Chemical and biological investigation of N-hydroxy-valdecoxib: An active metabolite of valdecoxib

Article abstract of DOI:10.1016/j.bmc.2008.02.088

The inhibition of cyclooxygenase enzymes plays an important role in the treatment of inflammatory diseases. N-Hydroxy-4-(5-methyl-3-phenylisoxazol-4-yl)benzenesulfonamide (3)-a primary metabolite of the highly selective COX-2 inhibitor valdecoxib-was synthesized and stabilized as its monohydrate (3a·H₂O). The anti-inflammatory properties of 3a·H₂O were investigated in carrageenan-induced edema and in acute and chronic pain models. Based on our biological investigation, we conclude that N-hydroxy-valdecoxib 3a is an active metabolite of valdecoxib.

Full text of DOI:10.1016/j.bmc.2008.02.088

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 5322–5330
Chemical and biological investigation of N-hydroxy-valdecoxib:
An active metabolite of valdecoxib
a,
a
b
c
´
*
´
´
´
´ ´
Peter Erdelyi, Tamas Fodor, Agnes Kis Varga, Matyas Czugler,
b
a
´
´
Aniko Gere and Janos Fischer
aMedicinal Chemistry Research Laboratory No. IV., Gedeon Richter Plc., PO Box 27, H-1475 Budapest 10, Hungary
bDepartment of Pharmacology, Gedeon Richter Plc., PO Box 27, H-1475 Budapest 10, Hungary
cX-ray Diffraction Department, Institute of Structural Chemistry, Chemical Research Center, Hungarian Academy of Sciences,
Pusztaszeri u´t 59-67, 1025 Budapest, Hungary
Received 14 October 2007; revised 25 February 2008; accepted 28 February 2008
Available online 4 March 2008
Abstract—The inhibition of cyclooxygenase enzymes plays an important role in the treatment of inflammatory diseases. N-Hydroxy-
4-(5-methyl-3-phenylisoxazol-4-yl)benzenesulfonamide (3)—a primary metabolite of the highly selective COX-2 inhibitor valdecox-
ib—was synthesized and stabilized as its monohydrate (3aÆH2O). The anti-inflammatory properties of 3aÆH2O were investigated in
carrageenan-induced edema and in acute and chronic pain models. Based on our biological investigation, we conclude that N-hydro-
xy-valdecoxib 3a is an active metabolite of valdecoxib.
ꢀ 2008 Elsevier Ltd. All rights reserved.
1. Introduction
sis of PGs is determined by COX enzymes, the amount
of which is mediated by the cytokines selected by the
stimulated cells. The cyclooxygenase enzyme exists in
two isoforms: cyclooxygenase-1 (COX-1) and cyclooxy-
genase-2 (COX-2).3,4 Further experiments including the
cloning of COX-2 led to the realization that COX-1 is
constitutively expressed in essentially all tissues whereas
COX-2 is induced by inflammatory stimuli.5–7 It is
known that the latter is responsible for the conversion
of arachidonic acid to prostaglandin H2, the key media-
tor of inflammation.8 The long term use of NSAIDs is
limited by gastrointestinal side effects such as dyspepsia
and abdominal pain, and less often, gastric or duodenal
perforation and bleeding.9 Coxibs, the highly COX-2
selective second generation of NSAIDs, were developed
to achieve a better gastrointestinal adverse effect profile;
however, these side effects could only be reduced. The
first member of coxibs is celecoxib an ortho-diaryl-het-
erocyclic structure was launched in 199910 (Scheme 1).
In the recent years, however most of the coxibs have
been withdrawn from the market because of their unex-
pected cardiovascular effects.
