Suppression of nitric oxide induction and pro-inflammatory cytokines by novel proteasome inhibitors in various experimental models
© Qureshi et al; licensee BioMed Central Ltd. 2011
Received: 26 August 2011
Accepted: 12 October 2011
Published: 12 October 2011
Inflammation has been implicated in a variety of diseases associated with ageing, including cancer, cardiovascular, and neurologic diseases. We have recently established that the proteasome is a pivotal regulator of inflammation, which modulates the induction of inflammatory mediators such as TNF-α, IL-1, IL-6, and nitric oxide (NO) in response to a variety of stimuli. The present study was undertaken to identify non-toxic proteasome inhibitors with the expectation that these compounds could potentially suppress the production of inflammatory mediators in ageing humans, thereby decreasing the risk of developing ageing related diseases. We evaluated the capacity of various proteasome inhibitors to suppress TNF-α, NO and gene suppression of TNF-α and iNOS mRNA, by LPS-stimulated macrophages from several sources. Further, we evaluated the mechanisms by which these agents suppress secretion of TNF-α, and NO production. Over the course of these studies, we measured the effects of various proteasome inhibitors on the RAW 264.7 cells, and peritoneal macrophages from four different strains of mice (C57BL/6, BALB/c, proteasome double subunits knockout LMP7/MECL-1-/-, and peroxisome proliferator-activated receptor-α-/- (PPAR-α-/-) knockout mice. We also directly measured the effect of these proteasome inhibitors on proteolytic activity of 20S rabbit muscle proteasomes.
There was significant reduction of chymotrypsin-like activity of the 20S rabbit muscle proteasomes with dexamethasone (31%), mevinolin (19%), δ-tocotrienol (28%), riboflavin (34%), and quercetin (45%; P< 0.05). Moreover, quercetin, riboflavin, and δ-tocotrienol also inhibited chymotrypsin-like, trypsin-like and post-glutamase activities in RAW 264.7 whole cells. These compounds also inhibited LPS-stimulated NO production and TNF-α secretion, blocked the degradation of P-IκB protein, and decreased activation of NF-κB, in RAW 264.7 cells. All proteasome inhibitors tested also significantly inhibited NO production (30% to 60% reduction) by LPS-induced thioglycolate-elicited peritoneal macrophages derived from all four strains of mice. All five compounds also suppressed LPS-induced TNF-α secretion by macrophages from C57BL/6 and BALB/c mice. TNF-α secretion, however, was not suppressed by any of the three proteasome inhibitors tested (δ-tocotrienol, riboflavin, and quercetin) with LPS-induced macrophages from LMP7/MECL-1-/- and PPAR-α-/- knockout mice. Results of gene expression studies for TNF-α and iNOS were generally consistent with results obtained for TNF-α protein and NO production observed with four strains of mice.
Results of the current study demonstrate that δ-tocotrienol, riboflavin, and quercetin inhibit NO production by LPS-stimulated macrophages of all four strains of mice, and TNF-α secretion only by LPS-stimulated macrophages of C57BL/6 and BALB/c mice. The mechanism for this inhibition appears to be decreased proteolytic degradation of P-IκB protein by the inhibited proteasome, resulting in decreased translocation of activated NF-κB to the nucleus, and depressed transcription of gene expression of TNF-α and iNOS. Further, these naturally-occurring proteasome inhibitors tested appear to be relatively potent inhibitors of multiple proteasome subunits in inflammatory proteasomes. Consequently, these agents could potentially suppress the production of inflammatory mediators in ageing humans, thereby decreasing the risk of developing a variety of ageing related diseases.
Modern industrialized societies are experiencing great increases in many age-related diseases such as diabetes, cardiovascular, neurodegenerative diseases, and certain types of cancer. Although numerous factors undoubtedly contribute to this trend, significant evidence implicates nitric oxide (NO), and inflammation, in the pathogenesis of several of these age-related diseases . A number of studies, using experimental animal models, have demonstrated that senescence is accompanied by increases in production of NO in response to a variety of microbial products. For example, lipopolysaccharide (LPS)-induced macrophages from 22 and 32 month old CBA/CA mice to produce approximately 5 fold and 15 fold more NO, respectively, than LPS-stimulated macrophages from young (2-month-old) CBA/CA mice . Through further exploration of innate inflammatory responses we have learned that the kinetics of NO production and TNF-α secretion differ in LPS-stimulated murine macrophages, that induction of these inflammatory products are regulated by two independent signaling pathways, and that cytoplasmic proteasomes are key regulators of LPS-induced inflammatory responses in macrophages [3–7].
We have recently reviewed the important role of proteasomes in inflammation and other macrophage functions, and hypothesized that inhibition of proteasome activity can suppress inflammatory responses that contribute to ageing . Many of our earlier experiments designed to delineate the role of proteasomes in innate inflammatory responses utilized lactacystin, a potent proteasome inhibitor . Lactacystin is a synthetic compound that contains a β-lactone moiety, which is responsible for lactacystin's capacity to block production of a number of pro-inflammatory cytokines by LPS-stimulated macrophages . Unfortunately, lactacystin is very expensive and toxic even at micromolar levels so, although it has been quite useful for in vitro experimentation, it is not suitable for clinical use . As reported recently, proteasomal activities are tightly regulated, and naturally-occurring compounds (γ-tocotrienol and δ-tocotrienol) are able to inhibit or activate these activities .
Consequently, we sought to identify other, non-toxic proteasome inhibitors with anti-inflammatory properties. Specifically, we have been evaluating a number of relatively inexpensive, commercially available naturally-occurring, synthetic, and FDA approved compounds for their capacity to inhibit proteasome activity, and the production of nitric oxide, certain pro-inflammatory cytokines (TNF-α, IL-1β, IL-6), and the iNOS enzyme. As part of this pursuit, we recently reported that two important inflammatory markers associated with ageing, TNF-α and NO, were effectively decreased in chickens whose diets were supplemented with a variety of naturally-occurring, synthetic or FDA approved compounds such as δ-tocotrienol, quercetin, riboflavin, (-) Corey lactone, amiloride, and dexamethasone . As described earlier, NO production increases during ageing process , which could be due to a diminished activation of NF-κB signaling [11, 12]. It was suggested that above mentioned compounds may also block the activation of NF-κB, thus resulting lowering of serum TNF-α and NO levels in chickens . The important role played by NF-κB in various biological functions has been described .
