fig. 2, C and D).

Effects of PARP-1 Inhibitors and the Antisense RNA for PARP-1 mRNA on the Expression of Pro-inflammatory Mediators—We also investigated the effects of PARP-1 inhibitors and the antisense vector on the synthesis of other pro-inflammatory mediators such as IL-1{beta}, IL-6, and nitric oxide, which is synthesized by inducible iNOS in LPS-treated glial cells (21, 25). Treatment of mixed glial cells and microglial cells with LPS for 24 h resulted in significant increases in iNOS expression levels (Fig. 3A, upper panel) and release of nitrite (lower panel). The treatment of mixed glial cells with benzamide and nicotinamide inhibited the LPS-induced increase in a concentration-dependent manner (Fig. 3A, lower panel). Transfection of the antisense vector into microglial cells also reduced the expression of iNOS and the release of nitrite (Fig. 3B). On the other hand, neither IL-1{beta} nor IL-6 release induced by LPS (1 µg/ml for 24 h) was affected by PARP-1 inhibitors and transfection of the antisense vector (data not shown). Thus, these results suggest that poly(ADP-ribosyl)ation by PARP-1 participates in the LPS-induced expression of inflammatory mediators such as TNF-{alpha} and iNOS in glial cells.



View larger version (25K):
[in this window]
[in a new window]
  		FIG. 3.
Effects of PARP-1 inhibitors (A) and the antisense RNA for PARP-1 mRNA (B) on LPS-induced release of nitrite and expression of iNOS in primary cultured mouse glial cells. A, mixed glial cells were pretreated with 0.1–10 mM benzamide (BZM; lanes 3–5) and 0.2–20 mM nicotinamide (NIC; lanes 6–8) for 30 min, followed by additional incubation with LPS (1 µg/ml) for 24 h. Lane 1 received no treatment (NT). Lane 2 received treatment with LPS alone. Culture supernatants were collected after 24 h, and the micromolar concentrations of nitrite were measured by the Griess method as described previously (25). Data are indicated as the means ± S.E. of 4–10 samples. **, p < 0.01 (analysis of variance, followed by Dunnett's multiple range test) compared with the LPS-treated control group (open bar). B, microglial cells (1 x 105/well) were transfected with pcDNA3 as a mock vector (MOCK; 0.2 µg/well; lane 2) or the antisense RNA vector for PARP-1 mRNA (AS; 0.2 µg/well; lane 3) and incubated for 24 h, followed by treatment with LPS (1 µg/ml) for 24 h. Lane 1 received no treatment. The levels of nitrite were measured by the Griess method. Data are indicated as the means ± S.E. of four to seven samples. **, p < 0.01 (Student's t test) compared with control-treated groups (open bar). In A and B, the expression of iNOS (130 kDa; upper panels) was assessed by Western blotting with anti-iNOS mAb. Anti-GAPDH mAb (36 kDa; middle panels) was used as a control.

