Non-thiazolidinedione antihyperglycaemic agents. Part 3: The effects of stereochemistry on the potency of α-methoxy-β-phenylpropanoic acids

Article abstract of DOI:10.1016/S0968-0896(99)00034-6

Rhizopus delemar lipase catalysed ester hydrolysis of the α-methoxy-β-phenylpropanoate 1 affords the (R)-(+) and (S)-(-) isomers in >84% enantiomeric excess. Absolute stereochemistry was determined by a single crystal X-ray analysis of a related synthetic analogue. The activity of these two enantiomers on glucose transport in vitro and as anti-diabetic agents in vivo is reported and their unexpected equivalence attributed to an enzyme-mediated stereospecific isomerisation of the (R)-(+) isomer. Binding studies using recombinant human PPARγ (peroxisomal proliferator activated receptor γ), now established as a molecular target for this compound class, indicate a 20-fold higher binding affinity for the (S) antipode relative to the (R) antipode. Copyright (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(99)00034-6

Bioorganic & Medicinal Chemistry 7 (1999) 821±830
Non-thiazolidinedione Antihyperglycaemic Agents.
Part 3: The E€ects of Stereochemistry on the Potency of
ꢀ-Methoxy-ꢁ-phenylpropanoic Acidsy
David Haigh,a,* Graham Allen,b Helen C. Birrell,a Derek R. Buckle,a
Barrie C. C. Cantello,a Drake S. Eggleston,a R. Curtis Haltiwanger,c Julie C. Holder,a
Carolyn A. Lister,a Ivan L. Pinto,a Harshad K. Rami,a John T. Sime,a
Stephen A. Smitha and John D. Sweeneyb
aSmithKline Beecham Pharmaceuticals, New Frontiers Science Park North, Third Avenue, Harlow, Essex, CM19 5AW, UK
bSmithKline Beecham Pharmaceuticals, The Frythe, Welwyn, Herts., AL6 9AR, UK
cSmithKline Beecham Pharmaceuticals, Physical and Structural Chemistry Department, PO Box 1539,
King of Prussia, PA, 19406-0939, USA
Received 26 April 1998; accepted 23 October 1998
AbstractÐRhizopus delemar lipase catalysed ester hydrolysis of the a-methoxy-b-phenylpropanoate 1 a€ords the (R)-(+) and (S)-
( ) isomers in >84% enantiomeric excess. Absolute stereochemistry was determined by a single crystal X-ray analysis of a related
synthetic analogue. The activity of these two enantiomers on glucose transport in vitro and as anti-diabetic agents in vivo is
reported and their unexpected equivalence attributed to an enzyme-mediated stereospeci®c isomerisation of the (R)-(+) isomer.
Binding studies using recombinant human PPARg (peroxisomal proliferator activated receptor g), now established as a molecular
target for this compound class, indicate a 20-fold higher binding anity for the (S) antipode relative to the (R) antipode. # 1999
Elsevier Science Ltd. All rights reserved.
Introduction
zone,7 troglitazone8 and rosiglitazone (BRL 49653),9,10
have been shown to increase insulin sensitivity and nor-
malize glycaemic control in rodent models of diabetes.
Similar results have also been reported with troglitazone
Type 2 (non-insulin-dependent) diabetes mellitus is a
chronic metabolic disorder characterised by a pro-
gressive resistance of peripheral tissues to the metabolic
e€ects of insulin.1 Insulin resistance interferes with glu-
cose homeostasis through impaired uptake and utilisa-
tion of glucose in skeletal muscle and adipose tissues
and through reduced control of hepatic glucose output.2
Impaired release of insulin in response to elevated
plasma glucose is a feature of Type 2 diabetes which
leads to persistent hyperglycaemia and long term dia-
betic complications.3 Moreover, several prospective
studies indicate that the development of insulin resis-
tance is an early event in the progression to Type 2 dia-
betes,4,5 suggesting that drugs which are capable of
improving insulin sensitivity should be useful in the
treatment of the disease. Indeed, a number of thiazoli-
dinedione derivatives typi®ed by ciglitazone,6 pioglita-
from clinical studies in human Type
patients.11
2 diabetic
Until recently, attempts to identify potent insulin sensi-
tisers not containing the thiazolidinedione pharmaco-
phore have been largely unsuccessful.10,12 Work from
our own laboratories, however, has recently shown that
this ring system may be e€ectively replaced by a variety
of a-functionalized carboxylic acid derivatives.13,14
Mechanistic studies suggest that, in common with the
thiazolidinediones, these compounds act by the speci®c
activation of a peroxisome proliferator-activated recep-
tor (PPARg), a member of the steroid nuclear hormone
receptor superfamily.15 Of particular interest were a-
alkoxy derivatives such as the a-methoxypropanoic acid
1, since many compounds of this series were up to ten-
fold more potent in the ob/ob mouse than the corre-
sponding thiazolidinedione, BRL 48482.10,13 In contrast
to the thiazolidinediones, which undergo a facile pH-
dependent racemization,16 it was anticipated that the
enantiomers of 1 would be less labile and thus permit an
Key words: Antihyperglycaemic; enzymes and enzyme reactions;
resolution; X-ray crystal structures.
*Corresponding author. Fax: +44-1279-627841; e-mail: david_haigh-1
@sbphrd.com
yDedicated to the memory of Professor Sir Derek H. R. Barton FRS.
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00034-6

822
D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
investigation into the enantiomeric dependence of bio-
logical activity. In this paper we discuss the synthesis of
the enantiomers of the a-methoxypropanoic acid 1 and
some related compounds and discuss their activity in
vitro and in in vivo models of diabetes.
hydrolysis was re-esteri®ed with methanolic hydrogen
chloride and rehydrolysed to 32% completion with R.
delemar lipase to furnish the (S)-acid 5 with 86% ee.
The absolute stereochemistry of 5 and 6 was initially
deduced by comparison with analogous reactions,19 but
was subsequently con®rmed by comparison with an
authentic sample of 5 (vide infra).
Chemistry
The corresponding a-benzyloxypropanoic acids 12 and
15 were prepared as illustrated in Scheme 2. In this
instance, the O-protected phenol 720 was hydro-
genolytically debenzylated to the phenol 8, which on
coupling with the mesylate 3 gave the racemic a-hy-
droxypropanoate 9. Treatment of 9 with vinyl acetate in
the presence of Pseudomonas ¯uorescens lipase21 selec-
tively acetylated the (S) isomer, leaving the unchanged
(R) enantiomer 10 in high purity (99% ee). Alkylation
of 10 with benzyl bromide then gave the ester 11, which
was hydrolysed to the requisite (R) acid 12 without loss
of chiral integrity. The absolute stereochemistry of this
acid was established by conversion of the ester 10 into
the crystalline oxazolidinedione 16 by reaction with (S)-
a-methylbenzyl isocyanate. A single crystal X-ray dif-
fraction study of this adduct (Fig. 1) clearly established
the stereochemistry of the unknown center as (R).
Rhodium acetate-mediated insertion of methanol into
diazotyrosine methyl ester following the procedure of
Moody et al.17 gave the a-methoxy precursor 2, which
on further reaction with the mesylate 310 a€orded the
racemic a-methoxypropanoate 4 in good yield (Scheme 1).
Since enzyme catalysed enantioselective ester formation
and hydrolysis is a well-established procedure for the
resolution of carboxylic acids,18 we decided to investi-
gate the hydrolysis of 4 using a variety of commercially
available lipases, having ®rst established an ecient
HPLC assay for the isomers of 4 based on their di€er-
ential elution on a chiral a1-acid glycoprotein column.