Materials and methods
Highly purified, deep rough chemotype LPS (Re LPS) from E. coli D31m4 was prepared as described . Dulbecco's Modified Eagle Medium (DMEM) heat-inactivated low-endotoxin fetal bovine serum (FBS), and gentamicin were purchased from Cambrex (Walkersville, MD, USA) for tissue culture studies. Thioglycolate was purchased from Sigma, Aldrich Chemical Co. (St. Louis, MO, USA) and RNeasy mini kit from QIAGEN Sciences (Germantown, MD, USA). RAW 264.7 cells (ATCC TIB 71) were purchased from American Type Culture Collection (Manassas, VA, USA). Mevinolin, pure α-tocopherol, (-) Corey lactone, amiloride-HCL, (-) riboflavin and dexamethasone were purchased from Sigma-Aldrich Chemical Co (St. Louis, MO, USA). Quercetin was purchased from Alfa Aesar (Johnson Matthey Co. Lancastor, UK). The 50% purified δ-tocotrienol fraction from annatto seeds was received as a gift from American River (Boston, MA, USA).
Purification of δ-tocotrienol from 50% purified fraction of annatto seeds
The δ-tocotrienol was purified from 50% purified fraction of annatto seeds, as described previously . The purity of δ-tocotrienol was established by high pressure liquid chromatography (HPLC) against its standard, as reported earlier .
Cell culture and maintenance
The RAW 264.7 cells or mouse thioglycolate-elicited peritoneal macrophages were maintained in DMEM supplemented with 10% heat inactivated FBS and 10 mg/mL gentamicin at 37°C in a humidified atmosphere with 5% CO2, as described previously . Cells were cultured in 6-well plates as described in the Tables or legends to the Figures.
The 6-week-old female, C57BL/6, Wild Type (WT) and BALB/c, were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). 6-week-old female proteasome double subunits knockouts (LMP7/MECL-1-/-) mice were obtained from Dr. John J Monaco (Department of Molecular Genetics, University of Cincinnati Medical Center, Cincinnati, OH, USA) and peroxisome proliferator-activated receptor-α-/- (PPAR-α-/-) were bred at UMKC's Animal Facility (Kansas City, MO, USA).
Mice used in this study received humane care in compliance with the principles of laboratory animal care formulated by the National Society of Health Guide for the "Care and Use of Laboratory Animals" by the US National Society of Health (NIH Publication No 85-23, revised 1996).
The experimental procedures involving animals were reviewed and approved by the "Institutional Animal Care and Use Committee of UMKC", Medical School, MO. All 6-week-old female mice (n = 20) were acclimatized to the new environment for 14 days before beginning experimentation. The mice were fed ad libitum regular commercial mouse diet and had free access to water throughout the experiment. A 12 h light and 12 h dark cycle was maintained during feeding period.
Chymotrypsin-like activity of 20S rabbit muscle proteasomes
Comparative effectiveness of different doses (5 μM-320 μM) of dexamethasone, mevinolin, α-tocopherol, δ-tocotrienol, riboflavin, quercetin-HCL, amiloride-HCL, and (-) Corey lactone on the suppression of chymotrypsin-like activity of 20S rabbit muscle proteasomes was evaluated. For these studies, different concentrations of each compound were prepared in media containing 0.5% DMSO. The proteasomal activities of the 20S rabbit muscle proteasomes (0.4 μg/mL) were assayed with synthetic peptide substrate in 0.02 M Tris-HCl buffer (pH 7.2). The substrate used for the chymotrypsin-like activity was 100 μM of succinyl-Leu-Leu-Val-Tyr-amino methyl coumarin. Fluorescence was measured (absorption at 360 nm and emission at 460 nm) using an Flx 800 microplate fluorescence reader (Bio-Tek Instruments, Winooski, VT, USA).
Effects of quercetin, riboflavin, and δ-tocotrienol (concentrations of 5 μM-40 μM) after 60 min treatment of different proteasomal activities (chymotrypsin-like, trypsin-like, post-glutamase) in RAW 264.7 whole cells
The comparative inhibitory effect of quercetin, riboflavin, and δ-tocotrienol on the chymotrypsin-like, trypsin-like, and post-glutamase activities of proteasome using RAW 264.7 whole cells were carried out essentially as reported recently . Briefly RAW 264.7 cells (10 × 103 cells/100 μL/well) were added in white plates (96-well, Fisher, 0877126), followed by the addition of various concentrations of quercetin, riboflavin, or δ-tocotrienol (5, 10, 20, or 40 μM in 100 μL; dissolved in 0.4% dimethyl sulfoxide (DMSO). The mixtures were incubated at 37°C in an incubator at 5% CO2 for 60 min. After incubation period, the cells in the 96-well plates were taken out 20 min prior to the addition of Caspase-Glo reagent (brought to room temperature before addition to the wells). Caspase-Glo reagent (100 μL) was added to each well to a total volume of 200 μL/well (tris buffer, pH 7.5; 0.02 M). The plate were covered with a plate sealer and incubated at room temp for 30 min. The relative luminescence units (RLU) of assays were read in a Promega Plate Luminometer. The chymotrypsin-like, trypsin-like, or post-glutamase activities were quantitated by measuring luminescence after stimulation of RAW 264.7 whole cells with various concentration of each compound in a Luminometer (Promega, Madison, Wisconsin USA), according to the directions of manufacturer.
TNF-α secretion and NO production by LPS-induced RAW 264.7 cells and LPS-induced thioglycolate-elicited peritoneal macrophages of four strains of mice
LPS-induced RAW 264.7 cells
RAW 264.7 cells (1 × 106 cells/500 μL/well) were adhered for 2 h in the wells. After 2 h cells were treated with dexamethasone, mevinolin (positive controls), α-tocopherol, δ-tocotrienol, riboflavin, or quercetin (100 μL; dissolved in 0.5% DMSO) for 1 h (pre-treatment). One hour later all wells were challenged with LPS (10 ng/well; 400 μL) or medium and incubated at 37°C in 5% CO2 for 4 h (TNF-α) or 36 h (NO). After incubation, the supernatants were collected and stored at -20°C.
The levels of TNF-α in supernatants were measured by Quantikine M ELISA kit (R&D System, Minneapolis, MN, USA) according to manufacturer's instructions. The lower limit of detection for TNF-a in this method is approximately, 5.0 pg/mL [4, 7].
