Design of inhibitors of scytalone dehydratase: Probing interactions with an asparagine carboxamide

Article abstract of DOI:10.1016/S0968-0896(02)00272-9

Among the active-site residues of scytalone dehydratase, the side-chain carboxamide of asparagine 131 has the greatest potential for strong electrostatic interactions. Structure-based inhibitor design aimed at enhancing interactions with this residue led to the synthesis of a series of highly potent inhibitors that have a five- or six-membered ring containing a carbonyl functionality for hydrogen bonding. To achieve a good orientation for hydrogen bonding, the inhibitors incorporate a phenyl substituent that displaces a phenylalanine residue away from the five- or six-membered rings. Without the phenyl substituent, inhibitor binding potency is diminished by three orders of magnitude. Larger K_i values of a site-directed mutant (Asn131Ala) of scytalone dehydratase in comparison to those of wild-type enzyme validate the design concept. The most potent inhibitor (K_i=15 pM) contains a tetrahydrothiophenone that can form a single hydrogen bond with the asparagine carboxamide. Inhibitors with a butyrolactam that can form two hydrogen bonds with the asparagine carboxamide demonstrate excellent in vivo fungicidal activity.

Full text of DOI:10.1016/S0968-0896(02)00272-9

Bioorganic & Medicinal Chemistry 10 (2002) 4143–4154
Design of Inhibitors of Scytalone Dehydratase: Probing
Interactions with an Asparagine Carboxamide
Gregory S. Basarab,a,* Douglas B. Jordan,b Troy C. Gehreta and Rand S. Schwartzc
aDuPont Central Research & Development, Experimental Station, Wilmington, DE 19880, USA
bDuPont Pharmaceuticals Company, Experimental Station, Wilmington, DE 19880, USA
cDuPont Agricultural Products, Experimental Station, Wilmington, DE 19880, USA
Received 10 January 2002; accepted 17 May 2002
Abstract—Among the active-site residues of scytalone dehydratase, the side-chain carboxamide of asparagine 131 has the greatest
potential for strong electrostatic interactions. Structure-based inhibitor design aimed at enhancing interactions with this residue led
to the synthesis of a series of highly potent inhibitors that have a five- or six-membered ring containing a carbonyl functionality for
hydrogen bonding. To achieve a good orientation for hydrogen bonding, the inhibitors incorporate a phenyl substituent that dis-
places a phenylalanine residue away from the five- or six-membered rings. Without the phenyl substituent, inhibitor binding
potency is diminished by three orders of magnitude. Larger Ki values of a site-directed mutant (Asn131Ala) of scytalone dehy-
dratase in comparison to those of wild-type enzyme validate the design concept. The most potent inhibitor (Ki=15 pM) contains a
tetrahydrothiophenone that can form a single hydrogen bond with the asparagine carboxamide. Inhibitors with a butyrolactam that
can form two hydrogen bonds with the asparagine carboxamide demonstrate excellent in vivo fungicidal activity.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
Three-dimensional structures of SD in complex with
each of the inhibitors shown in Fig. 1 have been deter-
mined by X-ray crystallography in an iterative fungicide
discovery program.11ꢀ13,16,17 The inhibitors are well
buried in the enzyme, and the various binding pocket
environments have been mapped and visualized as
Connolly surfaces. The steric surroundings revealed by
the surfaces have led to modifications of the inhibitors
to improve binding potency by better occupying the
available space.11 The steric environment has also been
used to focus the selection of inputs for a directed com-
binatorial library of inhibitors, which identified a novel
class of low picomolar Ki inhibitors.13,14 A purposeful
displacement of a water molecule associated with the
binding pocket by appropriate substitution on another
class of inhibitors has been shown to lower Ki’s 2- to 20-
fold.5,10 Structures of SD complexed with different
inhibitors have revealed conformationally mobile resi-
dues within the SD binding pocket leading to the design
of new generation inhibitors with modified substituents
that displace these residues outward.12,17
Scytalone dehydratase (SD) catalyzes the dehydration
of two physiological substrates, scytalone and verme-
lone, within the fungal melanin biosynthetic pathway
employed by certain invasive plant pathogens.1,2 Cata-
lytic activity of SD is absolutely required for the pro-
duction of melanin and the initiation of disease by the
fungi as genetic and inhibitor knockouts of the enzyme
function block melanin production and mitigate fungal
penetration into the cells of the host.3 Hence, SD has
been targeted for inhibitor design and optimization
programs directed towards discovering crop fungicides.
There is considerable information detailing the reac-
tions catalyzed by the enzyme4ꢀ8 and the inhibitors that
curtail its activity.9ꢀ15 The several structural classes of
inhibitors (e.g., see Fig. 1) that have been reported in
the literature accentuate a measure of promiscuity in the
SD inhibitor binding pocket, which portends the
potential for further structural refinements in the course
of inhibitor design.
Improvement in binding potency can be realized by
adjusting the fit of the inhibitors in the enzyme through
the addition of hydrophobic substituents; as the SD
binding pocket is mostly hydrophobic, inhibitor binding
*Corresponding author at present address: AstraZeneca R&D Boston
Inc., 35 Gatehouse Drive, Waltham, MA 02451, USA. Tel.: +1-781-
839-4689; fax: +1-781-839-4630; e-mail: greg.basarab@astrazeneca.
com
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00272-9