Non-steroidal anti-inflammatory drugs (NSAIDs) are
widely used in the treatment of diseases accompanied
by rheumatic inflammation, pain, and fever. The action
of these compounds are based on the inhibition of the
enzyme cyclooxygenase (COX), sometimes referred to
as prostaglandin synthase.1 Prostaglandins (PGs) are
lipid compounds derived from fatty acids that are
produced by most cells in the body and are present in
virtually all human organs and tissues. PGs are
considered to be mediators with a variety of strong
physiological effects such as platelet aggregation and
gastrocytoprotection. Another important function of
PGs is the response of the human tissue to inflammation
and injury. They play a significant role in symptoms
accompanied by inflammation: swelling, redness, and
pain because of the increased vascular permeability
and the sensitized peripheral nerve endings and addi-
tionally fever is also mediated by PGs. It was determined
that the levels of PGs are appreciably elevated in in-
flamed tissues compared to normal tissues.2 The synthe-
The scope of our research programme was to find novel
selective COX-2 inhibitors with a better side effect pro-
file by the method of analogue based drug discovery11
using valdecoxib12 (Scheme 1) as a lead compound.
Keywords: Valdecoxib analogues; Cyclooxygenase-2 inhibitor; COX-2;
Active metabolite; Non-steroidal anti-inflammatory drugs.
*
Corresponding author. Tel.: +36
p.erdelyi@richter.hu
1
889 8711; e-mail:
0968-0896/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.02.088

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5323
Scheme 1. Celecoxib and valdecoxib.
The close analogues of celecoxib and rofecoxib have al-
ready been reported in this journal by others.13,14
ylation at the sulfonamide moiety in mice15 and in
humans.16
There are no synthetic and biological data on this
metabolite of valdecoxib. The present work reports on
its synthesis and biological investigations.
2. Chemistry
Several novel direct analogues of valdecoxib were pre-
pared and their pharmacological properties were studied
in vitro and in vivo. N-Hydroxy-4-(5-methyl-3-phen-
ylisoxazol-4-yl)benzenesulfonamide (N-hydroxy-valdec-
oxib) (3) was found to be the only effective compound
among these analogues. It has been reported earlier that
valdecoxib transforms in primary oxidative metabolic
pathways to N-hydroxy-valdecoxib through N-hydrox-
Compound 3 was prepared according to the procedure
outlined in Scheme 2. 5-Methyl-3,4-diphenylisoxazole
(1)12 was chlorosulfonated to give a mixture of isomers
2a and 2b. Recrystallization from cyclohexane afforded
the pure 4-isomer, which was treated with hydroxyl-
amine hydrochloride in dioxane to give 3 in a good
yield.
Scheme 2. The synthesis of N-hydroxy-valdecoxib (3).
Scheme 3. The preparation of N-hydroxy-valdecoxib monohydrate (3aÆH2O).

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5324
Figure 1. (a) Molecular structure of 3aÆH2O with 50% probability anisotropic displacement ellipsoids. (b) Packing in the crystal structure of 3aÆH2O
as seen along the b crystallographic axis. Hydrophilic layers formed at near to x ꢀ 0 are visible while apolar (hydrophobic) surfaces are clearly
exposed at around x ꢀ 1/2.
Unfortunately, stability problems were observed with
N-hydroxy-valdecoxib (3) prepared as described above
via a fast decomposition to the sulfinic acid derivative
at room temperature. Other aromatic N-hydroxysulf-
onamides such as Piloty’s acid (N-hydroxy-benzenesul-
fonamide) decompose while nitroxyl (HNO) is
generated.17 If crude N-hydroxy-valdecoxib (3) was
dissolved in EtOAc and treated with Na2EDTA solution
to remove chelating metal ion traces, and crystallized
from a dilute solution of L-ascorbic acid in aqueous eth-
anol, 3 was isolated as a stable monohydrate (3aÆH2O)
whose water content remained constant at room temper-
atures (under standard humidity conditions) over a per-
iod of two years (Scheme 3). According to TGA and
DSC analysis 3aÆH2O decomposed to afford one mole
of water between 93 and 160 ꢁC and a fast decomposi-
tion occurred at 180 ꢁC. We have also found other stable
solvates of 3, whose preparation and structure determi-
nations by X-ray crystallography will be reported else-
where. The molecular structure of 3aÆH2O as seen
from the crystal structure is shown in Figure 1a. The
crystal structure is characterized by a double-layer like
arrangement of the hydrogen-bonded array of the asso-
ciated 3 molecules and water (cf. Fig. 1a and b) yielding
to such layers. These double-layer like formations with
their polar core and apolar exterior are then combined
into making the macroscopic crystal.
