No job name

J. Med. Chem. 2004, 47, 5555-5566
Increased Anti-P-glycoprotein Activity of Baicalein by Alkylation on the A Ring
Yashang Lee,†,| Hosup Yeo,†,‡,| Shwu-Huey Liu,§ Zaoli Jiang,§ Ruben M. Savizky,‡ David J. Austin,‡ andYung-chi Cheng*,† Department of Pharmacology, Yale University School of Medicine, Department of Chemistry, Yale University, andPhytoCeutica, Inc., New Haven, Connecticut 06520 The aqueous extract of Scutellariae baicalensis Georgi has inhibitory activity against P-gp 170,
a multiple drug resistant gene product. Baicalein, one of the major flavones, was found to be
responsible for this activity. The hydroxyl groups of the A ring of baicalein were systematically
alkylated in order to assess the effect of such modifications on the activity against P-gp 170.
The impact of the baicalein modifications on activity against the growth of a human
nasopharyngeal cancer cell line KB and its P-gp 170 overexpressing cell line KB/MDR were
also examined. The results indicate that alkylation of R5 of baicalein does not have a major
impact on the interaction with P-gp 170, whereas alkylation of R6 or R7 alone or both, could
enhance the interaction of baicalein with P-gp 170 as well as the amount of intracellular
accumulation of vinblastine, a surrogate marker for the activity of P-gp 170 pump of KB/MDR
cells. In this case, the optimal linear alkyl functionality is a propyl side chain. These
modifications could also alter the activity of compounds inhibiting cell growth. Among the
different compounds synthesized, the most potent molecule against P-gp 170 is 5-methoxy-
6,7-dipropyloxyflavone (23). Its inhibitory activity against P-gp 170 is approximately 40 times
better, based on EC50 (concentration of the compound enhancing 50% of the intracellular
vinblastine accumulation in the KB/MDR cells) and 3 times higher, based on Amax (the
intracellular vinblastine accumulation of the KB/MDR cells caused by the compound) as
compared to baicalein. Compound 23 is also a more selective inhibitor than baicalein against
P-gp 170, because its cytotoxicity is less than that observed for baicalein. The growth inhibitory
IC50 of compound 23 against KB and KB/MDR cells are about the same, suggesting that
compound 23 is unlikely to be a substrate of P-gp 170 pump. Acetylation of R6, R7 or both
could also decrease EC50 and increase Amax. Acetylated compounds are more toxic than baicalein,
and their potency against cell growth is compromised by the presence of P-gp 170, suggesting
that these compounds are substrates of P-gp 170. Benzylation of R6 or R7 but not both also
enhanced anti-P-gp170 activity and potency against cell growth; however, the presence of P-gp
170 in cells did not have an impact on their sensitivity to these molecules, suggesting that the
benzylated compounds are inhibitors but not substrates of P-gp 170, and perhaps have a
different mechanism of action. In conclusion, the substitutions of R6 and R7 hydroxyl groups
by alkoxy groups, acetoxy groups, or benzyloxy groups could yield compounds with different
modes of action against P-gp 170 with different mechanisms of action against cell growth.
Introduction
blocking absorption by the intestine, excretion of chemi-cals into the bile duct or kidney tubules, prevention of P-glycoprotein 170 (P-gp 170), a member of the ABC (ATP Binding Cassette) family, acts as an ATP-depend- chemicals taken into the brain through the blood-brain ent drug efflux pump, preventing intracellular ac- barrier, and efflux of steroid hormones and cholesterol cumulation of miscellaneous drugs.1,2 Overexpression of from feces.2,4,5 Developing drugs to inhibit P-gp 170 this protein is one of the mechanisms of multidrug activity is an important area of drug discovery. Such resistance (MDR) of cancer cells. This protein is ex- drugs could have use in facilitating the oral absorption pressed in a cell- and tissue-specific manner, with high of drugs through the intestine, or the uptake of chemi- levels detectable in the kidney, liver, blood-brain cals that are substrates of P-gp, into the brain. In barrier, and lining of the intestine.3 Studies using mdr1 addition, these compounds could also potentiate the knockout mice and P-gp 170 tissue distribution in action of antitumor drugs, which are substrates of P-gp humans suggested several physiological roles of P-gp 170 in cancer cells that overexpress the P-gp 170 170, including protection against toxic xenobiotics by A large number of compounds with major structural * Corresponding author. Dr. Yung-chi Cheng, Department of Phar- differences have been found to act as inhibitors or macology, Yale University School of Medicine, 333 Cedar St., SHM B254, New Haven, CT 06520. Tel: (203)-785-7119. Fax: (203)-785-7129. E-mail: [email protected].
(VRM), a calcium channel antagonist; trifluoperazine, † Department of Pharmacology, Yale University School of Medicine.
a calmodulin inhibitor; cyclosporin A (CSA), an immuno- ‡ Department of Chemistry, Yale University.
§ suppressant; and progesterone, a steroid hormone.
| The first two authors contributed equally to this manuscript.
Verapamil has been examined clinically in combination Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 with cancer chemotherapy.6,7 However, the results were action. Since baicalein (a 5,6,7-trihydroxyflavone) and rather unsatisfactory due to high plasma drug levels, wogonin (a 5,7-dihydroxy-8-methoxyflavone) have a very required to effectively reverse the MDR phenotype of similar structure, this raised the possibility of an cancer cells, which could cause cardiac toxicity. Com- interesting structure-activity relationship of the fla- pounds with higher potency against and selectivity for vone natural products for P-gp 170 inhibition.
P-gp 170 are needed. A second generation of MDR In this study, we synthesized and evaluated a series reversal agents has emerged and is based on the of baicalein analogues, focusing on the substitution chemical modification of the first generation of inhibi- pattern at positions 5, 6, and 7 of the A-ring. Our results tors. Among these, dexniguldipine8 and dexverapamil9 show that alkoxy groups on the A-ring of the flavone were found to be more selective against P-gp 170, but greatly increase the anti-P-gp 170 activity and alter they did not display improved potency. The acridone- their selectivity for the efflux pump.
carboxamide derivative GF120918 (GG918)10 and thecyclosporin A analogue PSC83311 both displayed an Chemistry
activity that was 10-30 times more potent than the first The synthesis of O-substituted baicalein derivatives generation of modulators, such as verapamil, tamoxifen, from commercially available reagents was carried out and cyclosporin A.12 A number of these compounds are using the general synthetic approach shown in Schemes currently under clinical evaluation.
1-4. Acetylation of baicalein (1) with acetic anhydride
Flavonoids are important class of natural products was performed in pyridine to give mono-, di-, and found in plants. With its polyphenolic structure, this triacetylated derivatives 2-4. Baicalein was treated
class of compounds has multiple actions, including with benzyl bromide in dry acetone with potassium interacting with estrogen receptor, serving as a free- carbonate to provide mono- and dibenzylated analogues radical scavenger, and inhibiting protein kinase, NF- 15 and 16. The O-methylated derivatives 6 and 8 of
κB and P-gp 170, among others.13 The biological activity baicalein were readily prepared by reaction with tri- found in herbal preparations is often attributed to its methylsilyldiazomethane (TMSCHN2) in a methanolic flavonoids. For example, the coadministration of grape- THF solution at room temperature. Since the hydroxyl fruit juice with various drugs has led to an increase in function on the C-5 position of baicalein makes an the plasma concentration of the drugs, which was intramolecular hydrogen bond with the 4-keto group, attributed to the bioflavonoids found in the grapefruit it is resistant to alkylation, and benzylation of baicalein occurred in the following order: 6 > 7 > 5.
Flavonoids have also been shown to act on multiple The O-methylated products 6 and 8 of baicalein
targets with different specificity. For example, the exhibited more potent anti-P-gp activities than that of flavonoid that binds to estrogen receptor requires hy- baicalein itself. This finding led us to design and droxyl groups at positions 2 and 3 of the B-ring, a double synthesize a series of alkylated baicalein analogues in bond at positions 2-3 of the C-ring, and the absence of order to examine their anti-P-gp activity. A variety of any hydrophobic prenylated substituent.16 This is mark- alkyl chains were attached to the hydroxyl groups in edly different than the flavonoids that inhibit various the baicalein A ring, and related flavonoids, through ATPases or protein kinases. Recognition of the ATP phenol alkylation with a number of alkyl halides. The binding pocket of these proteins requires the presence reaction of catechols on the baicalein A ring and of three hydroxyl groups at positions 5 and 7 on the flavonoids with bromochloromethane in DMF at 50 °C A-ring and position 3 of the C-ring, which favors some in the presence of cesium carbonate provided the cor- flavonols.17 Moreover, some protein kinases exhibit responding methylenedioxy derivative 19. Similarly,
different structural requirements for binding: an isofla- compound 13 was readily prepared by heating baicalein
vone structure has been demonstrated to inhibit ty- at 170 °C for 1 h with dichlorodiphenylmethane and rosine kinase activity,18 and a flavone substituted at then reaction with TMSCHN2 in methanolic THF at position 8 of A-ring with 4-(3-hydroxy-1-methylpiperidi- nyl) group has demonstrated activity against CDK2.19 The selective brominations at the C-8 position of The inhibition of P-gp 170 by flavonoids has also been baicalein and baicalein derivatives were performed investigated and two binding modes have been postu- directly with N-bromosuccinimide in the presence of a lated. The studies were performed using a truncated catalytic amount of concentrated sulfuric acid at room form of P-gp 170 in a membrane preparation from P-gp 170 overexpressing cells. The structural requirementsfor flavonoid activity have recently been summarized,20 Biological Results
and it appears that different classes of flavonoids have In recent studies, we found that the anti-P-gp activity different structural requirements for inhibitory activity of Scutellaria baicalensis Georgi could inhibit P-gp 170 and have attributed this activity to the high quantity In our previous study, we found that Radix scutel- of baicalein (compound 1) found in the extract. We
lariae (Scute), a well-known Chinese herb, has inhibi- therefore synthesized a number of baicalein-related tory activity against P-gp 170. The active component of compounds in order to evaluate their structure-activity the Scute herb was found to be the natural product relationship against P-gp 170 activity. Since flavonoids baicalein. However, the 7-glucuronyl form of baicalein, could have multiple sites of action that affect cell which is the most abundant component of Scute, did not growth, we also evaluated their growth inhibitory show anti-P-gp 170 activity. Other abundant Scute activity against KB, a human cancer cell line. If a components, such as wogonoside and wogonin, were also compound exhibits cytotoxicity through the inhibition found to lack inhibitory activity against P-gp 170 efflux of cell function, in addition to serving as a substrate of Anti-P-glycoprotein Activity of Baicalein Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 Scheme 1a
a Reagents and conditions: (a) Ac2O, pyridine, rt; (b) TMSCHN2, THF:MeOH (2:1), rt; (c) K2CO3, BnBr, acetone, reflux; (d) NBS, THF, concd H2SO4, rt; (e) K2CO3, KI, BnBr.
