Inflamm. Res. (2010) 59:711–721 DOI 10.1007/s00011-010-0182-8
Inflammation Research
ORIGINAL RESEARCH PAPER
Flavonoid combinations cause synergistic inhibition of proinflammatory mediator secretion from lipopolysaccharide-induced RAW 264.7 cells Omar A. Harasstani • Saidi Moin • Chau Ling Tham • Choi Yi Liew • Norazren Ismail • Revathee Rajajendram • Hanis H. Harith • Zainul A. Zakaria Azam S. Mohamad • Mohamad R. Sulaiman • Daud A. Israf
•
Received: 23 November 2009 / Revised: 15 January 2010 / Accepted: 17 February 2010 / Published online: 11 March 2010 Ó Springer Basel AG 2010
Abstract Objectives We evaluated several flavonoid combinations for synergy in the inhibition of proinflammatory mediator synthesis in the RAW 264.7 cellular model of inflammation. Methods The inhibitory effect of chrysin, kaempferol, morin, silibinin, quercetin, diosmin and hesperidin upon nitric oxide (NO), prostaglandin E2 (PGE2) and tumour necrosis factor-a (TNF-a) secretion from the LPS-induced RAW 264.7 monocytic macrophage was assessed and IC50 values obtained. Flavonoids that showed reasonable inhibitory effects in at least two out of the three assays were combined in a series of fixed IC50 ratios and reassessed for inhibition of NO, PGE2 and TNF-a. Dose– response curves were generated and interactions were analysed using isobolographic analysis. Results The experiments showed that only chrysin, kaempferol, morin, and silibinin were potent enough to produce dose–response effects upon at least two out of the three mediators assayed. Combinations of these four flavonoids showed that several combinations afforded highly significant synergistic effects. Conclusions Some flavonoids are synergistic in their antiinflammatory effects when combined. In particular chrysin and kaempferol significantly synergised in their inhibitory effect upon NO, PGE2 and TNF-a secretion. These findings
Responsible Editor: J. Skotnicki. O. A. Harasstani S. Moin C. L. Tham C. Y. Liew N. Ismail R. Rajajendram H. H. Harith Z. A. Zakaria A. S. Mohamad M. R. Sulaiman D. A. Israf (&) Department of Biomedical Science, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia e-mail:
[email protected];
[email protected]
open further avenues of research into combinatorial therapeutics of inflammatory-related diseases and the pharmacology of flavonoid synergy. Keywords Flavonoid Synergy Isobologram RAW 264.7 Nitric oxide Prostaglandin E2 Tumour necrosis factor-a Abbreviations NO Nitric oxide PGE2 Prostaglandin E2 TNF-a Tumour necrosis factor-a LPS Lipopolysaccharide IC50 Inhibitory concentration 50 DMSO Dimethyl sulfoxide DMEM Dulbecco’s modified eagle media MTT 3-[4, 5-Dimethyl-2-thiazolyl]-2,5-diphenyl tetrazolium bromide FBS Foetal bovine serum L-NAME N-nitro-L-arginine methyl ester MRSA Methicillin-resistant Staphylococcus aureus MIC Minimal inhibitory concentration EGCG Epigallocatechin gallate NSAID Non-steroidal anti-inflammatory drug MAPK Mitogen-activated protein kinase IRF-1 Interferon regulatory factor-1 ERK Extracellular-regulated kinase JNK c-Jun N-terminal kinase LDL Low density lipoprotein
Introduction The flavonoids comprise a very large group of compounds found in many herbal preparations and have been shown to
712
possess potent anti-inflammatory effects in both cellular and animal models of inflammation [1]. Herbal practitioners have always believed that effective phytomedicines on the drug market acquire their therapeutic efficacy via synergistic interactions between the components of individual or mixtures of herbs. The use of drug combinations is not confined to herbal products alone, and, for example, cancer chemotherapy, the treatment of HIV, neurological disorders and hypertension routinely employ drug combinations consisting of two or more substances [2]. The combinatorial use of drugs aims to induce a response upon multiple targets, multiple subpopulations, or multiple diseases simultaneously. Furthermore, combinatorial use of multiple drugs with dissimilar mechanisms or modes of action may also direct the effect against a single target or disease with a more effective therapeutic outcome. The possible favourable outcomes for synergism include (1) increased therapeutic efficacy, (2) decreased dosage and toxicity while increasing or maintaining the same efficacy, (3) reduction of the development of drug resistance, and (4) provision of selective synergism against target (efficacy synergism) versus host (toxicity antagonism). Indeed drug combinations have been widely used and have become the leading choice for treating chronic debilitating diseases [3]. Most pharmacological studies with anti-inflammatory flavonoids have described drug-like effects in cellular in vitro systems. Studies in man have evaluated flavonoids for prophylaxis of cancer [4], cardiovascular [5], gastrointestinal [6], and neurological [7] disease. These studies are mostly concerned with effect of dietary flavonoid consumption and prevention of disease. However, it is interesting to note that a micronized purified flavonoid fraction containing 90% diosmin and 10% hesperidin TM (Daflon 500 ) has been used successfully for the therapeutic management of chronic venous disease and haemorrhoidal disease [8]. Very few studies have shown that flavonoids can act synergistically when combined with other drugs. For instance, some flavones have weak antibacterial effects on methicillin-resistant Staphylococcus aureus (MRSA), but at sub-MIC concentrations they greatly increase the susceptibility of MRSA to b-lactam antibiotics [9]. In the area of cancer therapeutics silibinin, a flavanone, has shown synergistic antiproliferative effects upon normal and drugresistant ovarian and breast tumour cell lines when combined with cisplatin and doxorubicin [10]. A study on LDL oxidation showed that rutin, a quercetin-3-rutinoside, caused a synergistic prevention of LDL oxidation when combined with ascorbate and c-terpinene [11]. Several studies have described effects of flavonoids upon the central nervous system [12–14]. Several flavonoids possess a selective and relatively mild affinity for the central benzodiazepine
