Anti-AIDS agents 66: Syntheses and anti-HIV activity of phenolic and aza 3′,4′-di-O-(-)-camphanoyl-(+)-cis-khellactone (DCK) derivatives

Article abstract of DOI:10.1016/j.bmc.2006.12.013

New phenolic and aza 3′,4′-di-O-(-)-camphanoyl-(+)-cis-khellactone (DCK) analogs were synthesized and assayed for inhibition of HIV-1 IIIB replication in H9 lymphocytes. Compound 16, 4-methyl-1′-aza-DCK (4-methyl-aza-DCK), was less lipophilic than 4-methyl-DCK, and retained sub-micromolar anti-HIV activity with EC₅₀ and TI values of 0.77 μM and >42, respectively. Moreover, it showed moderately improved metabolic stability. Introduction of phenolic hydroxyl groups to 4-methyl-DCK decreased lipophilicity significantly, but did not improve metabolic stability and also decreased activity.

Full text of DOI:10.1016/j.bmc.2006.12.013

Bioorganic & Medicinal Chemistry 15 (2007) 6852–6858
Anti-AIDS agents 66: Syntheses and anti-HIV activity of
phenolic and aza 30,40-di-O-(ꢀ)-camphanoyl-(+)-
cis-khellactone (DCK) derivatives
Madoka Suzuki,a Donglei Yu,a Susan L. Morris-Natschke,a
Philip C. Smithb and Kuo-Hsiung Leea,*
aNatural Products Research Laboratories, School of Pharmacy, University of North Carolina, Chapel Hill, NC 27599-7360, USA
bDivision of Molecular Pharmaceutics, School of Pharmacy, University of North Carolina, Chapel Hill, NC 27599-7360, USA
Received 3 February 2006; revised 27 November 2006; accepted 11 December 2006
Available online 13 December 2006
Abstract—New phenolic and aza 30,40-di-O-(ꢀ)-camphanoyl-(+)-cis-khellactone (DCK) analogs were synthesized and assayed for
inhibition of HIV-1 IIIB replication in H9 lymphocytes. Compound 16, 4-methyl-10-aza-DCK (4-methyl-aza-DCK), was less
lipophilic than 4-methyl-DCK, and retained sub-micromolar anti-HIV activity with EC50 and TI values of 0.77 lM and >42, respec-
tively. Moreover, it showed moderately improved metabolic stability. Introduction of phenolic hydroxyl groups to 4-methyl-DCK
decreased lipophilicity significantly, but did not improve metabolic stability and also decreased activity.
ꢀ 2007 Published by Elsevier Ltd.
1. Introduction
R
5
4
30,40-Di-O-S-(ꢀ)-camphanoyl-30R,40R-(+)-cis-khellactone
(DCK, Fig. 1) is a potent anti-HIV agent that suppresses
HIV replication with sub-micromolar EC50 values and
acts by inhibiting reverse transcriptase DNA-dependent
DNA polymerization activity.1 Initial structure–activity
relationships (SAR) have been established for this
compound type.2,3 The previously synthesized DCK
derivatives were substituted with methyl [for instance,
4-methyl-DCK (Fig. 1)], methoxy, halogen, and
hydroxymethyl groups on the (+)-cis-khellactone nucle-
us. These non-polar groups were shown to be beneficial
for activity at positions 3, 4, and 5, but detrimental at
position-6. Slightly more polar hydroxymethyl deriva-
tives were moderately active, even when the substituent
was introduced at position-6. The previous studies also
indicated that, despite their high potencies, most DCK
derivatives show low oral bioavailability in male Spra-
gue–Dawley rats and are rapidly metabolized both in vi-
tro and in vivo.4 The most recent metabolic study of
DCKs suggested camphanoyl moieties as the target of
metabolism;5 however, they are also reported to be the
6
3
1'
O
O
O
DCK: R = H
4-Methyl-DCK: R = CH3
O
O
O
O
O
O
O
O
Figure 1. Structures of DCK and 4-methyl-DCK.
important determinants for the activity.6 Thus, we were
also concerned with improving these pharmaceutical
parameters, which are significant to efficient drug
design.7,8
In this study, we introduced a phenolic hydroxyl (a
moderately polar isostere of a methyl group) directly
on the khellactone nucleus or nitrogen (an isostere of
oxygen) in the dimethylpyran-ring in order to extend
SAR beyond the prior lipophilic substituents. In addi-
tion, introduction of polar substituents can also enhance
Keywords: DCK; Khellactone; HIV; aza-DCK.
*
Corresponding author. Tel.: +1 919 962 0006; fax: +1 919 966
3893; e-mail: khlee@unc.edu
0968-0896/$ - see front matter ꢀ 2007 Published by Elsevier Ltd.
doi:10.1016/j.bmc.2006.12.013

M. Suzuki et al. / Bioorg. Med. Chem. 15 (2007) 6852–6858
6853
aqueous solubility, an important parameter for oral bio-
availability. The new more polar analogs were postulat-
ed to have lower lipophilicity (in terms of logD)
compared to the non-polar DCK equivalents. The logD
values for a subset of these compounds were calculated
by using ACD/LogD Sol Suite software. The calculated
logD values (pH 7.4) of 13a, 15, and 16 are 0.52, 1.91,
and 3.97, respectively, while our lead compound 4-meth-
yl-DCK has a calculated logD value of 4.82. According-
ly, compounds 13a, 15, and 16, and related compounds
were synthesized and evaluated in vitro against HIV-1
replication in H9 lymphocytes. The in vitro metabolic
stabilities of these compounds were also tested with hu-
man liver microsomes under oxidative conditions.
refluxing pyridine (Scheme 1B). Commercially available
4-methylesculetin (4) was used as the substrate for the
initial SAR investigation. The structural arrangement
of the pyran ring of the product seselin 5 was confirmed
to be 6-hydroxyseselin based on NOESY data, and was
distinguished from its isomer. 4-Methyl-10-azaseselin (9)
was synthesized from the corresponding 7-aminocouma-
rin (7) by Aza-Claisen rearrangement (Scheme 1 C).