From this range of enzymes, Rhizopus delemar lipase
was found to e€ect preferential hydrolysis of the (S)
ester under conditions of partial hydrolysis (66%) when
the pH was maintained at 7.0 by the addition of sodium
hydroxide. By this procedure unchanged (R) ester was
recovered in 84% ee and was converted into the (R) acid
6 (84% ee) by exhaustive hydrolysis using R. delemar
lipase. The (S) acid 5 (40% ee) isolated from the initial
Exhaustive hydrolysis of the (S)-acetate 13 (produced
on treatment of hydroxyester 9 with vinyl acetate and P.
¯uorescens lipase) followed by re-esteri®cation a€orded
Scheme 1. Reagents: (i) NaH, DMF, 80 ꢀC, 16 h; (ii) R delemar lipase, 0.1 M NaOH, aq acetone.

D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
823
Scheme 2. Reagents: (i) NaH, DMF, 80 ꢀC, 16 h; (ii) P. ¯uorescens lipase, vinyl acetate; (iii) PhCH2Br, NaH, Ag2O, DMF, RT, 8 h; (iv) HCl, dioxan,
H2O, 80 ꢀC, 4.5 h; (v) MeOH, 4-MePhSO3H, 70 ꢀC, 4 h.
the (S) isomer 14 which was converted into the (S)
enantiomer 15 (96% ee) by benzylation and hydrolysis.
Enantiomeric purity was again assessed by HPLC using
a chiral a1-acid glycoprotein column. Furthermore,
methylation and subsequent hydrolysis of 14 provided
an authentic sample of the (S) a-methoxy acid, which
was identical with compound 5, thus con®rming the
stereochemistry of this series.
Results and Discussion
It is well-documented that anti-diabetic thiazolidine-
diones readily racemize in solution16 and it is this labi-
lity that has largely confounded attempts to establish
whether they show enantiomeric dependence at their site
of action. The identi®cation of a series of a-alkox-
ypropanoic acids related to one of the more potent
compounds, BRL 48482, but lacking the thiazolidine-
dione moiety,13 suggested that this intriguing question
could now be answered. Rather surprisingly, however,
resolution of 1, the a-methoxy member of this series,
failed to show any distinction between the two enantio-
mers, 5 and 6 in vivo in the ob/ob mouse (Table 1),
despite their apparent aqueous stability.22 Furthermore,
essentially identical activity was observed for both
compounds in their ability to stimulate 2-deoxyglucose
uptake in di€erentiated 3T3-L1 adipocytes (Table 1), an
in vitro assay of glucose transport23 which correlates
with in vivo potency.24 In the belief that steric integrity
was indeed maintained under the assay conditions we
prepared the a,a-dimethoxy compound 21 in an attempt
to determine whether there was steric tolerance for both
(R) and (S) methoxy groups at this position, but this
compound was inactive in vivo (Table 1).
Finally, the a,a-dimethoxypropanoic acid 21 was pre-
pared following the route shown in Scheme 3. Reaction
of the benzyl chloride 17 with the silyl enolate of methyl
dimethoxyacetate gave a modest yield of the acetal 18,
which was hydrogenolytically debenzylated to the phe-
nol 19. Subsequent reaction of the anion of 19 with
mesylate 3 followed by mild hydrolysis then gave a good
yield of the requisite acid 21.
In order to establish whether the similarity between
compounds 5 and 6 was a general phenomenon we pre-
pared the (R) and (S) enantiomers, 12 and 15, respec-
tively, of the related a-benzyloxy compound, which was
previously shown to be a potent anti-hyperglycaemic
agent.13 By contrast with the a-methoxy antipodes, 5
and 6, compounds 12 and 15 were readily distinguish-
able in vivo in that the (S) isomer 15 was 100-fold more
potent as an antihyperglycaemic agent (Table 1).
In view of these results we examined the possibility that
the enantiomers of 1 may racemize under the conditions
Figure 1. X-ray structure of oxazolidinedione 16.

824
D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
Scheme 3. Reagents: (i) LDA, TMSCl, (MeO)2CHCO2Me, then ZnCl2, CH2Cl2; (ii) H2, Pd/C, MeOH; (iii) NaH, DMF, then 3; (iv) LiOH, MeOH,
THF, H2O.
Table 1. Evaluation of a-alkoxy-b-phenylpropanoic acidsa
Compound
ee (%)
In vivo potency
ob/ob mouse ED25
(mmol kg diet)
Glucose transport
3T3-L1 cells
EC50 (nM)
PPARg Binding E. coli expressed
GST-hPPARg LBD
IC50 (nM)
1
1 (R,S)
5 (S)
6 (R)
12 (R)
15 (S)
21
Ð
86.0
84.0
98.0
96.0
Ð
3
3
3
30
3.1‹0.1 (n=3)
6.3‹0.4 (n=3)
4.9‹0.2 (n=3)
ndb
15 (n=1)
3.9 (n=1)
82‹26 (n=3)
ndb
0.3
Inactive @ 100
3
ndb
ndb
ndb
1.8‹0.2 (n=5)
ndb
22 (n=1)
BRL48482
Ð
aSee Experimental for de®nitions.
bNot determined.
of the 3T3-L1 and ob/ob mouse experiments, while
retaining at least partial steric integrity during the
shorter binding studies. Each enantiomer was therefore
independently incubated with di€erentiated 3T3-L1
adipocytes at nominal initial concentrations of 1.08 mM
1
(400 ng mL (Fig. 2) and 5.40 mM (2000 ng mL 1) at
37 ꢀC over a 96-h period. At both concentrations, the
(S)-enantiomer 5 was una€ected, whereas the (R)-enan-
tiomer 6 showed almost complete (>90%), unidirec-
tional stereoinversion to 5. Neither compound showed
any change when incubated in the cell culture medium
in the absence of 3T3-L1 cells. The time for 50% con-
version of the (R)- into the (S)-enantiomer was esti-
mated at ca. 36 h at the lower concentration and ca. 42 h
at the higher concentration. The sigmoidal shape of the
enantiomer interconversion curve is consistent with the
probable involvement of a two-stage enzyme mediated
reaction. In a similar experiment, 5 and 6 were incu-
bated with cultured rat hepatocytes over a 3-h period at
initial concentrations of 20 mM and 50 mM. Again, uni-
directional stereoinversion of the (R)-enantiomer into
the (S)-antipode was observed only in the presence of
hepatocytes.
Similar unidirectional stereoinversion is well docu-
mented for non-steroidal anti-in¯ammatory drugs such
as (R)-( )-ibuprofen, which have been shown to invert
via the formation and epimerization of a coenzyme-A
thioester.25 Given the demonstrated inversion of the
(R)-enantiomer 6 in vitro, it is likely that a similar
enzyme mediated conversion occurs in vivo during the
evaluation of the enantiomers in the ob/ob mouse, thus
accounting for their identical anti-hyperglycaemic
activity. However, this epimerisation mechanism
Figure 2. Incubation of 5 and 6 with 3T3-L1 Adipocytes. Compounds
5 and 6 were incubated individually with 3T3-L1 adipocytes at a
concentration of 1.08 mM. Samples were withdrawn at intervals
and derivatised for GC/MS assay by the method described in the
Experimental.

D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
825
appears to be highly substrate dependent, since the (R)
and (S) enantiomers of the a-benzyloxy acid, 12 and 15
respectively, did not undergo stereoinversion when
incubated with rat hepatocytes for 24 h under condi-
tions similar to those described above. Very recently a
related a-ethoxy compound was also reported to be
con®gurationally stable under conditions of the 3T3-L1
assay,26 adding support to the structural dependence
observed in this work.
nitrobenzyl alcohol±sodium acetate (NOBA±Na) matrix.
Compounds characterised by high resolution mass
measurement were homogeneous by TLC. 1H NMR
spectra were recorded at 270 MHz in CDCl3 solution.
Chemical shifts are given in d (ppm) relative to TMS
and coupling constants J are given in Hz. [a]2d5 values
are given in deg cm2 g 1. Dry solvents refers to the use
of Aldrich Sure-Seal2 dried solvents. All organic solu-
tions obtained from aqueous extractions were dried
over MgSO4. Chromatography refers to ¯ash chroma-
tography on silica gel.
