2,3,5-Substituted tetrahydrofurans: COX-2 inhibitory activities of 5-hydroxymethyl-/carboxyl-2,3-diaryl-tetrahydro-furan-3-ols

Article abstract of DOI:10.1016/j.ejmech.2007.12.017

5-Hydroxymethyl-/carboxyl-2,3-diaryl-tetrahydro-furan-3-ols have been investigated for their COX-1 and COX-2 inhibitory activities. Compounds 17, 18 and 20 have been identified as showing appreciable COX-2 inhibition and selectivity. The group present at

Full text of DOI:10.1016/j.ejmech.2007.12.017

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 43 (2008) 2792e2799
http://www.elsevier.com/locate/ejmech
Original article
2,3,5-Substituted tetrahydrofurans: COX-2 inhibitory activities
of 5-hydroxymethyl-/carboxyl-2,3-diaryl-tetrahydro-furan-3-ols
a
b
a
Palwinder Singh a, , Anu Mittal , Satwinderjeet Kaur , Subodh Kumar
*
a Department of Chemistry, Guru Nanak Dev University, Amritsar 143005, Punjab, India
b Department of Botanical and Environmental Sciences, Guru Nanak Dev University, Amritsar 143005, Punjab, India
Received 28 August 2007; received in revised form 6 December 2007; accepted 11 December 2007
Available online 28 December 2007
Abstract
5-Hydroxymethyl-/carboxyl-2,3-diaryl-tetrahydro-furan-3-ols have been investigated for their COX-1 and COX-2 inhibitory activities. Com-
pounds 17, 18 and 20 have been identified as showing appreciable COX-2 inhibition and selectivity. The group present at C-5 of tetrahydrofuran
and the substituents at the two phenyl rings, through their interactions with active site amino acid residues, significantly affect the activities of
these molecules. The quantitative structureeactivity relationship studies indicate the role of log P, TPSA, molecular connectivity and valence
connectivity towards the activities of these molecules.
Ó 2007 Elsevier Masson SAS. All rights reserved.
Keywords: Cyclooxygenase-2; 2,3-Diaryl-tetrahydrofurans; In vitro studies; COX-2 inhibitors; QSAR
1. Introduction
[14], carry aryl rings on chiral, sp3 hybridized C-2 and C-3
carbons of tetrahydrofuran. Preliminary investigations have
identified 5-hydroxymethyl-/carboxyl-/ethylsulfanylmethyl-2,3-
diphenyl-tetrahydrofuran-3-ols as highly potent and selective
COX-2 inhibitors [24]. In the present contribution, tetrahydrofu-
rans with CH2OH (A, Fig. 1), COOH (B, Fig. 1) group at C-5 and
various substituents at the two aryl rings have been investigated
for their COX-1 and COX-2 inhibitory activities. Experimental
data and the docking studies indicate the role of substituents at
the two phenyl rings of tetrahydrofuran-3-ols in affecting the
COX-1 and COX-2 inhibitory activities of these molecules.
Quantitative structureeactivity relationship (QSAR) studies, to
some extent, specify the contributions of log P, total polar surface
area, molecular connectivity and valence molecular connectivity
for the activities of these molecules.
The establishment of the role of COX-2 (out of the two iso-
forms of cyclooxygenase [1,2] viz. COX-1 and COX-2) in
causing inflammation, promotion of cancer [3e7] and activa-
tion of P-glycoprotein [3,6,8e10] (transporter protein respon-
sible for multidrug resistance) has made this enzyme as the
cellular target of a number of chemical entities for the treat-
ment of inflammation [11e15], minimizing the growth of can-
cer [16e19] and reducing multidrug resistance [20e23].
While COX-2 inhibitors are primarily used as anti-inflamma-
tory agents, their applications for the treatment of cancer
and modulating multidrug resistance are still at the infancy.
Due to the multiple role of COX-2, we are aiming at the devel-
opment of COX-2 inhibitors and their evaluations as anti-cancer
agents also. Based uponthe chiral nature and flexibility of COX-2
active site, we have designed 5-substituted 2,3-diaryl-tetrahydro-
furan-3-ols which, in contrast to the reported COX-2 inhibitors
2. Results
2.1. Chemistry
Benzoins 1e6 on treatment with indium metal, allyl bro-
mide in THFeH2O (2:1) at 0 ꢀC provided the corresponding
allylic alcohols 7e12. Treatment of 12 with oxone
* Corresponding author. Tel.: þ91 183 2258802x3495; fax: þ91 183
2258819.
E-mail address: palwinder_singh_2000@yahoo.com (P. Singh).
0223-5234/$ - see front matter Ó 2007 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2007.12.017

P. Singh et al. / European Journal of Medicinal Chemistry 43 (2008) 2792e2799
2793
X
X
separately) which probably has been formed by the in-situ lac-
tonization of 22. However, the same reactions of compounds
18 and 19 provided, respectively, an unstable compound 24
and a mixture of two products from which 25 could not be sep-
arated. The relative stereochemistries at various chiral centres
of these molecules have also been defined on the basis of NOE
experiments and X-ray structure of 14.