 
Examination of Poly(ADP-ribosyl)ation in Primary Cultured Glial Cells Induced by LPS—Because the data described above suggest that PARP-1 activity has an important function in the LPS-activated signaling pathway in glial cells, we used Western blotting with anti-PAR antibody to investigate poly(ADP-ribosyl)ation at the cellular level. As shown in Fig. 4A, two bands of poly(ADP-ribosyl)ated proteins were observed in the unstimulated cells (lane 1), and LPS treatment (1 µg/ml for 1 h) had no effect on the level of poly(ADP-ribosyl)ation (lane 2). The same results were obtained with three other anti-PAR antibodies (see the legend of Fig. 4) (data not shown). The 114-kDa band was identified as AM-PARP since immunoprecipitation with either anti-PARP-1 (Fig. 4B, lane 2) or anti-PAR (lane 3) antibody depleted the band. On the other hand, the 65-kDa band was nonspecific because rabbit control IgG depleted it (Fig. 4B, lane 1). Thus, the one major poly(ADP-ribosyl)ated protein in glia was identified as AM-PARP. LPS treatment did not affect the amounts of AM-PARP (Fig. 4, C and D). To confirm this result, we measured the incorporation of a substrate of PARP-1, [3H]NAD+ (26), into the permeabilized glial cells in the presence of LPS. As shown in Fig. 4E, LPS treatment had no effect on incorporation (open bars), but incorporation was increased ~14-fold by 10 mM H2O2 (shaded bar), a PARP-1 activator that works by making DNA strand breaks (33). These results suggest that the LPS treatment did not affect the poly(ADP-ribosyl)ation of PARP-1 itself. Next, we examined the effects of PARP-1 inhibitors on the automodification of PARP-1. As shown in Fig. 4F, benzamide reduced the automodification in a concentration-dependent manner. Nicotinamide, 3-aminobenzamide, and 1,5-dihydroxyisoquinoline (data not shown) also reduced it, but the inactive analogs benzoic acid and nicotinic acid had no effect (Fig. 4G). The inhibitory effects of PARP-1 inhibitors on the automodification of PARP-1 correlated with those shown in LPS-stimulated TNF-{alpha} release (r2 = 0.83) (see also Fig. 1, A–C).



View larger version (42K):
[in this window]
[in a new window]
  		FIG. 4.
Examination of poly(ADP-ribosyl)ation in primary cultured mouse glial cells. A, effect of LPS treatment on poly(ADP-ribosyl)ation of total glial proteins. Total cell extracts (30 µg) prepared from control (lane 1) or LPS (1 µg/ml for 1 h)-treated cells (lane 2) mixed glial were subjected to Western blotting with rabbit anti-PAR pAb. The same results were obtained by other Western blot analyses with anti-PAR mAb (4335-MC and 10H) and guinea pig anti-PAR pAb (data not shown). B, detection of AM-PARP in glial cells. Total cell extracts (100 µg) prepared from mixed glial cells treated with 1 µg/ml LPS for 1 h were immunodepleted with antibody-bound protein A-Sepharose as follows: rabbit control IgG (lane 1), anti-PARP-1 pAb (lane 2), or rabbit anti-PAR pAb (lane 3). The unbound fractions were collected and subjected to Western blotting with anti-PAR mAb (4335-MC). C and D, effect of LPS treatment on the automodification of PARP-1. C, time course. Total cell extracts (50 µg) prepared from LPS (1 µg/ml)-treated mixed glial cells at the indicated times were subjected to Western blotting with rabbit anti-PAR pAb (upper panel) or anti-PARP-1 mAb (lower panel). D, concentration dependence of LPS. Total cell extracts (30 µg) prepared from mixed glial cells treated with LPS (0.001–10 µg/ml) for 1 h were subjected to Western blotting as described above. E, time course of incorporation of [3H]NAD+ into digitonin-permeabilized mixed glial cells in the presence of 1 µg/ml LPS (open bars) and the effect of treatment with 10 mM H2O2 (a PARP-1 activator) for 10 min on the same parameter (shaded bar). Data are indicated as the means ± S.E. (n = 7). F, effects of benzamide in the presence of LPS on the automodification of PARP-1. Mixed glial cells were pretreated with benzamide (BZM; 0.1–10 mM) for 0.5 h. Total cell extracts (50 µg) prepared from cells treated with LPS (1 µg/ml) for 1 h were subjected to Western blotting with rabbit anti-PAR pAb. G, effects of 10 mM 3-aminobenzamide (3AB; lane 2), 20 mM nicotinamide (NIC; lane 3), 10 mM benzoic acid (BA; lane 4), and 20 mM nicotinic acid (NA; lane 5) in the presence of LPS on the automodification of PARP-1. Mixed glial cells were pretreated with these compounds for 0.5 h. Total cell extracts (50 µg) prepared from cells treated with LPS (1 µg/ml) for 1 h were subjected to Western blotting with rabbit anti-PAR pAb. Note that the inhibitory effects of PARP-1 inhibitors on the extent of the automodification of PARP-1 were fairly consistent with those shown in the LPS-stimulated release of TNF-{alpha} (r2 = 0.83) (see also Fig. 1, A–C).