3. Pharmacology
The effect of N-hydroxy-valdecoxib monohydrate
(3aÆH2O) on human COX-2 inhibition was 28 times
Table 1. In vitro efficacity of valdecoxib and 3aÆH2O in the TMPD
assay
Compound
IC50 SEM (lM)
Human
recombinant
COX-2
Ovine COX-1
COX-1/ COX-2
selectivity
Valdecoxib
3a
0.04 0.02
1.1 0.2
157.2 6.2
96.2 10.2
3930
88

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5325
Figure 4. Effect of the test compounds in the LPS evoked PGE2
production assay.
Figure 2. Inhibition effect on ovine COX-1 enzyme.
lower than that of valdecoxib, therefore, the COX-1/
COX-2 selectivity of 3aÆH2O was much lower than that
of valdecoxib (Table 1 and Figs. 2 and 3). Compound
3aÆH2O inhibited the LPS-induced PGE2 elevation sig-
nificantly weaker than valdecoxib (Fig. 4).
Compound 3aÆH2O had a dose dependent significant
edema inhibition effect in the carrageenan-induced paw
edema assay. It had an anti-edema effect at a dose of
0.3 mg/kg po whereas valdecoxib showed no inhibition
at the same dose (Figs. 5 and 6).
The analgetic and anti-inflammatory activity was deter-
mined in rats in acute and chronic experiments. The car-
rageenan provoked mechanical hyperalgesia assay
indicated that both compounds—valdecoxib and
3aÆH2O—had considerable analgesic activity (Fig. 7).
Compound 3aÆH2O showed a significant analgesic activ-
ity for 3 h after treatment and it has a longer duration of
action as compared with valdecoxib.
Figure 5. Effect of valdecoxib in paw edema.
In the subchronic pain model, valdecoxib reached the
efficacy of 3aÆH2O only at a higher dose level (10 mg/
kg vs 3 mg/kg), (Fig. 8). Valdecoxib and 3aÆH2O have
Figure 6. Effect of 3aÆH2O in paw edema.
strong analgesic effect in the carrageenan-induced
mechanical allodynia assay (Figs. 9 and 10).
Figure 3. Inhibition effect on human COX-2 enzyme.

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5326
Figure 7. Effects of test compounds in mechanical hyperalgesia.
Figure 10. Effect of 3aÆH2O in mechanical allodynia.
Figure 11. Effect of test compounds in monoarthritis model.
Figure 12. Effect of valdecoxib in writhing test in mice.
Figure 8. Effects of test compounds in subchronic pain model.
Figure 9. Effect of valdecoxib in mechanical allodynia.
According to the carrageenan–kaolin-induced monoar-
thritis model both compounds produced almost total
inhibition (complete reversal). The analgesic activity
had a duration of 2 h (Fig. 11).
In the phenylquinone-induced visceral pain model val-
decoxib and 3aÆH2O are very effective. Test compounds
show dose dependent significant analgesic activity:
ED50(valdecoxib): 15.5 mg/kg; ED50(3aÆH2O): 3.8 mg/
kg po (Figs. 12 and 13).

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5327
of compound 3 from a water–ethanol mixture and elim-
inating the traces of metal ions and free radicals with the
help of EDTA and L-ascorbic acid to afford a stable
monohydrate 3aÆH2O. This monohydrate had a weak
in vitro COX-2 activity as compared to valdecoxib.
Based on in vivo analgesic and anti-inflammatory tests
3aÆH2O is more potent and has a longer duration of ac-
tion than the reference compound and significantly in-
creases the coronary blood flow in isolated rabbit
heart. The nitric oxide (NO) donor properties of 3aÆH2O
could play a role in its mechanism of action.