O. A. Harasstani et al.
binding site in the GABAA receptor. Furthermore, synergy studies have shown that several flavonoids such as methylapigenin and hesperidin cause synergistic anxiolytic effects when combined with diazepam [15]. To our knowledge, isobolographic analysis of synergistic effects of flavonoid combinations upon proinflammatory mediator secretion have not been reported. Although one study [16] showed that genistein and epigallocatechin gallate (EGCG) enhanced effects of several NSAIDs, the analysis of the data did not employ appropriate isobolographic methodology. In this communication we show that the potency of several flavonoids can be enhanced following combinatorial treatment of monocytic macrophage cells induced to synthesize excess proinflammatory mediators. In addition to the objective of enhancing potency while reducing toxicity, we also took into account the cost of the flavonoids used when devising our objectives. The cost is usually determined by the complexity of the synthetic procedure and cost of starting material. We realized that potent anti-inflammatory flavonoids such as apigenin, luteolin, and EGCG are extremely costly and therefore we focused our approach upon flavonoids that may exert equivalent, if not better potencies when combined, with reduced costs. This communication describes an isobolographic approach to the study of combinatorial treatment of antiinflammatory flavonoids in cellular assays of inflammation. We demonstrate several flavonoid combinations that cause a significantly high degree of synergy in their inhibition of proinflammatory mediator secretion. These findings strengthen claims by herbalists pertaining to the synergism of components of herbal cocktails to induce a therapeutic effect and the rationale for multimodal therapeutics in experimental and clinical medicine. Additionally, these findings open further avenues of research into combinatorial therapeutics of inflammatory-related diseases and the pharmacology of flavonoid synergy.
Methods Reagents and kits Lipopolyssacharide (LPS from Escherichia coli serotype 055:B5), NS-398, sulfanilamide, naphthylenediamine, dexamethasone, 3-[4,5-Dimethyl-2-thiazolyl]-2,5-diphenyl tetrazolium bromide (MTT) and N-nitro-L-arginine methyl ester (L-NAME), were purchased from Sigma-Aldrich (St Louis, MO, USA). Antibiotics (5,000 U/ml penicillin and 5,000 lg/ml streptomycin), foetal bovine serum (FBS) and Dulbecco’s Modified Eagle Medium (DMEM) were purchased from Hyclone (UT, USA). TNF-a immunoassay kit was purchased from BD Pharmingen (San Diego, CA,
Flavonoid combinations cause synergistic inhibition of proinflammatory mediators Fig. 1 Chemical structure of the flavonoids used
713
OH
HO
HO
O
O
OH OH
OH
O
O
Chrysin
Kaempferol O OH HO
O
O O
OH
OH OH
Silibinin
O
HO
OH
HO
O
HO
OH
O
OH
OH OH
OH
O
OH
O
Morin
Quercetin
OCH 3 OH
O
O O
CH 3
O
O
OH
OH OH
OH OH OH
OH
O
Hesperidin OH
O
OCH 3
O O
CH 3
O
O
OH
OH OH
OH OH OH
OH
O
Diosmin
USA). Prostaglandin E2 enzyme immunoassay kit was purchased from Cayman Chemicals, USA. Dimethyl sulfoxide (DMSO) was purchased from Amresco (Solon, OH, USA). Flavonoids Chrysin, kaempferol, morin, silibinin, hesperidin, diosmin, and quercetin were purchased from Sigma-Aldrich (St Louis, MO, USA). Figure 1 shows the structure of these
flavonoids. These compounds were dissolved in 100% DMSO as a stock of 0.2 M and diluted to appropriate concentrations for assays. Assays with single compounds were performed with concentrations ranging from 0 to 200 lM. For combinatorial experiments concentration ranges varied depending on the IC50 of each compound. The final concentration of DMSO in all assays was kept constant at 0.1%. This concentration allows for the solubilization of the compounds in aqueous solution without toxic effects upon cells.