 
Effects of PARP-1 Inhibitors and the Antisense RNA for PARP-1 mRNA on the LPS-induced DNA Binding of NF-{kappa}B and Translocation into the Nucleus—It has been reported that NF-{kappa}B has a key role in transcription in response to LPS (28). The rapid nuclear translocation of NF-{kappa}B (p65-p50 heterodimer) by LPS activates genes that are related to inflammatory mediators such as TNF-{alpha} and iNOS (28, 34, 35). Abundant AM-PARP exists in the nucleus, and endotoxin-induced expression of some inflammatory mediators is impaired in PARP-1 null mice in vivo (13). Thus, we hypothesized that AM-PARP regulates the gene expression of TNF-{alpha} and iNOS by interacting with NF-{kappa}B in the nucleus. To test this possibility, we determined whether PARP-1 inhibitors and transfection of the antisense vector affect LPS-induced activation of NF-{kappa}B. EMSA analysis showed that LPS (1 µg/ml) increased the binding of NF-{kappa}B to its consensus DNA (Fig. 5A, lanes 3 and 8), and binding reached a maximum at 1 h (data not shown). Pretreatment with benzamide (Fig. 5A, lanes 4–6) and nicotinamide (lanes 9–11) reduced binding in a concentration-dependent manner. Maximal inhibition by benzamide and nicotinamide was 34 ± 7.0% at 10 mM (p < 0.01) and 77 ± 0.8% at 20 mM (p < 0.01), respectively (Fig. 5B). Their inactive analogs, benzoic acid and nicotinic acid, had no effect (Fig. 5, A and B). Transfection of the antisense vector into isolated microglial cells also significantly reduced the DNA binding of NF-{kappa}B to 62 ± 1.1% of the control levels (Fig. 5, D and E). The effects of PARP-1 inhibitors and transfection of the antisense vector on the nuclear translocation of NF-{kappa}B were also examined. Pretreatment with benzamide and nicotinamide had no effects on translocation of NF-{kappa}B p65 into the nucleus (Fig. 5C). Transfection of the antisense vector also did not affect translocation (Fig. 5F).