Addition of alkyne to 7 was catalyzed by copper and cu-
prous chloride in water and triethylamine.12 The prod-
uct 8 was then heated in the presence of copper
chloride to obtain 4-methyl-10-azaseselin (9).13 Accord-
ing to Scheme 2, the asymmetric dihydroxylation of
3a–c, 6, and 9 was accomplished using a catalytic sharp-
less asymmetric dihydroxylation (SAD) with AD-mix-a
[K2OsO2(OH)4 (cat.), (DHQ)2Pyr, K2CO3, and
K3FeCN6]. Esterification of the khellactones with
camphanoyl chloride gave the final products 4-hy-
droxy-DCKs 13a–c and 4-methyl-10-aza-DCK 16.
6-Hydroxy-4-methyl-DCK 15 was obtained by
deprotecting intermediate 14.
2. Synthesis
Seselins (3a–c) were synthesized as shown (Scheme 1). In
order to introduce a hydroxyl at the 4-position of DCK,
the 20,20-dimethylpyran-ring was added prior to couma-
rin formation (Scheme 1A), in contrast to prior synthet-
ic methodology for DCK analogs. Chromenes 2a–c were
synthesized by refluxing 1a–c, which were commercially
available (1a–b) or prepared by Friedel Crafts acetyla-
tion of resorcinol9 (1c), with 4,4-dimethoxy-2-methyl-
2-butanol in pyridine at 150 ꢁC. Cyclization using
diethyl carbonate gave the desired 4-hydroxyseselins
(3a–c).10,11 Introduction of hydroxyl at the seselin
6-position was achieved by condensing a 6,7-dihydroxy-
coumarin with 4,4-dimethoxy-2-methyl-2-butanol in
3. Results and discussion
The logD values of target compounds at pH 7.4 were
calculated by using ACD/LogD Sol Suite software and
are listed in Table 1. These DCK analogs were incubat-
ed with human liver microsomes under oxidative condi-
tions, and substrate depletion rates from the microsomal
incubation were analyzed by high-performance liquid
R2
R2
OH
O
O
A
R2
O
R1
O
R1
i
ii
R1
O
O
OH
HO
OH
1a: R1=R2=H
1b: R1=Me, R2=H
1c: R1=H, R2=Me
2a: R1=R2=H
2b: R1=Me, R2=H
2c: R1=H, R2=Me
3a: R1=R2=H
3b: R1=Me, R2=H
3c: R1=H, R2=Me
B
iii
v
iv
TBSO
O
HO
HO
O
O
HO
O
O
O
O
O
4
7
5
8
6
9
C
vi
HN
O
O
H2N
O
O
HN
O
O
Scheme 1. Synthesis of seselins. Reagents and conditions: (A) Synthesis of 4-hydroxyseselins. (i) 4,4-Dimethoxy-2-methyl-2-butanol
[HO(CH3)2CCH2CH(OMe)2], pyridine 150 ꢁC; (ii) diethyl carbonate, NaH, 220 ꢁC, 30 min. (B) Synthesis of 6-hydroxyseselin. (iii) 4,4-
Dimethoxy-2-methyl-2-butanol, pyridine 150 ꢁC; (iv) imidazole, TBSCl. (C) Synthesis of 10-azaseselin. (v) 3-Chloro-3-methyl-1-butyne, CuCl, Cu,
Et3N, H2O, reflux; (vi) CuCl, THF, reflux.

6854
M. Suzuki et al. / Bioorg. Med. Chem. 15 (2007) 6852–6858
R3
R2
O
R4
R1
O
R3
R2
O
R3
R2
O
X
R4
X
R1
O
R4
X
R1
O
vii
viii
O
O
O
O
O
OH
OH
O
O
O
10a: R1=R3=R4=H,
R2=OH, X=O
10b: R1=Me, R2=OH,
R3= R4=H, X=O
10c: R1=R4=H, R2=OH,
R3=Me, X=O
11: R1= R3=H, R2=Me,
R4=OTBS, X=O
13a: R1=R3=R4=H,
3a: R1=R3=R4=H,
R2=OH, X=O
3b: R1=Me, R2=OH,
R3= R4=H, X=O
3c: R1=R4=H, R2=OH,
R3=Me, X=O
6: R1= R3=H, R2=Me,
R4=OTBS, X=O
9: R1= R3=R4=H,
R2=Me, X=NH
R2=OH, X=O
13b: R1=Me, R2=OH,
R3= R4=H, X=O
13c: R1=R4=H, R2=OH,
R3=Me, X=O
14: R1= R3=H, R2=Me,
R4=OTBS, X=O
15: R1= R3=H, R2=Me,
R4=OH, X=O
ix
12: R1= R3=R4=H,
R2=Me, X=NH
16: R1= R3=R4=H,
R2=Me, X=NH
Scheme 2. Syntheses of DCK analogs. Reagents and condition: (vii) SAD: K2OsO2(OH)4, K2CO3, K3Fe(CN)6, (DHQ)2Pyr, t-BuOH/H2O (v/v, 1:1),
0 ꢁC; (viii) (S)-(ꢀ)-camphanoyl chloride, pyridine, CH2Cl2; (ix) CeF, THF.