714
Cell culture and induction of mediator secretion The murine monocytic macrophage cell line (RAW 264.7) was purchased from American Type Culture Collection (Manassas, VA, USA). RAW 264.7 cells were maintained in DMEM supplemented with 10% FBS, 100 U/mL penicillin, 100 lg/mL streptomycin, 4.5 g/L glucose, 1 mM sodium pyruvate, 2 mM L-glutamine, and 10 mM HEPES. Cells were maintained at 37°C, with 5% CO2 in fully humidified air and split 3 times a week. Cells at a confluency of 80–90% were scraped out and centrifuged at 120 g at 4°C for 7 min. The concentration was adjusted to (1 9 106 cells/mL) and cell viability was always more than 90%, as determined by Trypan blue dye exclusion. A total of 50 lL of cell suspension was dispensed into wells of a tissue-culture-grade 96-well microplate (1 9 105 cells/ well) and incubated for 2 h at 37°C, 5% CO2 to attach the cells. Blank wells with media alone were included. After 2 h, unattached cells were gently discarded. Attached cells were then induced with 10 lg/mL of LPS in the presence or absence of the compounds at a final volume of 100 lL/ well. Untreated cells and drug controls were stimulated with LPS and also had the same amount of DMSO in culture medium. Cells were then incubated for 18 h at 37°C in 5% CO2. Cell viability Cell viability was assessed following the removal of spent media by addition of MTT. Briefly, 100 lL of DMEM containing 5% FBS was added to each well followed by 20 lL of MTT (5 mg/mL in PBS). The formazan crystals were dissolved with 100 lL of 100% DMSO per well after 3 h. The absorbance was measured at k 570 nm with a microplate reader (UVM 340, ASYS Hitech GmbH, Austria) by using a reference wavelength of 650 nm. Cell viability was determined as the percentage of untreated induced cells. Determination of nitrite secretion The concentration of nitrite (NO2-) in spent media was determined by the Griess reaction. Briefly, an equal volume of the Griess reagent (1% sulfanilamide/0.1% naphtylethyenediamine dihydrochloride in 2.5% H3PO4) was mixed with spent media and colour development was assessed at k 550 nm with a microplate reader (UVM 340, ASYS Hitech GmbH, Austria). Fresh culture medium was used as the blank in all the experiments. The concentration of nitrite in the samples was calculated from a sodium nitrite standard curve (0–100 lmol/L). L-NAME was used at 250 lM as a drug control. Dose–response curves and IC50 values were calculated using GraphPad Prism version 5.00 for
O. A. Harasstani et al.
Windows (GraphPad Software, San Diego, California, USA, http://www.graphpad.com). All experiments were performed in triplicate. Determination of PGE2 secretion Spent media was collected and stored at -30°C prior to assay. PGE2 concentrations were determined with an enzyme immunoassay kit (Cayman Chemicals, Ann Arbor, MI, USA; Cat no: 514010) according to the manufacturer’s instructions. NS-398 was used at 50 lM as a drug control. Dose–response curves and IC50 values were calculated using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, California, USA, http:// www.graphpad.com). All experiments were performed in triplicate. Determination of TNF-a secretion Spent media was collected and stored at -30°C prior to determination of cytokine concentrations. TNF-a concentrations were assayed with an enzyme immunoassay kit (BD Pharmingen, San Diego, CA, USA, Cat No. 555268) according to the manufacturer’s instructions. Dexamethasone was used at 10 lM as a drug control. Dose–response curves and IC50 values were calculated using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, California, USA, http://www.graphpad.com). All experiments were performed in triplicate. Isobolographic analysis The interaction of the various flavonoids was evaluated by the simultaneous administration of fixed proportions of flavonoids, and performing an isobolographic analysis for the different combinations, as described by [17]. The isobologram was constructed by connecting the IC50 of the corresponding flavonoids plotted on the abscissa and ordinate to obtain the additivity line. For each flavonoid mixture, the IC50 and its associated 95% confidence intervals were determined by linear regression analysis of the log dose–response curve (five experiments in triplicate at each of nine doses) and compared by a t-test to a theoretical additive IC50 obtained from the calculation: IC50 add = IC50 flavo1/(P1 ? RP2), where R is the potency ratio of the flavonoid 1 alone, P1 is the proportion of flavonoid 1 and P2 is the proportion of flavonoid 2 in the total mixture. In the present study, fixed-ratio proportions were selected by first combining the IC50 of each compound and then constructing a dose–response curve in which IC50 fractions (following twofold serial dilutions) of flavonoid combinations were used; in the equation above, IC50 add is the total dose and the variance of IC50 add was calculated from the
Flavonoid combinations cause synergistic inhibition of proinflammatory mediators
715
Table 1 50% inhibitory concentration (IC50) of the various flavonoids tested for the inhibition of NO, PGE2, and TNF-a biosynthesis in LPSinduced RAW 264.7 cells Subclass
Name
IC50 (lM) NO
Flavonol Flavone
Flavanone
PGE2
TNF-a
Kaempferol
9.83 ± 1.55
16.66 ± 3.50
58.87 ± 1.61
Morin
44.85 ± 1.05
17.47 ± 0.64
nd
Quercetin Chrysin
35.90 ± 1.24 7.50 ± 1.84
nd 4.75 ± 1.25
nd 120.90 ± 16.12
Diosmin
nd
nd
nd
Silibinin
48.01 ± 2.42
5.98 ± 0.91
44.84 ± 1.09
Hesperidin
nd
5.00 ± 1.08
nd
Assays were conducted in triplicate as described in Methods and repeated on five different occasions The data represents the mean ± SEM of five independent experiments nd not determined (none of the tested doses exceeded 50% inhibition)
fraction of the IC50’s in the combination as: Var IC50 add = (fraction)2 Var IC50 flavo1 ? (fraction)2 Var IC50 flavo2 [18]. From these variances, confidence limits are calculated and resolved according to the ratio of the individual drugs in the combination. Supra-additivity or synergism is determined as the effect of a flavonoid combination which is higher and statistically different (IC50 significantly lower) than the theoretical calculated equieffect of a flavonoid combination with the same proportions. When the flavonoid combination gives an experimental IC50 not statistically different from the theoretically calculated IC50, the combination has an additive effect and additivity means that each constituent contributes to the effect in accord with its own potency and the less potent flavonoid is acting as though it is merely a diluted form of the other [19]. Statistical analysis Results are presented as IC50 values with 95% confidence limits. Student’s t test for independent means was used to assess statistical significance between observed and theoretical IC50s.
