76
A. Boumendjel et al. / Bioorg. Med. Chem. Lett. 11 (2001) 75±77
Hydroxyl groups at positions 3 and 5 are essential for
high-anity binding to P-gp. In a recent study con-
ducted with chalcones, we have shown that the presence
of a hydrophobic substituent on the B-ring (at 4-posi-
tion; Fig. 1) considerably enhanced the binding a-
nity.5,6 However, the instability of chalcones (due to
their cyclization to give inactive ¯avanones) makes them
less desirables for further studies. In order to prepare
high-anity and much more stables ¯avonols to be used
for in vivo studies, we have targeted and synthesized a
series of ¯avonols (3,5,7-trihydroxy¯avones) with
dierent hydrophobic substituents at the 40-position.
with NBD2 with a KD of 5.9 Æ 0.2 mM. Addition of a
hydroxyl group at position 40 of the B-ring produced a
slightly negative eect when comparing kaempferol (6.7
Æ 0.3 mM) to galangin. On the contrary, the presence of
a hydrophobic methoxy group (in kaempferide) slightly
increased the binding anity as monitored by quenching
(KD=5.0 Æ 0.2 mM).
As seen from Table 1, no signi®cant eect was produced
by introducing a ¯uorine in 4a (KD=6.8 Æ 0.28 mM),
whereas two chlorine atoms at positions 20 and 40 in 4b
produced a positive eect (KD=4.0 Æ 0.23 mM). A
much higher increase in anity was observed by intro-
ducing an iodine in 4d (KD=1.06 Æ 0.08 mM), which
was probably due to its hydrophobic character as con-
cluded previously with chalcone derivatives.5 Interest-
ingly, an intermediate 3-fold increase in anity was
observed by replacing the complete B-ring with a
diphenylmethyl substituent in 4c. Finally, the most
striking increase in binding anity was observed by an
octyl group which produced a 93-fold higher anity in
4e as compared to galangin. The choice of this sub-
stituent relies on our previous study conducted on chal-
cones which shows that an octyl group is the most
optimal substituent, and further increase in the chain
length began to alter the binding. The fact that the
maximal quenching of ¯uorescence was signi®cantly
reduced (62.1 Æ 3.3%) as compared to all other com-
pounds (ꢀ84.6 Æ 2.0%) might indicate a dierently
orientated binding that would dierentially modify
tryptophan environment.
Chemistry
Flavonols 4 were synthesized in two steps starting from
2,4,6-trihydroxy-o-methoxyacetophenone 1 (1 was pre-
pared from phloroglucinol and methoxyacetonitrile
according to Wahala and Hase).7 Condensation of 1
with an appropriately substituted benzoyl chloride in
the presence of K2CO3 in acetone aorded directly the
diketone 2 which was not isolated but cyclized by treat-
ment with AcOH to give 5,7-dihydroxy-3-methoxy-
¯avones 3. In one case (R=F), the expected 5,7-
dihydroxy-3-methoxy¯avone 3a was not obtained. The
main product was the 7-¯uorobenzoyl-5-hydroxy-3-
methoxy¯avone. The latter was then treated with KOH
in MeOH to give 3a in a quantitative yield. Demethyla-
tion of methoxy¯avones 3 with BBr3 in CH2Cl2 gave
¯avonols 4 (Scheme 1).
Table 1. Role of ¯avonol B-ring substituents on the binding anity
for NBD2
Results and Discussion
Compound
R
KD
(mM)
ÁFmax
(%)
The binding of ¯avonols to the puri®ed C-terminal
nucleotide-binding domain (NBD2) of P-gp was directly
measured by the quenching of protein intrinsic ¯uores-
cence due to a single tryptophan residue, as previously
described for other classes of ¯avonoids.4 The results of
this study are listed in Table 1.
Galangin
Kaempferol
Kaempferide
4a
4b
4c
4d
4e
H
40-OH
40-OCH3
40-F
5.9 Æ 0.2
6.7 Æ 0.3
95.0 Æ 1.6
97.6 Æ 1.0
96.7 Æ 0.6
91.6 Æ 1.2
85.3 Æ 1.3
84.6 Æ 2.0
95.5 Æ 2.0
62.1 Æ 3.3
5.0 Æ 0.2
6.8 Æ 0.28
4.0 Æ 0.23
2.0 Æ 0.22
1.06 Æ 0.08
0.057 Æ 0.02
20, 40-Cl
PhR= CH(Ph)2
40-I
Our lead compound was galangin (3,5,7-trihydroxy-
¯avone) which was used as a reference and interacted
40-n-C8H17
Scheme 1.