Scheme 2a
a Reagents and conditions: (a) Cs2CO3, BrCH2Cl, DMF, 50 °C; (b) TMSCHN2, THF:MeOH (2:1), rt.
P-gp 170, then the compound would be expected to 5, 6, and 7, since our initial studies indicated that these demonstrate less activity against a cell line overex- positions might play a crucial role in P-gp 170 inhibition.
pressing P-gp 170 than the parent cell line. Therefore, The KB/MDR cells, which overexpress human P-gp we employed a multi-drug resistant cell line (KB/MDR) 170 protein, were used to evaluate the anti-P-gp 170 in addition to the parent KB cell line, to assess the activity of drugs. In this manner, the intracellular susceptibility of a compound to act as a substrate or amount of vinblastine, a substrate of P-gp 170 pump, inhibitor of the P-gp 170 efflux pump. Our modification was measured in the presence of the known P-gp of the natural products focused primarily on the func- inhibitors cyclosporin A and verapamil and compared tional groups of the baicalein A ring, especially positions to our synthetic flavones. The concentration of com- Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 Scheme 3a
This suggests that the bromo substitution prevents
compound 3 from acting as a substrate of P-gp 170.
By adding one benzyloxy group to the A ring of baicalein (Table 1), the flavone became a very potent
inhibitor of P-gp 170, irrespective of the position or the
presence of other functional groups. All the flavones 6,7-
diacetoxy-5-benzyloxyflavone (9), 7-acetoxy-6-benzyloxy-
5-hydroxyflavone (10), 5,6-diacetoxy-7-benzyloxyflavone
(14), and 6-benzyloxy-5,7-dihydroxyflavone (15) exhib-
ited a low EC50 but higher Amax in comparison with
baicalein. On the other hand, the benzyloxy group
transformed the flavone into a more toxic compound
toward both KB and KB/MDR cells than baicalein. This
suggests that these compounds have a different site of
action in addition to P-gp 170, for which activity is
required to maintain cell growth. The fact that these
compounds have the same IC50 for KB and KB/MDR
cells, suggests that they are inhibitors but not sub-
strates of the P-gp 170 efflux pump. When the A ring
of baicalein is substituted with two benzyloxy groups,
the flavone shows decreased cytotoxicity and anti-P-gp
activity, without regard to being cyclic (13) or not (16).
Experimental results for the alkylated baicalein compounds are shown in Table 2. The introduction of a
methoxy group onto the A ring decreased the EC50 and
increased the Amax. The presence of two (8) or three (6)
a Reagents and conditions: (a) Ac2O, pyridine, rt; (b) TMSCHN2, methoxy groups on the A ring of baicalein dramatically THF:MeOH (2:1), rt; (c) K2CO3, BnBr, acetone, reflux; (d) NBS, increased the anti-P-gp activity, as evidenced by their EC50 of 4.6 μM and 5.5 μM respectively, without afurther increase in the A pound required to achieve 50% maximum accumulation position of the A ring decreased the cytotoxic activity of vinblastine is presented as the EC50 value. The maximum accumulation of intracellular vinblastine max, but did not alter the EC50 against P-gp 170 activity as compared to compounds 8 and 6, regardless
caused by the compounds in 1 h is expressed as Amax of whether the substitution at R5 is a hydroxy (19) or
(picomoles/106 cells). The cell growth inhibitory activity methoxy (20) group.
of compounds is presented as the concentration required Alkyl substitution (methoxy) turned out to be a far to inhibit 50% growth of KB or KB/MDR cell lines (IC50) more favorable substitution than either the acetoxy or following 3 days of compound treatment.
benzyloxy group for flavone anti-P-gp efflux activity.
As shown in Table 1, the number of acetoxy groups This is most likely due to the fact that this substitution on the baicalein A ring altered the EC50 of anti-P-gp prevents the compound from acting as a substrate for activity. Compounds with one (2) or two (3) acetoxy
P-gp 170 or renders it more specific against P-gp 170.
groups on position 6 and 7 of the A ring exhibit an EC50 For this reason, we further explored the potential of this that is one-fourth that of the parental compound. The type of substitution in a search for an optimal linear activity of the flavone with three acetoxyl groups (4) at
alkoxy group. The ethoxy group was shown to be better position 5, 6, and 7 did not differ from compound 2 or
than the methoxy group. The EC50 values of 6-ethoxy- 3. The Amax values of these three compounds were also
5,7-dihydroxyflavone (32) and 6,7-diethoxy-5-hydroxy-
similar, but higher than that found for baicalein. The flavone (33) were 2.3 μM and 1.8 μM, respectively, which
flavones with acetoxy groups were more toxic to KB cells is about the same as that of CSA. In addition, the ethoxy than the KB/MDR cells, indicating that the substitu- compound 33 had a higher Amax and less cytotoxicity
tions of hydroxyl groups by acetoxy groups could render than either CSA or compound 32. While there is no
the flavone a better substrate for P-gp 170 efflux pump.
significant difference between one and two ethoxy Substitution of the acetoxy groups in compound 4 with
groups on the EC50, there is a big difference in their one (7) or two (5) methoxy groups did not alter the EC50
cytotoxicity. Substitution of the R5 hydroxyl group in or Amax substantially, but increased the IC50 value compound 33 with methoxy (44) did not change the EC50
against cell growth. We also evaluated the impact of a significantly. Substitution of the R7 ethoxy group of bromo group on position 8 on anti-P-gp 170 efflux compound 33 with a methoxy (38) did not alter the IC50
activity. Compound 25 (8-bromobaicalein) decreased the
EC50 to 15 μM, but the Amax did not change, as compared Similar to the benzyloxy group substitution, the to baicalein. Compound 24 (the 8-bromo derivative of
presence of the propyloxy substitution on any position compound 3) also showed less favorable activity against
of the A ring has a significant effect on the potency and P-gp 170 activity than compound 3. Both of the com-
degree of anti-P-gp 170 efflux activity. One propyloxy pounds with an 8-bromo group were toxic to KB and group on R6 (17) decreases the EC50 to 2 μM and
KB/MDR cells and showed the same IC50 as compound increases the Amax 9-fold higher than that of the control; 3 to KB cells, but lower than that for the KB/MDR cells.
two propyloxy groups on R6 and R7 (18) decreases the
Anti-P-glycoprotein Activity of Baicalein Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 Scheme 4a
a Reagents and conditions: (a) K2CO3, CH3(CH2)nX (X ) I or Br, n ) 1 for 32, 33, n ) 2 for 17, 18, n ) 3 for 45, n ) 4 for 40, n ) 5
for 42, n ) 7 for 34), acetone, reflux; (b) TMSCHN2, THF:MeOH (2:1), rt.
Table 1. Anti-P-gp Activity and Cytotoxicity of Modified Baicalein Compounds
a clogD was calculated using the following equation: clogD ) clogP - log[1 + 10(pH-pKa)], where clogP is found using ChemDraw ULTRA, version 6.0.1, for 1-octanol/water system, the pH of the experiment was 7, and the pKa was assumed to be 10 (values ranging from 8 to11 made insignificant changes to the clogD value). b The anti-P-gp efflux activity is represented by intracellular vinblastine accumulationin 1 h with or without drug treatment; see Experimental Section for details. c Maximum vinblastine accumulation (pmoles) of flavones(<100 μM) treated cells in the 106 cells in 1 h. d EC50 calculated as the concentration (μM) that causes 50% of maximum vinblastineaccumulation in the cells in 1 h. e Cytotoxicity IC50 calculated as the concentration (μM) required for 50% inhibition of cell growth, foreach respective cell line, after 72 h of drug exposure. f Control cells were treated with vinblastine only. All values represent the mean (SD of at least three identical experiments.
EC50 to 1.4 μM, and the Amax is 10-fold higher than the 5,6,7-trimethoxyflavone (6), compound 41 to 40 and
control. In addition, the EC50 of 5-hydroxy-7-methoxy- compound 43 to 42. Finally, the most potent flavone in
6-propyloxyflavone (21) was further decreased to 1.2 μM
this series was found to be 5-methoxy-6,7-dipropyloxy- and the Amax was 9-fold higher than that of the control.
flavone (23), with two propyloxy groups on R6 and R7
The molecule 5,7-dimethoxy-6-propyloxyflavone (22),
and a methoxy group on R5. The EC50 of compound 23
with an added methoxy on the R5 hydroxyl group, was found to be 0.9 μM and the Amax is 10-fold higher showed the same EC50 and Amax as compound 21. The
than that of the control. This is more efficient than CSA, cytotoxicity of 6,7-diethoxy-5-methoxyflavone (44) was
which shows an activity that is 7-fold higher than higher than that of 6,7-diethoxy-5-hydroxyflavone (33),
control. The presence of a methoxy group on the R5 which only contains a single R5 position change from position of compound 23 makes it slightly more toxic
hydroxy to methoxy. The same phenomenon is found than compound 18 to KB and KB/MDR cells, but this
when comparing 5-hydroxy-6,7-dimethoxyflavone (8) to
cytotoxicity is not altered in the presence of P-gp 170 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 Table 2. Anti-P-gp Activity and Cytotoxicity of Alkylated Baicalein Compounds
a clogD was calculated using the following equation: clogD ) clogP - log[1 + 10(pH-pKa)], where clogP is found using ChemDraw ULTRA, version 6.0.1, for 1-octanol/water system, the pH of the experiment was 7, and the pKa was assumed to be 10 (values ranging from 8 to11 made insignificant changes to the clogD value). b The anti-P-gp efflux activity is represented by intracellular vinblastine accumulationin 1 h with or without drug treatment; see Experimental Section for details. c Maximum vinblastine accumulation (pmoles) of flavones(<100 μM) treated cells in the 106 cells in 1 h. d EC50 calculated as the concentration (μM) that causes 50% of maximum vinblastineaccumulation in the cells in 1 h. e Cytotoxicity IC50 calculated as the concentration (μM) required for 50% inhibition of cell growth, foreach respective cell line, after 72 h of drug exposure. f Control cells were treated with vinblastine only. All values represent the mean (SD of at least three identical experiments.
activity, which suggests that compound 23 is not a
for binding affinity. The assay used in these studies, substrate of P-gp 170. Compounds with alkoxy substitu- however, directly measures the binding affinity of tions of longer chain length lead to a decrease in Amax, compounds for the cytosolic nucleotide-binding domain without altering the EC50 substantially. No obvious of P-gp 170 protein and does not provide a measure of correlation between a molecule’s clog D value and anti- anti-P-gp 170 functional activity in living cells.