Results Cytotoxicity All compounds (single and in combination) except for chrysin had no cytotoxic effect upon cells. Chrysin reduced viability to 72% at the highest dose tested (200 lM), however all other doses were not toxic. Generally cell viability was always above 90%.
Selection of flavonoids for combinatorial experiments In order to determine the doses to be used in combinatorial treatments, we assayed single flavonoids for inhibitory effects upon NO, PGE2, and TNF-a secretion. Table 1 summarizes the IC50 values of each flavonoid following dose–response analysis. All flavonoids produced dose– response curves, however, several were weak and failed to cause inhibitory effects of more than 50% in some assays. In particular, hesperidin, diosmin and quercetin were the weakest and the IC50 could not be calculated in several assays. Flavonoids that were poor inhibitors of two or more mediators were not included in combinatorial experiments. Therefore, we selected four of the seven flavonoids, namely chrysin, kaempferol, morin, and silibinin for combination studies. Synergy and additivity upon NO inhibition Figure 2 shows the various dose response curves generated following treatment of induced RAW 264.7 cells with combinations of flavonoids. The most potent response was observed when kaempferol was mixed with chrysin. This combination reduced the IC50 to 2.2 lM. Other combinations reduced the IC50 except for morin/silibinin, in which the combined IC50 of 53.7 lM was close to the IC50 of both compounds (44.85 and 48.01 lM, respectively). Isobolograms are shown in Fig 3. All combinations except morin/ silibinin showed a significant (P \ 0.001) synergistic effect. Morin/silibinin was additive. Synergy and additivity upon PGE2 inhibition Figure 4 shows the various dose response curves generated following treatment of induced RAW 264.7 cells with
120
IC 50 = 2.27 ± 1.01 µM
100
R2 = 0.88
Inhibition (%)
Fig. 2 Nitric oxide inhibitory dose–response curves following treatment of induced RAW 264.7 cells with various flavonoid combinations. Spent media was assayed in triplicate following 18 h treatments. The data represents the mean ± SEM of five independent experiments
O. A. Harasstani et al.
Inhibition (%)
716
80 60 40
120
IC 50 = 5.25 ± 0.87 µM
100
R2 = 0.84
80 60 40
20
20
0
0 -1.0 -0.5
0.0
0.5
1.0
1.5
2.0
0.0
2.5
1.0
120
IC 50 = 7.68 ± 1.07 µM
120
R2 = 0.87
IC 50 = 17.86 ± 1.66 µM
100
100
R2 = 0.88
80 60 40 20
2.5
80 60 40
0.5
1.0
1.5
2.0
2.5
0.0
Log [Kaempferol+Silibinin] IC 50 = 53.70 ± 1.17 µM R
2
= 0.88
80 60 40
IC 50 = 15.01 ± 1.20 µM
100
R2 = 0.86
2.0
2.5
3.0
Log [Morin+Silibinin]
combinations of flavonoids. Again, the most potent response was observed when kaempferol was mixed with chrysin. This combination reduced the IC50 to 2.28 lM. Other combinations also reduced the IC50 in comparison to single compound treatments except for morin/silibinin and silibinin/chrysin. The combined IC50 of morin/silibinin was 14.19 lM and was between individual IC50s of 17.47 and 5.98 lM respectively. The combined IC50 of silibinin/ chrysin was 7.1 lM which was slightly higher than that of both compounds when used singularly (5.98 and 4.75 lM, respectively). Isobolograms are shown in Fig 5. All combinations except morin/silibinin and silibinin/chrysin showed a significant synergistic effect. However, morin/ silibinin and silibinin/chrysin where not antagonistic. Synergy and additivity upon TNF-a inhibition Figure 6 shows the various dose response curves generated following treatment of induced RAW 264.7 cells with
2.5
40
0 1.5
2.0
60
0 1.0
1.5
80
20
0.5
1.0
120
20
0.0
0.5
Log [Chrysin+Morin]
Inhibition (%)
100
2.0
0 0.0
120
1.5
20
0
Inhibition (%)
0.5
Log [Kaempferol+Morin]
Inhibition (%)
Inhibition (%)
Log [Kaempferol+Chrysin]
0.0
0.5
1.0
1.5
2.0
2.5
Log [Chrysin+Silibinin]
combinations of flavonoids. Since morin was a weak inhibitor of TNF-a secretion, in which the IC50 could not be determined, this compound was not included in these experiments. Combinations with kaempferol were potent and reduced the IC50 far below the IC50 values of the compounds when used alone. However, chrysin/silibinin were only additive in their effect. Isobolograms are shown in Fig 7. Combinations with kaempferol showed a significant synergistic effect, however, chrysin/silibinin were only additive.