HO
HO
COOH
CH2OH
O
O
A
B
X
X
X = H, 4-Cl, 2-CI,
4-F, 4-OMe,
4SO2Me
X = H, 4-Cl,
2-CI, 4-F,
2.2. Biology
Fig. 1.
In vitro COX-2 inhibiting activities of these compounds
have been evaluated using ‘COX (ovine) inhibitor screening
assay’ kit with 96-well plates (Caymen Chemicals, catalogue
no. 560101). This screening assay directly measures PGF2a
produced by SnCl2 reduction of COX-derived PGH2. COX-1
and COX-2 initial activity tubes were prepared by taking
950 ml of reaction buffer, 10 ml of heme, 10 ml of COX-1
and COX-2 enzymes in respective tubes. Similarly, COX-1
and COX-2 inhibitor tubes were prepared by adding 20 ml of
inhibitor (compound under test) in each tube in addition to
the above ingredients. The background tubes correspond to in-
activated COX-1 and COX-2 enzymes obtained after keeping
the tubes containing enzymes in boiling water for 3 min. Re-
actions were initiated by adding 10 ml of arachidonic acid in
each tube and quenched with 50 ml of 1 M HCl. PGH2 thus
formed was reduced to PGF2a by adding 100 ml of SnCl2.
The prostaglandin produced in each well was quantified using
broadly specific prostaglandin antiserum that binds with major
prostaglandins and reading the 96-well plate at 405 nm. The
wells of the 96-well plate showing low absorption at 405 nm
indicate the low level of prostaglandins in these wells and
hence the less activity of the enzyme. Therefore, the COX in-
hibitory activities of the compounds could be quantified from
the absorption values of different wells of the 96-well plate.
The results of these studies have been represented in terms
of the percentage inhibition of COX-1 and COX-2 enzymes.
The anti-cancer activities were determined on 59 human
tumor cell lines at five concentrations (10ꢁ4 M, 10ꢁ5 M,
10ꢁ6 M, 10ꢁ7 M, 10ꢁ8 M) following the standard procedure
[26e28] at NIH. The results are given in terms of the GI50
values and the percentage inhibition of growth of tumor cells
at some specific cell lines.
transformed its SMe group to SO2Me in 13 (Scheme 1). The
reactions of compounds 7e13 with m-CPBA in chloroform
furnished the respective compounds 14e19.
All these compounds are well characterised by various
spectroscopic data. The observation of NOE between 2-H
and 5-H indicates their syn orientation. X-ray crystal structure
of 14 (Fig. 2) shows an almost anti-orientation of the two phe-
nyl rings and syn-positioning of 2-H and 5-H while 3-OH and
CH2OH groups are projecting in the same direction. On the
basis of NOE experiments and X-ray crystal structure, the rel-
ative configurations at the three chiral centres have been
assigned (Scheme 1) and the molecules with these geometries
have been used in docking studies.
Treatment of 14 with pyridinium chloro chromate (PCC)
[25] in dichloromethane after usual work up provided the cor-
responding carboxyl derivative of tetrahydrofuran (20). Under
the same reaction conditions, compounds 15e17 when treated
with PCC gave the corresponding compounds 21e23
(Scheme 2) The appearance of an absorption peak in the re-
gion 1800 cmꢁ1 in the IR spectra of compounds 21e23 in
comparison to a peak at 1710 cmꢁ1 in the IR spectrum of
20 indicates the presence of an ester group in compounds
21–23. Later investigations are indicating structure 22A for
compound 22 (X-ray crystal structure will be reported
X
O
X
HO
X
i
X
OH
OH
13
7-13
1-6
iii
ii
12
2.3. Molecular modeling and QSAR studies
X
1, 7, 14: X = H
2, 8, 15: X = 4-Cl
3, 9, 16: X = 2-Cl
4, 10, 17: X = 4-F
5, 11, 18: X = 4-OMe
6, 12: X = 4-SMe
13, 19: X = 4-SO2Me
For dockings, the crystal structure of COX-2 with Sc-558 as
the ligand present in the active site (mouse COX, pdb ID
6COX) and COX-1 with ibuprofen in the active site (ovine
COX, pdb ID 1EQG) were downloaded from protein data
bank. The crystal structures of the enzymes were refined,
docking procedure was validated and dockings were per-
formed using ‘dock into active site module’ of BioMed Cache
7.5.0.85 following the previously described procedure [29].
Quantitative structureeactivity relationship studies were
carried out using ‘Project Leader’ module with the available
descriptors in BioMed Cache 7.5.0.85.
OH
4
3
H
H
OH
5
2
O
X
14-19
Reaction conditions and reagents:
i) In, THF-H2O, allyl bromide, stir
ii) Oxone, THF-H2O, stir
iii) m-CPBA, CHCl3, 0 °C, stir
Scheme 1.

2794
P. Singh et al. / European Journal of Medicinal Chemistry 43 (2008) 2792e2799
Fig. 2. The ORTEP diagram of compound 14.