View larger version (33K):
[in this window]
[in a new window]
  		FIG. 5.
Effects of PARP-1 inhibitors (A–C) and the antisense RNA for PARP-1 mRNA (D–F) on the LPS-induced DNA binding of NF-{kappa}B and translocation into the nucleus in primary cultured mouse glial cells. A, mixed glial cells were pretreated with 0.1–10 mM benzamide (BZM; lanes 4–6), 10 mM benzoic acid (BA; lane 7), 0.2–20 mM nicotinamide (NIC; lanes 9–11), or 20 mM nicotinic acid (NA; lane 12) for 0.5 h, followed by additional incubation with LPS (1 µg/ml) for 1 h. The DNA binding of NF-{kappa}B in the nuclear extracts (5 µg) was assessed by EMSA. Lane 1 indicates the activity of NF-{kappa}B in HeLa cell nuclear extract as a positive control. Lanes 3 and 8 indicate the LPS treatment alone. Competition with a 100-fold excess of unlabeled probes (lane 13) was examined with the nuclear extract of the LPS-treated cells alone (lane 2). Arrows indicate the NF-{kappa}B·DNA complex, which was also assessed by supershift assay with anti-p65 pAb (data not shown). B, the density of the shifted bands of NF-{kappa}B shown in A was quantified using a densitometer as described under "Experimental Procedures." Data are indicated as the means ± S.E. of four samples from duplicate experiments. The values are indicated as percent of each LPS-treated control group (closed bars). **, p < 0.01 (analysis of variance, followed by Dunnett's multiple range test) compared with each LPS-treated control group. C, shown is the nuclear translocation of the p65 subunit of NF-{kappa}B in extracts of isolated nuclei from PARP-1 inhibitor-treated mixed glial cells. Mixed glial cells were pretreated with 0.1–10 mM benzamide (lanes 3–5) or 0.2–20 mM nicotinamide (lanes 6–8) for 0.5 h, followed by additional incubation with LPS (1 µg/ml) for 1 h. The p65 subunit in the nucleus (5 µg) was detected by Western blotting with anti-p65 pAb. Lane 1 received no treatment (NT). Lane 2 received LPS treatment alone. D, microglial cells (1 x 105/well) were transfected with pcDNA3 as a mock vector (MOCK; 0.2 µg/well, lane 2) or the antisense RNA vector for PARP-1 mRNA (AS; 0.2 µg/well; lane 3) and then incubated for 24 h, followed by treatment with LPS (1 µg/ml) for 1 h. Lane 1 received LPS treatment alone. The DNA binding of NF-{kappa}B in nuclear extracts (10 µg) was assessed by EMSA. E, the density of the shifted bands of NF-{kappa}B shown in D was quantified using a densitometer as described under "Experimental Procedures." Data are indicated as the means ± S.E. of four samples from duplicate experiments. The values are indicated as percent of LPS treatment alone (open bar), as they were in B. **, p < 0.01 (Student's t test) compared with the mock vector-treated group. F, shown is the nuclear translocation of the p65 subunit of NF-{kappa}B in extracts of isolated nuclei from the antisense-treated microglial cells. Microglial cells (1 x 105/well) were transfected with pcDNA3 as a mock vector (0.2 µg/well; lane 2) or the antisense RNA vector for PARP-1 mRNA (0.2 µg/well; lane 3) and then incubated for 24 h, followed by treatment with LPS (1 µg/ml) for 1 h. Lane 1 received LPS treatment alone. The p65 subunit in the nucleus (2 µg) was detected by Western blotting with anti-p65 pAb.

 
Effect of the Automodification of PARP-1 on the Physical Association of PARP-1 with the p50 Subunit of NF-{kappa}B in Vitro and in Vivo—Our data described above suggest that the automodification of PARP-1 regulates the transcriptional activities of NF-{kappa}B. Several reports have suggested that PARP-1 may regulate the expression of some specific genes by physical association with NF-{kappa}B (12, 14, 18). Therefore, we further examined whether automodification affects the physical association between PARP-1 and NF-{kappa}B. To do this, we introduced a cell-free system with purified recombinant His-tagged PARP-1 and the purified recombinant p50 subunit of NF-{kappa}B. Purified recombinant PARP-1 was poly(ADP-ribosyl)ated in the presence of NAD+ (Fig. 6A, panel a, right panel, second lane), whereas it was not poly(ADP-ribosyl)ated in the absence of NAD+, and only NM-PARP was present (left panel, first lane). The degree of automodification of purified recombinant PARP-1 (120 kDa) was almost identical to that of AM-PARP (114 kDa) in glial nuclear extracts (Fig. 6, A, panel b). The p50 subunit of NF-{kappa}B was incubated with AM- or NM-PARP and then immunoprecipitated with a p50-specific antibody. The co-immunoprecipitation assay revealed the association between the p50 subunit and NM-PARP (Fig. 6B, lane 2), whereas AM-PARP did not coprecipitate with the p50 subunit (lane 3) since the faint band in lane 3 (WB: PARP-1 in panel a) was not poly(ADP-ribosyl)ated. On the contrary, most of the AM-PARP was detected in the supernatants (Fig. 6B, panel b, lane 3), suggesting that the automodification of PARP-1 restricts the association of PARP-1 with the p50 subunit in vitro. These results introduced us to the hypothesis that the decrease in AM-PARP by a PARP-1 inhibitor might increase NM-PARP and the association with NF-{kappa}B in vivo. We confirmed this hypothesis using primary glial cells. As expected, the association between NF-{kappa}B p50 and PARP-1 in nuclear extracts was dramatically increased by treatment with 3-aminobenzamide (Fig. 6C, panel a, lane 2). On the other hand, the poly(ADP-ribosyl)ation of PARP-1 was detected in the supernatants from the control sample (Fig. 6C, panel b, lane 1).