5. Methods
In vitro and in vivo testing were performed as given in
the literature cited.19–21 Only modifications are de-
scribed in this article.
Figure 13. Effect of 3aÆH2O in writhing test in mice.
5.1. Spectrophotometric assays with human recombinant
COX-2 and bovine COX-1
Enzymatic activities of the human recombinant COX-2
and bovine COX-1 were measured using a spectropho-
tometric assay based on the oxidation of TMPD during
the reduction of PGG2 to PGH2. Valdecoxib and
3aÆH2O inhibited human recombinant COX-2 in a con-
centration dependent manner. Inhibition data of valdec-
oxib and 3aÆH2O are shown in Table 1 and Figures 2
and 3.
5.2. LPS evoked PGE2 production in human whole blood
The production of PGE2 by LPS-challenged human
whole blood has been used to evaluate the efficacy of
COX-2 inhibitors in vitro. PGE2 production was mea-
sured in heparinized whole blood samples, incubated
with LPS for 24 h, as a reflection of the inducible
COX activity of monocyte COX-2 (Fig. 4).
Figure 14. Cardiac effect of valdecoxib and 3aÆH2O on isolated rabbit
heart.
5.3. Carrageenan-induced paw edema
Compound 3aÆH2O significantly increases coronary
blood flow dose dependently without altering other
parameters (QTS, HR, LPV) in rabbits. Valdecoxib
has no effect on coronary blood flow (Fig. 14). The nitric
oxide (NO) donor properties of 3aÆH2O could explain
this phenomenon.18 Neither valdecoxib nor its analogue
3aÆH2O modified the resting blood pressure of the con-
scious animals as compared with the solvent-treated
group and the values were 84.5 3.54, 82.4 7.96,
and 86.45 6.38 mmHg, respectively.
The edema was induced in male Wistar rats (130–150 g)
by intraplantar injection of carrageenan (CARR; 50 ll
of 1% solution) into the right hind paw 1 h after the oral
administration of a drug (Figs. 5 and 6). The irritant in-
duced a rapid swelling of the right paw being maximal at
3 h and lasting for 6 h. Right hind paw volume was mea-
sured with a plethysmometer (Ugo Basile, type 7150) be-
fore and 1, 2, 3, 4, 5, and 6 h after the CARR injection.
The increase of paw volume is calculated as a percentage
compared with the volume measured immediately be-
fore injection of the irritant for each animal. Effectively
treated animals show much less edema. The difference of
average values between treated animals and vehicle con-
trol groups is calculated for each time interval and sta-
tistically evaluated.22
4. Conclusions
Direct analogues of valdecoxib have been prepared to
obtain a drug analogue with better properties. N-Hydro-
xy-valdecoxib (3a) proved to be an active compound. It
is a known metabolite of valdecoxib in human and in
mice. The synthesis and the stabilization of 3a are de-
scribed. The unsolvated N-hydroxy-valdecoxib 3 proved
to be unstable. It could be stabilized by recrystallization
5.4. Carrageenan-induced mechanical hyperalgesia
CARR-induced hyperalgesia is one of the most fre-
quently applied model of inflammatory pain. The mac-
romolecule induces local inflammation resulting in

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5328
persisting hyperalgesia and allodynia. This assay moni-
tors the capacity of the test drugs to reverse hyperalgesia
(i.e., decreased threshold to painful stimuli) produced by
carrageenan (CARR)-induced inflammation. Male albi-
no rats (Wistar strain—TOXI-COOP Ltd, Hungary–
weighing 140–190 g) were used (n = 6/group). The
mechanical pain threshold of the inflamed hind paw
was measured according to the Randall-Selitto method
with an ‘analgesimeter’ (Ugo Basile). The hind paw
was compressed with a progressively increasing pres-
sure. The nociceptive threshold is defined as the force,
in grams when the animal first vocalized or made a vig-
orous attempt to remove the paw. Data are presented as
the absolute threshold values (means SEM) in grams.
