Triptolide, an active component of the Chinese herbal remedy Tripterygium wilfordii Hook F, inhibits production of nitric oxide by decreasing inducible nitric oxide synthase gene transcription

Authors

  • B. Wang,

    1. Autoimmunity Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, NIH, Bethesda, Maryland
    Search for more papers by this author
  • L. Ma,

    1. Autoimmunity Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, NIH, Bethesda, Maryland
    Current affiliation:
    1. China-Japan Friendship Hospital, Beijing, China
    Search for more papers by this author
  • X. Tao,

    1. Autoimmunity Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, NIH, Bethesda, Maryland
    Search for more papers by this author
  • P. E. Lipsky

    Corresponding author
    1. Autoimmunity Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, NIH, Bethesda, Maryland
    • Autoimmunity Branch, National Institute of Arthritis and Musculoskeletal Skin Diseases, NIH, 9000 Rockville Pike, Building 10, Room 9N228, Bethesda, MD 20892
    Search for more papers by this author

Abstract

Objective

The ethyl acetate (EA) extract of Tripterygium wilfordii Hook F (TWHF) and its major active component, triptolide, have been reported to be effective in the treatment of rheumatoid arthritis and other autoimmune inflammatory diseases. Nitric oxide (NO) has been recognized as an important mediator of inflammation. This study was therefore undertaken to examine the effects of the EA extract and triptolide on the production of NO and inducible NO synthase (iNOS) gene expression and transcription in vivo and in vitro.

Methods

Peritoneal macrophages from C57BL/6J mice treated orally with the EA extract of TWHF were assayed for NO production and iNOS messenger RNA (mRNA) expression by reverse transcriptase–polymerase chain reaction. The murine fibroblast cell line NIH3T3 was also assessed for NO production and iNOS mRNA expression, as well as for iNOS promoter activation, Oct-1 nuclear binding capacity, and Oct-1 protein content by transient transfection, electrophoretic mobility shift assay, and immunoblotting, respectively.

Results

NO production and iNOS mRNA expression by macrophages from C57BL/6J mice immunized with trinitrophenyl–bovine serum albumin in Freund's complete adjuvant were significantly inhibited by oral administration of the EA extract (52.3% and 59.8% of control, respectively, at one-eighth of the dose that is lethal for 50% of the animals [LD50] and 21.0% and 38.1% of control, respectively, at one-fourth the LD50). Moreover, the EA extract and triptolide significantly inhibited NO production in vitro in activated peritoneal macrophages, which reflected a decreased level of iNOS mRNA. Finally, triptolide inhibited promoter activity of the iNOS gene and induction of the activity of the regulator of iNOS transcription, Oct-1.

Conclusion

The EA extract of TWHF and triptolide inhibit transcription of the iNOS gene. This may contribute to the antiinflammatory effects of this traditional herbal remedy.

Tripterygium wilfordii Hook F (TWHF) has been used as an herbal remedy to treat arthritis and other autoimmune inflammatory disorders for several centuries in China. The ethyl acetate (EA) extract is one of the most popular preparations of TWHF because it causes fewer adverse effects than cruder preparations (1, 2). Both phase I and phase II studies of the EA extract of TWHF in rheumatoid arthritis (RA) patients showed that it appeared to be safe and clinically beneficial (3, 4). Previous studies showed that the antiinflammatory and immunosuppressive effects of the EA extract could be accounted for by the content of specific diterpenoids, including triptolide and tripdiolide (5, 6). Recently, it has been reported that the EA extract of TWHF inhibited prostaglandin E2 production by in vitro lipopolysaccharide (LPS)–stimulated human monocytes and synovial fibroblasts derived from RA patients and by carrageenan-induced air pouch lining cells in rats (7, 8). However, the precise mechanism by which TWHF inhibits inflammation has not been completely delineated.

Nitric oxide (NO) is known to be involved in the support of normal physiologic functions, such as vasodilation and neurotransmission, and also contributes to the killing of intracellular pathogens and host defense against tumor cells. On the other hand, sustained and excess production of NO can play a role in inflammation and tissue damage. The rate-limiting enzyme in the production of NO is NO synthase (NOS). There are various isoforms of NOS. The constitutive and inducible (iNOS) forms are responsible for regulating the physiologic and pathologic roles of NO, respectively (9–11).