The levels of NO were determined by measuring the amount of nitrite, a stable metabolic product of nitric oxide, as previously reported . The assay mixture contained medium (100 μL) plus Griess reagent (100 μL), and absorption was measured at 570 nm using a "Microplate Reader" (MR 5000; Dynatech Labs, Inc. USA). The amount of nitrite was determined by comparison of unknowns using a NaNO2 standard curve. The NO detection limit was 0.20 nM .
LPS-induced thioglycolate-elicited peritoneal macrophages
The procedures described above are identical to those utilized for peritoneal macrophages except that thioglycolate-elicited peritoneal macrophages were adhered to the bottom of 100 mm tissue culture plates (1 × 107 cells/well in 1.0 mL media) for 4 h, the supernatants were removed, and cells were washed extensively with medium three times. The cells were cultured overnight in fresh media after the final wash. After overnight incubation at 37°C, the cells were treated with various proteasome inhibitors and LPS as described above. Viability of peritoneal macrophages treated with various inhibitors plus LPS were also determined by trypan blue dye exclusion or a quantitative colorimetric assay with 3-(4,5)-dimethylthiozol-2,5-diphenyl-tetrazolium bromide (MTT) as described previously [4, 7].
Effects of various compounds on the inhibition of NF-κB in LPS-stimulated RAW 264.7 cells
RAW 264.7 cells were pretreated with dexamethasone, mevinolin, α-tocopherol, δ-tocotrienol, riboflavin, or quercetin for 1 h, then treated with LPS (10 ng/well; 400 μL) for 4 h. Nuclear protein was extracted and the NF-κB activation was measured using "electronic mobility shift assay" (EMSA) according to the reported procedure . Nuclear protein was extracted following the recommendation of the manufacturer (QIAGEN Sciences; Germantown, MD, USA) . NF-κB was measured in nuclear extracts with the respective ELISA-based commercial kits (NF-κB p65). Nuclear protein (5 μg), was added to each well coated with an oligonucleotide containing the consensus binding site for NF-κB incubated for I h. Activated NF-κB was detected after 1 h with primary antibody, i.e. Anti-NF-κB, which specifically recognizes an epitope (p65) accessible only when the factor is activated and bound to its target DNA. A secondary anti-IgG horseradish peroxidase conjugate allows detection of the activated NF-κB by a colorimetric reaction. Absorption was read within 5 min at 450 nm with a reference wavelength of 655 nm .
Degradation of P-IκB, protein in LPS-stimulated RAW 264.7 cells (Western blots analyses)
RAW 264.7 cells (1 × 106 cells/500 μL/well) were treated with dexamethasone, mevinolin, α-tocopherol, δ-tocotrienol, riboflavin, or quercetin for 1 h, then stimulated with LPS (10 ng/well; 400 μL) for 36 h. Macrophages were washed with phosphate-buffered saline, and cytoplasmic extracts were prepared using cell extraction buffer (Biosource, Camarillo, California, USA) supplemented with a protease inhibitor cocktail, containing phenylmethylsulfonyl fluoride and phosphatase inhibitors, according to the manufacturer's directions . Protein concentrations were measured with BCA protein assay kits, and Western blots analyses were used to measure IκB-α, (Santa Cruz). Each well of the gel was loaded with 40 μg of protein and gels were electrophoresed at a constant 150 V 1 × Tris glycine buffer for 50 minutes. Proteins in gels were transferred into the "Immobilon Transfer Membranes" (IPVH 15150; Millipore, Bedford, Mass, USA) using the semidry transfer cell and, after appropriate antibody treatments, the bands were visualized with an enhanced chemiluminescence detection kit (Pierce) as described previously [7, 20].
Expression of TNF-α, and iNOS genes (Southern blots analyses)
All proteasome inhibitors (α-tocopherol, δ-tocotrienol, riboflavin, or quercetin) were dissolved in media containing 0.2% DMSO. Thioglycolate-elicited peritoneal macrophages were prepared from 8-week-old mice as described previously (7,17). The macrophages (1 × 107) were adhered for 2 h in the wells, with α-tocopherol (100 μM), δ-tocotrienol (10 μM), riboflavin (40 μM), or quercetin (40 μM) for 2 h. Then all the wells were challenged with LPS (10 ng/well; 400 μL), and incubated at room temperature for 4 h.
After 4 h, assay mixtures were centrifuged at 2,000 rpm for 20 min. Cells were harvested, and total cellular RNA was extracted from each pellet with RNeasy mini kit (QIAGEN Sciences; Germantown, MD, USA) according to the instructions of the manufacturer. The RNA of each treatment was transcribed and resulting data was amplified and analyzed by real-time polymerase chain reaction (RT-PCR) to quantitate gene expression of TNF-α, and iNOS by using 1-step RT-PCR kit (QIAGEN, Chatsworth, CA, USA; Southern blots analyses) according to the manufacturer's instructions [6, 7, 21].
Detection of cell viability
Viability of peritoneal macrophages treated with and without LPS plus dexamethasone, mevinolin, α,-tocopherol, δ-tocotrienol, riboflavin, and quercetin-HCL was determined by trypan blue dye exclusion or a quantitative colorimetric assay with 3-(4,5)-dimethyl-thiozol-2, 5-diphenyltetrazolium bromide (MTT) as reported [4, 7].
Stat View software (version 4.01, Abacus Concepts, Berkeley, CA) was used for the analyses of treatment-mediated effects as compared to control group. Treatment-mediated differences were detected with a two-way ANOVA, and when the F test indicated a significant effect, differences between the means were analyzed by a Fisher's protected least significant difference test. Data were reported as means ± SD in text and Tables. The statistical significance level was set at 5% (P< 0.05).