Table 2. Metabolic stability values of selected DCK analogs
Table 1. Anti-HIV activity data and calculated lipophilicity values of
DCK analogs
Compound
t1/2 in vitroa (min)
a
CLint (L/min/kg)
Compound
logDa
EC50 (lM)
IC50 (lM)
TId
b
c
13a
13b
NDb
ND
2.0 0.3
ND
3.5 0.5
ND
13a
13b
0.52
1.11
1.50
1.91
3.97
4.82
15
>40
31
>3.0
15
13c
15
2.1
1.2
3.1 0.2
9.6 0.7
3.4 0.3
2.3 0.3
0.7 0.1
2.0 0.2
13c
15
14
12
16
4-Methyl-DCK
1.5
>38
>32
>10
1900
>25
>42
>500
38,000
16
4-Methyl-DCK
AZT
0.77
0.02
0.05
a See calculations in Section 5.
b ND, not determined.
a From ACD/LogD Sol Suite software.
b EC50, concentration required to inhibit the replication of HIV in cell
culture by 50%.
c IC50, concentration required to reduce host cell viability by 50%.
ance utilized scaling factors applied to the observed in vi-
tro data, and allowed comparison with data obtained
using different protein concentrations. The Clint value
of 16 (0.7 L/min/kg) is still larger than reported values
of high clearance drugs;17 however, in consideration of
the high protein binding (unpublished data), 16 is likely
to show more moderate clearance in vivo. Additionally,
although hydroxymethyl DCK is reported to show mod-
erate bioavailability and in vitro half life in rats,4 com-
pound 16 is the only known analog so far that has
comparable in vitro half life in human liver microsomes.
d TI, in vitro therapeutic index, ratio of IC50:EC50
.
chromatography (HPLC) linked with a UV detector. In
vitro half life and intrinsic clearance values (Clint), which
are frequently used in both academia and the pharma-
ceutical industry to estimate in vivo metabolic stabilities
of drug candidates,14–16 were also calculated (Table 2).
In vitro half-lives of the synthesized phenolic DCKs
(13b and 15) were 3.1–3.5 min, similar to half-lives
reported previously for methyl and methoxy DCKs,5
under the same conditions. The results suggest that
modifications of the coumarin ring do not influence in vi-
tro oxidative metabolism, regardless of the type of mod-
ification, although impact on the anti-HIV activity is
significant. The findings also would predict similarly ra-
pid in vivo metabolism and low bioavailability. In con-
trast, the 10-aza-DCK analog (16) differed from the
other DCK analogs and had a longer in vitro half life,
close to 10 min. The estimate of in vivo intrinsic clear-
The synthesized DCK analogs were assayed for anti-
HIV activity using the HIV-1 IIIB strain in H9 lympho-
cytes (Table 1). Introduction of a phenolic hydroxyl at
position-4 (13a) lowered the logD value significantly,
but was also highly detrimental to the activity as 13a
showed an EC50 value of only 15 lM. Addition of a
3- or 5-methyl (13 b, 13c) to 4-hydroxy-DCK (13a)
increased lipophilicity on the coumarin nucleus and im-
proved activity by almost 10-fold, which suggested that

M. Suzuki et al. / Bioorg. Med. Chem. 15 (2007) 6852–6858
6855
lipophilic modification on a polar DCK was beneficial,
similar to phenomena observed with other DCK
analogs.
thoxy-2-methyl-2-butanol were heated together at
150 ꢁC for 5 h. The same quantity of the 4,4-dime-
thoxy-2-methyl-2-butanol was added and the heating
was continued for additional 5 h. The solvent was evap-
orated and the brown residue was isolated on silica gel
using hexane–EtOAc to obtain the product.
Among all compounds synthesized, the 10-aza-DCK
analog (16) showed the most promising anti-HIV activ-
ity, even though the introduction of the secondary
amine in the dimethylpyran-ring only slightly lowered
the lipophilicity (compare logD values of the bioisoster-
ic isomer 16 and 4-methyl-DCK, Table 1). In addition,
while 4-methyl-DCK and the phenolic DCKs 13b and
15 showed similar metabolic stabilities, 4-methyl-10-
aza-DCK (16) was moderately more stable (Table 2).
These results indicate that, although many publications
suggest that lipophilicity contributes to metabolic labil-
ity, the position of polar modification is more important
than the overall lipophilicity for the metabolic stability
of this compound type. Combined with the previous re-
ports, this study indicates that the pyran ring may be a
good target for altering metabolic stability of DCKs.
5.1.2.1. 6-Acetyl-5-hydroxy-2,2-dimethyl-2H-chromene
(2a). Yield: 46% (starting with 1.52 g 1a); mp 90–92 ꢁC;
MS-ESI+ (m/z, %): 241 (M++Na, 100), 219 (M++1, 78);
1H NMR (CDCl3, 300 MHz) d 1.44 (6H, s, 2-(CH3)2),
2.53 (3H, s, COCH3), 5.58 (1H, d, J = 10.1 Hz, H-4),
6.33 (1H, d, J = 9.0 Hz, H-8), 6.70 (1H, d, J = 10.1 Hz,
H-3), 7.51 (1H, d, J = 9.0 Hz, H-7), 12.86 (1H, s,
OH).