An increase in NO production has been noted in patients with RA or other inflammatory joint diseases (12–15). Increases in NO production and iNOS expression by peripheral blood monocytes or synovium were found to be more significant in patients with RA than in those with osteoarthritis (15, 16). NO production in RA patients was significantly correlated with clinical manifestations of inflammation, including morning stiffness and the number of tender and swollen joints, as well as with the serum level of C-reactive protein (17, 18). Glucocorticoids and other antirheumatic drugs inhibit NO production in RA patients (12, 18). Notably, administration of the EA extract of TWHF caused a reduction in production of NO in an animal model of arthritis that paralleled improvement in the degree of arthritis (19). These results indicate that NO is a mediator of inflammation that may be involved in the pathogenesis of RA and other forms of arthritis. Therefore, it was of interest to determine whether the EA extract of TWHF and its active component, triptolide, could affect production of NO in vivo and in vitro.

Results from the current study show that treatment of mice with the EA extract both in vivo and in vitro inhibited NO production and iNOS messenger RNA (mRNA) expression. Triptolide suppressed iNOS gene expression at the transcriptional level by inhibiting induction of the activity of Oct-1, which is known to regulate iNOS transcription. The data suggest that the clinical activity of the EA extract of TWHF in RA patients may relate to its ability to suppress iNOS expression.

MATERIALS AND METHODS

Drugs and animal treatment regimen

The EA extract of TWHF was prepared from peeled roots of TWHF (obtained from Fujian Province in China) by sequential extraction with ethanol and ethyl acetate, as previously described (5, 20). Triptolide was isolated from the EA extract using silica gel chromatography followed by preparative high-performance liquid chromatography with a Nova-Pak C-18 column (Waters, Milford, MA), and was identified by ultraviolet and infrared spectroscopy, proton nuclear magnetic resonance spectroscopy, and mass spectroscopy (21). The triptolide content of the EA extract was identified and quantitated by high-performance liquid chromatography, as previously described (20), and was found to be 1.08 ng per mg of the EA extract. For in vivo treatment, the EA extract was dissolved in a solution of ethanol, Tween 20, and water (1:1:8). For in vitro study, the EA extract, triptolide, and dexamethasone (Sigma, St. Louis, MO) were first dissolved in a designated volume of ethanol. The stock solutions were further diluted to the indicated concentration with culture medium or phosphate buffered saline (PBS).

As shown in Figure 1, male C57BL/6J mice were divided into 3 groups with 6 mice in each group. Trinitrophenyl–bovine serum albumin (TNP-BSA) (50 μg) emulsified in 0.1 ml of Freund's complete adjuvant (CFA) (22) was administered to each mouse on day 0. Animals were treated by daily gavage for 6 days with vehicle only (98 mg/kg daily) or the EA extract (196 mg/kg daily), beginning on day 14. These doses were equivalent to one-eighth and one-fourth of the dose that caused death in 50% of animals after a single oral administration of the EA extract of TWHF (LD50). The animals received a booster of TNP-BSA by peritoneal injection on the last day of treatment (day 19). The animals exhibited no signs of distress after receiving vehicle or the EA extract of TWHF, as manifested by weight loss, ruffled fur, anorexia, or diarrhea.

Figure 1.

Diagram of administration of ethyl acetate (EA) extract to mice. Eighteen mice were randomly divided into 3 groups of 6 each. The mice were immunized with trinitrophenyl–bovine serum albumin (TNP-BSA) mixed with Freund's complete adjuvant (CFA; 50 μg/0.1 ml/mouse) on day 0. The mice were treated orally with vehicle or 98 mg/kg or 196 mg/kg of the EA extract of Tripterygium wilfordii Hook F by gavage daily for 6 days starting on day 14. The animals received a peritoneal booster injection with TNP-BSA mixed with CFA on the last day of treatment (day 19), were killed by an overdose of metofane 24 hours later, and peritoneal exudates were harvested.

Cell sources and cell culture.

Four sources of cells were used: 1) peritoneal macrophages prepared from the C57BL/6J mice that had been immunized with TNP-BSA (50 μg/0.1 ml/animal) 2 weeks before and treated orally for 6 days with vehicle or the EA extract of TWHF at a daily dose of one-eighth the LD50 or one-fourth the LD50, 2) peritoneal macrophages from normal C57BL/6J mice, prepared 4 days after intraperitoneal injection with 1.5 ml of 3% thioglycollate, 3) murine macrophage cell line RAW 264.7 cells, and 4) NIH3T3 cells. The exudate cells were collected and incubated on glass petri dishes for 3 hours. The macrophages were harvested and counted after the nonadherent cells were removed by washing 3 times with RPMI 1640. Both of the cell lines were obtained from American Type Culture Collection (Rockville, MD).