Towards the completion of our work a strong correla-
tion was demonstrated between the anti-hyperglycaemic
activity of the thiazolidinediones in rodents and their
binding anity and agonist potency on PPARg, and
there is now compelling evidence that these agents exert
their action predominantly through the activation of
this class of nuclear hormone receptors.15 What had still
to be established, however, was direct evidence that
PPARg receptors elicit enantiomeric discrimination,
although such discrimination has previously been
demonstrated at related PPARa receptors.27 We there-
fore evaluated the binding anity of 5 and 6 at recom-
binant human PPARg whereby, in clear contrast to the
3T3-L1 and ob/ob mouse experiments, a marked di€er-
entiation between the two enantiomers was observed.
Thus at this receptor the (S) enantiomer 5 had an IC50
of 3.9 nM (n=1), whereas the (R) antipode 6 had an
IC50 of 82‹26 nM (n=3). It is possible that the binding
anity of the latter enantiomer 6 (84% ee) is due to the
presence of the small amount of 5 in the sample, how-
ever this 20-fold greater binding anity of 5 over 6,
taken together with our recently published work on a
related a-(tri¯uoroethoxy) analogue,28 implicitly con-
®rms enantiomer discrimination at this receptor.
Determination of enantiomeric excess
Enantiomeric excess of chiral compounds was deter-
mined by HPLC using a chiral a1-acid glycoprotein
(chiral AGP) column. HPLC methods and conditions,
including eluents used, solvent ¯ow rate and detection
wavelength were:- (A): MeCN-0.01M aq Na3PO4, pH
7.0 (12:88 v/v), 0.8 mL/min, l=245 nm; (B): MeCN-
0.01M aq Na3PO4, pH 7.0 (5:95 v/v), 0.5 mL/min,
l=225 nm; (C): MeCN-0.01 M aq Na3PO4, pH 7.0
(10:90 v/v), 1 mL/min, l=245 nm and (D): MeCN-
0.01 M aq Na3PO4, pH 5.7 (12:88 v/v), 1 mL/min,
l=245 nm. The method appropriate to the analysis of
each chiral compound is indicated within the text.
Biological procedures
Determination of anti-hyperglycaemic activity. C57 Bl/6
ob/ob mice were obtained at 9±10 weeks of age from
Harlan Olac Ltd., Oxon, U.K. and were maintained at
26‹2 ꢀC on a 12 h light/12 h dark cycle. The animals
were provided with powdered RM3 diet (Special Diet
Services, Witham, Essex, U.K.) and water ad libitum.
Compounds were administered by dietary admixture for
8 days. Diets supplemented with compound were stored
at 4 ꢀC, and mice were provided with fresh food each
day. Six to eight mice were used in each treatment group
and none of the compounds tested caused signi®cant
changes in body weight gain during the treatment period.
After 8 days of compound administration, mice were
fasted for 5 h from 08.00 hrs after which time a blood
sample (10 mL) was taken for glucose analysis from the
cut tip of the tail. Glucose (3 g/kg of body weight) was
then administered by oral gavage. Additional blood
samples (10 mL) for glucose analysis were then taken at
45, 90 and 135 min after glucose challenge. Blood sam-
ples were added to 1 mL of haemolysis reagent (50 mg/L
digitonin, 100 mg/L maleimide), and the glucose con-
centration in the haemolysed sample was determined
spectrophotometrically with hexokinase/glucose-6-
phosphate dehydrogenase29 using a Ciba-Corning 550
express clinical chemistry analyser.
Conclusion
From detailed experiments in rodent adipocytes we have
unequivocally demonstrated that the unexpected
equivalence of the two enantiomers 5 and 6 of the
potent insulin sensitiser 1 on glucose transport in vitro
results from an enzyme mediated stereospeci®c inver-
sion of the (R)-(+) antipode 6. In view of this observa-
tion it is likely that the equivalence of these enantiomers
as anti-diabetic agents in vivo is attributable to a similar
enzyme mediated conversion. By contrast, the related a-
benzyloxy acids 12 and 15 did not show bioequivalence
in vivo, a result which is consistent with the ®nding that
they did not undergo stereoinversion in vitro. We have
further shown that the (S)-( ) isomer 5 (IC50=3.9 nM)
shows a 20-fold greater binding than the (R)-(+) isomer
at human PPARg, therefore providing con®rmatory
evidence for stereospeci®c binding at this receptor.
Anti-hyperglycaemic activity is de®ned as the percentage
of reduction in the area under the blood glucose versus
time curve (AUC) relative to control animals. The AUC
was calculated trigonometrically using the trapezoid
rule.30 The ED25 dose is that dose of compound that
produces an approximate 25±30% reduction in the
AUC and results in a glucose tolerance pro®le equiva-
lent to that of normoglycaemic (+/?) lean littermates.
Experimental
General experimental details
Mass spectroscopy was conducted using electron impact
(EI), chemical ionisation (CI), with ammonia as the
reagent gas, or fast atom bombardment (FAB) in a 3-

826
D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
3T3-L1 cell experiments
Screen of enzymes for ester hydrolysis activity. A mix-
ture of the test enzyme (2±5 mg; below), potassium
phosphate bu€er (0.1 M, pH 7.0; 0.475 mL) and a solu-
tion of racemic ester 4 in acetone (7 mg/mL; 0.025 mL)
was shaken at ambient temperature and monitored by
TLC for evidence of ester hydrolysis.
Glucose uptake. The 3T3-L1 ®broblast cell line was
obtained from the American Type Culture Collection.
Cells were grown to con¯uence, di€erentiated according
to established procedures31 and used 8±10 days post
di€erentiation. Compounds were dissolved in DMSO
(®nal concentration of DMSOꢁ0.2%) and incubated
with 3T3-L1 cells for 48 h prior to assay of 2-deoxy-1-
[14C]-glucose transport.31,32 The EC50 is de®ned as the
concentration of test compound giving a 50% stimula-
tion of glucose transport.
Enzymes screened. Candida cylindracea lipase, pancrea-
tic lipase, Mucar javanicus lipase, Rhizopus javanicus
lipase, Rhizopus arrhizus lipase, Pen. roqueforte lipase,
Pen. cyclopium lipase, Pseudomonas ¯uorescens lipase.
Rhizopus niveus lipase, Rhizopus delemar lipase, Candida
lipolytica lipase, Rhizopus sp. LIP F3, Rhizopus sp. LIP
F4, acetylesterase (orange peel) and Aspergillus sp.
lipase.
Bioconversion experiments. On the day of assay, cells
were washed in serum free medium and then exposed to
either compound 5 or compound 6 at nominal con-
centrations of 400 and 2000 ng mL 1. Samples were
then incubated at approximately 37 ꢀC over a period of
96 h with aliquots being removed at speci®c times for
isomer analysis. The separate isomers were also incu-
bated at the same nominal concentrations in cell culture
medium in the absence of 3T3-L1 cells over a period of
120 h at approximately 37 ꢀC to assess their potential to
interconvert under cell-free conditions. Control samples
(non-incubated) were also prepared in the cell culture
medium and analysed immediately prior to the incuba-
tion samples to con®rm that no interconversion occur-
red during the period of analysis. These were prepared
in triplicate and randomly placed throughout the run
sequence.
Methyl 3-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]ethoxy]-
phenyl]-2-methoxypropanoate 4. Sodium hydride (60%
dispersion in mineral oil; 0.92 g, 23 mmol) was added to
a solution of phenol 217 (4.34 g, 21 mmol) in dry DMF
(50 mL) at room temperature under argon. The mixture
was stirred for 30 min at room temperature prior to the
addition of a solution of mesylate 310 (5.67 g, 21 mmol)
in DMF (50 mL) and then heated at 80 ꢀC for 16 h
before being cooled and diluted with water (1.4 L). The
mixture was acidi®ed to pH 2 with hydrochloric acid
then extracted with ethyl acetate (3Â400 mL). The
combined ethyl acetate solutions were washed with
water (4Â1 L) and brine (1 L), dried and evaporated.