3. Discussion
with 4-Cl and 4-SO2Me groups at the two phenyl rings show
comparable inhibitions of COX-2 (43% and 40% at 10ꢁ6 M)
and COX-1 (56% and 55% at 10ꢁ5 M). Compound 16 (o-
chlorophenyl rings) exhibits moderate inhibition of COX-2
at 10ꢁ6 M and 10ꢁ5 M concentrations (67% and 68%) and
73% inhibition of COX-1 at 10ꢁ5 M concentration. Com-
pounds 17 and 18 (with p-fluorophenyl and p-methoxyphenyl
rings) show significant inhibition of COX-2 at 10ꢁ6 M concen-
tration (88% and 65%, respectively) and a selectivity order of
>10 for COX-2 over COX-1.
The percentage inhibitions of COX-1 and COX-2 by each
compound are given in Table 1. All the compounds have
been evaluated in duplicates and Table 1 shows the average
results. For COX-2, two concentrations of the compounds
viz. 10ꢁ5 M and 10ꢁ6 M were used while COX-1 inhibitory
activities were evaluated at 10ꢁ5 M concentration only.
Amongst compounds 14e19 (with CH2OH group at C-5 of
tetrahydrofuran), the presence of substituents at the phenyl
rings considerably affects the inhibitory activities and selectiv-
ity of these compounds for COX-2. Compounds 15 and 19
The dockings of compounds 14 (unsubstituted phenyl
rings) and 17 ( p-fluoro-substituted phenyl rings) show the dif-
ference in the interactions of two molecules in the active site
of COX-2. In compound 17, the fluorines present on C-2 and
C-3 phenyl rings show significant interactions with Y385 and
H90 residues, respectively. Fluorine at C-2 phenyl is at a dis-
X
H
NOE
X
˚
tance of 3.0 A from Y385 while fluorine at C-3 phenyl has
PCC, DCM
rt, 20-24 h
˚
approached NH of H90 at a distance of 2.36 A (Fig. 3). The
blockage of Y385 is advantageous due to the fact that tyrosyl
radical initiates the cyclooxygenase reaction by abstracting the
˚
13-pro(S ) hydrogen of arachidonic acid (placed 2.3 A from
the phenolic oxygen on Y385) [31,32]. The oxygen of
OH
OH
4
H
H
3
H
OH
O
5
O
COOH
2
14-19
X
20-25
X
˚
CH2OH group of 17 lies at a distance of 2.67 A from the H
Yield
31%
35%
40%
40%
30%
of guanidine moiety of R120, the amino acid active during
the turnover phase of COX-2 in arachidonic acid metabolism.
However, such interactions are not observed during the dock-
ing of compound 14 in the active site of COX-2 (Fig. 4). In-
stead, the CH2OH group of 14 participates in intramolecular
H-bonding with its own OH group present at C-3.
20: X = H
O
O
Cl
21: X = 4-Cl
22: X = 2-Cl
23: X = 4-F
H
H
O
24: X = 4-OCH3
25: X = 4-SO2CH3 31%
Cl
22A
In comparison to compound 20, compounds 21e23 exhibit
better COX-2 inhibition at 10ꢁ5 M concentration but
Scheme 2.

P. Singh et al. / European Journal of Medicinal Chemistry 43 (2008) 2792e2799
2795
Table 1
In vitro percentage inhibition and IC50 values for COX-1 and COX-2 enzymes
Compound
% Inhibition
COX-2
IC50 (mM)
COX-2 selectivitya
Anti-cancer activity GI50
COX-1
COX-2
COX-1
1 mM
10 mM
10 mM
14
15
28
43
67
88
65
40
68
30
37
37
75
50
77
68
2.8
56
5.1
3
>10
<10
<10
>10
>10
<10
>10
<10
>10
>10
40
>1.96
<3.3
e
>10
>10
nd
1.99 ꢂ 10ꢁ5 M
2.20 ꢂ 10ꢁ5 M
9.50 ꢂ 10ꢁ5 M
9.33 ꢂ 10ꢁ5 M
1.0 ꢂ 10ꢁ4 M
nd
16
17
68
82
73
22
<1
<1
<1
6.5
<1
3.8
3.5
3.2
0.3
1.2
18
19
67
57
22
55
<1.5
>10
20
21
69
94
ꢁ1.15
55
34
<2.6
>2.8
>3.1
w133
w10
nd
72%c
22
82
90
23
30
75
60%d
Rofecoxibb
Celecoxiba
a
100
100
15%e
65
14
4.7 ꢂ 10ꢁ5 Mf
Cox-2 selectivity ¼ IC50(COX-1)/IC50(COX-2).
b
c
d
e
f
Reported [30], nd: not determined.
Inhibition of tumor cells at NCI-H522 cell line.
Inhibition of tumor cells at SN 12C cell line.
Rofecoxib exhibits 15% inhibition (at 10ꢁ6 M concentration) of tumor cells at PC3 cell line of prostate cancer.
GI50 at PC3 cell line of prostate cancer.
a decrease in their activity has been observed at 10ꢁ6 M con-
centration. The dockings of compounds 20e23 in the active
site of COX-2 indicate that the aryl-substituted compounds
(21e23) fit in the active site of COX-2 in the same fashion
as compound 20 (Fig. 5) and interact with R120 through H-
bonding between carboxylate group and guanidine moiety
(Fig. 5).