5.1.2.2. 6-Propionyl-5-hydroxy-2,2-dimethyl-2H-chro-
menene (2b). Yield: 48% (starting with 415 mg of 1b); mp
70–72 ꢁC; MS-ESI+ (m/z, %): 233 (M++1,100); 1H
NMR (CDCl3, 300 MHz) d 1.20 (3H, t, J = 7.3 Hz,
CH3), 1.43 (6H, s, 2-(CH3)2), 2.91 (2H, q, J = 7.3 Hz,
CH2), 5.56 (1H, d, J = 9.9 Hz, H-4), 6.31 (1H, d,
J = 9.0 Hz, H-8), 6.70 (1H, d, J = 9.9 Hz, H-3), 7.53
(1H, d, J = 9.0 Hz, H-7), 13.05 (1H, s, OH).
4. Conclusion
Phenolic and aza DCK derivatives were synthesized and
evaluated for anti-HIV activity. The most promising
analog in this study, 4-methyl-10-aza-DCK (16), had
lower lipophilicity than the parent compound (4-meth-
yl-DCK), retained anti-HIV activity in the sub-micro-
molar range, and exhibited improved metabolic
stability. From the activity profiles of our synthesized
DCK analogs, lipophilicity around the coumarin ring
appeared to be required for potent anti-HIV activity.
Our extended SAR of the DCK derivatives will be valu-
able in the design and development of DCK derivatives.
Further studies to develop derivatives with improved
potency and pharmaceutical profiles are in progress in
our laboratory.
5.1.2.3. 6-Acetyl-5-hydroxy-2,2,7-trimethyl-2H-chro-
mene (2c). Yield: 38% (starting with 238 mg of 1c); mp
1
53–55 ꢁC; MS-ESIꢀ (m/z, %): 232 (Mꢀ,100); H NMR
(CDCl3, 300 MHz) d 2.60 (3H, s, CH3), 1.43 (6H, s, 2-
(CH3)2), 2.53 (3H, s, COCH3), 5.52 (1H, d, J = 9.9 Hz,
H-4), 6.19 (1H, s, H-8), 6.69 (1H, d, J = 9.9 Hz, H-3),
13.65 (1H, s, OH).
5.1.2.4. 6-Hydroxy-4-methylseselin (5). Yield: 58%
(starting with 481 mg of 4); mp 190–192 ꢁC; MS-ESIꢀ
1
(m/z, %): 258 (Mꢀ, 30), 257 (Mꢀꢀ1, 100); H NMR
(CDCl3, 300 MHz) d 1.50 (6H, s, 20-(CH3)2), 2.34 (3H,
s, CH3), 5.56 (1H, s, OH), 5.73 (1H, d, J = 10.3 Hz, H-
30), 6.14 (1H, H-3), 6.89 (1H, d, J = 10.3 Hz, H-40),
6.99 (1H, s, H-5).
5. Experimental
5.1. Chemistry
5.1.3. General procedure for synthesis of 4-hydroxysese-
lins (3a–c). Sodium hydride (60% dispersion in oil) was
added slowly to a solution of the chromene derivative
in dry diethyl carbonate. The mixture was slowly heated
to reflux and stirred for 20 min. After cooling, MeOH
was added carefully. Diethyl ether was added and the
solution was extracted with water. The aqueous extracts
were acidified with HCl (10%) and then extracted with
EtOAc. The combined organic extracts were dried and
recrystallized using EtOAc to obtain the product.
5.1.1. General procedures. Melting points were measured
with a Fisher Johns melting point apparatus without
correction. The proton nuclear magnetic resonance
(1H NMR) spectra were measured on a Varian
300 MHz spectrometer using TMS as internal standard.
Mass spectra were obtained on an Agilent 1100 MSD
ion trap mass spectrometer or PE-SCIEX API-3000
with turbo ion spray source. Thin-layer chromatogra-
phy (TLC) was performed on a precoated silica gel
GF plate purchased from Analtech, Inc. Silica gel
(200–400 mesh) from Natland (Durham, NC) was used
for column chromatography. All other chemicals were
obtained from Aldrich, Inc., unless otherwise noted.
5.1.3.1. 4-Hydroxyseselin (3a). Yield: 38% (starting
with 436 mg of 2a); mp >210 ꢁC sublimed; MS-ESIꢀ
1
(m/z, %): 244 (Mꢀ, 40), 243 (Mꢀꢀ1, 100); H NMR
(CDCl3, 300 MHz) d 1.43 (6H, s, 20-(CH3)2), 5.43 (1H,
s, H-3), 5.67 (1H, d, J = 9.9 Hz, H-30), 5.71 (1H, d,
J = 9.0 Hz, H-6), 6.83 (1H, d, J = 9.9 Hz, H-40), 7.60
(1H, d, J = 9.0 Hz, H-5).