Ahcene Boumendjel et al.20 incorporated modified chrysin in their studies and achieved results similar to Discussion and Conclusion
our own. Their experiments showed that the increase Other groups have explored the anti-P-gp activity of in hydrophobicity of chrysin by alkylation with methyl, balcalein, which exhibits moderate instability in cell isopropyl, benzyl, 3,3-dimethylallyl, or geranyl substit- culture. In drug screens involving molecules isolated uents correlated with an increase in affinity for in vitro from Chinese herbal medicines, Thomas Efferth et al.
binding to the P-gp cytosolic domain. Their anti-P-gp found several compounds exhibiting anti-P-gp activity.21 activity is therefore solely dependent on the number of However, baicalein was not included in their list. With isopropyl groups, irrespective of the A-ring position (6, adjacent trihydroxyl groups on the A ring, baicalein is 7, or 8). Our results indicate that the number of carbons not a stable compound. It is easily oxidized and forms permitted in the alkoxy group is limited to three. Alkoxy a green precipitate within a few hours under assay groups in all three positions 5, 6, or 7 make a contribu- conditions. In our 1-h vinblastine accumulation experi- tion to anti-P-gp activity. Potency is dependent on the ment, we observed a moderate loss of baicalein. Since alkoxy(l) group and decreases in the following order: this molecule is so unstable, it is not surprising that it propyloxy > ethoxy > methoxy. In our studies, all was not identified in the original 4-10 day anti-P-gp flavones with alkoxy groups have the same or better anti-P-gp activity than that of cyclosporin A (measured Our previous studies have shown that flavone has the by vinblastine accumulation) and much higher than that highest anti-P-gp activity among all of the flavonoid of the flavones with C-alkyl groups, which only ac- subclasses, which include flavone, flavonol, isoflavone, cumulate about 20-30% of drug as compared to cy- flavanone, and glycosylated flavone. This conclusion is closporin A. The impact on anti-P-gp activity caused by consistent with that observed by Gwenaelle Conseil et n-propyloxy substituents is greater than having C- al.22 who demonstrated that the 2,3-double bond on C isopropyl or C-dimethylallyl substituents on these com- ring and the hydroxyl group on the A ring are important pounds. Perhaps the high potency observed for their affinity determinants for flavonoid binding to P-gp. In best compound, which has an O-dimethylallyl substitu- addition, Jose M. Perez-victoria et al.23 showed that no tion pattern, results from the effect of oxygen alkylation matter how large a substitution is made, its effect is and not from the dimethylallyl group itself. Moreover, not as important as the 2,3-double bond in the C ring the Amax of compound 23 is 167% of cyclosporin A and
Anti-P-glycoprotein Activity of Baicalein Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 236% of verapamil. One possible interpretation of these flavones more toxic to cells as compared to the hydroxyl results is that the alkoxyflavone is more specific for P-gp 170 and has no other cellular effect, which is not the In the case of a fixed functional group on the R5 and case of cyclosporin A or verapamil. In this manner, it R7 position, the molecules can be divided into three can reach a higher level of apparent anti-P-gp activity groups, those with two hydroxy groups (compounds 1,
2, 15, 17, and 32), two methoxy groups (compounds 5,
Interestingly, all of the flavones with an alkoxy group 6, and 22), and a hydroxy group on the R5 and methoxy
of 4-8 carbons exhibited the same level of EC50 com- group on the R7 (compounds 8, 21, and 38). The anti-
pared to compound 23, indicating that maybe the
P-gp activity of the compounds with two hydroxyl groups binding affinity of this series of alkoxyflavones remain was largely dependent on the functional group at the the same, since the long chain substituents did not R6 position, from high (compounds 1 and 2) to very low
decreased the anti-P-gp activity. One possibility is that (compounds 15 and 32) EC50. The anti-P-gp efflux
the long carbon chain of the alkoxy group changes the activity of compounds with two methoxy groups is also orientation of the flavone (e.g., insertion into plasma dependent on the functional group at the R6 position, membrane) relative to the P-gp 170 binding site, so that showing both medium EC50 (compounds 5 and 6) and
they still bind to P-gp but do not block pump activity.
very low (compound 22) activity. The anti-P-gp 170
The benzyl group substituent also showed a large activity of compounds with a hydroxy group on R5 and impact on anti-P-gp activity both in the studies of a methoxy group on R7 do not show obvious dependence Ahcene Boumendjel’s20 and ours. The benzyloxy group on the functional group at R6, since neither the methoxy has a much stronger effect on the A ethoxy or propyloxy groups changed the EC group in both systems. On the other hand, both their noticeably, however, all three were very good inhibitors.
and our benzyloxy groups are connected to the A-ring Their inhibitory effectiveness is shows the following at positions 6 and 7. With the same benzyloxy group, decreasing order: propyloxy (21) > ethoxy (38) >
the two acetoxy groups of our compound (5,6-diacetoxy- methoxy (8).
7-benzyloxyflavone, the Amax is 97% as compared to Compounds with the same functional groups on R5 cyclosporin A) may have some supplementary effect and R6 could also be divided into four categories. With compared to their compound (7-benzyloxy-5-hydroxy- a hydroxy group on R5 and a benzyloxy group on R6 flavone, the Amax is 25% as compared to cyclosporin A).
(compounds 10 and 15), the EC50 values decrease and
The addition of two benzyloxy groups on the A-ring are not affected by the substituent on R7, except for the caused the flavone (6, 7-dibenzyloxy-5-hydroxyflavone) molecule containing two benzyloxy groups (16), which
to lose all of its anti-P-gp activity in our experiments, lost both anti-P-gp 170 activity and cytotoxicity. The but (6-benzyl-7-benzyloxy-5-hydroxyflavone) showed an anti-P-gp 170 activity of compounds with a hydroxy increase in both binding affinity and drug accumulation group on R5 and an ethoxy group on R6 was not in their experiments. The only difference between these dependent on the functional group at R7 either (com- two molecules is the benzyl or benzyloxy group on pounds 32, 38, and 33). The anti-P-gp 170 efflux activity
position 6. It should also be noted that cytotoxicity of compounds with a hydroxy group on R5 and a increases greatly with benzyloxy substitution on the propyloxy group on R6 was also not dependent on the functional group at the R7 (compounds 17, 21, and 18).
To achieve a more comprehensive overview of the Even the anti-P-gp activity of the most potent com- structure-activity relationship of the flavones, we sub- pounds, containing a methoxy group on R5 and a categorized them based on the position of their substit- propyloxy group on R6 was not obviously dependent on uents. We first organized the same functional group on the chemical function at R7 (compounds 22 and 23). It
R6 and R7, to see the impact of R5 on anti-P-gp efflux appears that small substituent changes on position 7 activity. Compounds with the same substitutions on R6 do not have an impact on anti-P-gp activity.
and R7 were divided into four groups: compounds 3, 4,
The inhibition of P-gp 170 efflux activity itself should 7, and 9 with two acetoxy groups on R5 and R7;
not cause cytotoxicity, and our data support this theory compounds 8 and 6 with two methoxy groups; com-
since there appears to be no correlation between P-gp pounds 33 and 44 with two ethoxy groups; and com-
inhibition and cytotoxicity in the compounds evaluated pounds 18 and 23 with two propyloxy groups on the R6
in this study. Therefore, the increased toxicity noted in and R7 position. The EC50 required to inhibit P-gp 170 some of the molecules is likely due to affinity of the activity of these four groups decrease in the following compounds for additional biochemical targets with ATP- order: two acetoxys > two methoxys > two ethoxys > binding sites. Several flavonoids have been reported two propyloxys. In addition, the composition of the R5 to be good inhibitors for a variety of ATP-binding substituents is not as dominant as the composition of proteins such as plasma membrane ATPases,24,25 pro- the same two groups on R6 and R7. The anti-P-gp tein kinase A,26 protein kinase C,27 serine/threonine activity of flavones with two acetoxy groups on R6 and protein kinases,28 tyrosine protein kinase,29 and topo- R7 is dependent on the functional group in the R5 isomerase II.30 Staurosporine produces high intrinsic position. The anti-P-gp activity of flavones with two cytotoxicity in human cells in addition to its anti-P-gp methoxy or ethoxy groups on R6 and R7 did not vary activity. On the basis of a structural analysis, R. B.
with a change in the R5 functional group. Flavones with Wang et al.31 suggested that the isobenzopyrrolidone of two propyloxy groups on R6 and R7 were the most staurosporine meets the binding requirement of the potent inhibitors in this series of compounds, regardless adenosine moiety of ATP and prevents ATP binding to of whether R5 is hydroxy or methoxy group. However, the ATP-binding site. In this comparison, however, the the methoxy group on the R5 position renders the A ring of galangin (3, 5, 7-trihydroxyflavone) does not Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 overlap well with the five-membered ring of adenosine, phobic region of P-gp rather than the ATP-binding site.