All peritoneal macrophages and RAW 264.7 cells were cultured in RPMI 1640 supplemented with penicillin (100 units/ml), streptomycin (100 μg/ml), L-glutamine (0.3 mg/ml), and 10% fetal bovine serum (FBS). For NO production and RNA extraction, cells were cultured in complete RPMI 1640 with or without stimulation as indicated and in the absence or presence of the indicated concentrations of inhibitors for the indicated periods. NIH3T3 cells were cultured in Iscove's modified Dulbecco's medium (IMDM) supplemented with penicillin (100 units/ml), streptomycin (100 μg/ml), L-glutamine (0.3 mg/ml), and 10% FBS. On the day before the experiments, NIH3T3 cells were cultured in IMDM and 1% FBS.

Assay of NO

NO production was evaluated by measuring the nitrite content in culture supernatants. Nitrite content in duplicate diluted samples was measured by adding 100 μl of freshly prepared Griess reagent (equal volumes 0.2% naphthylethylenediamine and 2% sulfanilamide in 5% phosphoric acid) to 100 μl of the samples in 96-well plates and reading the optical density (OD) at 540 nm (23). The concentration of nitrite was determined by comparison with the OD curves of serial dilutions of sodium nitrite.

RNA preparation and iNOS mRNA assay

Total RNA was extracted from mitogen-stimulated cells, as previously described (7, 8). Briefly, after the culture supernatants were removed, total RNA was extracted from the remaining cells with guanidinium thiocyanate–phenol–chloroform (Ultraspec; Biotecx, Houston, TX). Then iNOS mRNA was measured by reverse transcriptase–polymerase chain reaction (RT-PCR), as previously described (7). Briefly, mRNA was transcribed into complementary DNA (cDNA) using the GeneAmp RNA PCR kit (Perkin-Elmer, Branchburg, NJ) (7). First-strand cDNA was reverse-transcribed from 1 μg total RNA using an oligo(dT)16 primer and murine leukemia virus RT at 42°C for 15 minutes. The reaction was stopped by heating to 99°C for 5 minutes, followed by cooling to 5°C for 5 minutes. Then 10 μl of the synthesized cDNA from each sample was used in a 50-μl PCR amplification with 2.5 units/100 ml AmpliTaq DNA polymerase and specific primers for mouse iNOS or β-actin. Amplification of cDNA sequences was carried out in a PTC-100 96V programmable thermal controller (MJ Research, Waltham, MA). The primers used were as follows: for mouse iNOS, sense 5′-TTTGTGCGAAGTGTCAGTGGC-3′ and antisense 3′-TGCCCTTTTTTGCCCCATAGG-5′; for β-actin, sense 5′-GTGGGCCGCTCTAGGCACCAA-3′ and antisense 3′-CTCTTTGAGTCACGCACGACTTC-5′, designed using the GeneWorks program. The amplification consisted of an initial step at 95°C for 2 minutes, denaturation at 95°C for 1 minute, annealing-extension at 60°C for 1 minute, and extension at 72°C for 7 minutes.

PCR product (10 μg) from each sample was separated by electrophoresis on a 1.2% agarose gel containing ethidium bromide and visualized by ultraviolet-induced fluorescence. The gel was denatured, and the PCR product was transferred onto a nylon membrane. The membranes were prehybridized in hybridization solution for 45 minutes, followed by hybridization with a γ32P-labeled iNOS probe (CTACGTTCAGGACATCCTGA) or a β-actin probe (CCCAGATCATCATGTTTGAGACCTTCAACACCC) for 45 minutes. The membranes were washed and then exposed to x-ray film (Eastman Kodak, Rochester, NY). Signal intensity was quantitated by densitometry using an image analyzer (AMBIS Systems, San Diego, CA). Signal intensity of the specific mRNA was normalized by comparison with that of β-actin mRNA and expressed as the ratio of the specific mRNA to the corresponding β-actin mRNA.