However, during the dockings of these compounds (14e23)
in the active site of COX-1, it has been observed that none of
them enters in the active site of COX-1 (all have positive dock-
ing score) except compound 16 in which the C-3 phenyl ring
enters in the active site cavity of COX-1 and the chlorine
present on this phenyl ring approaches Y385 at a distance of
˚
3.07 A. In contrast to other compounds, 16 gave a ꢁve dock-
ing score when docked in the active site of COX-1 (ꢁ40 kcal/
mol while it is ꢁ66 kcal/mol for COX-2). This behaviour of
16 is also evident from the experimental data where it shows
73% inhibition of COX-1 at 10ꢁ5 M concentration.
Therefore, the nature of the substituent at C-2 and C-3 phe-
nyl rings of tetrahydrofuran-3-ols considerably affects the
efficacy of these compounds for COX-2 inhibition and also
their selectivity for COX-2 over COX-1. Out of the 10 com-
pounds for which the results are presented here, compounds
16e18 and 20 exhibit considerable inhibition of COX-2 which
is better than the COX-2 inhibitory activity of celecoxib and
comparable to that of rofecoxib. However, the reluctance of
these molecules to air oxidation provides an advantage over
Fig. 3. Compound 17 docked in the active site of COX-2. Hydrogens are omit-
˚
ted for clarity. Distances of 17 from R120 (2.67 A), Y385 (3.0 A) and H90
˚
(2.36 A) are visible.
˚
Fig. 4. Compound 14 docked in the active site of COX-2. Hydrogens are
omitted for clarity.

2796
P. Singh et al. / European Journal of Medicinal Chemistry 43 (2008) 2792e2799
structureeactivity relationship studies [35]. Out of these prop-
erties, not a single one shows linear relationship with the
COX-2 inhibitory activities.
Taking log P and TPSA as the descriptors, an unpredictive
equation with low statistical parameters both as fitting (r2) and
cross validation (r2CV), was obtained for the activity of these
compounds. Combining TPSA, connectivity index and valence
connectivity index of 0th order, Eq. (1) was obtained for the
activity of these compounds.
C ¼ ꢁ1:01TPSA þ 15:30c0 ꢁ 4:78j0 ꢁ 71:29
ð1Þ
r2CV ¼ 0.28, r2 ¼ 0.55, c0 ¼ connectivity index (0th order),
j0 ¼ valence connectivity index (0th order).
The predictiveness of the equation gets decreased when the
dipole moment or polarizability parameters were introduced.
However, introducing log P in Eq. (1), a reliable model for
the activities of these compounds (Eq. (2)) was obtained.
Fig. 5. Compounds 20 and 23 docked in the active site of COX-2. A close
overlapping has been observed in the docking of two compounds.
C ¼ ꢁ11:78 log P ꢁ 1:37 TPSA þ 13:93c0 ꢁ 3:28j0 ꢁ 12:47
ð2Þ
rofecoxib which has been demonstrated to show toxicity due
to its oxidation to maleic anhydride derivative [33].
r2CV ¼ 0.43, r2 ¼ 0.61.
Significant anti-cancer activities of compounds 15e19 have
been observed at some specific cell lines of 59 human tumor
cell line panels [34]. Amongst compounds 21e23, compounds
22 and 23 show, respectively, 82% and 90% inhibition of
COX-2 at 10ꢁ5 M concentration and exhibit 72% and 60%
growth inhibitions of tumor cells at NCI-H522 (Non Small
Cell Lung Cancer) and SN 12C (Breast Cancer) cell lines, re-
spectively, at 10ꢁ5 M concentrations (Table 1). Remarkably,
the anti-cancer activities of these compounds are better than
those reported for rofecoxib and comparable to that of
celecoxib.
The activities of compounds predicted from Eq. (2) show
close resemblance with the experimental activities (particu-
larly the compounds carrying CH2OH group at C-5 of tetrahy-
drofuran, 14e19) except in the case of compound 20 (Fig. 6).
Irrespective of trying a number of other combinations with the
available descriptors (Table 2), no further improvement in the
model was observed. These studies indicate the partial depen-
dency of the COX-2 inhibitory activities of these molecules on
lipophilicity, total polar surface area, molecular connectivity
and valance molecular connectivity of the molecules.
The COX-2 inhibitory activities of these compounds have
been correlated with various physico-chemical properties. A
number of descriptors viz. partition coefficient (log P), total
polar surface area (TPSA), polarizability, dipole moment, con-
nectivity index of 0th, 1st and 2nd order, valence connectivity
index of 0th, 1st and 2nd order, shape index of order 1, 2 and
3, solvent accessibility surface area (Table 2) and their combi-
nations were taken into consideration for quantitative
4. Conclusion
The results of these investigations indicate that
(i) 2,3-Diaryltetrahydrofuran-3-ols with CH2OH group at
C-5 (15e19) exhibit better COX-2 inhibitory activities
than compounds 21e23.