Effects of various compounds on the chymotrypsin activity of 20S rabbit muscle proteasomes
Effects of various compounds on the chymotrypsin-like activity of 20S rabbit muscle proteasomes1
(-) Cor Lact
20S Prot (20S P)
Tris + 20S P = A
A + DMSO3 = B
B + 5 μM
B + 10 μM
B + 20 μM
B + 40 μM
B + 80 μM
B + 160 μM
B + 320 μM
Effect of quercetin, riboflavin, and δ-tocotrienol (concentrations of 5 μM-40 μM) in RAW 264.7 cells on proteasomal activities (chymotrypsin, trypsin, post-glutamase) for a duration of 1 h1
Avg RLU value
Avg RLU value
Avg RLU value
1. Media and Cells
2. DMSO Control3
Effects of various proteasome inhibitors on production of TNF-α and NO by LPS-stimulated RAW 264.7 cells
We also tested the ability of proteasome inhibitors to suppress TNF-α secretion by LPS-stimulated macrophages under conditions in which macrophages were treated simultaneously with proteasome inhibitors and LPS. We found that the degree of inhibition of TNF-α secretion by simultaneous treatment with LPS and quercetin (74%), dexamethasone (66%), riboflavin (65%), δ-tocotrienol (16%), or mevinolin (11%) (data not presented) was comparable to the level of inhibition attained when macrophages were pre-treated with proteasome inhibitors. Consequently, further studies were carried out only under pre-treatment condition.
Effects of various proteasome inhibitors on the inhibition of NF-κB in LPS-stimulated RAW 264.7 cells
Effects of various proteasome inhibitors on levels of P-IκB in LPS-stimulated RAW 264.7 cells
Effects of various proteasome inhibitors on the secretion of TNF-α by LPS-stimulated peritoneal macrophages from BALB/c mice
Effects of various compounds on the secretion of TNF-α (pg/mL) in LPS-stimulated (pre-treatment) thioglycolate-elicited peritoneal macrophages of 8-week-old BALB/c female mice1
Secretion of TNF-α (pg/mL)
Media + Cells = A
A + LPS (10 ng/mL) = B
B + 0.2% DMSO2 = C
C + 5 μM
C + 10 μM
C + 20 μM
C + 40 μM
C + 80 μM
C + 160 μM
C + 320 μM
C + 640 μM
Effects of various proteasome inhibitors on secretion of TNF-α secretion and NO production by LPS-stimulated peritoneal macrophages from C57BL/6 versus BALB/c mice
All the inhibitors, except α-tocopherol, inhibited LPS-induced secretion of TNF-α by macrophages from C57BL/6 mice. Dexamethasone was the most potent inhibitor, yielded a 68% decrease in TNF-α secretion (P< 0.02), whereas, quercetin (37%), riboflavin (31%), mevinolin (24%), and δ-tocotrienol (20%) resulted in only moderate, but significant decreases (P< 0.05), compared to controls (Figure 7A). Similarly, the results with macrophages derived from BALB/c mice paralleled those of macrophages from C57BL/6 mice; decreases in the levels of TNF-α were observed with dexamethasone (72%; P< 0.02), quercetin (28%), riboflavin (25%), mevinolin (17%), and δ-tocotrienol (15%), compared to controls (Figure 7B). The extent to which TNF-α secretion by LPS-stimulated macrophages was blocked by these various proteasome inhibitors was slightly greater for C57BL/6 mice than for BALB/c mice (~ 20%), though these differences were not statistically significant (Figure 7B).
Production of NO by LPS-stimulated macrophages from C57BL/6 and BALB/c mice was significantly (P< 0.02) reduced by dexamethasone (60% and 48%, respectively), δ-tocotrienol (46% and 41%, respectively), and quercetin (51% and 36%, respectively), compared to respective controls (Figure 8A, B). However, mevinolin (27% and 22%) and riboflavin (23% and 25%) produced only moderate reductions in NO production compared to controls (Figure 8B). As with TNF-α the extent to which NO production by LPS-stimulated macrophages was blocked by these various proteasome inhibitors was slightly greater for C57BL/6 mice than for BALB/c mice (~ 20%), though these differences were not statistically significant (Figure 8B). Thus, these studies clearly demonstrate that the proteasome inhibitors tested suppress secretion of TNF-α and production of NO by LPS-stimulated thioglycolate-elicited peritoneal macrophages derived from C57BL/6 and BALB/c mice.
Comparative effects of various proteasome inhibitors on TNF-α secretion and NO production by LPS-stimulated thioglycolate-elicited peritoneal macrophages from C57BL/6, BALB/c, double knockout LMP7/MECL-1-/-, and PPAR-α-/- knockout mice
The next series of experiments were conducted with macrophages from mice that have aberrant responses to LPS. It has previously been well documented that macrophages from PPAR-α-/- mice produce an unusually robust and inflammatory response to LPS [14, 15]. Similarly, we have been studying the effect of role of various proteasome subunit knockouts on the inflammatory response to LPS [22, 23]. Of particular relevance, we reported that LPS-stimulated peritoneal macrophages derived from double knockout LMP7/MECL-1-/- mice generated a relatively normal TNF-α response, but a markedly reduced NO response, compared to control (C57BL/6) mice . Consequently, we thought it would be of value to determine the effect of proteasome inhibitors on TNF-α, and NO production by LPS-stimulated macrophages from these mouse strains. The experiments were carried out only with three lead, naturally-occurring non-toxic, commercially available compounds (δ-tocotrienol, riboflavin, and quercetin), which inhibit the secretion of TNF-α and production of NO, which are increased during ageing process . Therefore, the anti-inflammatory properties of δ-tocotrienol (10 μM), riboflavin (40 μM), quercetin (40 μM), and α-tocopherol (100 μM; as a control)) were explored in LPS-induced thioglycolate-elicited peritoneal macrophages derived from knockout LMP7/MECL-1-/- and PPAR-α-/- mice using identical conditions for this series of experiments as described in the previous paragraph for the measurement of TNF-α (4 h) or nitrite (36 h) in culture supernatants, respectively, after LPS-stimulation with the expectation that these experiments would provide further insight into the mechanisms by which these proteasome inhibitors suppress inflammatory responses; macrophages from C57BL/6 and BALB/c mice were used as controls.