The percentage reversal of mechanical hyperalgesia was
calculated from mean threshold values of a given treat-
ment group as follows:
cavity of the right knee (in 0.1 mL sterile saline) under
ether anesthesia (n = 8–10/group). Control rats were in-
jected with vehicle.
In therapeutic type experiment drugs were given 4 h
after CarrKao injection. Time of measurements: pre-
injection of CarrKao, immediately before drug treat-
ment, and 1 and 2 h after the treatment. Change in
weight load in the injected and contra-lateral hind limb
was measured by ICT device in grams.
The incapacitance was calculated as follows:
mll ꢁ mrl
% incapacitance ¼
ꢂ 100;
mll þ mrl
wherein mll and mrl are the left and right limb weight
load (g).
T xh ꢁ T0h
The efficacy of treatment was characterized by percent-
age reversal of IC% for various time points. Intraarticu-
lar administration of CarrKao induced a significant
reduction of weight load on the affected leg, with maxi-
mum effect at 3–5 h after the injection (Fig. 11).22
% reversal ¼
ꢂ 100;
T lh ꢁ T0h
wherein Txh and T0h are the post- and pre-dose thresh-
old and Tlh is the baseline threshold.
• Acute experiments: On the day of the experiment the
baseline threshold was determined first. After base-
line measurements the animals received CARR
(100 ll of 2% solution) into the plantar surface of
right hind paw. The injected hind paw became maxi-
mally inflamed by 2.5–3 h after injection, character-
ized by edema and decreased mechanical pain
threshold. Test compounds were administered orally
immediately after 1 h reading. Changes in the thresh-
old were checked at 1, 2, and 3 h following drug
administration (Fig. 7).
• Subchronic experiments: After baseline measurements
the animals received CARR (150 ll mL of 2% solu-
tion) ip of right hind paw. The effects of analgesic
drugs were studied 24 h after induction of inflamma-
tion (in subchronic phase). Test compound was
administered orally immediately after 24 h reading
(Fig. 8).
5.7. Visceral pain model in mice
The phenylquinone (PQ) abdominal contraction assay
was used to study the effects of increasing doses of test
compounds in a chemo-nociceptive assay.
COX-2 Expression can be induced by lipopolysaccha-
ride (LPS) pre-treatment with 24 h before the study.
Inducible COX-2 can be upregulated in acute inflamma-
tory processes. Male NMRI mice (23–30 g, TOXI-
COOP Ltd, Hungary) were orally administered with test
compound or vehicle 30 min before the challenge with
the intraperitoneal injection of 0.02% PQ. Immediately
after ip administration of the irritant the animals were
placed individually in boxes (12 · 12 · 12 cm) and the
number of abdominal contractions (writhing) were
counted for 5–20 min post-injection. Each dose groups
consisted of eight animals. The number of these abdom-
inal writhings (constriction of abdomen, turning of
trunk, and extension of hind legs) was counted during
a 20 min session, beginning 5 min after the injection of
PQ. The analgesic activity of the drugs at each dose
was determined as a percentage inhibition compared
to the degree of writhing in the vehicle controls. The
anti-nociceptive response was analyzed by analysis of
variance (ANOVA) followed by Tukey’s t-test to assess
the significance (Figs. 12 and 13).
5.5. Carrageenan-induced mechanical allodynia in rats
The Electronic von Frey Anesthesiometer was used for
the determination of pain sensitivity. The test is per-
formed by applying pressure to the paw through the
hand-held force transducer with a sensitive cotton
coated rigid tip of the probe. After baseline measure-
ments the animals received CARR (100 ll of 2% solu-
tion) into the plantar surface of right hind paw (Figs.
9 and 10).
5.8. Cardiac effect on isolated rabbit heart
5.6. Carrageenan–Kaolin-induced monoarthritis in rats
New Zealand white rabbits hearts were perfused via the
aorta with oxygenated, thermostated Krebs solution.