and that is the reason galangin does not exhibit cyto- Benzyloxyflavone, on the other hand, may likely be toxicity as a staurosporine analogue. A. D. Pietro et al.32 overlapping the ATP-binding site, in order to decrease indicated that among a total of 29 flavonoids examined, the efflux of P-gp 170 substrates such as vinblastine, only three flavonols were found to bind to the ATP- and be exhibiting a high intrinsic cytotoxicity through binding site, and the hydroxyl group at position 3 of the interaction with the ATP-binding sites of other vital flavonol is critical for binding. This requirement is similar to that observed for quercetin binding to the Hck In conclusion, acetylation, alkylation, or benzylation tyrosine kinase, as demonstrated by cocrystallization33 of hydroxyl groups on the A ring of baicalein can and for other ATPases by inhibition kinetics.34 On the enhance interaction with the P-gp 170 protein and basis of these observations, most of our synthetic prevent its substrate efflux activity. The mode of flavones are unlikely to be good candidates for ATP- interaction with these modified flavones appear be binding site affinity because they lacking the hydroxyl quite different, given the broad differences noted in group on R3, except the benzyloxy flavone which could EC50, Amax and cytotoxicity. The alkoxyflavones may potentially block ATP binding by fitting to the adenine interact with P-gp at a site other than the ATP-binding site in the ATP-binding pocket like staurosporine.
site, whereas the other modified flavonoids likely mimic When a benzyloxy group is connected to the A-ring of the adenosine moiety of ATP and block the ATP-binding baicalein, its structure resembles L868276 (5,7-dihydroxy- site. This suggests that the alkoxyflavones may have a 8-[4-(3-hydroxy-1-methyl)piperidinyl]-flavone), which fits lower propensity to interact with the ATP-binding site nicely to the adenine-binding pocket of CDK2,19 and this of other proteins, as observed by their lower cytotoxicity.
may be a source of its increased cytotoxicity. Interest- In summary, the alkoxyflavones appear to be quite ingly, the cytotoxicity is lost when we add another promising modulators of P-gp 170 function and warrant benzyloxy group to the A-ring of baicalein. This is similar to a pattern noted with staurosporine modifica-tion: the addition of one benzyl group causes a decrease Experimental Section
in cytotoxicity (CGP41251),35 and adding another totally General Chemistry Methods. All solvents and reagents
abolishes the cytotoxic effect (CGP42700).36 It is possible were obtained from commercial suppliers and were used that the increased steric bulk added to the molecule by without further purification. Unless otherwise specified, reac- virtue of the second benzyl group renders the com- tions were performed under a nitrogen atmosphere withexclusion of moisture. All reaction mixtures were magnetically pounds incapable of entering the adenine-binding pocket stirred and monitored by TLC using Si250F precoated plates of CDK2 due to steric hindrance. We are currently from J. T. Baker (0.25 mm). Flash column chromatography was performed on 32-63 D 60 Å silica gel from ICN SiliTech The nucleotide binding site of P-glycoprotein contains (ICN Biomedicals GmbH). Melting points were determined a region that interacts with hydrophobic steroid deriva- with an Electrothermal capillary melting point apparatus andare uncorrected. 1H NMR spectra were recorded on a Bruker tives, such as RU486, called the steroid binding hydro- AM-400, Bruker AM-500, or GE QE-plus 300 spectrometer.
phobic region (SBHR). This region is most likely located Chemical shifts are reported using chloroform-d (δ 7.24 ppm) in close proximity to the ATP binding site since RU486 or DMSO-d6 (2.50 ppm). All coupling constants are described completely prevents or displaces the hydrophobic nucle- in Hz. Mass spectra were conducted at the Mass Spectrometry otide derivative, 2(3)-methylanthraniloyl-ATP (MANT- Laboratory of the University of Illinois.
ATP).37 A tentative model for the interaction of fla- 6-Acetoxy-5,7-dihydroxyflavone (2). Baicalein 1 (54 mg,
vonoids with P-glycoprotein and related multidrug 0.2 mmol) was dissolved in acetic anhydride (1 mL) andpyridine (1 mL), and the solution was stirred at room tem- transporters has been proposed by A. D. Pietro et al.32 perature for 2 h. The reaction mixture was poured into ice- Galangin (5,7-dihydroxyflavonol), kaempferol (4′,5,7- water (10 mL), and the precipitate was collected by filtration trihydroxyflavone), kaempferide (3,5,7-trihydroxy-4′- and purified by flash chromatography on a column of silica methoxyflavone), and dehydrosilybin (3, 5, 7-trihydroxy- gel eluted with CH2Cl2/MeOH (20:1) to yield compound 2 (35
3′-monolignolflavone) appear to interact with the cytosolic mg, 56%) as a yellow powder. mp 205-207 °C; 1H NMR nucleoside binding domain; the hydroxyl groups at (DMSO-d6) δ 2.28 (s, 3H), 6.67 (s, 1H), 7.01 (s, 1H), 7.57 (m,3H), 8.79 (m, 2H), 11.28 (s, 1H), 12.97 (s, 1H,); MS (EI) m/z positions 3 and 5, in addition to the ketone at position 4, are proposed to bind the ATP-binding site, whereas 6,7-Diacetoxy-5-hydroxyflavone (3) and 5,6,7-Triac-
other parts of the molecules bind in the vicinal SBHR etoxyflavone (4). Baicalein 1 (216 mg, 0.8 mmol) was
region. Prenylation of the A-ring would therefore in- dissolved in acetic anhydride (40 mL) and pyridine (12 mL), crease the hydrophobic interactions with both the cy- and the solution was stirred at room temperature for 48 h.
tosolic steroid-interacting region and the drug-binding The reaction mixture was poured into ice-water (100 mL), and site. This would potentially produce a significant shift the precipitate was collected by filtration and purified by flashchromatography on a column of silica gel eluted with CH in flavonoid positioning, in such a way that overlap with MeOH (40:1) to give compounds 3 (72 mg, 25%) and 4 (220
the ATP binding site can no longer occur. Such a prenyl- mg, 69%) as a pale yellow powder, respectively. (3): mp 204-
flavonoid positioning appears to be efficient enough to 206 °C; 1H NMR (CDCl3) δ 2.36 (s, 6H), 6.74 (s, 1H), 6.98 (s, directly inhibit P-gp 170 substrate binding and trans- 1H), 7.55 (m, 3H), 7.89 (m, 2H), 12.95 (s, 1H); MS (EI) m/z port, while indirectly interfering with ATP hydrolysis 354 [M]+, 312, 270 (base). (4): mp 194-195 °C; 1H NMR
and energy transduction. In our case, flavones without (CDCl3) δ 2.36, 2.37, 2.46 (each s, 9H), 6.67 (s, 1H), 7.52 (s,1H), 7.54 (m, 3H), 7.87 (m, 2H); MS (EI) m/z 396 [M]+, 354, the 3-hydroxyl group may be acting in a similar fashion, since the alkoxyl group helps trihydroxyflavone (baica- 6-Acetoxy-5,7-dimethoxyflavone (5). To a stirred solution
lein) to be a better P-gp modulator with less cytotoxicity of 2 (25 mg, 0.08 mmol) in a mixture of MeOH (4 mL) and
and may be interacting with the steroid binding hydro- THF (8 mL) was added trimethylsilyldiazomethane (TM- Anti-P-glycoprotein Activity of Baicalein Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 SCHN2, 2 M in hexanes, 0.4 mL, 0.8 mmol). The reaction and anhydrous K2CO3 (30 mg) in acetone (15 mL) was refluxed mixture was stirred at room temperature for 12 h and then for 24 h with stirring. The reaction mixture was filtered, and evaporated. Flash chromatography of the residue, eluting with the solvent was evaporated under reduced pressure. Flash n-hexane/EtOAc (1:1), gave compound 5 (5 mg, 18%) as a pale
chromatography of the residue, eluting with n-hexane/EtOAc yellow powder. mp 208-210 °C; 1H NMR (CDCl3) δ 2.40 (s, (3:2), afforded compound 14 (20 mg, 52%) as a white solid. mp
3H), 3.96 (s, 6H), 6.70 (s, 1H), 6.88 (s, 1H), 7.54 (m, 3H), 7.88 174-175 °C; 1H NMR (CDCl3) δ 2.27, 2.48 (s, 6H), 5.14 (s, 2H), (m, 2H); MS (EI) m/z 340 [M]+, 298, 280 (base).
6.53 (s, 1H), 6.97 (s, 1H), 7.38 (m, 5H), 7.47 (m, 3H), 7.76 (m, 5,6,7-Trimethoxyflavone (6). To a stirred solution of 1 (54
2H); MS (EI) m/z 444 [M]+, 402, 360, 269 (base).
mg, 0.2 mmol) in a mixture of MeOH (6 mL) and THF (12 mL) 6-(Benzyloxy)-5,7-dihydroxyflavone (15) and 6,7-(Diben-
was added TMSCHN2 (2 M in hexanes, 1.2 mL, 2.4 mmol).
zyloxy)-5-hydroxy-flavone (16). A mixture of 1 (54 mg, 0.2
The reaction mixture was stirred at room temperature for 36 mmol), benzyl bromide (0.12 mL), and anhydrous K2CO3 (83 h and evaporated. Flash chromatography of the residue, mg) in acetone (15 mL) was refluxed for 8 h with stirring. The eluting with CH2Cl2/acetone (15:1), gave compound 6 (30 mg,
reaction mixture was filtered, and the solvent was evaporated 48%) as a pale yellow powder. mp 164-165 °C; 1H NMR under reduced pressure. Flash chromatography of the residue, (CDCl3) δ 3.90, 3.97, 3.99 (s, 9H), 6.67 (s, 1H), 6.81 (s, 1H), eluting with CH2Cl2/MeOH (100:1 to 50:1), afforded compound 7.50 (m, 3H,), 7.86 (m, 2H); 13C NMR (CDCl3) δ 56.71, 61.96, 15 (30 mg, 42%) as a yellow powder and compound 16 (24 mg,
62.60, 96.67, 108.84, 126.39, 129.38, 131.68, 132.03, 154.98, 27%) as a pale yellow powder. (15): mp 195-197 °C; 1H NMR
158.20; MS (EI) m/z 312 [M]+, 297 (base). HRMS for C18H16O5 (CDCl3) δ 5.27 (s, 2H), 6.67 (s, 2H), 7.44 (m, 5H), 7.52 (m, 3H), [M]+: calculated, 312.0998; found, 312.0995.