The resulting gum was chromatographed using metha-
nol-dichloromethane (1.5/98.5 v/v) as eluent to a€ord
the racemic methoxy-ester 4, (5.77 g, 72%), a colourless
gum. [Found (M+H)+ (CI) 385.1759. C21H24N2O5
Assay procedure. Samples of the incubation media
(500 mL) containing the test compound were acidi®ed
with hydrochloric acid (1.0 M, 100 mL) and extracted
with toluene (2 mL). The mixture was centifuged
(ꢂ3000 g, 10 min) and the toluene layer was separated
and evaporated in a gentle stream of nitrogen at 60 ꢀC.
The resulting material was derivatised by treatment
with oxalyl chloride (50 mL, 15 min) followed, after
re-evaporation, by (R)-1-cyclohexylethylamine (1% v/v
in toluene, 100 mL). After brief whirlimixing the
resulting amides were analysed by capillary GC/MS
1
requires [M+H]+ 385.1763]; nmax (Film)/cm
1740
(CO); dH (270 MHz, CDCl3) 2.95 (2H, m, ArCH2), 3.33
(3H, s, NMe), 3.34 (3H, s, OMe), 3.70 (3H, s, OMe),
3.93 (3H, m, NCH2 and ArCH2CH), 4.23 (2H, t, J 5.2,
OCH2), and 6.70±7.40 (8H, m, aryl-H); m/z (CI) 385
[(M+H)+, 100%].
Rhizopus Delemar lipase resolution of 4: (S)-( )-3-[4-
[2-[N-(2-Benzoxazolyl)-N-methylamino]ethoxy]phenyl]-2-
methoxypropanoic acid 5 and (R)-(+)-3-[4-[2-[N-(2-
Benzoxazolyl)-N-methylamino]ethoxy]phenyl]-2-methoxy-
propanoic acid 6. Rhizopus delemar lipase (0.680 g,
ex Biocatalysts Ltd.) was stirred in deionised water
(380 mL) at 23 ꢀC and the mixture adjusted to pH 7.0. A
solution of racemic methoxy-ester 4 (1.0 g, 2.6 mmol) in
acetone (20 mL) was added and the mixture stirred and
maintained at pH 7.0 by autotitration with 0.1 M
sodium hydroxide solution until a 66% molar equiva-
lent of base had been added. The solution was acidi®ed
to pH 2.0 with dilute hydrochloric acid (2 M) and
extracted with dichloromethane (400 mL). The dichlor-
omethane solution was washed with 50% saturated
sodium bicarbonate solution (200 mL) and water, dried
and evaporated to a€ord the enriched (R)-(+)-methoxy-
ester [0.350 g, 34%; a colourless oil; 90% ee by HPLC
(Method A, retention time 15.7 min)]. Acidi®cation of
the sodium bicarbonate solution and re-extraction with
dichloromethane a€orded the enriched (S)-( )-meth-
oxy-acid (0.530 g, 55%; a colourless solid) which was
dissolved in methanol presaturated with hydrogen
using
a
conventional non-chiral silica column
(25 mÂ0.32 mm, HP-1 {methyl silicone}, 0.52 mm phase
thickness).
PPARꢂ binding assays. Binding assays were performed
in 96-well plates on crude extracts of XL-1 blue E.Coli.
cells expressing a fusion protein comprising GST and
the ligand binding domain of human PPARg (GST-
hPPARgLBD). 1.43 mg of total protein was present in
a ®nal volume of 50 mL. Radioligand [125I]-SB 236636
{3-[4-[2-[N-(2-benzoxazolyl)-N-methylamino]ethoxy]-3-
iodophenyl]-2-ethoxypropanoic acid,28 Kd 8 nM at
1
GST-hPPARgLBD, ED25 3 mmol.kg
diet in ob/ob
mice} was present at 145 pM. Competing compounds
were dissolved in DMSO (concentration of
DMSOꢁ0.1%) and binding was allowed to reach equi-
librium by incubation for 18 h at 4 ꢀC. Bound ligand
was separated from free on mixed cellulose ester ®lters.
Each assay was performed in triplicate.

D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
827
chloride gas, stirred at room temperature for 3 h and
then concentrated. The residue was partitioned between
water and ethyl acetate and the organic solution was
dried and evaporated to a€ord the enriched (S)-( )-
methoxy-ester (0.362 g, 36% based on initial racemic
ester used; 40% ee by HPLC (Method A, retention time
12.5 min).
(270 MHz, CDCl3) 2.90 (1H, dd, J=14.0 and 6.3 Hz,
ArCHH), 3.05 (1H, dd, J=14.0 and 4.4 Hz, ArCHH),
3.77 (3H, s, OMe), 4.42 (1H, dd, J=6.3 and 4.4 Hz,
ArCH2CH) 5.50 (2H, br, exchanges with D2O, 2ÂOH),
6.70 (2H, d, J=8.8 Hz, aryl 3-H & 5-H) and 7.05 (2H, d,
J=8.8 Hz, aryl 2-H & 6-H); m/z (CI) 214 [(M+NH4)+,
100%], 197 [(M+H)+, 6], 179 (5) and 107 (34).
The reaction was repeated and the respective enriched
(R)-(+)-ester (0.334 g, 33%; 80% ee) and enriched (S)-
( )-ester (0.590 g, 59%; ee not determined) combined
with those isolated above.
Methyl 3-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]-
ethoxy]phenyl]-2-hydroxypropanoate 9. Phenol 8 (2.38 g,
12 mmol) was reacted with mesylate 3 (3.24 g, 12 mmol)
in DMF in a manner analogous to that described for the
preparation of a-methoxy-ester 4 (above). The resulting
gum was chromatographed using ethyl acetate/hexane
(50/50 v/v) as eluent to a€ord the racemic hydroxy-ester
9, (1.52 g, 37%), a colourless solid, mp 130±132 ꢀC.
[Found C, 65.0; H, 6.1; N, 7.5%; M+ (EI) 370.1526.
C20H22N2O5 requires C, 64.85; H, 6.0; N, 7.6%; M
(S)-( )-3-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]-
ethoxy]phenyl]-2-methoxypropanoic acid 5. A solution
of the combined enriched (S)-( )-esters (0.950 g,
2.4 mmol) in acetone (20 mL) was treated with R. dele-
mar lipase (0.500 g) in deionised water (380 mL) at pH
7.0 as before and autotitrated with 0.1 M sodium hy-
droxide solution until a 32% molar equivalent of base
had been added. The mixture was acidi®ed to pH 2.0
with hydrochloric acid and extracted into dichlor-
omethane. The dichloromethane solution was extracted
with 50% saturated sodium bicarbonate solution and
the dichloromethane solution discarded. The sodium
bicarbonate solution was re-acidi®ed and re-extracted
with dichloromethane to a€ord the crude product which
was triturated with hexane to a€ord the (S)-( )-meth-
oxy-acid 5 (0.256 g, 13% based on total racemic ester
used), a colourless solid, mp 116±119 ꢀC. [Found C,
64.5; H, 5.9; N, 7.5%; M+ (EI) 370.1530. C20H22N2O5
requires C, 64.85; H, 6.0; N, 7.6%; M 370.1529]; [a]d25
1
370.1529]; nmax (KBr)/cm 3400 (OH) and 1730 (CO);
d
H (270 MHz, CDCl3) 2.88 (1H, dd, J=14.0 and 6.6 Hz,
ArCHH), 3.02 (2H, m, ArCHH and OH), 3.31 (3H, s,
NMe), 3.73 (3H, s, OMe), 3.90 (2H, t, J=5.2 Hz,
NCH2), 4.20 (2H, t, J=5.2 Hz, OCH2), 4.42 (1H, m,
ArCH2CH) and 6.70±7.40 (8H, m, aryl-H); m/z (FAB)
393 [(M+Na)+, 75%] and 371 [(M+H)+, 100].