Transient transfection of iNOS promoter reporter constructs into NIH3T3 cells

The day before transfection, 5 × 105 NIH3T3 cells in 5 ml of IMDM with 1% FBS were seeded in 6-well plates (24). After the cells grew to ∼80% confluence, they were washed with serum-free medium. Four micrograms of a human iNOS, promoter construct linked to luciferase cDNA (pGL3-iNOS, kindly provided by Dr. Arnold S. Kristof, NHLBI, NIH, Bethesda, MD) (25) and 2 μg of pSV-β-galactosidase (Clontech, Palo Alto, CA) were suspended in 50 μl of medium without serum for a few seconds, and 20 μl of the superfect transfection reagent (Qiagen, Chatsworth, CA) was added and mixed to allow complex formation. Ten minutes later, the complexes were diluted with 0.6 ml of IMDM containing 10% FBS and transferred to the cells in the 6-well plate. The cells were then incubated for 3 hours at 37°C. After the medium containing the complexes was removed, the transfected cells were cultured with or without LPS (5 μg/ml) plus phorbol 12-myristate 13-acetate (PMA; 10 ng/ml) and in the absence or presence of the indicated concentrations of inhibitors for 24 hours. Lysates in each sample were obtained by 3 cycles of freeze-thawing. Luciferase and β-galactosidase activity from the cell extracts were assayed by the chemiluminescence method according to the instructions of the manufacturer (Promega, Madison, WI). The iNOS gene promoter activity was quantitated by luciferase/β-galactosidase activity.

Electrophoretic mobility shift assay (EMSA)

Oct-1 binding activity was measured by EMSA, as previously described (26). NIH3T3 cells were grown to 80% confluence, then seeded at a density of 4 × 106 cells in a culture flask containing 30 ml of IMDM with 1% FBS and incubated overnight. After stimulation with or without LPS (5 μg/ml) plus PMA (10 ng/ml) and in the absence or presence of the indicated concentrations of inhibitors for 3 hours, the cells were washed twice with prechilled PBS, harvested by scraping, and nuclear protein extracts were prepared using a nuclear extract kit (Active Motif, Carlsbad, CA). Briefly, the cells were pelleted and the membranes were dissolved with lysis buffer containing Nonidet P40. The cytoplasmic fraction was then removed, nuclei were further lysed, and nuclear proteins were harvested. After the protein concentration of nuclear extract was measured by an assay based on the Bradford method (27), the nuclear extract was examined for binding activity by EMSA.

A double-stranded oligonucleotide probe prepared to match the human Oct-1 binding site (forward 5′-TGTCGAATGCAAATCACTAGAA-3′, complement 3′-ACAGCTTACGTTTAGTGATCTT-5′) was labeled with T4 kinase and γ32P-ATP and purified with a G-25 spin column. Nuclear extract (10 μg) was then incubated with gel shift loading buffer (Promega) at room temperature for 10 minutes, and 32P-labeled Oct-1 consensus oligonucleotide was added, incubated at room temperature for another 20 minutes, and subjected to 5% polyacrylamide gel electrophoresis at 190V for 90 minutes. To examine the specificity of the Oct-1 binding protein, the gel shift was performed in parallel in the presence of a 100-fold excess of unlabeled wild-type Oct-1 or mutant Oct-1 oligonucleotide as competitors. The gel was then dried and exposed to x-ray film overnight at −70°C. The binding band was quantitated by densitometry.

Immunoblot analysis

Nuclear extracts of NIH3T3 cells were prepared, and the protein concentration of the extracts was determined by an assay based on the Bradford method. The extracts (20 μg) were fractionated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes. Oct-1 was detected using rabbit polyclonal IgG primary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at a 1:200 dilution for 2 hours at room temperature. After extensive washing, a horseradish peroxidase–conjugated goat anti-rabbit second antibody at a 1:2,000 dilution was applied for 1 hour at room temperature. The blot was washed and exposed to x-ray film with enhanced chemiluminescence (Pierce, Rockford, IL). Immunoblotting of actin was also performed on the same membrane after stripping.

Statistical analysis

Student's t-test was used to evaluate the significance of the differences between groups.

RESULTS

Effects of in vivo treatment with the EA extract of TWHF on NO production and iNOS mRNA expression in murine peritoneal macrophages.