We previously reported that NO production by LPS-stimulated peritoneal macrophages was markedly reduced in LMP7/MECL-1-/- knockout mice; TNF-α production, in contrast, was not markedly affected by the LMP7/MECL-1-/- genotype . We also demonstrated that in order to suppress TNF-α, production by LPS stimulated macrophages with proteasome inhibitors, both chymotrypsin- and trypsin-like proteasome activities must be suppressed . Thus, the capacity of δ-tocotrienol, riboflavin, and quercetin to block TNF-α production in C57BL/6 and BALB/c, but not in LMP7/MECL-1-/-, would be explained if these agents inhibited primarily the LMP2, LMP7 and MECL-1 subunits of mouse immunoproteasomes, with comparatively lower suppressive effects on X, Y, and Z subunits of constitutively expressed proteasomes. LPS-stimulation of macrophages from C57BL/6 and BALB/c would induce production of immunoproteasomes in which X, Y and Z subunits were partially replaced by LMP2, LMP7 and MECL-1-/-, decreased TNF-α production would occur if these latter subunits were potently suppressed by δ-tocotrienol, riboflavin, and quercetin. In contrast, the × and Z components of LPS-stimulated macrophages from LMP7/MECL-1-/- mice, could not be replaced by LMP7 and MECL-1-/-. If the chymotrypsin-like or trypsin-like activities of × and Z proteasome subunits were comparatively resistant to the inhibitory effects of δ-tocotrienol, riboflavin, and quercetin, TNF-α production by LPS-stimulated macrophages from LMP7/MECL-1-/- mice would be unaffected, since inhibition of both chymotrypsin- and trypsin-like proteasomal activities are required to suppress TNF-α secretion.
The capacity of δ-tocotrienol, riboflavin, and quercetin to inhibit NO production by LPS-stimulated macrophages from LMP7/MECL-1-/- mice might be explained by the high sensitivity of NO production to modulation of immunoproteasome subunits [22, 23]. Thus, we previously demonstrated that LPS-stimulated macrophages from LMP2-/- knockout mice produce substantially less NO than control littermates, and concluded from those experiments that inducible immunoproteasome subunits play a critical role in NO production . Thus, NO production by LPS-stimulated macrophages from LMP7/MECL-1-/- mice is very low, therefore IFN-γ has to be added concurrently to observe NO production and the defect is reversed. The present inhibitors suppress the LPS-induced NO production in the presence of IFN-γ (Figure 10A, B)
Effect of various proteasome inhibitors on gene expression of TNF-α and iNOS by LPS-stimulated thioglycolate-elicited peritoneal macrophages from C57BL/6, BALB/c, double knockout LMP7/MECL-1-/-, and PPAR-α-/- knockout mice
The experiments described above demonstrated that δ-tocotrienol, riboflavin, and quercetin inhibited secretion of TNF-α by LPS-stimulated thioglycolate-elicited peritoneal macrophages from C57BL/6 and BALB/c, but not LMP7/MECL-1-/- or PPAR-α-/- knockout mice (Figures 9A, B). In contrast, these compounds all inhibited NO production by LPS-stimulated thioglycolate-elicited peritoneal macrophages from all four strains of mice (Figure 10A, B). In order to determine whether these changes resulted from alterations in transcription of the relevant genes, we measured the effect of various proteasome inhibitors on mRNA levels for TNF-α and iNOS in LPS-stimulated thioglycolate-elicited peritoneal macrophages from 8-week-old female C57BL/6, BALB/c, double knockout LMP7/MECL-1-/-, and PPAR-α-/- mice. The concentrations of each inhibitor and condition were similar to those used in earlier experiments. Cells were treated with δ-tocotrienol (10 μM), riboflavin (40 μM), and quercetin (40 μM) for 1 h, then LPS (10 μL of 1.0 μg/mL) was added to each well and incubated for an additional 4 h. Total cellular RNA was then extracted and reverse-transcribed, and gene analyses were carried out by RT-PCR and Southern blots analyses.
The primary objectives in the present study were to further evaluate several naturally-occurring proteasome inhibitors for their capacity to suppress inflammatory processes, and to define mechanisms responsible for these anti-inflammatory effects. Anti-inflammatory properties of the proteasome inhibitors were evaluated in macrophages derived from several sources (e.g. the RAW 264.7 cell line, and thioglycolate-elicited peritoneal macrophages from four strains of mice (C57BL/6, BALB/c, double knockout LMP7/MECL-1-/- and PPAR-α-/-). As a result of these studies we have identified several naturally-occurring proteasome inhibitors that could potentially decrease levels of inflammatory cytokines and NO that may contribute to the development of diseases associated with ageing.
First, we demonstrated that dexamethasone, mevinolin, δ-tocotrienol, riboflavin and quercetin are potent inhibitors of chymotrypsin-like activity of 20S rabbit muscle proteasomes, and that (-) Corey lactone and amiloride enhanced this activity. We also demonstrated that dexamethasone, mevinolin, δ-tocotrienol, riboflavin and quercetin inhibited the secretion of TNF-α and NO production by LPS-stimulated RAW 264.7 macrophage like murine cell cultures. Further, levels of NF-κB within the nucleus were decreased, whereas cellular P-IκB levels were increased, by pre-treatment of LPS-stimulated RAW 264.7 cells with dexamethasone, mevinolin, δ-tocotrienol, riboflavin and quercetin.
NF-κB is maintained in an inactive state in the cytoplasm of cells when it is bound to IκB. LPS induces a series of events that results in phosphorylation and ubiquitination of IκB with subsequent degradation by the proteasome. These actions result in NF-κB activation, translocation to the nucleus, and increased transcription of several genes encoding pro-inflammatory cytokines. Consequently, the capacity of dexamethasone, mevinolin, δ-tocotrienol, riboflavin and quercetin to inhibit proteasome activity in conjunction with their capacity to increase cellular levels of P-IκB and decrease nuclear translocation of NF-κB, suggests that the mechanism by which these agents suppress production of TNF-α, and NO involves decreased degradation of ubiquinated P-IκB by the proteasome, resulting in depressed translocation of NF-κB to the nucleus. Therefore, ultimately, these proteasome inhibitors suppress production of TNF-α, and NO, and exert their anti-inflammatory effects by inhibiting NF-κB, activation.
This conclusion was further supported by experimental testing of these inhibitors (dexamethasone, mevinolin, δ-tocotrienol, riboflavin and quercetin) in LPS-stimulated thioglycolate-elicited peritoneal macrophages derived from four different strains of mice. All inhibitors significantly inhibited LPS-induced secretion of TNF-α by macrophages derived from C57BL/6 and BALB/c mice (Figure 9A). Although, macrophages derived from C57BL/6 mice compared to BALB/c mice yielded slightly better (20%) inhibition in LPS-stimulated secretion of TNF-α (Figure 9A). In marked contrast to the results attained with C57BL/6 and BALB/c mice, TNF-α, secretion was essentially unaffected by treatment of LPS-stimulated macrophages derived from LMP7/MECL-1-/- knockout mice with δ-tocotrienol, riboflavin, and quercetin (9B). Similarly, these compounds failed to suppress TNF-α secretion by LPS-stimulated macrophages derived from PPAR-α-/- knockout mice (since activation of PPAR-α-/- mice normally reduced inflammation); δ-tocotrienol and riboflavin treatment actually enhanced TNF-α secretion (Figure 9B). In marked contrast to the observations with TNF-α, δ-tocotrienol, riboflavin, and quercetin suppressed NO production by LPS-stimulated macrophages from C57BL/6, BALB/c, LMP7/MECL-1-/-, and PPAR-α-/- knockout mice (Figure 10A, B). The effects of these proteasome inhibitors on mRNA production by LPS-stimulated macrophages were generally consistent with at the protein levels of TNF-α, secretion and NO production.