Discussion Investigations into the effect of dietary flavonoids upon inflammation have demonstrated significant anti-inflammatory effects due to the inhibition of major proinflammatory mediators. We notice that the majority of reports that show inhibitory effects of flavonoids in similar
Flavonoid combinations cause synergistic inhibition of proinflammatory mediators
10.00 ± 1.69 µM 2.27 ± 1.01 µM P < 0.02
10
5
60
Morin IC50 (µM)
15
Chrysin IC 50 (µM)
Fig. 3 Isobolograms of various flavonoid combinations used to inhibit NO secretion. Theoretical (open square) and observed (filled square) IC50 values and degree of significance are shown on each isobologram. Comparisons were made with the Student’s t test
717
30.00 ± 1.77 µM 5.25 ± 0.87 µM P < 0.001
50
30
0
0 0
5
10
0
15
Kaempferol IC50 (µM) 30.00 ± 1.98 µM 7.68 ± 1.07 µM P < 0.001
50
30
15
30.00 ± 1.91 µM 17.86 ± 1.66 µM P < 0.01
10
5
0
0 0
5
10
0
15
60
15
Chrysin IC50 (µM)
50.00 ± 2.21 µM 53.70 ± 1.17 µM
50
30
50
60
Morin IC 50 (µM)
Kaempferol IC 50 (µM)
Silibinin IC 50 (µM)
10
15
Chrysin IC50 (µM)
Silibinin IC 50 (µM)
60
5
Kaempferol IC50 (µM)
30
30.00 ± 2.12 µM 15.01 ± 1.20 µM P < 0.005
10
5
0
0 0
30
50
Morin IC 50 (µM)
in vitro systems to ours, employed a pretreatment approach and others used different cell lines of similar or different species. Whether this approach was chosen to mimic preventive effects in complete biological systems is not clear, however, we have noticed that this approach has provided data that conflicts with ours. For instance, in human mononuclear leucocytes it was shown that silibinin inhibited PGE2 secretion with an IC50 of 43 lM [20] whereas in our assay the IC50 was 5.98 lM, a sevenfold difference. Studies on quercetin showed that it strongly inhibits NO and TNF-a secretion via inhibition of nuclear factor-jB (NF-jB) and mitogen-activated protein kinase (MAPK) pathways [21, 22]. Conversely, we did not find quercetin to be a good choice for combinatorial treatment since its inhibition of NO was moderate and failed to give a strong inhibition of TNF-a and PGE2 secretion and calculable IC50. We showed that morin and hesperidin produced
60
0
30
50
60
Silibinin IC 50 (µM)
reasonable inhibition of PGE2 in this report whereas [23] showed no inhibitory effect upon PGE2 secretion by rat peritoneal macrophages that received treatment prior to LPS induction, however, our IC50values for chrysin, kaempferol, and quercetin were not far off from theirs. Furthermore, kaempferol and morin were poor inhibitors of NO secretion from LPS-induced J774 cells [24] whereas we found both compounds to be strong inhibitors of NO. An example of variation due to cell line and/or stimulus is observed in the effect of both kaempferol and quercetin in which both compounds had no effect upon NF-jB translocation and DNA-binding activity in PMA-induced human mast cells [25] whereas both flavonoids inhibited NF-jB binding when applied to HUVEC lines induced by a mixture of cytokines [26]. It is obviously clear that cell type, species and assay system employed significantly affect the inhibition of
718
100
120
IC 50 = 2.28 ± 1.72 µM R2 = 0.96
Inhibition (%)
120
Inhibition (%)
Fig. 4 Prostaglandin E2 inhibitory dose–response curves following treatment of induced RAW 264.7 cells with various flavonoid combinations. Spent media was assayed in triplicate following 18 h treatments. The data represents the mean ± SEM of five independent experiments