View larger version (25K):
[in this window]
[in a new window]
  		FIG. 6.
Effect of the automodification of PARP-1 on the physical association between PARP-1 and NF-{kappa}B p50 in vitro and in vivo. A, panel a, Western blotting of NM and AM-PARP. Purified His-PARP-1 (120 kDa) was incubated with or without NAD+ (1 µM) in the reaction mixture and then submitted to Western blotting (WB) with anti-PARP-1 pAb (left panel) or anti-PAR mAb (right panel). Panel b, glial AM-PARP (114 kDa). B, co-immunoprecipitation (IP) of NM- or AM-PARP with the p50 subunit of NF-{kappa}B. NM- or AM-PARP (200 ng) was preincubated with the purified p50 subunit of NF-{kappa}B (350 ng) for 30 min at 4 °C. Subsequently, the mixture was incubated with anti-p50 pAb (2 µg). After immunoprecipitation assay, the eluted proteins (panel a) and the first supernatants (panel b) were submitted to Western blotting with anti-p50 mAb, anti-PARP-1 pAb, or anti-PAR mAb (10H). Input lanes represent 10% of the input. The results are representative of three independent experiments. C, association of PARP-1 with NF-{kappa}B p50 in vivo. Mixed glial cells were pretreated with 10 mM 3-aminobenzamide (3AB; lane 2) for 0.5 h, followed by additional incubation with LPS (1 µg/ml) for 1 h. Glial nuclear extracts (1 mg/assay) prepared as described under "Experimental Procedures" were initially incubated with both rabbit control IgG (5 µg) and protein G-Sepharose beads for 30 min at 4 °C with rocking to block nonspecific binding. After immunoprecipitation assay, the eluted proteins (panel a) and the first supernatants (panel b) were obtained, and the same procedure as described for B was followed. The results are representative of three independent experiments.

 
Effect of the Automodification of PARP-1 on the DNA Binding of the p50 Subunit of NF-{kappa}B in Vitro and in Vivo—Next, we investigated whether the automodification of PARP-1 affects the DNA binding of the p50 subunit. As shown in Fig. 7A, preincubation of the p50 subunit with various amounts of NM-PARP resulted in a reduction in its binding to the consensus DNA: 20 ng of NM-PARP (lane 2) slightly, 100 ng (lane 3) moderately, and 200 ng (lane 4) completely reduced the DNA binding of the p50 subunit (Fig. 7B). In contrast, preincubation with any amount of AM-PARP did not affect the DNA binding of the p50 subunit (Fig. 7, A, lanes 5–7; and B). Similar results were also obtained in experiments with nuclear extracts prepared from LPS-stimulated mixed glial cells (Fig. 7, C and D). When NM-PARP (200 ng) was added to the extracts, a significant inhibition of the DNA binding of the p50 subunit was observed (48 ± 3.8% of the control levels; p < 0.01), whereas the same amount of AM-PARP had no effect on binding (93.2 ± 7.3%).



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 7.
Effect of NM- or AM-PARP on the DNA binding of the p50 subunit of NF-{kappa}B in the cell-free system and glial nuclear extracts. A, shown are the effects of NM- and AM-PARP on the DNA binding of the p50 subunit of NF-{kappa}B. NM-PARP (lanes 2–4) or AM-PARP (lanes 5–7) was preincubated with the p50 subunit (350 ng) for 30 min at 4 °C, and the complex was then resolved by EMSA. Lane 1 is the control (p50 only). B, the density of the shifted bands of NF-{kappa}B shown in A was quantified using a densitometer as described under "Experimental Procedures." Data are indicated as the means ± S.E. of eight samples from four independent experiments. The values are indicated as percent of the control group. {circ}, NM-PARP; , AM-PARP. C, shown are the effects of NM- or AM-PARP on the DNA binding of NF-{kappa}B in nuclear extracts in mixed glial cells. Mixed glial cells were treated with LPS (1 µg/ml) for 1 h. The DNA binding of NF-{kappa}B in the presence of NM-PARP (200 ng; lane 2) or AM-PARP (200 ng; lane 3) was assessed by EMSA. The results are representative of three independent experiments. D, the density of the shifted bands of NF-{kappa}B shown in C was quantified using a densitometer as described under "Experimental Procedures." Data are indicated as the means ± S.E. of eight samples from four independent experiments (closed bars). Data are also indicated as the means ± S.E. of the percent of the control group (open bar). **, p < 0.01 (Student's t test, control versus NM-PARP); ##, p < 0.01 (Student's t test, NM-PARP versus AM-PARP).