The effect of the EA extract of TWHF on in vivo NO production was first examined using peritoneal macrophages obtained from C57BL/6J mice that were immunized with TNP-BSA in CFA (Figure 1). The nitrite that accumulated in the culture supernatant of macrophages stimulated in vivo was used as an index for NO synthesis by these cells. Both doses of the EA extract administered orally to the animals significantly suppressed NO production by the activated peritoneal macrophages assessed immediately ex vivo (Figure 2). NO production was inhibited by 47.7% or 79% by dosing with one-eighth the LD50 or one-fourth the LD50 of the EA extract, respectively.

Figure 2.

Effects of in vivo treatment with the ethyl acetate (EA) extract of Tripterygium wilfordii Hook F on nitrite production by murine peritoneal macrophages. C57BL/6J mice (n = 18) were immunized with 50 μg of trinitrophenyl–bovine serum albumin (TNP-BSA) emulsified in 0.1 ml of Freund's complete adjuvant (CFA). Mice were treated orally with vehicle or with the EA extract at a daily dose of 98 mg/kg (one-eighth of the dose that is lethal for 50% of the animals [LD50]) or 196 mg/kg (one-fourth of the LD50) for 6 days. The animals were challenged with TNP-BSA in CFA intraperitoneally on the last day of treatment, killed the next day, and peritoneal exudate cells were collected. After the nonadherent cells were removed to petri dishes, the macrophages were harvested and counted. Macrophages (2 × 106 cells/ml) were incubated for 24 hours. Cell-free supernatants were collected for determination of nitric oxide (NO) content with the Griess reagent. Results are the mean and SEM from 6 mice in each group. ∗ = P < 0.05; ∗∗ = P < 0.01, versus the vehicle-treated group.

To examine the mechanism of the inhibition of NO production by the EA extract in vivo, the effect of the EA extract on iNOS mRNA expression was examined in the cells from the treated animals. As shown in Figure 3 (upper panel), both doses of the EA extract significantly reduced iNOS mRNA expression in the peritoneal macrophages. Treatment with one-eighth the LD50 or one-fourth the LD50 of the EA extract reduced relative iNOS mRNA levels, quantitated by densitometry and expressed as iNOS/β-actin ratios, by 40.2% and 61.9%, respectively (lower panel). The results suggest that in vivo treatment with the EA extract inhibits up-regulation of iNOS mRNA, which results in decreased NO production.

Figure 3.

Effect of in vivo treatment with ethyl acetate (EA) extract of Tripterygium wilfordii Hook F on inducible nitric oxide synthase (iNOS) mRNA expression by murine peritoneal macrophages. Macrophages were collected and incubated as described in Figure 2. After supernatants were removed, the cells were extracted for assay of iNOS mRNA by reverse transcriptase–polymerase chain reaction. Top, Representative autoradiography showing the expression of iNOS mRNA and β-actin mRNA from individual animals. A normal mouse with no treatment and no stimulation (Nil) was used as a negative control. Bottom, Relative iNOS mRNA levels quantitated by densitometry and expressed as iNOS/β-actin optical density ratios. Results are the mean and SEM from 6 mice in each group. ∗ = P < 0.05; ∗∗ = P < 0.01, versus vehicle-treated group. LD50 = median lethal dose.

In vitro effects of triptolide and the EA extract of TWHF on NO production and iNOS mRNA expression in activated macrophages.

The effects of triptolide and the EA extract on NO production and iNOS mRNA expression were examined in vitro using thioglycollate-elicited peritoneal macrophages. Since dexamethasone suppresses NO production and iNOS mRNA expression (28–30), we used it as a positive control. As shown in Figures 4 and 5, triptolide, the EA extract, and dexamethasone suppressed NO production and iNOS mRNA expression in a concentration-dependent manner. Similar effects on NO production and iNOS mRNA expression were observed in the LPS-stimulated murine macrophage cell line RAW 264.7 (data not shown). The dose-response characteristics were similar to those noted with peritoneal macrophages. Importantly, neither the EA extract nor triptolide affected cell viability, assessed by trypan blue exclusion assay (data not shown), indicating that inhibition of NO synthesis is not simply related to a cytotoxic effect. Since inhibition of iNOS mRNA paralleled the decrease in NO accumulation, the results suggest that triptolide and the EA extract inhibited NO production by reducing iNOS mRNA expression.

Figure 4.