7.88 (m, 2H); 13C NMR (CDCl3) δ 71.78, 92.25, 105.87, 106.69, 6,7-Diacetoxy-5-methoxyflavone (7). To a stirred solution
126.69, 127.99, 129.02, 129.28, 129.48, 130.39, 131.86, 132.18, of 3 (25 mg, 0.07 mmol) in a mixture of MeOH (2 mL) and
135.77, 146.35, 150.93, 152.22, 164.57, 183.09; MS(EI) m/z 360 THF (4 mL) was added TMSCHN2 (2 M in hexanes, 0.21 mL, [M]+, 269 (base). HRMS for C22H16O5 [M]+: calculated, 360.0998; 0.42 mmol). The reaction mixture was stirred at room tem- found, 360.0996. (16): mp 191-193 °C; 1H NMR (CDCl3) δ
perature for 12 h and evaporated. Flash chromatography of 5.17, 5.19 (s, 4H), 6.59 (s, 1H), 6.69 (s, 1H), 7.30-7.55 (m, 13H), the residue, eluting with CH2Cl2/acetone (30:1), gave compound 7.87 (m, 2H); 13C NMR (CDCl3) δ 71.36, 75.25, 92.38, 106.07, 7 (6 mg, 23%) as a pale yellow powder. mp 240-242 °C; 1H
126.68, 127.68, 128.36, 128.60, 128.68, 129.05, 120.10, 129.49, NMR (CDCl3) δ 2.35, 2.45 (s, 6H), 3.95 (s, 3H), 6.61 (s, 1H), 131.76, 132.22, 136.16, 137.89, 153.64, 153.94, 158.59, 164.38, 6.96 (s, 1H), 7.52 (m, 3H), 7.85 (m, 2H); MS (EI) m/z 368 [M]+, 183.16; MS (EI) m/z 450 [M]+, 359, 269, 91 (base). HRMS for C29H22O5 [M]+: calculated, 450.1467; found, 450.1465.
5-Hydroxy-6,7-dimethoxyflavone (8). To a stirred solu-
5,7-Dihydroxy-6-propyloxyflavone (17) and 5-Hydroxy-
tion of 1 (54 mg, 0.2 mmol) in a mixture of MeOH (6 mL) and
6,7-dipropyloxyflavone (18). A mixture of 1 (54 mg, 0.2
THF (12 mL) was added TMSCHN2 (2 M in hexanes, 0.6 mL, mmol), n-propyl iodide (0.06 mL), and anhydrous K2CO3 (110 1.2 mmol). The reaction mixture was stirred at room temper- mg) in acetone (20 mL) was refluxed with stirring for 24 h.
ature for 8 h and evaporated. Flash chromatography of the The reaction mixture was concentrated under reduced pres- residue, eluting with CH2Cl2/acetone (40:1 to 20:1), gave sure, diluted with water (30 mL), and extracted with CH2Cl2 compound 8 (8 mg, 13%) as a pale yellow powder. mp 159-
(3 × 30 mL). The extract was washed with water and dried 160 °C; 1H NMR (CDCl3) δ 3.94, 3.99 (s, 6H), 6.59 (s, 1H), 6.70 over MgSO4, and the solvent was evaporated in vacuo. The (s, 1H), 7.55 (m, 3H), 7.89 (m, 2H); 13C NMR (CDCl3) δ 56.78, residue was purified by flash chromatography on a column of 61.31, 91.04, 106.07, 126.68, 129.53, 131.73, 132.27, 153.46, silica gel eluted with CH2Cl2/MeOH (70:1 to 50:1) to yield 159.31, 164.37, 183.16; MS (EI) m/z 298 ([M]+, base), 283.
compounds 17 (7 mg, 11%) and 18 (44 mg, 62%) as a yellow
HRMS for C17H14O5 [M]+: calculated, 298.0841; found, 298.0844.
powder, respectively. (17): mp 162-163 °C; 1H NMR (CDCl3)
6,7-Diacetoxy-5-(benzyloxy)flavone (9) and 7-Acetoxy-
δ 1.11 (t, 3H, J ) 7.5 Hz), 1.95 (sextet, 2H, J ) 7.5 Hz), 4.13 6-(benzyloxy)-5-hydroxyflavone (10). A mixture of 3 (21
(t, 2H, J ) 7.5 Hz), 6.62 (s, 1H), 6.69 (s, 1H), 7.54 (m, 3H), mg, 0.06 mmol), benzyl bromide (0.03 mL), and anhydrous K 107.99, 108.54, 128.79, 131.62, 132.22, 134.04, 134.29, 148.21, 3 (26 mg) in acetone (15 mL) was refluxed for 8 h with stirring. The reaction mixture was filtered, and the solvent 153.22, 154.79, 165.56, 185.24; MS (EI) m/z 312 [M]+, 297, 283, was evaporated under reduced pressure. Flash chromatogra- 270 (base). HRMS for C18H16O5 [M]+: calculated, 312.0998; phy of the residue, eluting with n-hexane/EtOAc (3:2), afforded found, 312.0993. (18): mp 89-91 °C; 1H NMR (CDCl3) δ 1.07,
compounds 9 (18 mg, 68%) and 10 (8 mg, 33%) as a pale yellow
1.11 (t, 6H, J ) 7.5 Hz), 1.82, 1.92 (sextet, 4H, J ) 7.5 Hz), powder, respectively. (9): mp 175-177 °C; 1H NMR (CDCl
4.02, 4.05 (t, 4H, J ) 7.5 Hz), 6.54 (s, 1H), 6.66 (s, 1H), 7.53 δ 2.32, 2.47 (s, 6H), 5.22 (s, 2H), 6.61 (s, 1H), 7.02 (s, 1H), (m, 3H), 7.88 (m, 2H); 13C NMR (CDCl3) δ 10.86, 10.92, 22.78, 7.42 (m, 5H), 7.52 (m, 3H), 7.84 (m, 2H); MS (EI) m/z 444 [M]+, 23.81, 30.09, 71.06, 75.43, 91.67, 106.02, 106.54, 126.64, 402, 360, 269 (base). (10): mp 184-185 °C; 1H NMR (CDCl
129.47, 131.87, 132.13, 132.47, 153.63, 153.71, 159.22, 164.19, δ 2.37 (s, 3H), 5.22 (s, 2H), 6.64 (s, 1H), 6.70 (s, 1H), 7.42 (m, 183.13; MS (EI) m/z 354 ([M]+, base), 325, 311, 283, 270. HRMS 5H), 7.55 (m, 3H), 7.88 (m, 2H), 13.00 (s, 1H); MS (EI) m/z for C21H22O5 [M]+: calculated, 354.1467; found, 354.1464.
5-Hydroxy-6,7-(methylenedioxy)flavone (19). A mixture
6,7-(Diphenylmethylenedioxy)-5-methoxyflavone (13).
of 1 (81 mg, 0.3 mmol) and cesium carbonate (244 mg, 0.75
A mixture of 1 (27 mg, 0.1 mmol) and dichlorodiphenylmethane
mmol) in DMF (5 mL) was stirred at room temperature for 30 (0.02 mL, 0.1 mmol) was stirred under nitrogen at 170 °C for min. Bromochloromethane (0.05 mL, 0.75 mmol) was added 1 h. The reaction mixture was cooled to 30 °C and then to the DMF solution, and the mixture was stirred at 50 °C for dissolved in a minimum amount of CH2Cl2. The crude product 8 h then diluted with CH2Cl2. The dichloromethane solution was purified by flash chromatography on a column of silica was washed with water and brine, dried over MgSO4, filtered, gel eluted with CH2Cl2 to yield compound 11 (35 mg, 81%).
and concentrated in vacuo. The residue was purified by flash To a stirred solution of 11 (14 mg, 0.03 mmol) in a mixture of
chromatography on a column of silica gel eluted with CH2Cl2/ MeOH (2 mL) and THF (4 mL) was added TMSCHN2 (2 M in MeOH (100:1 to 50:1) to give compound 19 (29 mg, 33%) as a
hexanes, 0.1 mL, 0.2 mmol). The reaction mixture was stirred pale yellow powder. mp 213-215 °C; 1H NMR (CDCl3) δ 6.12 at room temperature for 24 h and then evaporated. Flash (s, 2H), 6.61 (s, 1H), 6.70 (s, 1H), 7.55 (m, 3H), 7.87 (m, 2H); chromatography of the residue, eluting with CH2Cl2/MeOH (40: MS (EI) m/z 282 ([M]+, base), 149.
1), gave compound 13 (13 mg, 90%) as a pale yellow powder.
5-Methoxy-6,7-(methylenedioxy)flavone (20). To a stirred
mp 238-240 °C; 1H NMR (CDCl3) δ 4.24 (s, 3H), 6.67 (s, 1H), solution of 19 (17 mg, 0.06 mmol) in a mixture of MeOH (3
6.81 (s, 1H), 7.42 (m, 6H), 7.50 (m, 3H), 7.61 (m, 4H), 7.85 (m, mL) and THF (6 mL) was added TMSCHN2 (2 M in hexanes, 2H); MS (EI) m/z 448 [M]+, 402, 371, 266, 167 (base).
0.3 mL, 0.6 mmol). The reaction mixture was stirred at room 5,6-Diacetoxy-7-(benzyloxy)flavone (14). A mixture of
temperature for 24 h and evaporated. Flash chromatography 4 (34 mg, 0.086 mmol), benzyl bromide (0.05 mL), KI (3.5 mg),
of the residue, eluting with CH2Cl2/acetone (30:1), afforded Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 compound 20 (12 mg, 68%) as a white solid. mp 202-204 °C;
34%) and 33 (12 mg, 12%) as a yellow powder, respectively.