Our previously published studies strongly support the concept that proteasomes are key regulators of LPS-stimulated inflammatory signaling pathways (3,4,6-8). Proteolytic activity of proteasomes is mediated by the 20S proteasomes, a hollow, cylindrical multi-protein complex consisting of three proteolytic subunits, X, Y, and Z, with chymotrypsin-like, trypsin-like, and post-glutamase activities, respectively. A variety of inflammatory stimuli induce alterations in newly assembled "immuno-proteasomes" in which X, Y and Z subunits are partially replaced by LMP7, LMP2, and MECL-1, respectively.
We previously demonstrated that low doses of lactacystin, which suppresses primarily chymotrypsin-like activity of the proteasome, potently suppressed production of NO, but not TNF-α by LPS stimulated macrophages. We further demonstrated that in order to suppress TNF-α secretion by LPS-stimulated macrophages with proteasome inhibitors, both chymotrypsin-like and trypsin-like proteasome activities must be suppressed (22). Subsequent experiments with proteasome subunit knockout (LMP7-/-, LMP2-/-, MECL-1-/-, and LMP7/MECL-1-/-) revealed that NO production by LPS-stimulated peritoneal macrophages was markedly reduced in LMP7-/-, LMP2-/-, MECL-1-/-, and LMP7/MECL-1-/-knockout mice; TNF-α production, in contrast, was not markedly affected by any of these knockout genotypes .
In the current study, the capacity of the investigated proteasome inhibitors to inhibit TNF-α secretion by LPS-stimulated macrophages from several sources (i.e. the RAW 264.7 cell line, and peritoneal macrophages from C57BL/6 and BALB/c mice) supports the conclusion that these inhibitors suppress both chymotrypsin-like and trypsin-like activities of the proteasomes, since both of these activities must be suppressed in order to inhibit TNF-α, secretion. These conclusions were supported by our analysis of the effects of these inhibitors on proteolytic activity of the proteasome.
The capacity of δ-tocotrienol, riboflavin, and quercetin to block TNF-α secretion in C57BL/6 and BALB/c, but not in LMP7/MECL-1-/- mice, is an intriguing result. LPS-stimulation of macrophages from C57BL/6 and BALB/c mice would be expected to induce production of immunoproteasomes in which X, Y and Z subunits were partially replaced by LMP2, LMP7 and MECL-1-/-. The × and Z components of LPS-stimulated macrophages from LMP7/MECL-1-/- mice, however, cannot be replaced by LMP7 and MECL-1-/-. The macrophages from knockout mice when induced produce robust amounts of TNF-α. Thus, the differential capacity of δ-tocotrienol, riboflavin, and quercetin to inhibit TNF-α secretion by LPS-stimulated macrophages from C57BL/6 and BALB/c vs LMP7/MECL-1-/- knockout mice would appear to be attributable to a differential susceptibility of × and Z vs LMP7 and MECL-1 proteasomal subunits to inhibition by these inhibitors.
Presumably δ-tocotrienol, riboflavin, and quercetin are potent inhibitors of LMP2, LMP7 and MECL-1-/- subunits of mouse immunoproteasomes, with comparatively lesser suppressive effects on X, Y, and Z subunits of constitutively expressed proteasomes. Thus, the chymotrypsin-like and trypsin-like activities of normal immunoproteasomes (containing LMP2, LMP7 and MECL-1-/-) from C57BL/6 and BALB/c mice would be suppressed by δ-tocotrienol, and quercetin, resulting in decreased TNF-α secretion. In contrast, the chymotrypsin-like and trypsin-like activity of proteasomes (containing X, Y, Z, and LMP2) from LMP7/MECL-1-/- knockout mice would not be suppressed by δ-tocotrienol, riboflavin, and quercetin, resulting in normal TNF-α secretion.
The finding that δ-tocotrienol, riboflavin, and quercetin inhibited NO production by LPS-stimulated macrophages from C57BL/6, BALB/c, LMP7/MECL-1-/-, and PPAR-α-/- mice was also quite interesting; the LMP7/MECL-1-/- genotype mice did not affect susceptibility to inhibition by these compounds. These macrophages from knockout mice do not induce very much NO, therefore IFN-γ had to be added along with LPS to induce NO in cells that have X, Y, Z, and LMP2 subunits .
PPAR-α-/- knockout mice have exaggerated inflammatory responses to a variety of stimuli, because activation of PPAR-α leads to anti-inflammatory effects . The mechanisms leading to these exaggerated inflammatory responses are not clearly understood, but are believed to be at least partially attributable to increased NFκB activity [24, 25]. Consequently, one would expect TNF-α secretion by LPS-stimulated macrophages from PPAR-α-/- knockout mice to be highly up-regulated with respect to TNF-α secretion and relatively resistant to inhibition by proteasome inhibitors that degrade IκB, and decrease NF-κB activity. Therefore, we found that δ-tocotrienol, riboflavin, and quercetin failed to suppress TNF-α, secretion by LPS-stimulated macrophages from PPAR-α-/- knockout mice. In fact TNF-α secretion was substantially enhanced by riboflavin and δ-tocotrienol. A clearer explanation of these results will be dependent on further elucidation of interactions between PPAR-α, and ubiquitin proteasome pathways .
Our present results demonstrate that δ-tocotrienol, riboflavin, and quercetin are naturally-occurring potent proteasome inhibitors for the inhibition of NO production tested in vitro. These results also confirms earlier report that γ-tocotrienol blocked LPS-stimulated activation of NF-κB, and also blocked TNF-α induced phosphorylation and degradation of IκBα, through the inhibition of IκBα kinase activation . Tocotrienols have been shown to modestly inhibit or activate the proteasomal activity depending on its concentrations (Tables 1, ). Therefore, blocking the proteasomal activity with low doses of tocotrienols could potentially reduce inflammatory responses, but at high doses of tocotrienols may cause apoptotic cell death in cancer [9, 28]. Quercetin, on the other hand is a natural proteasome inhibitor that can affect several proteasomal activities.