Effects of the ethyl acetate (EA) extract of Tripterygium wilfordii Hook F (TWHF), triptolide (Tript), and dexamethasone (Dex) in vitro on nitric oxide (NO) production by murine peritoneal macrophages. Macrophages were collected from the peritoneal cavity of C57BL/6J mice 4 days after intraperitoneal injection of 2.5 ml of 3% thioglycollate. Macrophages (2 × 106 cells/ml) were prepared and incubated for 24 hours with or without lipopolysaccharide (LPS; 2 μg/ml) in the presence or absence of the indicated concentrations of the EA extract of TWHF, Tript, or Dex. Cell-free supernatants were collected for determination of NO content with the Griess reagent. Results are the mean and SEM from 4 separate experiments with cells from different mice. ∗ = P < 0.05; ∗∗ = P < 0.01, versus LPS control.

Figure 5.

Effect of the EA extract of TWHF, Tript, and Dex in vitro on inducible nitric oxide synthase (iNOS) mRNA expression by murine peritoneal macrophages. Macrophages were prepared and cultured as described in Figure 4. After the supernatants were removed, the cells were extracted for assay of iNOS mRNA by reverse transcriptase–polymerase chain reaction. Results are from 1 of 4 separate experiments with similar results, using cells from individual animals. Top, Representative autoradiography showing iNOS and β-actin mRNA. Bottom, Relative iNOS mRNA levels quantitated by densitometry and expressed as iNOS/β-actin optical density ratios. Data are from the same experiment as in the upper panel. See Figure 4 for other definitions.

Effects of triptolide on iNOS promoter activity and induction of Oct-1 binding activity.

To determine whether the EA extract or triptolide inhibited transcription of iNOS, a promoter construct in which luciferase was driven by 8.3 kb of the human iNOS promoter was transfected into NIH3T3 cells, and pSV-β-galactosidase was cotransfected as a control. LPS and PMA induced iNOS promoter activity, as shown in Figure 6. Dexamethasone (5 μM) markedly inhibited iNOS promoter activity induced by LPS and PMA. The EA extract at concentrations of 8 μg/ml inhibited luciferase activity by up to 70% of the control response. Similarly, triptolide (11–44 nM) suppressed the enhanced iNOS promoter activity in a concentration-dependent manner. The degree of inhibition by the EA extract of TWHF and triptolide was comparable with that noted with 5 μM dexamethasone.

Figure 6.

Effects of the EA extract of TWHF, Tript, and Dex on human inducible nitric oxide synthase (iNOS) promoter activity in NIH3T3 cells. The pGL3-iNOS driving the luciferase gene and the pSV-β-galactosidase plasmid were cotransfected into NIH3T3 cells. Transfected cells were cultured for 24 hours with or without LPS (5 μg/ml) plus phorbol 12-myristate 13-acetate (PMA; 10 ng/ml) and in the absence or presence of inhibitors at the indicated concentrations. Luciferase and β-galactosidase activity from the cell extracts was assayed by the chemiluminescence method. All iNOS gene promoter activity was quantitated by luciferase/β-galactosidase activity. Results are the mean and SEM of 4 experiments. ∗ = P < 0.05; ∗∗ = P < 0.01, versus LPS + PMA control. See Figure 4 for other definitions.

Since Oct-1 is one of the transcription factors that may regulate iNOS transcription (31–33), the effect of triptolide on Oct-1 binding activity induced by LPS and PMA was analyzed by EMSA. As shown in Figure 7A, stimulation of NIH3T3 cells with LPS and PMA induced Oct-1 binding activity. Specificity was confirmed by competition with an excess molar concentration of wild-type Oct-1 or mutant Oct-1 oligonucleotide. Notably, triptolide (11–44 nM) or dexamethasone (1–10 μM) significantly inhibited Oct-1 binding activity induced by the combination of LPS and PMA. However, the addition of either triptolide or dexamethasone directly to the reaction mixture of Oct-1 oligonucleotides and the nuclear extract obtained from the LPS- and PMA-stimulated cells did not influence the binding capacity of Oct-1 protein to its oligonucleotide (Figure 7B). These results suggest that triptolide inhibited the up-regulation of Oct-1 binding activity induced by LPS and PMA. Moreover, immunoblotting showed that triptolide did not reduce the total amount of Oct-1 protein in the nucleus (Figure 8). These results indicate that triptolide inhibited the induction of Oct-1 binding activity that is required for the induction of iNOS transcription.

Figure 7.