1H NMR (CDCl3) δ 4.14 (s, 3H), 6.08 (s, 2H), 6.71 (s, 1H), 6.75 (32): mp 190-192 °C; 1H NMR (CDCl3) δ 1.55 (t, 3H, J ) 6.9
(s, 1H), 7.51 (m, 3H), 7.86 (m, 2H); MS (EI) m/z 296 [M]+, 268 Hz), 4.23 (q, 2H, J ) 6.9 Hz), 6.60 (s, 1H), 6.68 (s, 1H), 7.53 (m, 3H), 7.89 (m, 2H), 12.50 (s, 1H); 13C NMR (CDCl3) δ 14.04, 5-Hydroxy-7-methoxy-6-propyloxyflavone (21) and 5,7-
64.57, 90.44, 104.87, 105.42, 125.67, 128.50, 129.05, 130.91, Dimethoxy-6-propyloxy-flavone (22). To a stirred solution
131.17, 145.08, 150.12, 151.56, 163.46, 182.10; MS (EI) m/z of 17 (19 mg, 0.06 mmol) in a mixture of MeOH (4 mL) and
298 ([M]+, base), 283, 270, 269, 254. HRMS for C17H14O5 [M - H]+: calculated, 297.0763; found, 297.0762. (33): mp 132-
0.4 mmol). The reaction mixture was stirred at room temper- 133 °C; 1H NMR (CDCl3) δ 1.41, 1.53 (t, 6H, J ) 7.2 Hz), 4.13, ature for 24 h and evaporated. Flash chromatography of the 4.19 (q, 4H, J ) 7.2 Hz), 6.55 (s, 1H), 6.67 (s, 1H), 7.54 (m, 3H), 7.89 (m, 2H,), 12.65 (s, 1H); 13C NMR (CDCl 2Cl2/MeOH (30:1), gave compound 21
(4 mg, 20%) as a yellow powder and compound 22 (9.2 mg,
15.94, 65.18, 69.30, 91.65, 106.03, 106.55, 126.66, 129.50, 45%) as a pale yellow powder. (21): mp 112-113 °C; 1H NMR
131.83, 132.08, 132.19, 153.69, 153.79, 159.08, 164.24, 183.13; MS (EI) m/z 326 [M]+, 311, 297 (base), 269. HRMS for C 3) δ 1.12 (t, 3H, J ) 7.5 Hz), 1.95 (sextet, 2H, J ) 7.5 Hz), 3.93 (s, 3H), 4.09 (t, 2H, J ) 7.5 Hz), 6.57 (s, 1H), 6.69 (s, [M]+: calculated, 326.1154; found, 326.1155.
1H), 7.55 (m, 3H), 7.90 (m, 2H); 13C NMR (CDCl3) δ 10.91, 5-Hydroxy-6,7-(dioctyloxy)flavone (34). A mixture of 1
22.73, 61.23, 71.13, 91.77, 106.02, 106.54, 126.66, 129.49, (81 mg, 0.3 mmol), 1-iodooctane (0.16 mL), and anhydrous K2- 131.81, 132.19, 133.30, 153.55, 153.73, 159.00, 164.28, 183.11; CO3 (166 mg) in acetone (25 mL) was refluxed with stirring MS (EI) m/z 326 ([M]+, base), 283, 269. HRMS for C19H18O5 for 30 h. The reaction mixture was concentrated under reduced [M]+: calculated, 326.1154; found, 326.1149. (22): mp 135-
pressure, diluted with water (50 mL), and extracted with CH2- 136 °C; 1H NMR (CDCl3) δ 1.13 (t, 3H, J ) 7.5 Hz), 1.96 (sextet, Cl2 (50 mL × 3). The extract was washed with water and dried 2H, J ) 7.5 Hz), 3.92, 4.01 (s, 6H), 4.09 (t, 2H, J ) 7.5 Hz), over MgSO4, and the solvent was evaporated in vacuo. The 6.73 (s, 1H), 6.82 (s, 1H), 7.52 (m, 3H), 7.89 (m, 2H); 13C NMR residue was purified by flash chromatography on a column of (CDCl3) δ 10.93, 22.71, 58.92, 61.86, 62.63, 71.10, 97.28, 108.62, silica gel and eluted with CH2Cl2/MeOH (100:1 to 50:1) to give 113.62, 115.71, 121.02, 126.43, 129.38, 130.02, 131.72, 132.00, compound 34 (122 mg, 82%) as a pale yellow powder. mp 85-
141.00, 153.05, 155.02, 157.98, 161.66, 177.67; MS (EI) m/z 86 °C; 1H NMR (CDCl3) δ 0.89, 0.90 (t, 6H, J ) 6.9 Hz), 1.31- 340 [M]+, 325 (base), 283. HRMS for C20H20O5 [M]+: calcu- 1.52 (m, 20H), 1.79, 1.89 (t, 4H, J ) 6.9 Hz), 4.03, 4.07 (t, 4H, J ) 6.9 Hz), 6.53 (s, 1H), 6.64 (s, 1H), 7.51 (m, 3H), 7.86 (m, 5-Methoxy-6,7-dipropyloxyflavone (23). To a stirred
2H), 12.45 (s, 1H); MS (EI) m/z 494 [M]+, 382, 270 (base).
solution of 18 (28 mg, 0.08 mmol) in a mixture of MeOH (4
6-Ethoxy-5-hydroxy-7-methoxyflavone (38). To a stirred
mL) and THF (8 mL) was added TMSCHN2 (2 M in hexanes, solution of 32 (10 mg, 0.034 mmol) in a mixture of MeOH (3
0.32 mL, 0.64 mmol). The reaction mixture was stirred at room mL) and THF (6 mL) was added TMSCHN2 (2 M in hexanes, temperature for 24 h and evaporated. Flash chromatography 0.02 mL, 0.04 mmol). The reaction mixture was stirred at room of the residue, eluting with CH2Cl2/acetone (15:1), gave temperature for 24 h and evaporated. Flash chromatography compound 23 (23 mg, 78%) as a pale yellow powder. mp 109-
of the residue, eluting with CH2Cl2/MeOH (50:1), afforded 110 °C; 1H NMR (CDCl3) δ 1.08, 1.12 (t, 6H, J ) 7.5 Hz), 1.82, compound 38 (7.4 mg, 70%) as a pale yellow powder. mp 133-
1.95 (sextet, 4H, J ) 7.5 Hz), 3.99 (s, 3H), 4.00, 4.06 (t, 4H, J 134 °C; 1H NMR (CDCl3) δ 1.55 (t, 3H, J ) 6.9 Hz), 3.93 (s, ) 7.5 Hz), 6.72 (s, 1H), 6.81 (s, 1H), 7.52 (m, 3H), 7.89 (m, 3H), 4.21 (q, 2H, J ) 6.9 Hz), 6.57 (s, 1H), 6.69 (s, 1H), 7.56 2H); 13C NMR (CDCl3) δ 10.95, 22.76, 23.89, 62.46, 71.00, (m, 3H), 7.90 (m, 2H), 12.68 (s, 1H); 13C NMR (CDCl3) δ 15.01, 76.28, 97.19, 108.75, 113.16, 126.38, 129.36, 131.61, 132.11, 61.19, 65.25, 91.72, 106.04, 106.58, 126.66, 129.49, 131.81, 140.26, 153.17, 154.90, 158.13, 161.47, 177.69.; MS (EI) m/z 132.19, 133.29, 153.55, 153.72, 158.75, 164.30, 183.11; MS (EI) 368 [M]+, 325 (base), 283. HRMS for C22H24O5 [M]+: calcu- m/z 312 ([M]+, base), 283. HRMS for C18H16O5 [M]+: calculated, 6,7-Diacetoxy-8-bromo-5-hydroxyflavone (24). A mix-
5-Hydroxy-6,7-(dipentyloxy)flavone (40). A mixture of
ture of 4 (84 mg, 0.21 mmol) and N-bromosuccinimide (NBS,
1 (51 mg, 0.19 mmol), 1-iodopentane (0.076 mL), and anhy-
56 mg, 0.32 mmol) in THF (8 mL) and concd H2SO4 (10 μL) drous K2CO3 (110 mg) in acetone (20 mL) was refluxed with was stirred at room temperature for 48 h. The reaction mixture stirring for 24 h. The reaction mixture was concentrated under was extracted with EtOAc, washed with 10% aqueous NaHSO4 reduced pressure, diluted with water (40 mL), and extracted solution and water, dried over MgSO4, and then concentrated with CH2Cl2 (40 mL × 3). The extract was washed with water in vacuo. The residue was recrystallized from MeOH to give and dried over MgSO4 and the solvent evaporated in vacuo.
compound 24 (50 mg, 55%) as a yellow powder. mp 244-246
The residue was purified by flash chromatography on a column °C; 1H NMR (CDCl3) δ 2.38, 2.44 (each s, 6H), 6.82 (s, 1H), of silica gel and eluted with CH2Cl2/MeOH (60:1) to give 7.59 (m, 3H), 8.01 (m, 2H); MS (EI) m/z 434 [M + 2]+, 432 compound 40 (68 mg, 87%) as a pale yellow powder. mp 111-
112 °C; 1H NMR (CDCl3) δ 0.94, 0.97 (t, 6H, J ) 6.6 Hz), 1.39- 8-Bromo-5,6,7-trihydroxyflavone (25). A mixture of 1 (35
1.52 (m, 8H), 1.81, 1.91 (quintet, 4H, J ) 6.6 Hz), 4.05, 4.09 mg, 0.13 mmol) and NBS (33 mg, 0.19 mmol) in THF (4 mL) (t, 4H, J ) 6.6 Hz), 6.56 (s, 1H), 6.68 (s, 1H), 7.54 (m, 3H), 7.90 (m, 2H); MS (EI) m/z 410 [M]+.