P. Fluorescens lipase resolution of methyl 3-[4-[2-[N-(2-
benzoxazolyl)-N-methylamino]ethoxy]phenyl]-2-hydroxy-
propanoate 9. Pseudomonas Fluorescens lipase (7.1 g)
was added to a solution of the racemic hydroxy-ester 9
(5.08 g, 13.7 mmol) in vinyl acetate (350 mL) and the
mixture stirred at room temperature for 6 days, then
®ltered and concentrated. The residue was chromato-
graphed using ethyl acetate/hexane (gradient: from
50/50 v/v to 60/40 v/v) as eluent to a€ord ®rstly the (S)-
acetate 13 (2.43 g, 43%), followed by the (R)-(+)-hy-
droxy-ester 10 (1.97 g, 39%).
10ꢀ (c 0.55 in MeOH); 86% ee by HPLC (Method B,
1
retention time 9.4 min); nmax (KBr)/cm
2900±2500
(COOH) and 1710 (CO); dH (270 MHz, CDCl3) 2.95
(1H, dd, J=14.0 and 6.9 Hz, ArCHH), 3.07 (1H, dd,
J=14.0 and 5.0 Hz, ArCHH), 3.31 (3H, s, NMe), 3.40
(3H, s, OMe), 3.90 (2H, m, NCH2), 3.96 (1H, dd, J=6.9
and 5.0 Hz, ArCH2CH), 4.15 (2H, m, J=5.2 Hz,
OCH2), 6.30 (1H, br, exchanges with D2O, COOH) and
6.77±7.40 (8H, m, aryl-H); m/z (EI) 370 (M+, 100%).
(S)-(+)-Methyl 3-[4-[2-[N-(2-benzoxazolyl)-N-methyl-
amino]ethoxy]phenyl]-2-acetoxy-propanoate 13. An oil.
[Found M+ (EI) 412.1635. C22H24N2O6 requires M
412.1635]; [a]2d5 +1ꢀ (c 1.0 in CHCl3); 98% ee by HPLC
(Method C, retention time 24.7 min; (R)-enantiomer
retention time 33.7 min); dH (270 MHz, CDCl3) 2.06
(3H, s, COMe), 3.05 (1H, dd, J=14.3 and 8.2 Hz,
ArCHH), 3.10 (1H, dd, J=14.3 and 4.7 Hz, ArCHH),
3.33 (3H, s, NMe), 3.69 (3H, s, OMe), 3.92 (2H, t,
J=5.2 Hz, NCH2), 4.23 (2H, t, J=5.2 Hz, OCH2), 5.15
(1H, dd, J=8.2 and 4.7 Hz, ArCH2CH) and 6.70±7.40
(8H, m, aryl-H); m/z (FAB) 435 [(M+Na)+, 36%], 413
[(M+H)+, 100], 369 (4), 329 (10), 307 (7) and 175 (47).
(R)-(+)-3-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]-
ethoxy]phenyl]-2-methoxypropanoic acid, 6. In a similar
manner, the combined enriched (R)-(+)-esters (0.684 g,
1.78 mmol) were exhaustively hydrolysed by R. delemar
lipase (0.300 g) by addition of 0.1 M sodium hydroxide
solution to maintain pH 7.0 to a€ord, after extractive
workup, the (R)-(+)-methoxy-acid 6 (0.490 g, 25%
based on total racemic ester used), a colourless solid,
mp 121±123 ꢀC. [Found C, 64.6; H, 5.9; N, 7.5%.
C20H22N2O5 requires C, 64.85; H, 6.0; N, 7.6%]; [a]d25
+13ꢀ (c 0.5 in MeOH); 84% ee by HPLC (Method B,
retention time 8.2 min); spectroscopically identical with
5.
(R)-(+)-Methyl 3-[4-[2-[N-(2-benzoxazolyl)-N-methyl-
amino]ethoxy]phenyl]-2-hydroxypropanoate 10. A col-
ourless solid, mp 130±131 ꢀC. [Found C, 64.9; H, 6.05;
N, 7.6%. C20H22N2O5 requires C, 64.85; H, 6.0; N,
7.6%]; [a]2d5 +10ꢀ (c 0.85 in CHCl3); 99% ee by HPLC
(Method C, retention time 12.4 min); spectroscopically
identical with the racemic hydroxy-ester 9.
Methyl 2-hydroxy-3-(4-hydroxyphenyl)propanoate 8. A
solution of benzyl ether 720 (10.06 g, 35 mmol) in
methanol (100 mL) was hydrogenated over 10% Palla-
dium-charcoal (0.5 g) at room temperature and pres-
sure. The mixture was ®ltered and concentrated to
(R)-(+)-Methyl 3-[4-[2-[N-(2-benzoxazolyl)-N-methyl-
amino]ethoxy]phenyl]-2-benzyloxypropanoate 11. To a
solution of the (R)-(+)-hydroxy-ester 10 (99% ee,
a€ord the phenol 8 (6.60 g, 97%) as a colourless oil.
1
nmax (Film)/cm
3400 (OH) and 1725 (CO); dH

828
D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
0.72 g, 1.96 mmol) and benzyl bromide (3.35 g,
19.6 mmol) in DMF (14 mL) was added sodium hydride
(60% dispersion in mineral oil, 0.072 g, 1.8 mmol) fol-
lowed by silver(I) oxide (0.458 g, 1.96 mmol). The mix-
ture was stirred at room temperature for 8 h before
being poured into a mixture of brine (60 mL), ethyl
acetate (80 mL) and dilute hydrochloric acid (2 M,
5 mL). The organic solution was separated and washed
with brine (4Â60 mL), dried and concentrated. Chro-
matography of the residue using ether/hexane (67/33 v/
v) as eluent a€orded the (R)-(+)-benzyloxy-ester 11
(0.63 g, 76% based on sodium hydride used) as a col-
ourless gum. [Found M+ (EI) 460.2000. C27H28N2O5
acetate and saturated aqueous sodium bicarbonate
solution, the organic solution dried and concentrated
and the crude product puri®ed by chromatography
using ethyl acetate/hexane (60/40 v/v) as eluent to give
(S)-( )-hydroxy-ester 14 (1.54 g, 79%), a colourless
solid, mp 129±130 ꢀC. [Found C, 64.7; H, 6.1; N, 7.6%;
M+ (EI) 370.1528. C20H22N2O5 requires C, 64.85; H,
6.0; N, 7.6%; M 370.1529]; [a]2d5 9ꢀ (c 0.8 in CHCl3);
97% ee by HPLC (Method C, retention time 25.8 min);
spectroscopically identical with racemic alcohol 9.
(S)-( )-3-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]-
ethoxy]phenyl]-2-benzyloxypropanoic acid 15. The (S)-
( )-benzyloxy-acid 15 was prepared from the (S)-( )-
hydroxy-ester 14 (97% ee) in an analogous manner to
that described above for the (R)-(+)-enantiomer 12,
and was obtained as a colourless foam. [Found M+ (EI)
446.1844. C26H26N2O5 requires M 446.1842]; [a]2d5 23
(c 0.8 in CHCl3); 96% ee by HPLC (Method D, reten-
tion time 10.5 min); spectroscopically identical with the
(R)-(+)-enantiomer 12.
requires M 460.1999]; [a]2d5 +35ꢀ (c 0.85 in CHCl3); ee
1
not determined; nmax (Film)/cm
1740 (CO); dH
(270 MHz, CDCl3) 2.98 (2H, m, ArCH2CH), 3.35 (3H,
s, NMe), 3.69 (3H, s, OMe), 3.94 (2H, t, J=5.2 Hz,
NCH2), 4.10 (1H, dd, J=7.7 and 5.5 Hz, ArCH2CH),
4.24 (2H, t, J=5.2 Hz, OCH2), 4.35 (1H, d, J=11.9 Hz,
OCHHPh), 4.64 (1H, d, J=11.9 Hz, OCHHPh), 6.79
(2H, d, J=8.8 Hz, aryl 3-H and 5-H) and 6.95±7.40
(11H, m, aryl-H)]; m/z (FAB) 483 [(M+Na)+, 18%]
and 461 [(M+H)+, 100].