 
Effect of NAD+ on the DNA Binding of NF-{kappa}B in Nuclear Extracts—To investigate to what degree automodification of PARP-1 affects the formation of the NF-{kappa}B·DNA complex, the PARP-1 substrate NAD+ was added to the nuclear extracts, and the DNA binding of NF-{kappa}B was examined by EMSA. In Fig. 8A, Western blotting revealed the existence of AM-PARP in the nuclear extracts of mixed glial cells stimulated with LPS (lane 3), as evidenced by the disappearance of the poly(ADP-ribosyl)ated band (114 kDa) in the immunodepletion assay with anti-PARP-1 (lane 1) or anti-PAR (lane 2) antibody. As shown in Fig. 8 (B and C), the addition of NAD+ to the nuclear extracts significantly increased the formation of the NF-{kappa}B·DNA complex by up to ~2-fold (lane 3) compared with the control levels, and the increase was cancelled by co-incubation with 10 mM 3-aminobenzamide (lane 4). In accordance with this, a highly poly(ADP-ribosyl)ated band emerged in the region between AM-PARP (114 kDa) and the gel top in the presence of NAD+ (Fig. 8A, lane 4). This band was identified as a dimeric form of AM-PARP because this band was previously reported to disappear almost completely upon co-incubation with 10 mM 3-aminobenzamide (36).



View larger version (14K):
[in this window]
[in a new window]
  		FIG. 8.
Effect of NAD+ on the DNA binding of NF-{kappa}B in glial nuclear extracts. A, shown are the results from Western blotting of nuclear extracts prepared from mixed glial cells treated with LPS (1 µg/ml) for 1 h. The existence of AM-PARP in nuclear extracts was assessed by Western blotting with anti-PAR mAb (4335-MC; lane 3) and immunodepletion assay with Immunobead-bound anti-PARP-1 pAb (lane 1) or anti-PAR pAb (lane 2). An increase in AM-PARP was shown in the presence of NAD+ (10 µM; lane 4), and 10 mM 3-aminobenzamide (3AB; lane 5) abolished AM-PARP completely. n.s., nonspecific staining (3-aminobenzamide and anti-PAR pAb failed to decrease the bands of AM-PARP). B, NAD+ promotes the DNA binding of NF-{kappa}B in nuclear extracts. Mixed glial cells were treated with LPS (1 µg/ml) for 1 h, and nuclear extracts were prepared. The extracts were further incubated with 1 mM 3-aminobenzamide (lane 1), 10 µM NAD+ (lane 3), or a combination of 10 mM 3-aminobenzamide and 10 µM NAD+ (lane 4), and the DNA binding of NF-{kappa}B was analyzed by EMSA. Lane 2 indicates control. C, the density of the shifted bands of NF-{kappa}B shown in B was quantified using a densitometer as described under "Experimental Procedures." The results are representative of three independent experiments. Data are indicated as the means ± S.E. of three samples. * and **, p < 0.05 and 0.01 (Student's t test), respectively, compared with the control group (open bar); ##, p < 0.01 (Student's t test) compared with the NAD+-treated group (gray bar) and the NAD+/3-aminobenzamide-treated group (closed bar).