5.1.2. General procedure for synthesis of dimethylchrom-
enes (2a–c) and 6-hydroxy-4-methylseselin (5). Substrate
(20,40-dihydroxyacetophenone 1a, 20,40-dihydroxypropi-
ophenone 1b, 20,40-dihydroxy-60-methylacetophenone
1c, or 4-methylesculetin 4), dry pyridine, and 4,4-dime-
5.1.3.2. 4-Hydroxy-3-methylseselin (3b). Yield: 91%
(starting with 324 mg of 2b); mp >190 ꢁC sublimed;
MS-ESIꢀ (m/z, %): 258 (Mꢀ, 30), 257 (Mꢀꢀ1, 100);

6856
M. Suzuki et al. / Bioorg. Med. Chem. 15 (2007) 6852–6858
1H NMR (CDCl3, 300 MHz) d 1.46 (6H, s, 20-(CH3)2),
2.12 (3H, s, 3-CH3), 5.68 (1H, d, J = 9.9 Hz, H-30),
6.69 (1H, br s, OH), 6.74 (1H, d, J = 8.6 Hz, H-6), 6.87
(1H, d, J = 9.9 Hz, H-40), 7.59 (1H, d, J = 8.6 Hz, H-5).
(105 mg, 0.75 mmol), and 2,5-diphenyl-4,6-bis(9-O-
dihydroquinyl)-pyrimidine [(DHQ)2Pyr] (4.4 mg,
0.005 mmol), K2OsO2(OH)4 (0.005 mmol), was dis-
solved in 5 mL of t-BuOH/H2O (v/v, 1:1) at rt. The solu-
tion was cooled to 0 ꢁC and methanesulfonamide
(0.25 mmol) added with stirring. When the solution
turned from light yellow to orange, the substituted ses-
elin compound (0.25 mmol) was added. The mixture
was stirred at 0 ꢁC for 2–4 days. Reaction was moni-
tored using TLC, and at completion, Na2S2O5 (excess),
water, and CHCl3 were added. After being stirred for
0.5 h at rt, the mixture was extracted with CHCl3 three
times. The residue was separated by column chro-
matography to obtain the pure substituted (+)-cis-
khellactones.
5.1.3.3. 4-Hydroxy-5-methylseselin (3c). Yield: 80%
(starting with 232 mg of 2c); mp 185–187 ꢁC; MS-ESIꢀ
(m/z, %): 257 (Mꢀꢀ1, 100); 1H NMR (CDCl3,
300 MHz) d 1.41 (6H, s, 20-(CH3)2), 2.60 (3H, s, 5-
CH3), 5.61 (1H, s, H-3), 6.49 (1H, d, J = 6.8 Hz, H-30),
6.80 (1H, d, J = 6.8 Hz, H-40).
5.1.4. Synthesis of 6-(tert-butyldimethylsilanyloxy)-4-
methylseselin (6). Compound 5 (125 mg) was protected
with tert-butyldimethylsilyl (TBS) by reacting with
TBS-Cl and imidazole in THF at rt overnight. Yield:
90%; mp 135–136 ꢁC; MS-ESI+ (m/z, %): 373 (M++1,
100); 1H NMR (CDCl3, 300 MHz) d 0.20 (6H, s,
Si(CH3)2), 1.03 (9H, s, SiC(CH3)3)1.53 (6H, s, 20-
(CH3)2), 2.33 (3H, s, CH3), 5.70 (1H, d, J = 10.1 Hz,
H-30), 6.12 (1H, H-3), 6.92 (1H, d, J = 10.1 Hz, H-40),
6.87 (1H, s, H-5).
5.1.8.1. 4-Hydroxy-(+)-cis-khellactone (10a). Yield:
58% (starting with 133 mg of 3a); mp 186–187 ꢁC; MS-
ESIꢀ (m/z, %): 277 (Mꢀꢀ1, 100); 1H NMR d 1.42,
1.45 (each 3H, s, 20-(CH3)2), 3.86 (1H, d, J = 4.7 Hz,
H-30), 5.17 (1H, d, J = 4.7 Hz, H-40), 6.82 (1H, d,
J = 9.1 Hz, H-6), 7.85 (1H, d, J = 9.1 Hz, H-5).
5.1.5. Synthesis of 7-amino-4-methylcoumarin (7). Ami-
nophenol and ethyl acetoacetate were mixed on graph-
ite/K10 (2:1, w/w) and heated at 130 ꢁC for 30 min.
After cooling, the reaction mixture was filtered and the
crude residue was chromatographed on silica gel using
CH2Cl2–MeOH to obtain the product. Yield: 25%
(starting from 3 g aminophenol); mp 222–223 ꢁC, MS-
5.1.8.2. 4-Hydroxy-3-methyl-(+)-cis-khellactone (10b).
Without isolation, compound 10b was treated with cam-
phonyl chloride in the next step.
5.1.8.3. 4-Hydroxy-5-methyl-(+)-cis-khellactone (10c).
Yield: 63% (starting with 129 mg of 3c); mp 197–200 ꢁC;
MS-ESI+ (m/z, %): 293 (M++1, 100); MS-ESIꢀ (m/z,
1
1
ESIꢀ (m/z, %): 174 (Mꢀꢀ1, 100); H NMR (acetone-
%): 291 (Mꢀꢀ1, 100); H NMR (CD3OD) d 1.39, 1.42
d6, 300 MHz) d 2.38 (3H, d, J = 1.1 Hz, 4-CH3), 5.94
(1H, q, J = 1.1 Hz, H-3), 6.50 (1H, d, J = 2.2 Hz, H-8),
6.65 (1H, dd, J = 8.6, 2.2 Hz, H-6), 7.45 (1H,
J = 8.6 Hz, H-5).
(each 3H, s, 20-(CH3)2), 2.63 (3H, s, 5-CH3), 3.73 (1H,
d, J = 4.7 Hz, H-30), 5.01 (1H, d, J = 4.7 Hz, H-40),
6.56 (1 H, s, H-6).