Preparation of (S)-( )-3-[4-[2-[N-(2-benzoxazolyl)-N-
methylamino]ethoxy]phenyl]-2-methoxypropanoic acid 5
by alkylation and hydrolysis of (S)-( )-hydroxy-ester 14.
Methylation of the (S)-( )-hydroxy-ester 14 (98% ee)
with methyl iodide in an analogous manner to that
described above for the (R)-(+)-benzyloxy-ester 11
a€orded the (S)-( )-methoxy-ester [A colourless gum;
[a]2d5 10ꢀ (c 1.23 in CHCl3); 93% ee by HPLC (see
above); spectroscopically identical with the racemic
methoxy-ester 4]. Subsequent hydrolysis in the manner
described for the (R)-(+)-benzyloxy-acid 12 a€orded
the (S)-( )-methoxy-acid 5 as a colourless solid, mp
118±120 ꢀC (dichloromethane/hexane). [Found C, 64.5;
H, 5.9; N, 7.6%; M+ (EI) 370.1528. C20H22N2O5
requires C, 64.85; H, 6.0; N, 7.6%; M 370.1529]; [a]d25
12ꢀ (c 1.08 in MeOH); 92% ee by HPLC (see above);
spectroscopically identical with that prepared via the R.
delemar lipase procedure.
(R)-(+)-3-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]-
ethoxy]phenyl]-2-benzyloxypropanoic acid 12. A solution
of the (R)-(+)-benzyloxy-ester 11 (0.61 g, 1.33 mmol) in
dioxan (15 mL) and dilute hydrochloric acid (2 M,
15 mL) was heated at 80 ꢀC for 4.5 h, then cooled and
concentrated. The residue was partitioned between ethyl
acetate and brine and the organic solution dried and
evaporated. The crude acid thus obtained was dissolved
in ethyl acetate (15 mL) and (S)-( )-a-methylbenzyl-
amine (0.169 g, 1.39 mmol) was added. The mixture was
stirred for 2 h and the resulting white solid was ®ltered
o€, washed with ether, then partitioned between ethyl
acetate (50 mL), brine (50 mL) and dilute hydrochloric
acid (2 M, 3 mL). The ethyl acetate solution was dried
and concentrated to yield the (R)-(+)-benzyloxy-acid
12, a foam (0.54 g, 91%). Crystallisation from ether-
hexane gave a white solid, mp 76±78 ꢀC. [Found C, 69.1;
H, 5.8; N, 6.2%; M+ (EI) 446.1844. C26H26N2O5 0.25
.
[3(1S),5R]-5-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]-
ethoxy]benzyl]-3-(1-phenylethyl)oxazolidin-2,4-dione 16.
Sodium hydride (60% dispersion in mineral oil,
0.011 g, 0.28 mmol) was added to a solution of (S)-( )-
a-methylbenzylisocyanate (0.046 mL, 0.31 mmol) and
the (R)-(+)-hydroxy-ester 10 (99% ee, 0.105 g,
0.28 mmol) in THF (3 mL). The mixture was stirred for
5 min then acidi®ed with dilute hydrochloric acid (2 M),
extracted with ethyl acetate and the organic solution
dried and concentrated. The residue was puri®ed by
preparative silica gel thin layer chromatography using
diethyl ether as eluent to a€ord the product, a colourless
gum (0.115 g, 83%), as a 4/1 mixture of diastereoi-
somers. Crystallisation from hexane-ethyl acetate a€or-
ded the major [3(1S),5R]-diastereoisomer 16 as a
colourless solid, mp 107±108 ꢀC. [Found C, 69.2; H,
5.65; N, 8.7%; M+ (EI) 485.1949. C28H27N3O5 requires
C, 69.3; H, 5.6; N, 8.7%; M 485.1951]; dH (270 MHz,
CDCl3) 1.58 (3H, d, J=7.1 Hz, Me), 3.11 (1H, dd,
J=14.6 and 3.85 Hz, ArCHH), 3.23 (1H, dd, J=14.6
and 4.4 Hz, ArCHH), 3.37 (3H, s, NMe), 3.96 (2H, t,
J=5.0 Hz, NCH2), 4.20 (2H, t, J=5.0 Hz, OCH2), 4.92
H2O requires C, 69.2; H, 5.9; N, 6.2%; M 446.1842];
[a]2d5 +23ꢀ (c 0.7 in CHCl3); 98% ee by HPLC (Method
D, retention time 13.3 min); nmax (Nujol)/cm 2800±
2400 (COOH) and 1722 (CO); dH (270 MHz, CDCl3)
3.05 (2H, m, ArCH2CH), 3.31 (3H, s, NMe), 3.90 (2H,
t, J=5.4 Hz, NCH2), 4.15 (3H, m, ArCH2CH and
OCH2CH2N), 4.45 (1H, d, J=11.5 Hz, PhCHHO), 4.70
(1H, d, J=11.5 Hz, PhCHHO), 6.70±7.40 (13H, m, aryl-
H) and 9.80 (1H, br, exchanges with D2O, COOH); m/z
(EI) 446 (8%), 311 (36), 175 (83), 161 (38), 148 (100).
1
(S)-( )-Methyl 3-[4-[2-[N-(2-benzoxazolyl)-N-methyl-
amino]ethoxy]phenyl]-2-hydroxypropanoate 14. A solu-
tion of the (S)-acetate 13 (98% ee, 2.16 g, 5.24 mmol) in
dilute hydrochloric acid (2 M, 30 mL) and dioxan
(30 mL) was heated at 80 ꢀC for 4.5 h, cooled and con-
centrated and the residue suspended in benzene (50 mL)
and re-evaporated. The resulting solid was dissolved in
methanol (150 mL) containing p-toluenesulfonic acid
(ca. 0.1 g) and heated at re¯ux for 4 h, then cooled and
concentrated. The residue was partitioned between ethyl

D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
829
(1H, dd, J=4.4 and 3.85 Hz, oxazolidine 5-H), 5.07
(1H, q, J=7.1 Hz, PhCH) and 6.70±7.40 (13H, m, aryl-
H); m/z (CI) 486 [(M+H)+, 100%].
(8H, m, aryl-H); m/z (FAB) 437 [(M+Na)+, 100%] and
415 [(M+H)+, 62].
3-[4-[2-[N-(2-Benzoxazolyl)-N-methylamino]ethoxy]phenyl]-
2,2-dimethoxypropanoic acid 21. A mixture of the dime-
thoxy-ester 20 (0.30 g, 0.73 mmol), lithium hydroxide
monohydrate (0.033 g, 8.0 mmol), methanol (2 mL),
THF (2 mL) and water (2 mL) was stirred at room tem-
perature for 18 h. A further portion of lithium hydrox-
ide (8.0 mmol) was added and stirring continued for an
additional 18 h. The mixture was concentrated and the
residue diluted with water (15 mL), acidi®ed to pH 4.5
with dilute hydrochloric acid (0.001 M) and extracted
with ethyl acetate (2Â30 mL). The ethyl acetate solution
was dried and concentrated and the resulting gum tri-
turated with isohexane-ether to a€ord the dimethoxy-
acid 21 (0.262 g, 90%), a white solid, mp 134±135 ꢀC.
[Found C, 62.9; H, 6.2; N, 6.95%; M+ (EI) 401.1711.