O. A. Harasstani et al.
80 60 40
100
60 40 20
0
0 0
1
R2 = 0.91
80
20
-1
IC 50 = 5.13 ± 1.32 µM
2
-1
Log [Kaempferol+Chrysin]
100
120
IC 50 = 5.14 ± 3.63 µM R2 = 0.97
Inhibition (%)
Inhibition (%)
120
80 60 40
100
IC 50 = 5.46 ± 0.91 µM R2 = 0.85
40
0
0 1
-1
2
120
IC 50 = 14.19 ± 10.80 µM R2 = 0.68
Inhibition (%)
Inhibition (%)
100 80 60 40
100
0 1
proinflammatory mediators by flavonoid compounds. Many flavonoids act via inhibition of major proinflammatory signaling pathways and therefore pretreatment of cells will block signal transduction prior to the addition of inflammatory stimuli. It is important that studies aimed at evaluating potential therapeutic use of flavonoids should either treat cells simultaneously or following application of the inflammatory stimuli. This is critical since certain flavonoids with potent preventive activity may be weak inhibitors of mediator synthesis/secretion when used therapeutically. The spectrum of synergism over the assay systems employed varied between mediators assayed but in general showed that kaempferol/chrysin combinations were highly synergistic in all assays. Morin/chrysin and morin/kaempferol combinations were also strongly synergistic in their inhibition of PGE2. The explanation as to
2
IC 50 = 7.04 ± 7.68 µM R2 = 0.87
40
0 0
2
60
20
Log [Morin+Silibinin]
1
80
20
-1
0
Log [Morin+Chrysin]
Log [Kaempferol+Silibinin] 120
2
60
20
0
1
80
20
-1
0
Log [Kaempferol+Morin]
-1
0
1
2
Log [Silibinin+Chrysin]
why these synergisms occur remains obscure, however, existing information regarding the precise action of these flavonoids allows for speculation and generates more questions for further work. It is possible that several combinations may act upon different pathways which converge to activate mediator synthesis. For instance, in the case of chrysin/kaempferol the lack of inhibitory effect of chrysin upon NF-jB activation [27] may have been taken over by kaempferol [28], furthermore chrysin inhibits interferon regulatory factor 1 (IRF-1) [27] and thus may act to reinforce a stronger inhibitory response. Additionally kaempferol strongly inhibits extracellularregulated kinase 1/2 (ERK 1/2) and c-Jun NH2-terminal kinase (JNK) [7] which could further enhance the inhibitory effect. Indeed similar possibilities may occur in other combinations albeit via varying disruptive mechanisms.
Flavonoid combinations cause synergistic inhibition of proinflammatory mediators 8
12.00 2.28
6
20 18
2.38 M 1.72 M P < 0.05
Morin IC50 ( M)
Chrysin IC 50 ( M)
Fig. 5 Isobolograms of various flavonoid combinations used to inhibit PGE2. Theoretical (open square) and observed (filled square) IC50 values and degree of significance are shown on each isobologram. Comparisons were made with the Student’s t test
719
4 2
18.00 5.13
15 10 5
0
0 0
5
10
15
18 20
25
0
5
Kaempferol IC 50 ( M)
15
18 20
25
8
12.00 5.14
6
2.20 M 3.63 M
Chrysin IC 50 ( M)
Silibinin IC 50 ( M)
10
Kaempferol IC 50 ( M)
8
4 2 0
12.00 5.46
6
0.95 M 0.91 M P < 0.01
4 2 0
0
5
10
15
18 20
25
0
5
8
10
15
18 20
25
Morin IC 50 ( M)
Kaempferol IC 50 ( M)
12.00 0.78 M 14.19 10.80 M
10
Silibinin IC 50 ( M)
Silibinin IC 50 ( M)
2.07 M 1.32 M P < 0.01
6 4 2
6.00 7.04
8
1.08 M 7.68 M
6 4 2
0 0
5
10
15
18 20
Morin IC 50 ( M)
In some cases, it can be argued as to why two compounds with similar modes of action, albeit with different degrees of potency, synergise when combined. Could yet unidentified targets be able to clarify this enigma? Indeed, in the case of the role of antioxidants (including several flavonoids) upon disruption of the NFjB, pathway it has been suggested that as NF-jB pathways are more defined and more potential points of regulation described, that many antioxidants purported to inhibit NF-jB due to effects on reactive oxygen species (ROS) actually may have novel targets that are not a direct antioxidant effect. Rather, they may chemically modify proteins or activate kinases or phosphatases that modulate inhibitors of jB kinase (IKK) activity [29]. Similar speculations can be projected to other proinflammatory pathways such as MAPK, signal transducer and activator of transcription (STAT) and phosphoinositide-3-kinase (PI3K). Other questions that arise include:
25
2
4
6
8
10
Chrysin IC 50 ( M)
do the flavonoids work on a common target in a more intense manner? Do the compounds interact chemically to form a new ligand? Is there a common denominator in the subclasses that promotes synergy upon combination? These are questions that need be addressed by the employment of medicinal chemistry and molecular pharmacology tools. Since we have only evaluated synergy based on secreted mediators we can only speculate that the effects are due to combined inhibition of complex biosynthetic pathways, however, it should be noted that cellular secretory mechanisms may also be affected. In fact, we have not found any work that has described the effects of flavonoids on the secretory mechanism of proinflammatory mediators. It is possible that flavonoids not only inhibit mediator synthesis but also secretory mechanisms when combined. Experiments are underway to address this possibility.