5.1.8.4. 6-(tert-Butyldimethylsilanyloxy)-4-methyl-
(+)-cis-khellactone (11). Yield: 82% (starting with
38 mg of 6); mp 148–150 ꢁC; MS-ESIꢀ (m/z, %): 407
(M++1, 100); 1H NMR (CDCl3, 300 MHz) d 0.18,
0.20 (each 3H, s, Si(CH3)2), 1.03 (9H, s, SiC(CH3)3),
1.44, 1.47 (each 3H, s, 20-(CH3)2), 2.36 (3H, s, CH3),
3.16 (1H, d, J = 6.2 Hz, 30-OH), 3.85 (1H, t,
J = 5.1 Hz, H-30), 3.91 (1H, d, J = 5.1 Hz, 40-OH),
5.20 (1H, t, J = 6.2 Hz, 40-H), 6.14 (1H, s, H-3), 6.96
(1H, s, H-5).
5.1.6. Synthesis of 7-(1,1-dimethyl-prop-2-ynylamino)-4-
methylcoumarin (8). Compound 7 (700 mg), 3-chloro-3-
methylbut-1-yne, copper powder, and copper (I)
chloride in 50% aqueous acetone were heated to reflux
overnight. After cooling, the reaction mixture was
filtered and the crude residue was chromatographed on
silica gel using CH2Cl2–MeOH to obtain the product.
Yield: 84%; mp 181–183 ꢁC; MS-ESIꢀ (m/z, %): 240
(Mꢀꢀ1, 100); 1H NMR (CDCl3, 300 MHz) d 1.66
(6H, s, 1-(CH3)2), 2.35 (3H, s, 4-CH3), 2.43 (1H, s,
CCH), 4.24 (1H, s, NH), 6.02 (1H, s, H-2), 6.71 (1H,
d, J = 8.8 Hz, H-6 ), 6.94 (1H, s, H-8), 7.37 (1H, s,
J = 8.8 Hz, H-5).
5.1.8.5. (9S,10R)-9,10-Dihydroxy-4,8,8-trimethyl-7,8,
9,10-tetrahydro-2H-pyrano[2,3-f]quinolin-2-one (12).
Yield: 78% (starting with 82 mg of 9); mp 255–257 ꢁC;
MS-ESIꢀ (m/z, %): 274 (Mꢀꢀ1, 100); 1H NMR
(CDCl3, 300 MHz) d 1.29, 1.34 (each 3H, s, 20-
(CH3)2), 2.35 (3H, s, CH3), 3.75 (1H, d, J = 5.1 Hz, H-
30), 5.99 (1H, s, H-3), 5.19 (1H, d, J = 5.1 Hz, H-40),
6.43 (1H, d, J = 9.0 Hz, H-6), 7.31 (1H, d, J = 9.0 Hz,
H-5).
5.1.7. Synthesis of 4,8,8-trimethyl-7,8-dihydro-2H-pyr-
ano[2,3-f]quinolin-2-one (9). Amine 8 (120 mg) and cop-
per (I) chloride were heated in refluxing THF to
obtain the product azaseselin in 98% yield. Mp 125–
127 ꢁC, MS-ESIꢀ (m/z, %): 240 (Mꢀꢀ1, 100); 1H
NMR (CDCl3, 300 MHz) d 1.35 (6H, s, 2-(CH3)2),
2.31 (3H, s, CH3), 4.11 (1H, s, NH), 5.54 (1H, d,
J = 9.9 Hz, H-3), 5.95 (1H, H-6), 6.33 (1H, d,
J = 8.2 Hz, H-10), 6.87 (1H, d, J = 9.9 Hz, H-4), 7.18
(1H, d, J = 8.2 Hz, H-9).
5.1.9. General procedure for esterification of khellactones.
Excess (S)-(ꢀ)-camphanoyl chloride (greater than
0.5 mmol) and (+)-cis-khellactone were reacted in
pyridine/CH2Cl2 for 1–2 days at rt. The mixture was
extracted with EtOAc and separated by column
chromatography to obtain the substituted DCK
derivatives.
5.1.8. General procedure for asymmetric dihydroxylation.
A mixture of K3Fe(CN)6 (150 mg, 0.75 mmol), K2CO3

M. Suzuki et al. / Bioorg. Med. Chem. 15 (2007) 6852–6858
6857
5.1.9.1.
4-Hydroxy-30,40-di-O-(ꢀ)-camphanoyl-(+)-
(each 3H, s, 20-(CH3)2), 1.67, 1.91, 2.21, 2.49 (each 2H,
m, camphanoyl CH2), 2.36 (3H, s, CH3), 5.42 (1H, d,
J = 4.8 Hz, H-30), 5.49 (1H, s, 6-OH), 6.14 (1H, s, H-
3), 6.68 (1H, d, J = 4.8 Hz, 40-H), 7.16 (1H, s, H-5).
cis-khellactone (13a). Yield: 38% (starting with 100 mg
of 10a); mp 197–199 ꢁC; MS-ESIꢀ (m/z, %): 637.5
(Mꢀꢀ1, 100); 1H NMR (CDCl3, 300 MHz) d 0.94,
0.98, 1.06, 1.07, 1.11, 1.11 (each 3H, s, camphanoyl
CH3), 1.46, 1.50 (each 3H, s, 20-(CH3)2), 1.69, 1.88,
2.21, 2.46 (each 2H, m, camphanoyl CH2), 5.38 (1H,
d, J = 9.1 Hz, H-30 ), 5.68 (1H, s, H-3), 6.64 (1H, d,
J = 9.1 Hz, 40-H), 6.81 (1H, d, J = 4.7 Hz, H-6), 7.79
(1H, d, J = 4.7 Hz, H-5).