Effects of Tript and Dex on Oct-1 activity. A, NIH3T3 cells were cultured for 3 hours with or without LPS (5 μg/ml) and phorbol 12-myristate 13-acetate (PMA; 10 ng/ml) and in the absence or presence of inhibitors at the indicated concentrations. Nuclear extracts in each sample (10 μg) were then prepared, and electrophoretic mobility shift assays (EMSAs) were performed as described in Materials and Methods. Left, Representative autoradiography showing Oct-1 binding. Right, Oct-1 bands quantitated by densitometry. Results are from 1 of 2 separate experiments with similar results. B, NIH3T3 cells were cultured for 3 hours with LPS (5 μg/ml) and PMA (10 ng/ml). Nuclear extracts from the cells (10 μg) were then incubated with gel shift loading buffer in the absence or presence of inhibitors at the indicated concentrations at room temperature for 10 minutes, and 32P-labeled Oct-1 consensus oligonucleotide was added and incubated at room temperature for another 20 minutes, and EMSAs were then performed. The bands were quantitated by densitometry. Left, Representative autoradiography showing Oct-1 binding. Right, Oct-1 bands quantitated by densitometry. Results are from 1 of 2 separate experiments with similar results. N = negative control (without nuclear extract); C = 100-fold excess of unlabeled Oct-1 oligonucleotide; M = 100-fold excess of unlabeled mutant Oct-1 oligonucleotide. See Figure 4 for other definitions.

Figure 8.

Results of Oct-1 nuclear protein immunoblotting. Top, NIH3T3 cells were cultured for 3 hours with or without LPS (5 μg/ml) and phorbol 12-myristate 13-acetate (PMA; 10 ng/ml) and in the absence or presence of Tript or Dex at the indicated concentrations. Nuclear extracts (20 μg) were prepared and separated on 10% sodium dodecyl sulfate–polyacrylamide gels, followed by immunoblot analysis using the specific antibodies against Oct-1. Immunoblotting of actin was also performed on the same membrane after stripping. Bottom, Oct-1 bands quantitated by densitometry normalized by the intensity of actin, as described in Materials and Methods. Results are the mean and SEM of the ratio of Oct-1/actin from 3 separate experiments, all with similar results. See Figure 4 for other definitions.

DISCUSSION

The EA extract of TWHF has a potent antiinflammatory and immunosuppressive effect and has been used successfully for the treatment of rheumatic diseases. RA is a chronic autoimmune inflammatory disease characterized by persistent synovitis. There is a considerable body of evidence suggesting that NO is involved in the pathogenesis of a variety of inflammatory diseases, including RA (12–18), juvenile RA (33), ankylosing spondylitis (34), and inflammatory arthritis (15). One possibility, therefore, to explain the therapeutic effect of the EA extract on RA could relate to its ability to inhibit NO production. Consistent with this possibility, treatment of rats with the EA extract at doses equivalent to one-fifth to one-eighth of its LD50 was previously shown to suppress NO production by lining cells in the carrageenan-stimulated air pouch model of inflammation (8) and to inhibit the levels of urinary nitrite in rats with inflammatory arthritis (21). In these experiments, a decrease in NO generation was correlated with reduced inflammation. However, the mechanism by which the EA extract inhibited NO production in these models was not delineated. The current data indicate that this is likely to result from a direct inhibitory effect on iNOS transcription.

In the first experiments, the effect of the EA extract of TWHF on NO production in vivo was assessed using a model in which NO production by macrophages depends on activation of Th1 cells secreting interferon-γ (IFNγ) (35, 36). Although treatment with the EA extract significantly inhibited macrophage NO production in this model, it was possible that the effect was indirectly mediated through suppression of T cell activation. In this regard, the EA extract and triptolide are known to inhibit T cell activation as well as the production of IFNγ by T cells (5). In vitro experiments with elicited and directly stimulated macrophages and with a macrophage cell line were therefore carried out to document that the EA extract and triptolide could directly inhibit NO production by macrophages. Previously, we had found that the maximum blood concentration of triptolide in rats was 45–60 ng/ml after a single oral administration equivalent to one-fifth of the LD50 of the EA extract (data not shown). This range is similar to the concentration of triptolide found to inhibit in vitro NO production by the murine cell lines used in the current study. In RA patients, this range of triptolide concentrations may be reached since they are treated with multiple administrations (3 times per day) of the EA extract. A detailed analysis of the pharmacokinetics of triptolide that is currently in progress should provide necessary information to confirm this. The results of the in vitro experiments are consistent with the in vivo findings that treatment with the EA extract inhibited production of NO in models of nonimmune inflammation (8), and suggest that the EA extract may be a potential therapeutic inhibitor of NO synthesis in various pathologic conditions.