The present study describes the possible mechanism of inhibition of NO production (released during ageing processes or acute and chronic disease) by naturally-occurring proteasome inhibitors (quercetin, δ-tocotrienol, and riboflavin). Dexamethasone, mevinolin, δ-tocotrienol, riboflavin and quercetin were all found to be potent inhibitors of chymotrypsin-like activity of 20S rabbit muscle proteasomes. δ-Tocotrienol, riboflavin, and quercetin also inhibited chymotrypsin-like, trypsin-like and post-glutamase activities of the proteasomes in RAW 264.7 cells. These compounds also blocked LPS-stimulated secretion of TNF-α, NO production, activation of NF-κB, and degradation of P-IκB in RAW 264.7 murine macrophages. Similarly, these compounds suppressed TNF-α secretion and NO production by LPS-stimulated peritoneal macrophages derived from C57BL/6 and BALB/c mice. δ-Tocotrienol, riboflavin, and quercetin blocked the LPS-stimulated production of NO and had no effect on the secretion of TNF-α in macrophages derived from LMP7/MECL-1-/- and PPAR-α-/- knockout mice. These results indicate that δ-tocotrienol, riboflavin, and quercetin treatments differentially inhibit the secretion of TNF-α of LPS-stimulated macrophages derived from C57BL/6, BALB/c versus LMP7/MECL-1-/- and PPAR-α-/- knockout mice. Moreover, all the gene expression results of TNF-α, and iNOS genes are generally consistent with the results of protein levels of secretion of TNF-α, and production of NO observed with either C57BL/6, BALB/c, LMP7/MECL-1-/-, or PPAR-α-/- mice. All these results indicate that the production of NO may be modulated via NF-κB pathway, and δ-tocotrienol acts as a proteasome inhibitor at lower doses and as activator at higher doses (Table 1). These collective properties of δ-tocotrienol, riboflavin, and quercetin can be referred to as "novel proteasome inhibitors", which might be used to suppress the production of inflammatory mediators in ageing humans, thereby decreasing the risk of developing a variety of age-related diseases that appear to be at least partially attributable to aberrant control of inflammatory responses with increasing age.
aDepartment of Basic Medical Sciences, University of Missouri-Kansas City, 2411 Holmes Street, Kansas City, MO 64108, USA.
bDepartment of Medicine, University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA.
cDivision of Pharmacology and Toxicology, School of Pharmacy, University of Missouri-Kansas City, 2464 Charlotte Street, Kansas City, MO 64108, USA.
tumor necrosis factor-α
nitric oxide synthase
peroxisome proliferator-activated receptor-α-/-
- 1. LPS (control group):
media + cells + LPS (10 ng/mL) + 0.2%-0.4% dimethyl sulfoxide
- 2. Dexa:
- 3. Mev:
- 4. Toco:
- 5. Trie:
- 6. Ribo:
- 7. Quer:
We thank Ann N Thomas (Department of Microbiology, University of Kansas, Kansas City, KS 66160) for her technical help and replicating analyses of gene expression assays. We also thank Mr. Keith Glichrist (USDA, ARS, MWA, Cereals and Crops Research Laboratory, Madison, WI, 53726, USA) for carrying out statistical analyses of this data. This study was supported in part, by NIH grants GM-50870, AI- 54962, AI57168 (NQ) and AI-18797 and AI-44936 (SNV). The study was carried out under a FDA approved IND number 36906.
- Knott AB, Bossy-Wetzel E: Impact of nitric oxide on metabolism in health and age-related disease. Diabetes Obesity & Metabolism. 2010, 12 (2): 126-133.View ArticleGoogle Scholar
- Chen LC, Pace JL, Russell SW, Morrison DC: Altered regulation of inducible nitric oxide synthase expression in macrophages from senescent mice. Infection and Immunity. 1996, 64 (10): 4288-4298.PubMed CentralPubMedGoogle Scholar
- Shen J, Reis J, Morrison DC, Papasian CJ, Sreekumar R, Kolbert C, Qureshi AA, Vogel SN, Qureshi N: Key Inflammatory signaling pathways are regulated by the proteasome. Shock. 2006, 25: 472-484. 10.1097/01.shk.0000209554.46704.64View ArticlePubMedGoogle Scholar
- Qureshi N, Perera PY, Splitter G, Morrison DC, Vogel SN: The Proteasome as a LPS-binding protein in macrophages: Toxic lipopolysaccharide activates the proteasome complex. J Immunol. 2003, 171: 1515-1525.View ArticlePubMedGoogle Scholar
- Suckerman SH, Qureshi N: In vivo inhibition of lipopolysaccharides-induced lethality and tumor necrosis factor synthesis by Rhodobactor sphaeroides diphosphoryl lipid A is dependent on corticosterone induction. Infect Immun. 1992, 60: 2581-2587.Google Scholar
- Reis J, Tan X, Yang R, Qureshi AA, Papasian CJ, Vogel SN, Morrison DC, Qureshi N: A combination of proteasome inhibitors and antibiotics prevents lethality in a septic shock model. Innate Immunity. 2008, 14 (5): 319-329. 10.1177/1753425908096855PubMed CentralView ArticlePubMedGoogle Scholar
- Shen J, Gao JJ, Zhang G, Tan X, Morrison DC, Papasian CJ, Vogel SN, Qureshi N: Proteasome inhibitor, lactacystin blocks CpG DNA- and peptidoglycan induced inflammatory genes, cytokines and mitogen-activated protein kinases in macrophages. Shock. 2006, 25: 594-599. 10.1097/01.shk.0000209555.46704.2dView ArticlePubMedGoogle Scholar
- Qureshi N, Vogel SN, Way CV, Papasian CJ, Qureshi AA, Morrison DC: The proteasome: A central regulator of inflammation and macrophage function. Immunological Research. 2005, 31 (3): 243-260. 10.1385/IR:31:3:243.View ArticleGoogle Scholar