5.2. Biological studies
5.2.1. Anti-HIV assay. HIV growth inhibition assay was
performed as described previously.4
5.2.2. Metabolic stability studies
5.1.9.2. 4-Hydroxy-3-methyl-30,40-di-O-(ꢀ)-campha-
noyl-(+)-cis-khellactone (13b). Yield: 23% (starting with
50 mg of 3b); mp 183–185 ꢁC; MS-ESIꢀ (m/z, %):
651.3 (Mꢀꢀ1, 100); 1H NMR (CDCl3, 300 MHz) d
1.01, 1.04, 1.07, 1.11, 1.13, 1.14 (each 3H, s, camphanoyl
CH3), 1.48, 1.52 (each 3H, s, 20-(CH3)2), 1.73, 1.89, 2.21,
2.52 (each 2H, m, camphanoyl CH2), 2.20 (3H, s, 3-
CH3), 5.41 (1H, d, J = 4.7 Hz, H-30), 6.69 (1H, d,
J = 4.7 Hz, H-40), 6.87 (1H, d, J = 8.9 Hz, H-6), 7.36
(1H, d, J = 8.9 Hz, H-5).
5.2.2.1. Sample preparations. DCKs were brought to a
final concentration of 1 lM with 100 mM sodium phos-
phate buffer at pH 7.4, which contained 0.1 mg/mL hu-
man liver microsome and 5 mM MgCl2. Incubation
volumes were 1.5 mL. Reactions were started by adding
150 lL of NADPH (15 mM) after 10-min incubation at
37 ꢁC, and stopped by addition of 3 volumes of ice-cold
acetonitrile. Incubations of all samples were run in dupli-
cate, and for control incubations, NADPH was omitted.
For each sample, 200 lL aliquots were taken out at 0, 5,
10, 15, 30, and 60 min time points. After addition of ace-
tonitrile, 15 lL of 0.01 mg/mL acetonitrile solution of
DCK was added to supernatants as an internal standard
prior to centrifugation. The mixture was centrifuged at
15,000 rpm for 5 min. Supernatants were then evaporated
to dryness under nitrogen using an N-Evap analytic evap-
orator (Organomation Associates Inc., Berlin, MA), and
the residue was reconstituted with 150 lL of acetonitrile
and water with 0.1% HOAc (3:7, v/v). An aliquot
(75 lL) was injected onto the HPLC with UV detector.
5.1.9.3. 4-Hydroxy-5-methyl-30,40-di-O-(ꢀ)-campha-
noyl-(+)-cis-khellactone (13c). Yield: 20% (starting with
50 mg of 10c); mp 174–176 ꢁC, MS-ESIꢀ (m/z, %):
651.3 (Mꢀꢀ1, 100); 1H NMR (CDCl3, 300 MHz) d
0.96, 1.01, 1.04, 1.05, 1.07, 1.08, 1.11 (each 3H, s, cam-
phanoyl CH3), 1.35, 1.48, (each 3H, s, 20-(CH3)2),
1.73, 1.89, 2.21, 2.52 (each 2H, m, camphanoyl CH2),
2.55 (3H, s, 5-CH3), 4.04 (1H, d, J = 4.7 Hz, H-30),
5.64 (1H, s, H-3), 6.46 (1H, d, J = 4.7 Hz, H-40), 6.54
(1H, s, H-6).
5.2.2.2. HPLC-UV conditions. Analysis was carried
out on an HP Series 1100 (Hewlett-Packard, Palo Alto,
CA), consisting of an autosampler, a degasser, a binary
pump, and a UV detector. An Axxiom ODS column
150 · 4.6 mm ID, 5 lm (Thomson Instrument Co.,
Springfield, VA) connected to a RP-18 guard column
(15 · 3.2 mm ID, 7 lm) (Brownlee, San Jose, CA) was
isocratically eluted with acetonitrile and 25 mM HOAc
in water (55:45, v/v) at a flow rate of 1.5 mL/min. Detec-
tion wavelength was set at 320 nm. Chromatograms
were recorded on a Hewlett-Packard Chemstation
A.05.01. (Hewlett-Packard, Palo Alto, CA).
5.1.9.4. 6-(tert-Butyldimethylsilanyloxy)-4-methyl-30, 40-
di-O-(ꢀ)-camphanoyl-(+)-cis-khellactone (14). Yield:
59% (starting with 50 mg of 11); mp 138–139 ꢁC; MS-
1
ESI+ (m/z, %): 789.4 (M++Na, 100); H NMR (CDCl3,
300 MHz) d 0.20 (6H, s, Si(CH3)2), 0.98, 1.01, 1.03, 1.07,
1.10, 1.11 (each 3H, s, camphanoyl CH3), 1.03 (9H, s,
SiC(CH3)3), 1.48, 1.50 (each 3H, s, 20-(CH3)2), 1.66,
1.91, 2.21, 2.49 (each 2H, m, camphanoyl CH2), 2.34
(3H, s, CH3), 5.37 (1H, d, J = 4.8 Hz, H-30), 6.11 (1H,
s, H-3), 6.65 (1H, d, J = 4.8 Hz, 40-H), 7.02 (1H, s, H-5).
5.1.9.5.
9,10-tetrahydro-2H-pyrano[2,3-f]quinolin-2-one
9,10-Di-O-camphanoyl-4,8,8-trimethyl-7,8,
(16).