The test compounds were dissolved in the perfusion
solution to obtain the requested concentrations. The ba-
sic value of coronary blood flow was determined. After-
wards measured amounts of compounds were added to
the perfusion liquid and the perfusion was recorded
for 30 min at every 10 min. After 30 min perfusion per-
formed without compound the measurement was
An Incapacitance Tester (ICT) was used to measure the
change in weight load on the hind paws before and after
induction of monoarthritis, as a measure of functional
disability caused by arthritic inflammation and pain.
Male WISTAR rats (TOXI-COOP Ltd, Hungary—
weighing 120–190 g) were injected with mixture of 2%
carrageenan and 2% kaolin (CarrKao) into the joint

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5329
repeated with the medium and high amount of com-
pounds (Fig. 14).
fractometers, Philipps-Universita¨t Marburg, Germany.
The structure was elucidated by SHELXS: Sheldrick,
G. M. SHELXS-97, Program for Crystal Structure
Solution, University of Go¨ttingen, Germany, 1997.
Structure refinement was carried out by Sheldrick, G.
M. SHELXL-97, Program for Crystal Structure Refine-
ment, University of Go¨ttingen, Germany, 1997.
5.9. Synthetic procedure for compound 3a
1H NMR spectra were obtained on a Varian Unity Ino-
va 300 spectrometer. Chemical shifts are reported in
parts per million relative to TMS as internal standard.
MS spectra were measured on Finnigan Mat 95 SQ
and 95 XP devices.
A plate-like single crystal of 3aÆH2O was selected for X-
ray crystal structure determination. A crystal of
FT_WA was mounted on a glass fiber. Cell parameters
were determined by least-squares of the setting angles
of 25 (20.24 6 h 6 24.36ꢁ) reflections. Crystal data:
C16H16N2O5S, Fwt: 348.37, colorless, plate, size:
0.22 · 0.11 · 0.09 mm, monoclinic, space groupP21/c,
5.9.1. 4-(5-Methyl-3-phenylisoxazol-4-yl)benzenesulfonyl
chloride (2a). Compound 1 (8.00 g, 34.0 mmol) was
added to chlorosulfonic acid (22.6 mL, 340 mmol) at
5 ꢁC in small portions over 30 min and the mixture
was left to stand for 5 h at room temperature. The con-
version was followed by TLC: toluene/EtOAc 12:1.
After the completion of the reaction, the mixture was
poured into crushed ice and the resulting aqueous sus-
pension was extracted with CH2Cl2 (2 · 90 mL). The
combined organic layers were washed with brine, dried
over anhydrous MgSO4, filtered, and concentrated in
vacuo to give a crude product. Recrystallization from
cyclohexane to remove meta-isomer (2b) provided the
pure ortho-sulfonyl chloride 2a (7.26 g, 64%). TLC: tol-
uene/EtOAc 12:1, silica, Rf = 0.55. MS (EI) M+ = 333,
1H NMR (300 MHz, DMSO-d6, 30 ꢁC) d 7.71–7.64
(m, 2H); 7.49–7.34 (m, 5H); 7.24–7.17 (m, 2H); 2.45 (s,
3H) ppm.