Table 2
Various physico-chemical properties of compounds 14e23
2
˚
Compound log P TPSA (A ) Dipole moment Polarizability Connectivity index
(Debye)
Valence connectivity
0th 1st 2nd
Shape index order
Solvent
accessibility
surface area
0th
1st
2nd
1
2
3
14
15
16
17
18
19
20
21
22
23
2.502
3.538
3.538
2.781
1.996
49.69
49.69
49.69
49.69
68.15
2.476
2.608
2.091
2.676
3.480
6.057
6.369
6.194
4.173
6.319
31.07
35.48
35.17
31.42
36.91
39.52
31.30
35.73
35.33
31.65
14.00
9.73
8.55 11.14
9.62 13.25
9.80 13.25
9.80 11.74
6.83
5.22 14.91 6.40 2.97 286.63
6.38 16.84 6.85 3.44 321.88
6.26 16.84 6.85 3.16 307.24
5.50 16.84 6.85 3.44 300.74
5.95 18.78 8.13 3.90 343.24
15.74 10.55
15.74 10.52
15.74 10.52
7.79
7.78
7.02
7.87
17.15 11.60 10.14 13.80
0.792 117.97
20.74 12.94 13.71 17.07 12.32 11.11 22.68 8.26 4.85 400.49
2.611
3.647
3.647
2.890
66.76
66.76
66.76
66.76
14.87 10.11
9.28 11.34
6.82
7.78
7.77
7.02
5.24 15.87 6.63 3.20 289.73
6.40 17.81 7.08 3.66 329.01
6.28 17.81 7.08 3.38 309.21
5.53 17.81 7.08 3.66 306.07
16.61 10.93 10.35 13.45
16.61 10.89 10.53 13.45
16.61 10.89 10.53 11.94

P. Singh et al. / European Journal of Medicinal Chemistry 43 (2008) 2792e2799
2797
1H, 4-H), 4.90 (d, J ¼ 2.1 Hz, 1H, 5-H), 5.27 (s, 1H, 2-H),
6.91 (d, J ¼ 6.6 Hz, 2H, ArH), 7.07e7.45 (m, 8H, ArH); 13C
(normal/DEPT-135) (CDCl3) d 45.53 (ꢁve, C-4), 78.12
(þve, C-5), 85.65 (þve, C-2), 92.11 (ab, C-3), 125.47 (þve,
ArCH), 126.95 (þve, ArCH), 127.89 (þve, ArCH), 128.36
(þve, ArCH), 128.61 (þve, ArCH), 128.97 (þve, ArCH),
132.04 (ab, ArC), 133.57 (ab, ArC), 171.60 (ab, C]O). De-
coupling of doublet at d 4.90 converts double doublet at
d 3.02 into a doublet and doublet at d 2.46 remains unaffected.
Decoupling of double doublet at d 3.02 converts doublets at
d 2.46 and 4.90 into singlets. Decoupling clearly shows that
5-H couples with only one of the two 4-H protons. NOE exper-
iments: irradiation of singlet at d 5.27 (2-H) shows NOE with
signals at d 4.90 (5-H), 3.02 (4-H) and 6.91, 7.06e7.45 (ArH)
and irradiation of doublet at d 4.90 (5-H) shows NOE with dd
at d 3.02 (4-H); FAB mass m/z 267 (Mþ ꢁ OH); (found C,
72.13; H, 5.97; C17H16O4 requires C, 71.82; H, 5.67).
70
60
50
40
30
17
18
16
23
15
45
14
19
21
22
20
25
65
85
Exptl activity
Fig. 6. Graph between the activities (% inhibition of COX-2) of compounds
14e23 predicted from Eq. (2) and experimental activities.
(ii) Among all these compounds, 17, 18 and 20 show best
COX-2 inhibition and selectivity.
(iii) The results of the docking studies are in parallel with
the experimental data for COX-2 and COX-1
inhibitions.
(iv) QSAR studies indicate that the COX-2 inhibitory activ-
ities of these compounds depend upon various parame-
ters out of which log P, TPSA and connectivity indices
are the major ones.
5.2.2. (2R*,3S*,5R*)-4,5-Bis-(4-chloro-phenyl)-4-hydroxy-
tetrahydro-furan-2-carboxylic acid 21
Compound 15 was allowed to react with PCC as described in
the general procedure to give 21 as white solid, 35% yield; mp
80 ꢀC; nmax (CHCl3) 3450 (OH), 1801(C]O) cmꢁ1 1H
;
(CDCl3) d 1.59 (br s, 1H, OH, exchanges with D2O), 2.45 (dd,
2J ¼ 11.1 Hz, J ¼ 0.3 Hz, 1H, 4-H), 2.98 (dd, J ¼ 11.1 Hz,
3J ¼ 2.4 Hz, 1H, 4-H), 4.89 (d, J ¼ 1.8 Hz, 1H, 5-H), 5.20 (s,
1H, 2-H), 6.83 (d, J ¼ 8.1 Hz, 2H, ArH), 7.12 (d, J ¼ 9.0 Hz,
2H, ArH), 7.22 (d, J ¼ 8.4 Hz, 2H, ArH), 7.37 (d, J ¼ 8.7 Hz,
2H, ArH); 13C (normal/DEPT-135) (CDCl3) d 45.31 (ꢁve,
C-4), 78.10 (þve, C-5), 85.04 (þve, C-2), 91.36 (ab, C-3),
126.85 (þve, ArCH), 128.22 (þve, ArCH), 128.26 (þve,
ArCH), 129.05 (þve, ArCH), 130.27 (ab, ArC), 131.82 (ab,
ArC), 134.47 (ab, ArC), 135.22 (ab, ArC), 170.96 (ab, C]O);
MALDI (TOF) m/z 335.59 (Mþ ꢁ OH); (found C, 57.69; H,
3.91; C17H14Cl2O4 requires C, 57.81; H 4.00).