5.2.2.3. Calculations. In the determination of the in vi-
tro half life (t1/2), the analyte/internal-standard peak
height ratios were converted to percentage drug remain-
ing, using the initial time (0 min) peak height ratio val-
ues as 100%. The slope of the linear regression from
log percentage remaining versus incubation time rela-
tionships (ꢀkel) was used in the conversion to in vitro
t1/2; t1/2 = 0.693/kel.
Yield: 5% (starting with 100 mg of 12); mp 227–
229 ꢁC; MS-ESIꢀ (m/z, %): 634.7 (Mꢀꢀ1, 30); 1H
NMR (CDCl3, 300 MHz) d 0.97, 0.98, 1.06, 1.09, 1.09,
1.10 (each 3H, s, camphanoyl CH3), 1.25, 1.27 (each
3H, s, 20-(CH3)2), 1.67, 1.91, 2.21, 2.49 (each 2H, m,
camphanoyl CH2), 2.32 (3H, s, CH3), 5.26 (1H, d,
J = 4.6 Hz, H-30), 5.95 (1H, s, H-3), 6.46 (1H, d,
J = 8.6 Hz, H-6), 6.67 (1H, d, J = 4.6 Hz, 40-H), 7.36
(1H, d, J = 8.6 Hz, H-5).
Conversion to Clint (in units of L/min/kg) was done
using the following formulas for the enzyme kinetic
and half life methods, respectively:14
5.1.10. Synthesis of 6-hydroxy-4-methyl-30,40-di-O-(ꢀ)-
camphanoyl-(+)-cis-khellactone (15). Compound 14
(114 mg) was stirred with CeF in THF at rt for 5 min
to remove the silyl protecting group. Yield: 36%, mp
255–257 ꢁC; MS-ESIꢀ (m/z, %): 651.6 (Mꢀꢀ1, 100);
1H NMR (CDCl3, 300 MHz) d 0.98, 1.00, 1.07, 1.09,
1.10, 1.11 (each 3H, s, camphanoyl CH3), 1.51, 1.53
0:693
ml incubation
Clint
¼
ꢁ
invitro t1=2 mg microsomes
44:8 mg microsomes
20:7 g liver
ꢁ
ꢁ
g liver
kg per body weight

6858
M. Suzuki et al. / Bioorg. Med. Chem. 15 (2007) 6852–6858
7. Roberts, S. A. Curr. Opin. Drug Discov. Devel. 2003, 6,
66–80.
Acknowledgments
8. Thompson, T. N. Curr. Drug Metab. 2000, 1, 215–241.
9. Munro, H. D.; Musgrave, O. C.; Templeton, R. J. Chem.
Soc. (C) 1971, 95–98.
10. Bandaranayake, W. M.; Crombie, L.; Whiting, D. A.
J. Chem. Soc. (C) 1971, 811–816.
This investigation was supported by Grant AI-33066
from the National Institute of Allergy and Infectious
Diseases (NIAID) awarded to K.H. Lee.
11. Barton, D. H. R.; Donnelly, D. M. X.; Finet, J.-P.; Guiry,
P. J. J. Chem. Soc., Perkin. Trans. 1 1992, 1365–1375.
12. Jolidon, S.; Hansen, H. J. Helv. Chim. Acta 1977, 60, 978–
1032.
13. Hamann, L. G.; Higuchi, R. I.; Zhi, L.; Edwards, J. P.;
Wang, X. N.; Marschke, K. B.; Kong, J. W.; Farmer, L.
J.; Jones, T. K. J. Med. Chem. 1998, 41, 623–639.
14. Obach, R. S.; Baxter, J. G.; Liston, T. E.; Silber, B. M.;
Jones, B. C.; MacIntyre, F.; Rance, D. J.; Wastall, P.
J. Pharmacol. Exp. Ther. 1997, 283, 46–58.
15. Naritomi, Y.; Terashita, S.; Kimura, S.; Suzuki, A.;
Kagayama, A.; Sugiyama, Y. Drug Metab. Dispos. 2001,
29, 1316–1324.
References and notes
1. Huang, L.; Yuan, X.; Yu, D.; Lee, K. H.; Chen, C. H.
Virology 2005, 332, 623–628.
2. Xie, L.; Takeuchi, Y.; Cosentino, L. M.; Lee, K. H.
J. Med. Chem. 1999, 42, 2662–2672.
3. Yu, D.; Suzuki, M.; Xie, L.; Morris-Natschke, S. L.; Lee,
K. H. Med. Res. Rev. 2003, 23, 322–345.
4. Xie, L.; Yu, D.; Wild, C.; Allaway, G.; Turpin, J.; Smith,
P. C.; Lee, K. H. J. Med. Chem. 2004, 47, 756–760.
5. Suzuki, M.; Li, Y.; Smith, P. C.; Swenberg, J. A.; Martin,
D. E.; Morris-Natschke, S. L.; Lee, K. H. Drug Metab.
Dispos. 2005, 33, 1588–1592.
6. Huang, L.; Kashiwada, Y.; Cosentino, L. M.; Fan, S.;
Chen, C. H.; McPhail, A. T.; Fujioka, T.; Mihashi, K.;
Lee, K. H. J. Med. Chem. 1994, 37, 3947–3955.
16. Mahmood, I. Drug Metabol. Drug Interact. 2002, 19, 49–
64.
17. Obach, R. S. Drug Metab. Dispos. 1999, 27, 1350–
1359.