˚
˚
˚
a = 7.659(1) A, b = 23.510(1) A, c = 9.148(1) A, a = 90ꢁ,
3
˚
b = 95.65(1)ꢁ, c = 90ꢁ, V = 1639.2(3) A , T = 295(2) K,
Z = 4,
F(000) = 728,
Dx = 1.412 mg/m3,
l = 2.022 mmꢁ1. Intensity data were collected on a En-
raf-Nonius CAD4 diffractometer (graphite monochro-
mator; Cu-Ka radiation, k = 1.54184 A) at 295(2) K in
˚
the range 3.76 6 h 6 75.58ꢁ using x/2h scans. The inten-
sities of three standard reflections were monitored regu-
larly (every 60 min). The intensities of the standard
reflections remained constant within experimental error
throughout the data collection. A total of 3655 reflec-
tions were collected of which 3344 were unique
[R(int) = 0.0077, R(r) = 0.0528]; intensities of 2021
reflections were greater than 2r(I), completeness to
h = 0.981. No absorption correction was applied to the
data. The structure was solved by direct methods (and
subsequent difference syntheses). Anisotropic full-matrix
least-squares refinement on F2 for all non-hydrogen
atoms yielded R1 = 0.0406 for 2029 [I > 2r(I)] and
R1 = 0.0855 and wR2 = 0.1089 for all (3352) intensity
data. The maximum and minimum residual electron
density in the final difference map were 0.25 and
5.9.2. N-Hydroxy-4-(5-methyl-3-phenylisoxazol-4-yl)ben-
zenesulfonamide monohydrate (3aÆH2O). A suspension of
hydroxylamine hydrochloride (6.88 g, 99.0 mmol) in
dioxane (50 mL) was added to a solution of sodium ace-
tate (8.12 g, 99.0 mmol) in water (25 mL). To this mix-
ture, compound 2a (11.0 g, 33.0 mmol) in dioxane
(50 mL) was added at 10 ꢁC dropwise, over a period of
30 min and it was then stirred for additional 30 min at
room temperature. The reaction mixture was diluted
with water (500 mL) and shaken for 2 h. The suspension
was filtered off, the filtrate was dissolved in EtOAc
(200 mL), extracted with an aqueous solution (40 mL)
of 5 m/m% Na2EDTA, with water (40 mL) and finally
with brine (20 mL) and evaporated in vacuo. The resi-
due was dissolved in EtOH (90 mL), decolorized by acti-
vated carbon (1 g), filtered, and water (270 mL)
containing ascorbic acid (3 g) was added to the solution
at 60 ꢁC. The mixture was cooled to 5 ꢁC, the precipitate
formed was collected by filtration, washed with water
and dried to provide the title compound 3aÆH2O
ꢁ3
˚
ꢁ0.26 e A
. The weighting scheme applied was
2
w ¼ 1=½r2ðF 2oÞ þ ð0:055PÞ þ 0:0000Pꢃ
where
P ¼ ðF 2o þ 2F 2cÞ=3. Hydrogen atomic positions were cal-
culated from assumed geometries except –OH and –NH
H atoms that were located in difference maps. Hydrogen
atoms were included in structure factor calculations but
they were not refined. The isotropic displacement
parameters of the hydrogen atoms were approximated
from the U(eq) value of the atom they were bonded to.23
Acknowledgments
1
(7.80 g, 68%). Mp 95–110 ꢁC. MS (EI) M+ = 330, H
We thank Drs. Gabor Tarkanyi and Tamas Gati for the
measurements and elucidations of the NMR spectra. We
´
´
´
´
´
NMR (300 MHz, DMSO-d6, 30 ꢁC) d 9.66 (s, 2H);
7.87–7.80 (m, 2H); 7.50–7.32 (m, 7H); 2.49 (s, 3H) ppm.
´
also thank Sandor Levai for his contribution to the sta-
bility assays and HPLC measurements.
´
5.10. Crystal structure determination of 3aÆH2O
Crystallographic data (excluding structure factors) for
the structure in this paper have been deposited with
the Cambridge Crystallographic Data Centre as supple-
mentary publication No. CCDC 660035.
References and notes
1. Vane, J. R. Nature (New Biol.) 1971, 231, 232.
2. Raz, A.; Wyche, A.; Siegel, N.; Needleman, P. J. Biol.
Chem. 1988, 263, 3022.
Data reduction was obtained from Harms, K. (1996)
XCAD4 Data Reduction Programme for CAD4 Dif-
3. Fu, J. R.; Masferrer, J. L.; Seibert, K.; Raz, A.; Needle-
man, P. J. Biol. Chem. 1990, 265, 16737.

´
P. Erdelyi et al. / Bioorg. Med. Chem. 16 (2008) 5322–5330
5330
4. Battistini, B.; Botting, B.; Bakhle, Y. S.. Drug News
Perspect. 1994, 7, 501.
14. Zarghi, A.; Rao, P. N. P.; Knaus, E. E. Bioorg. Med.
Chem. 2007, 15, 1056.