3
2
5. Experimental
5.1. General details
Melting points were determined in capillaries and are
uncorrected. 1H and 13C NMR spectra were run on JEOL
300 MHz and 75 MHz NMR spectrometers, respectively, us-
ing CDCl3 as solvent. Chemical shifts are given in parts per
million with TMS as an internal reference. J values are given
in hertz. Chromatography was performed with silica 100e200
mesh and reactions were monitored by thin layer chromatog-
raphy (TLC) with silica plates coated with silica gel
HF-254. The experimental data for compounds 14e19 have
already been reported [34].
5.2.3. (2R*,3S*,5R*)-4,5-Bis-(2-chloro-phenyl)-4-hydroxy-
tetrahydro-furan-2-carboxylic acid 22
Compound 16 was allowed to react with PCC as described
in the general procedure to give 22 as white solid, 40% yield;
mp 168 ꢀC; nmax (CHCl3) 3431, 3418 (OH), 1797 (C]O)
5.2. Synthesis of compounds 20e25
1
cmꢁ1; H (CDCl3) d 1.60 (br s, 1H, OH, exchangeable with
The solution of pyridinium chloro chromate (PCC)
(2 mmol) in dry DCM (5 ml) was added to the solution of hy-
droxylmethyl tetrahydrofuran derivatives 14e19 (0.5 mmol)
in dry DCM and stirred for 20e24 h at 30 ꢃ 1 ꢀC (TLC mon-
itoring). The reaction mixture was washed with dry diethyl
ether. The organic layer was concentrated and column chroma-
tographed (silica gel 100e200) to isolate the products 20e25.
2
D2O), 2.47 (d, J ¼ 11.4 Hz, 1H, 4-H), 3.84 (dd, J ¼ 11.4 Hz,
3J ¼ 2.4 Hz, 1H, 4-H), 4.86 (d, J ¼ 2.4 Hz, 1H, 5-H), 6.24 (s,
1H, 2-H), 7.08e7.21 (m, 4H, ArH), 7.28e7.39 (m, 4H,
ArH); 13C (normal/DEPT-135) (CDCl3) d 44.75 (ꢁve, C-4),
77.73 (þve, C-5), 78.84 (þve, C-2), 91.68 (ab, C-3), 126.69
(þve, ArCH), 127.00 (þve, ArCH), 129.11 (þve, ArCH),
129.21 (þve, ArCH), 129.37 (þve, ArCH), 129.49 (þve,
ArCH), 129.72 (ab, ArC), 130.12 (þve, ArCH), 130.49 (þve,
ArCH), 131.56 (ab, ArC), 133.42 (ab, ArC), 170.83 (ab,
C]O). Decoupling of doublet at d 4.86 converts double
doublet at d 3.84 into a doublet and doublet at d 2.47 remains
unaffected. Decoupling of double doublet at d 3.84 converts
doublets at d 2.47 and d 4.86 into singlets. Here also the decou-
pling experiments show that 5-H couples with only one of the
5.2.1. (2R*,3S*,5R*)-4-Hydroxy-4,5-diphenyl-tetrahydro-
furan-2-carboxylic acid 20
Compound 14 was allowed to react with PCC as described
in the general procedure to give 20 as white solid, 31% yield;
mp 126 ꢀC; nmax (CHCl3) 3580 (OH), 1710 (C]O) cmꢁ1; 1H
(CDCl3) d 1.59 (br s, 1H, OH, exchangeable with D2O), 2.46
(d, J ¼ 11.1 Hz, 1H, 4-H), 3.02 (dd, 2J ¼ 11.1 Hz, 3J ¼ 2.1 Hz,

2798
P. Singh et al. / European Journal of Medicinal Chemistry 43 (2008) 2792e2799
two 4-H protons; FAB mass m/z 336 (Mþ ꢁ OH); (found C,
compound 25 were identified on the basis of the comparative
chemical shifts in compounds 20e24.
57.72; H, 3.89; C17H14Cl2O4 requires C, 57.81; H 4.00).
5.2.4. (2R*,3S*,5R*)-4,5-Bis-(4-fluoro-phenyl)-4-hydroxy-
tetrahydro-furan-2-carboxylic acid 23
Acknowledgements
Compound 17 was allowed to react with PCC as described
in the general procedure to give compound 23 as white solid,
40% yield; mp 135 ꢀC; nmax (CHCl3) 3481, 3418 (OH), 1805
(C]O) cmꢁ1; 1H (CDCl3) d 1.59 (br s, 1H, OH, exchangeable
with D2O), 2.46 (d, J ¼ 11.1 Hz, 1H, 4-H), 3.00 (dd,
We thank DST, New Delhi, CSIR, New Delhi and UGC,
New Delhi for financial assistance. Anu thanks CSIR, New
Delhi for Senior Research Fellowship. We also thank Dr.
V.L. Narayanan and his group at NIH, Bethesda, USA, for
anti-cancer screening. Authors are also thankful to the referees
for their encouraging comments, helpful in the revision of the
article.