2SO4 (5 μL) was stirred at room temperature for 12 h. The reaction mixture was extracted with EtOAc, washed 5-Methoxy-6,7-(dipentyloxy)flavone (41). To a stirred
with 10% aqueous NaHSO4 solution and water, dried over solution of 40 (33 mg, 0.08 mmol) in a mixture of MeOH (4
MgSO4, and then concentrated in vacuo. The residue was mL) and THF (8 mL) was added TMSCHN2 (2 M in hexanes, recrystallized from MeOH to give compound 25 (26 mg, 57%)
0.32 mL, 0.64 mmol). The reaction mixture was stirred at room as a yellow powder. mp 263-265 °C; 1H NMR (DMSO-d6) δ temperature for 24 h and evaporated. Flash chromatography 7.07 (s, 1H), 7.59 (m, 3H), 8.12 (m, 2H), 9.58, 10.92, 12.76 (s, of the residue, eluting with CH2Cl2/acetone (20:1), gave 3H); MS (EI) m/z 350 [M + 2]+, 348 ([M]+, base), 270.
compound 41 (32.6 mg, 98%) as a white solid. mp 107-108
6-Ethoxy-5,7-dihydroxyflavone (32) and 6,7-Diethoxy-
°C; 1H NMR (CDCl3) δ 0.95, 0.97 (t, 6H, J ) 6.6 Hz), 1.39- 5-hydroxyflavone (33). A mixture of 1 (81 mg, 0.3 mmol),
1.54 (m, 8H), 1.81, 1.93 (quintet, 4H, J ) 6.6 Hz), 3.99 (s, 3H), 4.03, 4.10 (t, 4H, J ) 6.6 Hz), 6.78 (s, 1H), 6.81 (s, 1H), 7.53 acetone (25 mL) was refluxed with stirring for 18 h. The (m, 3H), 7.90 (m, 2H); MS (EI) m/z 424 [M]+.
reaction mixture was concentrated under reduced pressure, 6,7-(Dihexyloxy)-5-hydroxyflavone (42). A mixture of 1
diluted with water (50 mL), and extracted with CH2Cl2 (50 (54 mg, 0.2 mmol), 1-bromohexane (0.084 mL), and anhydrous mL × 3). The extract was washed with water and dried over K2CO3 (110 mg) in acetone (20 mL) was refluxed with stirring MgSO4 and the solvent evaporated in vacuo. The residue was for 24 h. The reaction mixture was concentrated under reduced purified by flash chromatography on a column of silica gel and pressure, diluted with water (40 mL) and extracted with CH2- eluted with CH2Cl2/MeOH (50:1) to give compounds 32 (30 mg,
Cl2 (40 mL × 3). The extract was washed with water and dried Anti-P-glycoprotein Activity of Baicalein Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 over MgSO4 and the solvent evaporated in vacuo. The residue were transferred to scintillation vials and counted with a was purified by flash chromatography on a column of silica -counter (Beckman, model LS 5000) after adding 10 mL gel and eluted with CH2Cl2/MeOH (60:1) to give compound 42
scintillation fluids (SafeScint, American Bioanalytical Co., (71.4 mg, 82%) as a pale yellow powder. mp 96-97 °C; 1H NMR Natick, MA). Data presented are the mean of three indepen- (CDCl3) δ 0.92, 0.93 (t, 6H, J ) 6.6 Hz), 1.32-1.41 (m, 8H), 1.48-1.53 (m, 4H), 1.80, 1.90 (quintet, 4H, J ) 6.6 Hz), 4.05, Growth Inhibitory Assay. Approximately 104 KB or KB/
4.09 (t, 4H, J ) 6.6 Hz), 6.56 (s, 1H), 6.68 (s, 1H), 7.54 (m, MDR cells were seeded into 24-well tissue culture plates in 3H), 7.90 (m, 2H); MS (EI) m/z 438 [M]+.
RPMI 1640 medium plus 10% fetal bovine serum for 24 h, after 6,7-(Dihexyloxy)-5-methoxyflavone (43). To a stirred
which the cells were treated with various concentrations of solution of 42 (47 mg, 0.1 mmol) in a mixture of MeOH (4 mL)
synthetic flavones in culture medium for 3 days. The cells were then fixed and stained with methylene blue in 50% MeOH, mL, 0.8 mmol). The reaction mixture was stirred at room washed thoroughly with tap water, and dissolved with 0.5 mL temperature for 24 h and evaporated. Flash chromatography of 0.5% sarcosyl.39 The amount of cellular protein, which is proportional to the cell number, is estimated by the absorption compound 43 (41 mg, 91%) as a white solid. mp 93-95 °C; 1H
(OD595 nm). The growth inhibitory assay, which is presented 3) δ 0.92, 0.94 (t, 6H, J ) 6.6 Hz), 1.30-1.41 (m, 50 value, represents the concentration of compound 8H), 1.51-1.55 (m, 4H), 1.80, 1.92 (quintet, 4H, J ) 6.6 Hz), required to inhibit 50% of cell growth. The cell doubling time 3.99 (s, 3H), 4.03, 4.10 (t, 4H, J ) 6.6 Hz), 6.76 (s, 1H), 6.81 of KB and KB/MDR cells is about 20 to 24 h. The data (s, 1H), 7.52 (m, 3H), 7.90 (m, 2H); MS (EI) m/z 452 [M]+.
presented are the mean of three independent experiments.
6,7-Diethoxy-5-methoxyflavone (44). To a stirred solu-
Acknowledgment. The corresponding author is a
tion of 33 (33 mg, 0.1 mmol) in a mixture of MeOH (4 mL)
and THF (8 mL) was added TMSCHN
fellow of the National Foundation for Cancer Research.
mL, 0.8 mmol). The reaction mixture was stirred at room Part of this work was supported by National Foundation temperature for 24 h and evaporated. Flash chromatography for Cancer Research (YCC) and the National Institutes of the residue, eluting with CH2Cl2/acetone (10:1 to 5:1), afforded compound 44 (33.8 mg, 99%) as a white solid. mp
126-128 °C; 1H NMR (CDCl3) δ 1.42, 1.55 (t, 6H, J ) 6.9 Hz),
Supporting Information Available: HPLC analytical
4.00 (s, 3H), 4.13, 4.19 (q, 4H, J ) 6.9 Hz), 6.77 (s, 1H), 6.81 data. This material is available free of charge via the Internet (s, 1H), 7.52 (m, 3H), 7.90 (m, 2H); 13C NMR (CDCl3) δ 14.94, 16.05, 62.39, 65.13, 70.22, 97.15, 108.78, 113.18, 126.37,129.36, 131.61, 132.10, 139.93, 153.27, 154.92, 157.94, 161.46, References
177.67; MS (EI) m/z 340 [M]+. HRMS for C20H20O5 [M]+: (1) Germann, U. A. P-glycoproteinsa mediator of multidrug resis- calculated, 340.1311; found, 340.1315.
tance in tumour cells. Eur. J. Cancer 1996, 32A, 927-944.
6,7-Dibutoxy-5-hydroxyflavone (45). A mixture of 1 (81
(2) Gottesman, M. M.; Pastan, I. Biochemistry of multidrug resis- mg, 0.3 mmol), 1-iodobutane (0.1 mL), and anhydrous K tance mediated by the multidrug transporter. Annu. Rev. (165 mg) in acetone (20 mL) was refluxed with stirring for 24 Biochem. 1993, 62, 385-427.
(3) Thiebaut, F.; Tsuruo, T.; Hamada, H.; Gottesman, M. M.; Pastan, h. The reaction mixture was concentrated under reduced I. et al. Cellular localization of the multidrug-resistance gene pressure, diluted with water (50 mL), and extracted with CH2- product P-glycoprotein in normal human tissues. Proc. Natl. Cl2 (3 × 50 mL). The extract was washed with water and dried Acad. Sci. U.S.A. 1987, 84, 7735-7738.
(4) Borst, P.; Schinkel, A. H. What have we learnt thus far from 4, and the solvent was evaporated in vacuo. The residue was purified by flash chromatography on a column of mice with disrupted P-glycoprotein genes? Eur. J. Cancer 1996,
32A, 985-990.
silica gel eluted with CH2Cl2/MeOH (60:1) to give compound (5) Luker, G. D.; Nilsson, K. R.; Covey, D. F.; Piwnica-Worms, D.
45 (86 mg, 75%) as a yellow powder. mp 116-117 °C; 1H NMR
Multidrug resistance (MDR1) P-glycoprotein enhances esterifi- (CDCl3) δ 0.99, 1.02 (t, 6H, J ) 6.6 Hz), 1.53-1.58 (m, 4H), cation of plasma membrane cholesterol. J. Biol. Chem. 1999, 274,
1.79, 1.89 (quintet, 4H, J ) 6.6 Hz), 4.06, 4.10 (t, 4H, J ) 6.6 Hz), 6.56 (s, 1H), 6.68 (s, 1H), 7.54 (m, 3H), 7.90 (m, 2H); 13C (6) Dalton, W. S.; Grogan, T. M.; Meltzer, P. S.; Scheper, R. J.; Durie, B. G. et al. Drug-resistance in multiple myeloma and non- 3) δ 14.17, 14.31, 19.55, 19.62, 31.40, 32.61, 69.26, Hodgkin’s lymphoma: detection of P-glycoprotein and potential 73.51, 91.65, 106.00, 106.52, 126.63, 129.47, 131.86, 132.13, circumvention by addition of verapamil to chemotherapy. J. Clin. 132.49, 153.63, 153.71, 159.25, 164.17, 183.11; MS (EI) m/z Oncol. 1989, 7, 415-424.
382 [M]+. HRMS for C23H26O5 [M]+: calculated, 382.1780; (7) Miller, T. P.; Grogan, T. M.; Dalton, W. S.; Spier, C. M.; Scheper, R. J. et al. P-glycoprotein expression in malignant lymphomaand reversal of clinical drug resistance with chemotherapy plus 6,7-Dibutoxy-5-methoxyflavone (46). To a stirred solu-
high-dose verapamil. J. Clin. Oncol. 1991, 9, 17-24.
tion of 45 (49 mg, 0.128 mmol) in a mixture of MeOH (5 mL)
(8) Nuessler, V.; Scheulen, M. E.; Oberneder, R.; Kriegmair, M.; and THF (10 mL) was added TMSCHN2 (2 M in hexanes, 0.5 Goebel, K. J. et al. Phase I and pharmacokinetic study of the mL, 1 mmol). The reaction mixture was stirred at room P-glycoprotein modulator dexniguldipine-HCL. Eur. J. Med. Res. temperature for 24 h and then evaporated. Flash chromatog- 1997, 2, 55-61.