O. A. Harasstani et al. 120
IC 50 = 20.53 ± 2.03 µM
100
R2 = 0.94
150
Chrysin IC 50 (µM)
Inhibition (%)
720
80 60 40 20
90.00 ± 8.86 µM 20.53 ± 2.03 µM P < 0.005
120 100
50
0 0
0.0
0.5
1.0
1.5
2.0
0
2.5
IC50 = 101.30 ± 12.31 µM
100
R2 = 0.85
80 60 40 20 0 0.0
0.5
1.0
1.5
2.0
100
50
0
R
20
40
50
60
Silibinin IC 50 (µM) 60
IC 50 = 12.67 ± 1.15 µM 2
80
0
2.5
= 0.88
Silibinin IC 50 (µM)
Inhibition (%)
100
60
85.55 ± 8.55 µM 101.30 ± 12.31 µM
120
Log [Chrysin+Silibinin] 120
40
150
Chrysin IC 50 (µM)
Inhibition (%)
120
20
Kaempferol IC 50 (µM)
Log [Kaempferol+Chrysin]
80 60 40 20
55.00 ± 1.35 µM 12.67 ± 1.15 µM P < 0.001
50 40
20
0 0.0
0.5
1.0
1.5
2.0
2.5
0 0
Log [Kaempferol+Silibinin] Fig. 6 Tumour necrosis factor-a inhibitory dose–response curves following treatment of induced RAW 264.7 cells with various flavonoid combinations. Spent media was assayed in triplicate following 18 h treatments. The data represents the mean ± SEM of five independent experiments
In conclusion, we have demonstrated that certain flavonoids do indeed act in a drug-like fashion in which combinations can yield significant synergistic effects in the inhibition of the secretion of major proinflammatory mediators. At present we can only speculate as to the reasons why certain combinations are synergistic, since the complexity of compound combinations in a cellular system complicates an explanation of the actual mechanism. Furthermore, the relatively modest and conflicting literature regarding flavonoid effects in cellular models of inflammation limit an in-depth mechanistic explanation. In view of this, further work is envisaged in the area of molecular
20
40
60
80
Kaempferol IC 50 (µM) Fig. 7 Isobolograms of various flavonoid combinations used to inhibit TNF-a. Theoretical (open square) and observed (filled square) IC50 values and degree of significance are shown on each isobologram. Comparisons were made with the Student’s t-test
pharmacology in order to unravel the exact mechanism for synergy amongst flavonoids. Acknowledgments We thank Zulkhairi Zainol, Abdul Rahman Hassan and Nora Asyikin Mohd Salim for technical assistance. This investigation was financially supported by Science Fund (06-01-04SF0973), Ministry of Science, Technology and Innovation, Malaysia.
References 1. Kim HP, Son KH, Chang HW, Kang SS. Anti-inflammatory plant flavonoids and cellular action mechanisms. J Pharmacol Sci. 2004;96:229–45.
Flavonoid combinations cause synergistic inhibition of proinflammatory mediators 2. Williamson EM. Synergy and other interactions in phytomedicines. Phytomed. 2001;8:401–9. 3. Ting-Chao Chou. Theoretical basis, experimental design, and computerized simulation of synergism and antagonism in drug combination studies. Pharmacol Rev. 2006;58:621–81. 4. Scott EN, Gescher AJ, Steward WP, Brown K. Development of dietary phytochemical chemopreventive agents: biomarkers and choice of dose for early clinical trials. Cancer Prev Res (Phila Pa). 2009;2:525–30. 5. Duarte J, Perez-vizcaino F, Jimenez J, Tamargo J, Zarzuelo A. Flavonoids and cardiovascular diseases. Stud Nat Prod Chem. 2001;25:565–605. 6. Galvez J, De Sanchez Medina F, Jimenez J, Zarzuelo A. Effects of flavonoids on gastrointestinal disorders. Stud Nat Prod Chem. 2001;25:607–49. 7. Schroeter H, Boyd C, Spencer JPE, Williams RJ, Cadenas E, Rice-Evans C. MAPK signaling in neurodegeneration: influences of flavonoids and of nitric oxide. Neurobiol Aging. 2002;23:861– 80. 8. Gohel MS, Davies AH. Pharmacological agents in the treatment of venous disease: an update of the available evidence. Curr Vasc Pharmacol. 2009;7:303–8. 9. Sato Y, Shibata H, Arai T, Yamamoto A, Okimura Y, Arakaki N, et al. Variation in synergistic activity by flavone and its related compounds on the increased susceptibility of various strains of methicillin-resistant Staphylococcus aureus to b-lactam antibiotics. Int J Antimicro Ag. 2004;24:28–35. 10. Scambia G, De Vincenzo R, Ranelletti FO, Benedetti Panici P, Ferrandina G, D’Agostino G, et al. Antiproliferative effect of silybin on gynaecological malignancies: synergism with cisplatin and doxorubicin. Eur J Canc. 1996;32A:877–82. 