C21H24N2O6 requires C, 63.0; H, 6.0; N, 7.0%; M
Methyl 3-(4-benzyloxyphenyl)-2,2-dimethoxypropanoate
18. A solution of methyl dimethoxyacetate (3.0 g,
22.4 mmol) in dry THF (5 mL) was added slowly to a
70 ꢀC solution of lithium diisopropylamide [prepared
from n-butyllithium (2.5
M in hexane, 8.95 mL,
22.4 mmol) and diisopropylamine (3.13 mL, 22.4 mmol)
in THF (10 mL)] under argon. The mixture was stirred
at 70 ꢀC for 1 h prior to the addition of chloro-
trimethylsilane (2.84 mL, 22.4 mmol) and stirring con-
tinued for an additional 50 min at 70 ꢀC, warmed to
room temperature and stirring continued for a further
30 min. The mixture was concentrated and the residue
suspended in dry diethyl ether (50 mL), ®ltered and the
®ltrate concentrated to a€ord the crude silyl enol ether
as a yellow oil (3.541 g). The oil was dissolved in dry
dichloromethane (15 mL) and 4-benzyloxybenzyl chlo-
ride (3.77 g, 22.4 mmol) and zinc(II) chloride (0.22 g,
2.24 mmol) added. The mixture was stirred at room
temperature overnight then washed with water
(2Â20 mL), dried and evaporated. The resulting gum
was chromatographed twice, ®rst using ethyl acetate/
hexane (gradient: from 5/95 v/v to 30/70 v/v) followed
by acetone-hexane (gradient: from 5/95 v/v to 10/90 v/v)
as eluent to a€ord the propanoate 18 (1.145 g, 16%), a
colourless solid, mp 49±50 ꢀC. [Found C, 67.9; H, 7.2%;
M+ (EI) 330.1467. C19H22O5 requires C, 67.95; H,
1
401.1713]; nmax (Nujol)/cm 3400±2500 (COOH) and
1740 (CO); dH (270 MHz, CDCl3) 3.16 (2H, s,
ArCH2C), 3.27 (3H, s, NMe), 3.42 (6H, s, 2ÂOMe),
3.80 (2H, t, J=5.2 Hz, NCH2), 3.93 (2H, t, J=5.2 Hz,
OCH2) and 6.72±7.75 (9H, m, reduces to 8H on
exchange with D2O, aryl-H and COOH); m/z (CI) 401
[(M+H)+, 100%].
X-ray crystallography of oxazolidinedione 16. Oxazolidi-
nedione 16 was crystallised from ethanol containing a
trace of acetonitrile. The molecule crystallised with two
independent molecules in the crystallographically inde-
pendent unit. The two molecules had identical stereo-
chemistry and di€ered conformationally only slightly.
The largest di€erence between molecules in any torsion
angle was 17ꢀ. One of the molecules is shown in Figure 1.
Lattice parameters were determined from the setting
angles of 25 re¯ections well distributed in reciprocal
space measured on an Enraf±Nonius CAD-4 dif-
fractometer. Intensity data were collected on the dif-
fractometer using graphite monochromated copper
radiation and an o-2y variable speed scan technique.
The structure was solved by direct methods using the
SHELXS program series and re®ned using the
SHELXL-93 re®nement program. Hydrogen atoms
were included in idealized positions which rode on the
atom to which they are attached.
1
7.2%; M 330.1467]; nmax (Nujol)/cm 1750 (CO); dH
(270 MHz, CDCl3) 3.12 (2H, s, ArCH2C), 3.37 (6H, s,
2ÂOMe), 3.61 (3H, s, COOMe), 5.02 (2H, s, ArCH2O)
and 6.87±7.44 (9H, m, aryl-H); m/z (FAB) 353
[(M+Na)+, 100%].
Methyl 3-(4-hydroxyphenyl)-2,2-dimethoxypropanoate 19.
A solution of benzyl ether 18 (1.016 g, 3.1 mmol) in
methanol (20 mL) was hydrogenated over 10% palla-
dium-charcoal (0.1 g) at room temperature and ambient
pressure. The mixture was ®ltered and concentrated to
a€ord the phenol 19 (0.583 g, 79%), a pale-yellow solid,
mp 115±116 ꢀC. [Found M+ (EI) 240.0998. C12H16O5
1
requires M 240.0998]; nmax (Nujol)/cm
3900 (OH)
and 1750 (CO); dH (270 MHz, CDCl3) 3.09 (2H, s,
ArCH2C), 3.35 (6H, s, 2ÂOMe), 3.63 (3H, s, COOMe),
5.50 (1H, br, exchanges with D2O, OH), 6.64 (2H, d,
J=7.98 Hz, aryl 3-H and 5-H) and 6.95 (2H, d,
J=7.98 Hz, aryl 2-H and 6-H); m/z (CI) 258
[(M+NH4)+, 52%], 226 (29) and 194 (100).
Crystal data
C28H27N3O5; M=485.53; T=223(2)K; l=1.54178 A;
crystal size 0.54Â0.50Â0.12 mm; monoclinic; space
group P21; unit cell dimensions a=7.652(4)A, b=
6.355(2)A, c=49.45(5)A, b=91.22(2)ꢀ; V=2404(3)A3;
Z=4; Dcalc=1.341 Mg/m3; m=0.761 mm 1; F(000)=1024;
y range for data collection 1.79 to 62.56ꢀ; index ranges
8ꢁhꢁ8, 0ꢁkꢁ7, 56ꢁlꢁ56; re¯ections collected 8451;
independent re¯ections 4231 (Rint=0.0421); re®nement
method full-matrix least-squares on F2; data/restraints/
parameters 4230/1/653; goodness-of-®t on F2 1.079;
®nal R indices: 4062 data; I>2s(I) R1=0.070, wR2=
0.194, all data R1=0.072, wR2=0.197; absolute struc-
Methyl 3-[4-[2-[N-(2-benzoxazolyl)-N-methylamino]-
ethoxy]phenyl]-2,2-dimethoxypropanoate 20. Phenol 19
(0.522 g, 2.2 mmol) was reacted with mesylate 3 (0.590 g,
2.2 mmol) in a manner analogous to that described pre-
viously for methoxy-ester 4. Dimethoxy-ester 20
(0.662 g, 73%) was obtained as a colourless oil which
crystallised on standing, mp 77±78 ꢀC. [Found M+ (EI)
414.1786. C22H26N2O6 requires M 414.1791]; nmax
1
(Nujol)/cm 1760 (CO); dH (270 MHz, CDCl3) 3.10
(2H, s, ArCH2C), 3.34 (3H, s, NMe), 3.36 (6H, s,
2ÂOMe), 3.61 (3H, s, COOMe), 3.93 (2H, t, J=5.2 Hz,
NCH2), 4.22 (2H, t, J=5.2 Hz, OCH2) and 6.78±7.37
ture parameter 0.1(4); largest di€. peak and hole 0.69
3
and 0.37 eA
.

830
D. Haigh et al. / Bioorg. Med. Chem. 7 (1999) 821±830
Brun, R. P.; Spiegelman, B. M.; Evans, R. M. Cell 1995, 83,
Acknowledgements
803. (c) Willson, T. M.; Cobb, J. E.; Cowan, D. J.; Wiethe, R.
W.; Correa, I. D.; Prakash, S. R.; Beck, K. D.; Moore, L. B.;
Kliewer, S. A.; Lehmann, J. M. J. Med. Chem. 1996, 39, 665.
16. Sohda, T.; Mizuno, K.; Kawamatsu, Y. Chem. Pharm.
Bull. 1984, 32, 4460.
The authors thank P. Coyle, K. Jennings, H. Oldham,
A. Ramaswamy, A. Selvaratnam, A. Shah, and C.
Shardlow for their expert technical assistance and R.
Hindley for helpful discussions and comments.
17. Cox, G. G.; Haigh, D.; Hindley, R. M.; Miller, D. J.;
Moody, C. J. Tetrahedron Lett. 1994, 35, 3139.