 
Effect of NAD+ or the Automodification of PARP-1 on NF-{kappa}B p50-dependent Transcription in Vitro—The results shown in Figs. 7 and 8 suggest that the DNA binding of NF-{kappa}B is dependent on the degree of automodification of PARP-1. To assess whether the automodification of PARP-1 is capable of mediating NF-{kappa}B-dependent transcription, an in vitro transcription assay using linearized pNF-{kappa}B-Luc plasmid was performed. The template contains five NF-{kappa}B-binding sites in the promoter region and the unique restriction site by BsrGI in the open reading frame of the luciferase gene (Fig. 9A). The addition of the p50 subunit to HeLa nuclear extracts led to a concentration-dependent increase in runoff transcripts, corresponding in size (489 nucleotides) to that predicted based on initiation sites for the luciferase gene (Fig. 9, B and C). Preincubation of the p50 subunit with NM-PARP completely reduced almost all of the specific transcription by p50 (Fig. 9D). In contrast, AM-PARP significantly enhanced p50-dependent transcription (p < 0.01) by up to ~1.5-fold (Fig. 9D). Enhancement of p50-dependent transcription was also detected upon the addition of NAD+ to HeLa nuclear extracts and was canceled by preincubation with 3-aminobenzamide (Fig. 9E, shaded bars). Interestingly, the enhanced transcription upon the addition of NAD+ and its cancellation by preincubation with 3-aminobenzamide were also detected in the absence of p50 (Fig. 9E, open bars), suggesting that increasing amounts of AM-PARP with NAD+ lead not only to the enhancement of p50-dependent transcription, but also to the promotion of basal transcriptional efficiency.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
The results from this study strongly suggest that PARP-1 is involved in the signaling cascade of LPS-induced expression of inflammatory mediators such as TNF-{alpha} and iNOS in glial cells and also that the automodification of PARP-1 regulates NF-{kappa}B-dependent transcriptional activation via restriction of the physical association between PARP-1 and NF-{kappa}B. These conclusions are based on the following findings. 1) The inhibitors of PARP-1 and the antisense RNA for PARP-1 mRNA decreased the amounts of AM-PARP and the expression of TNF-{alpha} and iNOS (Figs. 1, 2, 3) and the LPS-induced DNA binding of NF-{kappa}B (Fig. 5). 2) PARP-1 itself was the main acceptor of PAR in glial cells (Fig. 4). 3) The automodification of PARP-1 dissociated NF-{kappa}B p50 from PARP-1, whereas NM-PARP was associated with NF-{kappa}B p50 in vitro and in vivo (Fig. 6). 4) NM-PARP disturbed the formation of the NF-{kappa}B·DNA complex (Fig. 7). 5) AM-PARP led to the enhancement of p50-dependent transcription in vitro (Fig. 9). 6) NAD+ promoted the formation of the NF-{kappa}B·DNA complex and p50-dependent transcription in vitro concomitant with an enhancement of automodification (Figs. 8 and 9). These results are compatible with results from previous studies in which NF-{kappa}B-dependent inflammatory responses were impaired in PARP-1 null mice in vivo as well as in vitro (13, 14, 37) and in which the stress-induced gene expression of cytokines and adhesion molecules was reduced by PARP-1 inhibitors (37–39).

Activation of PARP-1 mainly causes the automodification of PARP-1 itself in vivo (36). We also confirmed that no poly(ADP-ribosyl)ated proteins other than PARP itself were detected by Western blot analysis (Fig. 4). Interestingly, LPS treatment did not affect the amounts of AM-PARP produced in glial cells (Fig. 4). On the other hand, PARP-1 inhibitors suppressed automodification and induced the subsequent decrease in AM-PARP (Fig. 4), probably because of the short half-life (<1 min) of PAR, which appears to be attributed to the high activity of poly(ADP-ribose) glycohydrolase (1, 2). These results suggest that the LPS-induced expression of TNF-{alpha} and iNOS is a prerequisite for the basal activity of PARP-1 and/or AM-PARP. Some studies have also shown that the activity of PARP-1 regulates transcriptional activation (29, 40, 41). In particular, AM-PARP regulates the transactivation of the Reg gene, a gene for insulin-producing {beta}-cell regeneration, by binding to the Reg promoter (29). These results suggest that among all of the PAR acceptor proteins, PARP-1 itself plays a predominant role in regulating some gene expression by its automodification.