3
2J ¼ 11.1 Hz, J ¼ 2.4 Hz, 1H, 4-H), 4.89 (d, J ¼ 1.8 Hz, 1H,
5-H), 5.20 (s, 1H, 2-H), 6.85e6.90 (m, 2H, ArH), 6.91e6.97
(m, 2H, ArH), 7.03e7.10 (m, 2H, ArH), 7.11e7.17 (m, 2H,
ArH); 13C NMR (normal/DEPT-135) (75 MHz, CDCl3)
d 45.21 (ꢁve, C-4), 78.06 (þve, C-5), 85.14 (þve, C-2),
91.42 (ab, C-3), 115.03 (þve, d, JCeF(ortho) ¼ 21.00 Hz,
ArCH), 115.83 (þve, d, JCeF(ortho) ¼ 21.68 Hz, ArCH),
127.42 (þve, d, JCeF(meta) ¼ 8.02 Hz, ArCH), 127.71 (ab, d,
References
[1] L.J. Marnett, S.W. Rowlinson, D.C. Goodwin, A.S. Kalgutkar,
C.A. Lanzo, J. Biol. Chem. 274 (1999) 22903e22906.
[2] W.L. Smith, R.M. Garavito, D.L. DeWitt, J. Biol. Chem. 271 (1996)
33157e33160.
JCeF(para) ¼ 3.68 Hz, ArC), 128.72 (þve, d, JCeF(meta)
¼
8.02 Hz, ArCH), 129.04 (ab, d, JCeF(para) ¼ 3.08 Hz, ArC),
162.80 (ab, d, JCeF ¼ 245.4 Hz, ArC), 162.94 (ab, d,
JCeF ¼ 246.6 Hz, ArC), 171.21 (ab, C]O); FAB mass m/z
303 (Mþ ꢁ OH); (found C, 63.70; H, 4.29; C17H14F2O4
requires C, 63.75; H 4.41).
[3] J.-B. Meric, S. Rottey, K. Olaussen, J.-C. Soria, D. Khayat, O. Rixe,
J.-P. Spano, Crit. Rev. Oncol. Hepatol. 59 (2006) 51e64.
[4] W. Dempke, C. Rie, A. Grothey, H.-J. Schmoll, J. Cancer Res. Clin.
Oncol. 127 (2001) 411e417.
[5] C.H. Liu, S.-H. Chang, K. Narko, O.C. Trifan, M.-T. Wu, E. Smith,
C. Haudenschild, T.F. Lane, T. Hla, J. Biol. Chem. 276 (2001)
18563e18569.
5.2.5. (2R*,3S*,5R*)-4-Hydroxy-4,5-bis-(4-methoxy-
phenyl)-tetrahydro-furan-2-carboxylic acid 24
[6] K. Subbaramaiah, A.J. Dannenberg, Trends Pharmacol. Sci. 24 (2003)
96e102.
[7] Y.T. Jeon, Y.S. Song, Mini Rev. Med. Chem. 6 (2006) 827e833.
[8] B. Miller, V.A. Patel, A. Sorokin, J. Am. Soc. Nephrol. 17 (2006)
977e985.
[9] V.A. Patel, M.J. Dunn, A. Sorokin, J. Biol. Chem. 277 (2002)
38915e38920 and references therein.
Compound 18 was allowed to react with PCC as described
in the general procedure to give 24 as thick liquid, 30% yield;
1H (CDCl3) d 1.26 (br s, 1H, OH, exchangeable with D2O),
2.42 (d, J ¼ 11.1 Hz, 1H, 4-H), 2.96 (dd, 2J ¼ 11.1 Hz,
3J ¼ 2.4 Hz, 1H, 4-H), 3.77 (s, 3H, OCH3), 3.80 (s, 3H,
OCH3), 4.83 (d, J ¼ 1.8 Hz, 1H, 5-H), 5.15 (s, 1H, 2-H), 6.78
(d, J ¼ 8.7 Hz, 2H, ArH), 6.84e6.89 (m, 4H, ArH), 7.07 (d,
J ¼ 9.0 Hz, 2H, ArH). MALDI (TOF) m/z 327.71 (Mþ ꢁ OH).
[10] A. Sorokin, Curr. Pharm. Des. 10 (2004) 647e657.
[11] G. Dannhardt, W. Kiefer, Eur. J. Med. Chem. 36 (2001) 109e126.
[12] C.R. Rodrigues, M.P. Veloso, H. Verli, C.A.M. Fraga, A.L.P. Miranda,
E.J. Barreiro, Med. Chem. Rev. 1 (2004) 73e90.
[13] S.M. Sondhi, M. Dinodia, J. Singh, R. Rani, Curr. Bioact. Compd. 3
(2007) 91e108.
[14] P. Singh, A. Mittal, Mini Rev. Med. Chem. 8 (2008) 73e90.
[15] E.F.F. da Cunha, T.C. Ramalho, D. Josa, M.S. Caetano, T.C.S. de Souza,
Recent Pat. Inflamm. Allergy Drug Disc. 1 (2007) 108e123.
[16] N. Pommery, T. Taverne, A. Telliez, L. Goossens, C. Charlier,
J. Pommery, J.-F. Goossens, R. Houssin, F. Durant, J.-P. Henichart, J.
Med. Chem. 47 (2004) 6195e6206.