- Qureshi AA, Reis JC, Papasion CJ, Morrison DC, Qureshi N: Tocotrienols inhibit the lipopolysacharides-induced pro-inflammatory cytokines in macrophages of female mice. Lipids in Health and Diseases. 2010, 9: 143-10.1186/1476-511X-9-143. 10.1186/1476-511X-9-143View ArticleGoogle Scholar
- Qureshi AA, Reis JC, Qureshi N, Papasion CJ, Morrison DC, Schaefer DM: δ-Tocotrienol and quercetin reduce serum levels of nitric oxide and lipid parameters in female chickens. Lipids in Health and diseases. 2011, 10: 39-10.1186/1476-511X-10-39. 10.1186/1476-511X-10-39View ArticleGoogle Scholar
- Chung HY, Cesari M, Anton S, Marzetti E, Giovannini S, Seo AY, Carter C, Yu BP, Leewenburgh C: Molecular inflammation: underpinnings of ageing and age-related diseases. Ageing Res Rev. 2009, 8: 18-30. 10.1016/j.arr.2008.07.002PubMed CentralView ArticlePubMedGoogle Scholar
- Nair MP, Mahajan S, Reynolds JL, Aalinkeel R, Nair H, Schwartz SA, Kandaswami C: The flavonoid quercetin inhibits pro-inflammatry cytokine (tumor necrosis factor alpha) gene expression in normal peripheral blood mononuclear cells via modulation of NF-κB system. Clinical and Vaccine Immunology. 2006, 13 (3): 319-328. 10.1128/CVI.13.3.319-328.2006PubMed CentralView ArticlePubMedGoogle Scholar
- Nam N: Naturally-occurring NF-κB inhibitors. Mini-Reviews in Medicinal Chemistry. 2006, 6: 945-951. 10.2174/138955706777934937View ArticlePubMedGoogle Scholar
- Shulman AI, Mangelsdorf DJ: Retinoid × receptor heterodimers in the metabolic syndrome. N Endl J Med. 2005, 353 (6): 604-615. 10.1056/NEJMra043590.View ArticleGoogle Scholar
- Chao C, Youssef J, Rezaiekhaleigh M, Birnbaum LS, Badr M: Senescence-associated decline in hepatic peroxisomal enzyme activities with diminished levels of retinoid × receptor alpha, but not peroxisome proliferator-activated receptor alpha. Mechanisms of ageing and development. 2002, 123: 1469-1476. 10.1016/S0047-6374(02)00086-6View ArticlePubMedGoogle Scholar
- Qureshi AA, Mo H, Packer L, Peterson DM: Isolation and structural identification of novel tocotrienols from rice bran with hypocholesterolemic, antioxidant and antitumor properties. J Agr Food Chem. 2000, 48: 3130-3140. 10.1021/jf000099t.View ArticleGoogle Scholar
- Stuehr DJ, Nathan CF: Nitric oxide: a macrophage product responsible for cytostasis and respiratory inhibition in tumor target cells. J Exp Med. 1989, 169: 1543-1555. 10.1084/jem.169.5.1543View ArticlePubMedGoogle Scholar
- Abraham E: NF-kappa B activation. Crit Carte Med. 2000, 28: N100-N104. 10.1097/00003246-200004001-00012.View ArticleGoogle Scholar
- Yubero S, Ramudo L, Manso MA, Dios ID: Mechanism of dexamethasone-mediated chemokine down-regulation in mild and severe acute pancreatitis. Biochimica et Biophysica Acta. 2009, 1792: 1205-1211.View ArticlePubMedGoogle Scholar
- Haddad JJ, Land SC: Amiloride blockades lipopolysaccharide-induced pro-inflammatory cytokine biosynthesis in an IκB/NF-κB-dependent mechanism. Evidence for the amplification of an anti-inflammatory pathway in the alveolar epithelium. Am J Respir Cell Mol Biol. 2002, 26: 114-126.View ArticlePubMedGoogle Scholar
- Chomczynski P, Saachi N: Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Ana Biochem. 1987, 162: 156-159.View ArticleGoogle Scholar
- Reis J, Guam XQ, Kisselev AF, Papasian CJ, Qureshi AA, Morrison DC, Van Way CW, Vogel SN, Qureshi N: LPS-induced formation of immunoproteasome: TNF-α and nitric oxide production are regulated by altered composition of proteasome-active sites. Cell Biochemistry and Biophysics. 2011, 60: 77-88. 10.1007/s12013-011-9182-8PubMed CentralView ArticlePubMedGoogle Scholar
- Reis J, Hassan F, Guam XQ, Shen J, Monaco JJ, Papasian CJ, Qureshi AA, Van Way CW, Vogel SN, Morrison DC, Qureshi N: The immunoproteasome regulate LPS-induced TRIF/TRAM signaling pathway in murine macrophages. Cell Biochemistry and Biophysics. 2011, 60: 119-126. 10.1007/s12013-011-9183-7PubMed CentralView ArticlePubMedGoogle Scholar
- Babaev VR, Ishiguro H, Ding L, Yancey PG, Dove DE, Kovacs WJ, Semenkovich CF, Fazio S, Linton MF: Macrophage expression of peroxisome proliferator-activated receptor-a reduces atherosclerosis in low-density lipoprotein receptor-deficient mice. Circulation. 2007, 116 (12): 1404-1412. 10.1161/CIRCULATIONAHA.106.684704View ArticlePubMedGoogle Scholar
- Bensinger SJ, Tontonoz P: Integration of metabolism and inflammation by lipid-activated nuclear receptors. Nature. 2008, 454 (7203): 470-477. 10.1038/nature07202View ArticlePubMedGoogle Scholar
- Genini D, Catapano CV: Control of peroxisome proliferator-activated receptor fate by the ubiquitin proteasome system. Journal of Receptors & Signal Transduction. 2006, 26 (5-6): 679-692. 10.1080/10799890600928202View ArticleGoogle Scholar
- Ahn KS, Sethi G, Krishnan K, Aggarwal BB: γ-Tocotrienol inhibits nuclear-κB signaling pathway through inhibition of receptor- interacting protein and TAK1 leading to suppression of antiapoptotic gene products and potentiation of apoptosis. The J Biol Chem. 2007, 282 (1): 809-820.View ArticlePubMedGoogle Scholar
- Lin kI, Baraban JM, Ratan RR: Inhibition versus induction of apoptosis by proteasome inhibitors depend on concentration. Cell Death and Differentiation. 1998, 5: 577-583. 10.1038/sj.cdd.4400384View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.