18. (a) Boland, W.; Frossl, C.; Lorenz, M. Synthesis, 1991,
1049. (b) Santaniello, E.; Ferraboschi, P.; Grisenti, P.; Man-
zocchi, A. Chem. Rev. 1992, 92, 1071. (c) Sih, C. J.; Wu, S-H.
Topics in Stereochemistry, 1989, 19, 63.
19. (a) Delinck, D. L.; Margolin, A. L. Tetrahedron Lett.
1990, 31, 6797. (b) Sih, J. C.; Gu, R. L. Tetrahedron: Asym-
metry, 1995, 6, 357. (c) Lefker, B. A.; Hada, W. A.; McGarry,
P. J. Tetrahedron Lett. 1994, 35, 5205. (d) Kato, Y.; Ohta, H.;
Tsuchihashi, G. Tetrahedron Lett. 1987, 28, 1303.
20. Haigh, D.; Jefcott, L. J.; Magee, K.; McNab, H. J. Chem.
Soc. Perkin Trans 1 1996, 2895.
21. (a) Preparative BiotransformationsÐWhole Cell and Iso-
lated Enzymes in Organic Synthesis; Roberts, S. M., Ed.; John
Wiley: New York. 1992. Section 1:6.41±1:6.48. (b) Wong, C.
H.; Whitesides, G. M. Enzymes in Synthetic Organic Chem-
istry; Oxford: Pergamon, 1994; p 82.
References and Notes
1. Reaven, G. M. Diabetes 1988, 37, 1595.
2. Moller, D. E.; Flier, J. S. N. Engl. J. Med. 1991, 325, 938.
3. DeFronzo, R. A. Diabetes 1988, 37, 667.
4. Warram, J. H.; Martin, B. C.; Krolewski, A. S.; Soeldner,
J. S.; Kahn, C. R. Ann. Intern. Med. 1990, 113, 909.
5. Lillioja, S.; Mott, D. M.; Spraul, M., Ferraro, R.; Foley, J.
E.; Ravussin, E.; Knowler, W. C.; Bennett, P. H.; Bogardus,
C. N. Engl. J. Med. 1993, 329, 1988.
6. (a) Sohda, T.; Mizuno, K.; Imamiya, E.; Sugiyama, Y.;
Kawamatsu, Y. Chem. Pharm. Bull. 1982, 30, 3580. (b) Fujita,
T.; Sugiyama, Y.; Taketomi, S.; Sohda, T.; Kawamatsu, Y.;
Iwatsuka, H.; Suzuoki, Z. Diabetes 1983, 32, 804.
7. Momose, Y.; Meguro, K.; Ikeda, H.; Ot, S.; Sohda, T.
Chem. Pharm. Bull. 1991, 39, 1440.
22. The failure to demonstrate signi®cant base-induced deu-
terium exchange of the a-hydrogen atom under NMR condi-
tions has been used to con®rm the stability of this stereo center
in the racemate 1; Haigh, D. Unpublished results.
23. (a) Frost, S. C.; Lane, M. D. J. Biol. Chem. 1985, 260,
2646. (b) Gibbs, E. M.; Leinhard, G. E.; Gould, G. W. Bio-
chemistry 1988, 27, 6681.
8. Yoshioka, T.; Fujita, T.; Kanai, T.; Aizawa, Y.; Kur-
umada, T.; Hasegawa, K.; Horikoshi, H. J. Med. Chem. 1989,
32, 421.
9. Cantello, B. C. C.; Cawthorne, M. A.; Haigh, D.; Hindley,
R. M.; Smith, S. A.; Thurlby, P. L. Bioorg. Med. Chem. Lett.
1994, 4, 1181.
10. Cantello, B. C. C.; Cawthorne, M. A.; Cottam, G. P.;
Du€, P. T.; Haigh, D.; Hindley, R. M.; Lister, C. A.; Smith, S.
A.; Thurlby, P. L. J. Med. Chem. 1994, 37, 3977.
24. Holder, J. C.; Lister, C. A.; Smith, S. A.; Young, P. W.
Unpublished results.
25. (a) Knihinicki, R. D.; Williams, K. M.; Day, R. O. Bio-
chem. Pharmacol. 1989, 38, 4389. (b) Wechter, W. J. J. Clin.
Pharmacol. 1994, 34, 1036.
11. (a) Nolan, J. J.; Ludvik, B.; Beerdsen, P.; Joyce, M.;
Olefsky, J. N. Engl. J. Med. 1994, 331, 1188. (b) Kumar, S.;
Boulton, A. J. M.; Beck-Nielsen, H.; Berthezene, F.; Muggeo,
M.; Persson, B.; Spinas, G. A.; Donoghue, S.; Lettis, S.;
Stewart-Long, P. Diabetologia 1996, 39, 701.
12. The thiazolidinedione moiety has been replaced by a vari-
ety of other heterocyclic systems, although few of these mod-
i®cations have given compounds with high anti-hyperglycaemic
potency. See for example: Hulin, B.; McCarthy, P. A.; Gibbs,
E. M. Curr. Pharmaceut. Design 1996, 2, 85.
13. Buckle, D. R.; Cantello, B. C. C.; Cawthorne, M. A.;
Coyle, P. J.; Dean, D. K.; Faller, A.; Haigh, D.; Hindley, R. M.;
Jefcott, L. J.; Lister, C. A.; Pinto, I. L.; Rami, H. K.; Smith,
D. G.; Smith, S. A. Bioorg. Med. Chem. Lett. 1996, 6, 2121.
14. Buckle, D. R.; Cantello, B. C. C.; Cawthorne, M. A.;
Coyle, P. J.; Dean, D. K.; Faller, A.; Haigh, D.; Hindley, R.
M.; Jefcott, L. J.; Lister, C. A.; Pinto, I. L.; Rami, H. K.;
Smith, D. G.; Smith, S. A. Bioorg. Med. Chem. Lett. 1996, 6,
2127.
26. Hulin, B.; Newton, L. S.; Lewis, D. M.; Genereux, P. E.;
Gibbs, E. M.; Clark, D. A. J. Med. Chem. 1996, 39, 3897.
27. (a) Boie, Y.; Adam, M.; Rushmore, T. H.; Kennedy, B. P.
J. Biol. Chem. 1993, 268, 5530, (b) Yu, K.; Bayona, W.; Kal-
len, C. B.; Harding, H. P.; Ravera, C. P.; McMahon, G.;
Brown, M.; Lazar, M. A. J. Biol. Chem. 1995, 270, 23975.
28. Young, P. W.; Buckle, D. R.; Cantello, B. C. C.; Chap-
man, H.; Clapham, J. C.; Coyle, P. J.; Haigh, D.; Hindley, R.
M.; Holder, J. C.; Kallender, H.; Latter, A. L.; Laurie, K. W.
W.; Mossakowska, D.; Murphy, G. J.; Cox, L. R.; Smith, S.
A. J. Pharmacol. Exp. Ther. 1998, 284, 751.
29. Bergmeyer, H.-U; Bernt, E.; Schmidt, F. H.; Stork, H. In
Methods of Enzymatic Analysis; Academic: New York, 1974; p
1196.
30. Phillips, G. M.; Taylor, P. J. Theory and Applications of
Numerical Analysis; Academic: New York 1973; p 131.
31. Frost, S. C.; Lane, M. D. J. Biol. Chem. 1985, 260, 2646.
32. Gibbs, E. M.; Leinhard, G. E.; Gould, G. W. Biochemistry
1988, 27, 6681.
15. (a) Lehmann, J. M.; Moore, L. B.; Smith-Oliver, T. A.;
Wilkison, W. O.; Willson, T. M.; Kliewer, S. A., J. Biol. Chem.
1995, 270, 12953. (b) Forman, B. M.; Tontonoz, P.; Chen, J.;