5.2.6. (2R*,3S*,5R*)-4-Hydroxy-4,5-bis-(4-methane-
sulfonyl-phenyl)-tetrahydro-furan-2-carboxylic acid 25
Compound 19 was allowed to react with PCC as described
in the general procedure to give 25 as yellowish solid, 31%
1
yield (based upon H NMR integration); mp 180e85 ꢀC;
nmax (CHCl3) 3407 (OH), 1800 (C]O), 1307 (S]O) cmꢁ1
;
[17] R.M. Peek Jr., Cancer Chemother. Pharmacol. 54 (2004) S50eS56.
[18] L.J. Marnett, R.N. DuBois, Annu. Rev. Toxicol. 42 (2002) 55e80.
[19] F.H. Sarkar, S. Adsule, Y. Li, S. Padhye, Mini Rev. Med. Chem. 7 (2007)
599e608.
1H (CDCl3) d 1.81 (br s, 1H, OH, exchangeable with D2O),
1
2.53 (d, J ¼ 11.1 Hz, 1H, 4-H), 3.04 (s, 3H, SO2CH3), 3.05
(dd, 2J ¼ 11.1 Hz, 3J ¼ 1.8 Hz, 1H, 4-H), 3.09 (s, 3H,
SO2CH3), 5.01 (d, J ¼ 1.8 Hz, 1H, 5-H), 5.44 (s, 1H, 2-H),
7.08 (d, J ¼ 8.4 Hz, 2H, ArH), 7.48 (d, J ¼ 8.7 Hz, 2H,
ArH), 7.81 (d, J ¼ 8.4 Hz, 2H, ArH), 8.00 (d, J ¼ 8.4 Hz,
2H, ArH); 13C NMR (normal/DEPT-135) (75 MHz, CDCl3)
d 44.34 (þve, CH3), 44.34 (þve, CH3), 46.06 (ꢁve, C-4),
77.42 (þve, C-5), 78.32 (ab, C-3), 84.78 (þve, C-2), 118.23
(ab, ArC), 126.42 (þve, ArCH), 127.31 (þve, ArCH),
127.61 (þve, ArCH), 128.19 (þve, ArCH), 131.61 (ab,
ArC), 139.55 (ab, ArC), 188.82 (ab, C]O); MALDI (TOF)
m/z 477.8 (Mþ ꢁ 1 þ Kþ). From the 1H and 13C NMR spectra,
which show the signals due to two components, the signals of
[20] S. Touhey, R. O’Connor, S. Plunkett, A. Maguire, M. Clynes, Eur. J.
Cancer 38 (2002) 1661e1670.
[21] A.R. Maguire, S.J. Plunkett, S. Papot, M. Clynes, R. O’Connor,
S. Touhey, Bioorg. Med. Chem. 9 (2001) 745e762.
[22] S.-K. Kim, S.-Y. Lim, K.-C. Wang, Y.-Y. Kim, J.E.G. Chi, Y.L. Choi,
H.J. Shin, B.-K. Cho, Oncol. Rep. 12 (2004) 403e409.
[23] H.-K. Kang, E. Lee, H. Pyo, S.-J. Lim, Mol. Cancer Ther. 4 (2005)
1358e1363.
[24] P. Singh, A. Mittal, S. Kaur, S. Kumar, Bioorg. Med. Chem. 14 (2006)
7910e7916.
[25] E.J. Corey, W. Suggs, Tetrahedron Lett. 31 (1975) 2647e2650.
[26] M.C. Ally, D.A. Scudiero, P.A. Monks, M.L. Hursy, M.J. Czerwinski,
D.A. Fine, B.J. Abbott, J.G. Mayo, R.H. Showmaker, M.R. Boyd, Cancer
Res. 48 (1988) 589e601.

P. Singh et al. / European Journal of Medicinal Chemistry 43 (2008) 2792e2799
2799
[27] M.R. Grever, S.A. Schepartz, B.A. Chabner, Semin. Oncol. 19 (1992)
622e638.
[31] A.L. Blobaum, L.J. Marnett, J. Med. Chem. 50 (2007) 1425e1441 and
references therein.
[28] M.R. Boyd, K.D. Paull, Drug Dev. Res. 34 (1995) 91e109.
[29] P. Singh, P. Kaur, Anu, S. Kumar, Indian J. Chem. 45B (2006) 1692e
1698.
[32] C. Michaux, C. Charlier, Mini Rev. Med. Chem. 4 (2004) 603e615.
[33] L.R. Reddy, E.J. Corey, Tetrahedron Lett. 46 (2005) 927e929.
[34] P. Singh, A. Mittal, S. Kumar, Bioorg. Med. Chem. 15 (2007)
3990e3996.
[30] R.R. Ranatunge, R.A. Earl, D.S. Garvey, D.R. Janero, L.G. Letts,
A.M. Martino, M.G. Murty, S.K. Richardson, D.J. Schwalb,
D.V. Young, I.S. Zemtseva, Bioorg. Med. Chem. Lett. 14 (2004)
6049e6052.
[35] log P and total polar surface area were calculated using molinspiration
(www.molinspiration.com). Other descriptors were taken from BioMed
Cache 7.5.0.85.