In the current study, murine macrophages were used to analyze the impact of the EA extract and triptolide on NO production, because human monocytes do not generate significant amounts of NO (data not shown). However, to delineate the inhibitory effect of the EA extract on NO production in detail, we assessed the effect of triptolide on the mitogen-induced activation of the human iNOS promoter. We reasoned that this might be the more relevant analysis to understand the potential actions of this agent in human patients. Some differences in human and murine iNOS promoter regulation have been reported. Chu and colleagues reported that the cytokine-response elements of the human iNOS promoter that contain 2 matched activator protein 1 (AP-1) sites are not present in the murine iNOS promoter region (37). However, the human iNOS promoter is active in murine cell lines in response to stimulation (38). The data indicate that triptolide inhibited human iNOS promoter activity. This paralleled the decrease in iNOS mRNA expression and NO production observed both in vivo and in vitro.

Sequence analysis has indicated that there are many binding sites for transcription factors, including NF-κB, AP-1, and Oct-1, in the human and mouse iNOS promoters (39–41). Previously, it has been reported that triptolide inhibits T cell interleukin-2 (IL-2) expression by reducing NF-κB–mediated transcriptional activation (42). We have also found that triptolide inhibited the transcriptional activities not only of NF-κB and the nuclear factor of activated T cells (NF-AT), but also of AP-1 (data not shown). Therefore, it was possible that inhibition of NF-κB, NF-AT, or AP-1 could play a role in suppressing iNOS transcription. It has been reported that mutation of the Oct-1 binding motif in the murine iNOS promoter completely inhibited iNOS transcription in BNL CL2 cells, whereas alteration of other binding sites had no effect on the promoter activity (27, 31). However, the role of Oct-1 in regulating the human iNOS promoter has not been analyzed in detail, and therefore the importance of Oct-1 nuclear binding in the regulation of human iNOS activity is unknown. Notably, however, comparison of murine and human iNOS promoters showed that the Oct-1 sequences are identical in the 2 species (38). Therefore, it is possible that regulation of the human iNOS promoter may be similar to that of the murine iNOS promoter, but this requires confirmation. Importantly, the current data show that elevated Oct-1 nuclear binding is correlated with increases in iNOS gene expression, suggesting that Oct-1 may play a role in human iNOS transcription. It is important to emphasize that no previous studies have examined the effects of triptolide on the activation of the Oct-1 transcription factor.

Our results showed that triptolide significantly inhibited the Oct-1 binding activity induced by the combination of LPS and PMA. Triptolide had no influence on the reaction of Oct-1 oligonucleotide with the nuclear extract from the cells induced by LPS and PMA in the absence of the inhibitor. These results suggest that triptolide inhibits the up-regulation of Oct-1 binding activity induced by LPS and PMA. The Oct-1 binding activity is regulated by protein modification such as phosphorylation or S-nitrosylation (31, 43). Our results show that triptolide did not reduce the abundance of Oct-1 protein in the nuclear extract, suggesting that it inhibited Oct-1 binding activity induced by LPS and PMA by blocking Oct-1 protein modification. Taken together, the current results clearly show that triptolide inhibits the up-regulation of Oct-1 binding activity, which could account for the inhibition of iNOS transcription.

NO synthesized by iNOS is known to be an important mediator of inflammation. An increase in NO production has been noted in patients with RA. Moreover, NO directly combines with the superoxide anion to form peroxynitrite. In addition, NO induces the production of cytokines, including IL-1 and tumor necrosis factor, and inflammatory products of the cyclooxygenase pathway in vitro, suggesting that it might mediate inflammation and joint destruction in RA (12–17). The present results indicate that the EA extract and triptolide inhibit NO production and iNOS mRNA expression in vivo and in vitro. Triptolide inhibition of iNOS promoter activity may result from suppression of up-regulation of Oct-1 binding activity, although an effect of triptolide on other transcription factors involved in regulation of iNOS expression is also possible. These data therefore suggest that the EA extract may exert its beneficial effect in RA by its ability to inhibit NO production by reducing iNOS transcription.

Acknowledgements

We thank Dr. Arnold S. Krist of the National Heart, Lung, and Blood Institute, NIH, for providing the Luciferase/iNOS promoter plasmid DNA.

Ancillary