(9) Wilson, W. H.; Jamis-Dow, C.; Bryant, G.; Balis, F. M.; Klecker, raphy of the residue, eluting with CH2Cl2/acetone (15:1), gave R. W. et al. Phase I and pharmacokinetic study of the multidrug compound 46 (46 mg, 91%) as a white solid. mp 103-105 °C;
resistance modulator dexverapamil with EPOCH chemotherapy.
1H NMR (CDCl3) δ 0.99, 1.03 (t, 6H, J ) 6.6 Hz), 1.55-1.59 J. Clin. Oncol. 1995, 13, 1985-1994.
(m, 4H), 1.78, 1.91 (quintet, 4H, J ) 6.6 Hz), 3.99 (s, 3H), 4.04, (10) Hyafil, F.; Vergely, C.; Du Vignaud, P.; Grand-Perret, T. In vitro 4.11 (t, 4H, J ) 6.6 Hz), 6.81 (s, 1H), 6.82 (s, 1H), 7.53 (m, and in vivo reversal of multidrug resistance by GF120918, an acridonecarboxamide derivative. Cancer Res. 1993, 53, 4595-
19.63, 31.37, 32.69, 62.44, 69.19, 74.33, 97.17, 108.78, 113.18, (11) Boesch, D.; Gaveriaux, C.; Jachez, B.; Pourtier-Manzanedo, A.; 126.36, 129.34, 129.63, 130.21, 131.58, 132.13, 140.25, 153.18, Bollinger, P. et al. In vivo circumvention of P-glycoprotein- 154.89, 158.11, 161.41, 177.68; MS (EI) m/z 396 [M]+. HRMS mediated multidrug resistance of tumor cells with SDZ PSC 833.
Cancer Res. 1991, 51, 4226-4233.
calculated, 396.1937; found, 396.1939.
(12) Ramu, A.; Spanier, R.; Rahamimoff, H.; Fuks, Z. Restoration of Vinblastine Uptake by the KB/MDR Cells. KB/MDR
doxorubicin responsiveness in doxorubicin-resistant P388 mu- cells,38 which overexpress human P-gp 170 protein, were rine leukaemia cells. Br. J. Cancer 1984, 50, 501-507.
seeded into 24-well tissue culture plates in RPMI 1640 medium (13) Middleton, E., Jr.; Kandaswami, C.; Theoharides, T. C. The plus 10% fetal bovine serum for 24 h without the selecting effects of plant flavonoids on mammalian cells: implications for agent doxorubicin (37 nM). The cells were then treated with inflammation, heart disease, and cancer. Pharmacol. Rev. 2000,
52, 673-751.
[3H]vinblastine plus synthetic flavones in HBSS for 60 min.
(14) Bailey, D. G.; Malcolm, J.; Arnold, O.; Spence, J. D. Grapefruit The cells were washed twice with ice-cold PBS, harvested with juice-drug interactions. Br. J. Clin. Pharmacol. 1998, 46, 101-
1 N NaOH, and neutralized with 1 N HCl. The cell aliquots Journal of Medicinal Chemistry, 2004, Vol. 47, No. 22 (15) Ducharme, M. P.; Warbasse, L. H.; Edwards, D. J. Disposition (27) Ferriola, P. C.; Cody, V.; Middleton, E., Jr. Protein kinase C of intravenous and oral cyclosporine after administration with inhibition by plant flavonoids. Kinetic mechanisms and structure- grapefruit juice. Clin. Pharmacol. Ther. 1995, 57, 485-491.
activity relationships. Biochem. Pharmacol. 1989, 38, 1617-
(16) De Vincenzo, R.; Scambia, G.; Benedetti Panici, P.; Ranelletti, F. O.; Bonanno, G. et al. Effect of synthetic and naturally (28) Hagiwara, M.; Inoue, S.; Tanaka, T.; Nunoki, K.; Ito, M. et al.
occurring chalcones on ovarian cancer cell growth: structure- Differential effects of flavonoids as inhibitors of tyrosine protein activity relationships. Anticancer Drug Des. 1995, 10, 481-490.
kinases and serine/threonine protein kinases. Biochem. Phar- (17) Murakami, S.; Muramatsu, M.; Tomisawa, K. Inhibition of macol. 1988, 37, 2987-2992.
gastric H+, K(+)-ATPase by flavonoids: a structure-activity (29) Perez-Victoria, J. M.; Chiquero, M. J.; Conseil, G.; Dayan, G.; study. J. Enzyme Inhib. 1999, 14, 151-166.
Di Pietro, A. et al. Correlation between the affinity of flavonoids (18) Akiyama, T.; Ishida, J.; Nakagawa, S.; Ogawara, H.; Watanabe, binding to the cytosolic site of Leishmania tropica multidrug S. et al. Genistein, a specific inhibitor of tyrosine-specific protein transporter and their efficiency to revert parasite resistance to kinases. J. Biol. Chem. 1987, 262, 5592-5595.
daunomycin. Biochemistry 1999, 38, 1736-1743.
(19) De Azevedo, W. F., Jr.; Mueller-Dieckmann, H. J.; Schulze- (30) Lo, A.; Burckart, G. J. P-glycoprotein and drug therapy in organ Gahmen, U.; Worland, P. J.; Sausville, E. et al. Structural basis transplantation. J. Clin. Pharmacol. 1999, 39, 995-1005.
for specificity and potency of a flavonoid inhibitor of human (31) Wang, R. B.; Kuo, C. L.; Lien, L. L.; Lien, E. J. Structure-activity CDK2, a cell cycle kinase. Proc. Natl. Acad. Sci. U.S.A. 1996,
relationship: analyses of p-glycoprotein substrates and inhibi- tors. J. Clin. Pharm. Ther. 2003, 28, 203-228.
(20) Boumendjel, A.; Di Pietro, A.; Dumontet, C.; Barron, D. Recent (32) Di Pietro, A.; Conseil, G.; Perez-Victoria, J. M.; Dayan, G.; advances in the discovery of flavonoids and analogues with high- Baubichon-Cortay, H. et al. Modulation by flavonoids of cell affinity binding to P-glycoprotein responsible for cancer cell multidrug resistance mediated by P-glycoprotein and related multidrug resistance. Med. Res. Rev. 2002, 22, 512-529.
ABC transporters. Cell Mol. Life Sci. 2002, 59, 307-322.
(21) Efferth, T.; Davey, M.; Olbrich, A.; Rucker, G.; Gebhart, E. et (33) Chambers, T. C.; Pohl, J.; Raynor, R. L.; Kuo, J. F. Identification al. Activity of drugs from traditional Chinese medicine toward of specific sites in human P-glycoprotein phosphorylated by sensitive and MDR1- or MRP1-overexpressing multidrug- protein kinase C. J. Biol. Chem. 1993, 268, 4592-4595.
resistant human CCRF-CEM leukemia cells. Blood Cells Mol. (34) Callaghan, R.; Higgins, C. F. Interaction of tamoxifen with the Dis. 2002, 28, 160-168.
multidrug resistance P-glycoprotein. Br. J. Cancer 1995, 71,
(22) Conseil, G.; Baubichon-Cortay, H.; Dayan, G.; Jault, J. M.; Barron, D. et al. Flavonoids: a class of modulators with (35) Smith, C. D.; Zilfou, J. T. Circumvention of P-glycoprotein- bifunctional interactions at vicinal ATP- and steroid-binding mediated multiple drug resistance by phosphorylation modula- sites on mouse P-glycoprotein. Proc. Natl. Acad. Sci. U.S.A. 1998,
tors is independent of protein kinases. J. Biol. Chem. 1995, 270,
(23) Perez-Victoria, J. M.; Perez-Victoria, F. J.; Conseil, G.; Maitre- (36) Conseil, G.; Perez-Victoria, J. M.; Jault, J. M.; Gamarro, F.; jean, M.; Comte, G. et al. High-affinity binding of silybin Goffeau, A. et al. Protein kinase C effectors bind to multidrug derivatives to the nucleotide-binding domain of a Leishmania ABC transporters and inhibit their activity. Biochemistry 2001,
tropica P-glycoprotein-like transporter and chemosensitization of a multidrug-resistant parasite to daunomycin. Antimicrob (37) Dayan, G.; Jault, J. M.; Baubichon-Cortay, H.; Baggetto, Agents Chemother 2001, 45, 439-446.
L. G.; Renoir, J. M. et al. Binding of steroid modulators to (24) Thiyagarajah, P.; Kuttan, S. C.; Lim, S. C.; Teo, T. S.; Das, N.
recombinant cytosolic domain from mouse P-glycoprotein in close P. Effect of myricetin and other flavonoids on the liver plasma proximity to the ATP site. Biochemistry 1997, 36, 15208-15215.
membrane Ca2+ pump. Kinetics and structure-function rela- (38) Chen, H. X.; Bamberger, U.; Heckel, A.; Guo, X.; Cheng, Y. C.
tionships. Biochem. Pharmacol. 1991, 41, 669-675.
BIBW 22, a dipyridamole analogue, acts as a bifunctional (25) Hirano, T.; Oka, K.; Akiba, M. Effects of synthetic and naturally modulator on tumor cells by influencing both P-glycoprotein and occurring flavonoids on Na+, K+-ATPase: aspects of the nucleoside transport. Cancer Res. 1993, 53, 1974-1977.
structure-activity relationship and action mechanism. Life Sci. (39) Park, S. Y.; Lam, W.; Cheng, Y. C. X-ray repair cross- 1989, 45, 1111-1117.
complementing gene I protein plays an important role in (26) Jinsart, W.; Ternai, B.; Polya, G. M. Inhibition of rat liver cyclic camptothecin resistance. Cancer Res. 2002, 62, 459-465.
AMP-dependent protein kinase by flavonoids. Biol. Chem. Hoppe
Seyler 1992, 373, 205-211.

Source: http://engfac.cooper.edu/pages/rsavizky/uploads/Savizky_JMedChem_20040.pdf