Recently, Chiarugi and Moskowitz (38) demonstrated a critical role of poly(ADP-ribosyl)ation in NF-{kappa}B-dependent transcription during microglial activation. The results in our study are compatible with their conclusion. Several novel findings further clarified the essential role of PARP-1 in gene expression via activation of NF-{kappa}B in primary cultured mouse glial cells. First, among the PARP family members, PARP-1 has a central role in NF-{kappa}B-dependent transcription. It has been reported that cellular poly(ADP-ribosyl)ation by some novel members of the PARP family (e.g. tankylase-1/2) is also reduced by PARP-1 inhibitors (3). Therefore, inhibitory effects of PARP-1 inhibitors on other PARPs were not completely excluded. The results obtained with the antisense RNA for PARP-1 mRNA (Figs. 1, 2, 3 and 5) reinforce the link between PARP-1 and LPS-induced gene expression because this technique has the advantage of specifically depleting PARP-1 among the PARPs due to full-length RNA expression (23). Second, the automodification of PARP-1 regulates the DNA binding activity of NF-{kappa}B in glial cells. As shown in Fig. 5, pretreatment with PARP-1 inhibitors and transfection of the antisense vector reduced the specific binding of NF-{kappa}B to the consensus DNA via a decrease in the amount of AM-PARP without affecting the translocation of NF-{kappa}B into the nucleus. The automodification domain of PARP-1 is located in the central region of the enzyme (between amino acids 374 and 525 of human PARP-1) (42). NF-{kappa}B (p50 and p65) preferentially interacted with the region containing the automodification domain (between amino acids 341 and 531) (14), forming a stable immunocomplex with PARP-1 (12, 14, 18). In this study, we also found that NM-PARP associated with the p50 subunit of NF-{kappa}B, whereas AM-PARP could not form a complex with the p50 subunit in vitro and in vivo (Fig. 6) because the net charge at the automodification domain of AM-PARP may be too negative to allow their association (2, 43). In addition, NM-PARP disturbed the formation of the NF-{kappa}B·DNA complex in both the cell-free system and the nuclear extracts (Fig. 7), whereas NAD+ promoted the formation concomitant with an enhancement of automodification (Fig. 8). These results suggest that the automodification of PARP-1 might dissociate PARP-1 from NF-{kappa}B in the nucleus, and thus, the DNA binding of NF-{kappa}B is protected against the inhibitory action by NM-PARP. A recent report indicating that the automodification of PARP-1 facilitates the DNA binding of NF-{kappa}B in vitro supports this idea (18).

This study also suggests that not only the automodification, but also the AM-PARP molecule itself is required for transcriptional regulation because the DNA binding of NF-{kappa}B was inhibited by depletion of AM-PARP with the antisense vector. In fact, genetic deletion of PARP-1, which theoretically leads to the disappearance of AM-PARP, abrogates some NF-{kappa}B-dependent gene expression in mice (13). However, the exact role of AM-PARP in in vivo transcriptional activation remains unclear. A recent study suggested functional relevance between PARP-1 and DNA structures in eukaryotic gene expression (44). For example, heteroduplex DNA with hairpins provides a potent binding site for PARP-1 in the gene-regulating sequence in the absence of damaged DNA (45). Therefore, as the auxiliary protein, AM-PARP may regulate NF-{kappa}B-dependent transcription by automodification via interaction with a DNA secondary structure. Indeed, PARP-1 binds specifically to the 5'-flanking MCAT-1 element to regulate muscle-specific transcription in a poly(ADP-ribosyl)ation-dependent manner (41).

As shown in Fig. 7, NM-PARP reduced the DNA binding of NF-{kappa}B in both the cell-free system and nuclear extracts, whereas AM-PARP had no effect. On the other hand, AM-PARP had the ability to promote NF-