11. Milde J, Elstner EF, Graßmann J. Synergistic inhibition of lowdensity lipoprotein oxidation by rutin, c-terpinene, and ascorbic acid. Phytomed. 2004;11:105–13. 12. Wang Q, Han Y, Xue H. Ligands of the GABAA receptor benzodiazepine binding site. CNS Drug Rev. 1999;5:124–44. 13. Medina JH, Viola H, Wolfman C, Marder M, Wasowski C, Calvo D, et al. Overview. Flavonoids: a new family of benzodiazepine receptor ligands. Neurochem Res. 1997;22:419–25. 14. Marder M, Paladini AC. GABAA-receptor ligands of flavonoid structure. Curr Top Med Chem. 2002;2:853–67. 15. Fernandez SP, Wasowski C, Paladini AC, Marder M. Synergistic interaction between hesperidin, a natural flavonoid, and diazepam. Eur J Pharmacol. 2005;512:189–98. 16. Murakami A, Takahashi D, Hagihara K, Koshimizu K, Ohigashi H. Combinatorial effects of nonsteroidal anti-inflammatory drugs and food constituents on production of prostaglandin E2 and tumour necrosis factor–a in RAW 264.7 murine macrophages. Biosci Biotechnol Biochem. 2003;67:1056–62. 17. Tallarida RJ, Porreca F, Cowan A. Statistical analysis of drugdrug and site-site interactions with isobolograms. Life Sci. 1989;45:947–61.
721
18. Pinardi G, Sierralta F, Miranda HF. Interaction between the antinociceptive effect of ketoprofen and adrenergic modulatory systems. Inflamm. 2001;25:233–9. 19. Tallarida RJ. Drug synergism: its detection and applications. J Pharmacol Exp Ther. 2001;298:865–72. 20. Dehmlow C, Murawski N, De Groot H. Scavenging of reactive oxygen species and inhibition of arachidonic acid metabolism by silibinin in human cells. Life Sci. 1996;58:l591–1600. 21. Wadsworth TL, McDonald TL, Koop DR. Effects of Ginkgo biloba extract (EGb 761) and quercetin on lipopolysaccharideinduced signaling pathways involved in the release of tumor necrosis factor-a. Biochem Pharmacol. 2001;62:963–74. 22. Cho SY, Park SJ, Kwon MJ, Jeong TS, Bok SH, Choi WY, et al. Quercetin suppresses proinflammatory cytokines production through MAP kinases and NF-jB pathway in lipopolysaccharidestimulated macrophage. Mol Cell Biochem. 2003;243:153–60. 23. Takano-Ishikawa Y, Goto M, Yamaki K. Structure–activity relations of inhibitory effects of various flavonoids on lipopolysaccharide-induced prostaglandin E2 production in rat peritoneal macrophages: Comparison between subclasses of flavonoids. Phytomed. 2006;13:310–7. 24. Hamalainen M, Nieminen R, Vuorela P, Heinonen M, Moilanen E. Anti-inflammatory effects of flavonoids: genistein, kaempferol, quercetin, and daidzein inhibit STAT-1 and NF-jB activations, whereas flavone, isorhamnetin, naringenin, and pelargonidin inhibit only NF-jB activation along with their inhibitory effect on iNOS expression and NO production in activated macrophages. Mediators Inflamm 2007;45673–45683. 25. Park HH, Lee S, Son HY, Park SB, Kim MS, Choi EJ, et al. Flavonoids inhibit histamine release and expression of proinflammatory cytokines in mast cells. Arch Pharm Res. 2008;31:1303–11. 26. Crespo I, Garcı´a-Mediavilla MV, Gutie´rrez B, Sa´nchez-Campos S, Tun˜o´n MJ, Gonza´lez-Gallego J. A comparison of the effects of kaempferol and quercetin on cytokine-induced pro-inflammatory status of cultured human endothelial cells. Br J Nutr. 2008;100:968–76. 27. Hecker M, Priess C, Klemm P, Brusse R. Inhibition by antioxidants of nitric oxide synthase expression in murine macrophages; role of nuclear factor kappa B and interferon regulatory factor 1. Br J Pharmacol. 1996;118:21178–84. 28. Chen CC, Chow MP, Huang WC, Lin YC, Chang YJ. Flavonoids inhibit tumor necrosis factor-a-induced up-regulation of intercellular adhesion molecule-1 (ICAM-1) in respiratory epithelial cells through activator protein-1 and nuclear factor-jB: structureactivity relationships. Mol Pharmacol. 2004;66:683–93. 29. Bowie A, O’Neill LAJ. Oxidative stress and nuclear factor-jB activation: a reassessment of the evidence in the light of recent discoveries. Biochem Pharmacol. 2000;59:13–23.