5. Xie, W.; Chipman, J. G.; Robertson, D. L.; Erikson, R.
L.; Simmons, D. L. Proc. Natl. Acad. Sci. U.S.A. 1991, 88,
2692.
6. Kujubu, D. A.; Fletcher, B. S.; Varnum, B. C.; Lim, R.
W.; Herschman, H. R. J. Biol. Chem. 1991, 266, 12866.
7. O’Banion, M. K.; Sadowski, H. B.; Winn, V.; Young, D.
A. A. J.. Biol. Chem. 1991, 266, 23261.
15. Zhang, J. Y.; Yuan, J. J.; Wang, Y. F.; Bible, R. H.;
Breau, A. P. Drug Metab. Dispos. 2003, 31, 491.
16. Yuan, J. J.; Yang, D. C.; Zhang, J. Y.; Bible, R.; Karim,
A.; Findlay, J. W. A. Drug Metab. Dispos. 2002, 30, 1013.
17. Bonner, F. T.; Ko, Y. Inorg. Chem. 1992, 31, 2514.
18. Zamora, R.; Grzesiok, A.; Weber, H.; Feelisch, M.
Biochem. J. 1995, 312, 333.
8. Vane, J. R.; Botting, R. M. . Adv. Prostaglandin, Throm-
boxane, Leukot. Res. 1995, 23, 41.
19. Gierse, K.; Hauser, S. D.; Creely, D. P.; Koboldt, C. M.;
Rangwala, S. H.; Isakson, P. C.; Siebert, K. Biochem. J.
1995, 305, 479.
20. Drug Discovery and Evaluation Pharmacological Assays;
Vogel, H. G., Vogel, W. H., Eds.; Springer-Verlag: Berlin,
Heidelberg, 1997.
9. Allison, M. C.; Howatson, A. G.; Torrance, C. J.; Lee, F.
D.; Russell, R. I. G. N. Engl. J. Med. 1984, 310, 563.
10. Penning, T. D.; Talley, J. J.; Bertenshaw, S. R.; Carter, J.
S.; Collins, P. W.; Docter, S.; Graneto, M. J.; Lee, L. F.;
Malecha, J. W.; Miyashiro, J. M.; Rogers, R. S.; Rogier,
D. J.; Yu, S. S.; Anderson, G. D.; Burton, E. G.; Cogburn,
J. N.; Gregory, S. A.; Koboldt, C. M.; Perkins, W. E.;
Seibert, K.; Veenhuizen, A. W.; Zhang, Y. Y.; Isakson, P.
C. J. Med. Chem. 1997, 40, 1347.
11. Analogue-Based Drug Discovery; Fischer, J., Ganellin, C.
R., Eds.; Wiley-VHC Verlag GmbH & Co. KGaA, 2006.
12. Talley, J. J.; Brown, D. L.; Carter, J. S.; Graneto, M. J.;
Koboldt, C. M.; Masferrer, J. L.; Perkins, W. E.; Rogers,
R. S.; Shaffer, A. F.; Zhang, Y. Y.; Zweifel, B. S.; Seibert,
K. J. Med. Chem. 2000, 43, 775.
21. Atkinson, D. C. Arch. Int. Pharmacodyn. Ther. 1971, 193,
391.
22. Statistics: Data were analyzed by two-way ANOVA
followed by Tukey’s (HSD) post hoc test, comparing
drug-treated with vehicle-treated animals. Significance was
considered at p < 0.05 level (*p < 0.05; **p < 0.01;
***p < 0.001).
23. Neutral atomic scattering factors and anomalous scat-
tering factors are taken from: International Tables for X-
ray Crystallography; Wilson, A. J. C., Ed.; Kluwer
Academic Publishers: Dordrecht, 1992; Vol. C,, Tables
6.1.1.4, pp 500–502, 4.2.6.8, pp 219–222 and 4.2.4.2, pp
193–199.
13. Uddin, Md. J.; Rao, P. N. P.; Knaus, E. E. Bioorg. Med.
Chem. 2003, 11, 5273.