4144
G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
Figure 1. Inhibitors that have been co-crystallized with SD for structure determination by X-ray diffraction.
potency is improved, in part, because the increased
hydrophobicity favors partitioning away from the aqu-
eous environment.18 However, there is a limitation to
this means of exploiting the information on the binding
pocket in that once the hydrophobicity becomes too
high, bioavailability and the expression of fungicidal
activity in whole-organism assays is compromised.
Hence, it is of practical value to identify more polar
inhibitor functionalities that can offer an electrostatic
interaction with complementary residues of the SD
binding pocket in order to balance the hydrophobic
nature of the inhibitors.
hydrogen atom from the Asn131 carboxamide as shown
in Figure 2A. An inhibitor carboxamide can form two
hydrogen bonds with the Asn131 if it is constrained in a
ring forcing a cis orientation between the carbonyl and
NH (Fig. 2B). Inhibitors maintaining these properties
are potent inhibitors of SD and constitute effective and
novel fungicides.
Results
Molecular modelling
Eight SD residues have been demonstrated by kinetic
analysis of site-directed mutants to mediate the dehy-
drations catalyzed by the enzyme.4,8 Seven of the resi-
dues border the inhibitor binding pocket as delineated
by the various X-ray structure determinations; the
binding constants of a number of inhibitors to the site-
directed mutants have been reported as a means of
approximating the contributions of the individual resi-
dues to the binding potency.5,15 Overall, Asn131 exhib-
ited the largest influence on inhibitor recognition as
determined by the 11- to 1100-fold decrease in binding
affinity upon its mutation to alanine.5,15 Inhibitors such
as 1 and 3 display functionalities (the salicylamide
hydroxyl and the cinnoline heterocycle, respectively)
towards the Asn131 side-chain carboxamide for forma-
tion of strong hydrogen bonds, and their binding affi-
nities to SD are profoundly diminished by the mutation
to Ala131. In contrast, the binding affinities of inhibi-
tors that display groups (such as the cyclopropyl chlor-
ine atoms of 2 and the nitrile of 4) of poor hydrogen
bonding capability are less affected by the mutation.
Neither 1 nor 3 expresses good in vivo disease control in
greenhouse assays. There is evidence that 1 is metabo-
lized by both the plant and the fungus for effective
detoxification.19 High lipophilicity of 3 may compro-
mise its bioavailability.
The crystal structures of 411 and 512 in complex with SD
were used as the starting points for assembling inhibitor
binding models. Replacing the t-butyl cyanoacetamide of
4 with a cyclopentanone carboxamide does not allow for
the alignment of the ketone functionality for a favorable
hydrogen bond with the Asn131 carboxamide without an
unfavorable steric clash of the cyclopentanone ring and
the side chain of Phe53. This was modeled with a simple
MM3 optimization allowing the inhibitor and a 10A
shell of the surrounding binding pocket to relax.
Compound 5 was designed from the knowledge that
there is considerable conformational mobility for
Phe53: the phenyl substituent of 5 induced an outward
movement of the Phe53 side chain through a p–p
stacking arrangement.12,17 It was reasoned that, with
Phe53 displaced away from the cyclopentanone, the
impediment towards the alignment of the cyclopenta-
none carbonyl and the Asn131 carboxamide would be
eliminated. Figure 2A depicts the relevant interactions
of the cyclopentanone analogue after its replacement for
5 in the crystal structure model, and optimization of a
10A shell. The orientation and distances between the
cyclopentanone and the Asn131 carboxamide are
favorable for alignment of the hydrogen bond: the het-
eroatom–hydrogen–heteroatom angle that makes up the
hydrogen bond is 157ꢁ while the heteroatom–hetero-
atom bond distance is 3.2 A.
We investigate herein the capability of a more polar
carbonyl functionality to share a hydrogen bond with
the Asn131 carboxamide in order to boost binding
potency. A properly oriented ketone can accept a
Optimization of a butyrolactam carboxamide replacing
compound 5 from its respective crystal structure

G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
4145
Figure 2. Schematic representation of cyclopentanone (A) and butyrolactam (B) binding interactions with Asn131. Phe53 is displaced away from the
cyclopentanone and butyrolactam rings by the inhibitor phenyl substituent.
allowed for formation of two hydrogen bonds. In the
model, the Asn131 side chain carboxamide and the lactam
carboxamide are nearly co-planar: the root mean square
deviation from co-planarity of the eight atoms of the
hydrogen bond array of Figure 2B is 0.1 A. Heteroatom–
heteroatom bond distances are acceptable for hydrogen
bonding as are the associated hydrogen bond angles.
10 via Mitsunobu reaction with dichlorophenol
(Scheme 2).
The syntheses of cyclopentanone, cyclohexanone and
tetrahydrothiophenone carboxamides proceeded by
protection of the carbonyl of the corresponding keto-
esters followed by hydrolysis to the acid and coupling to
amines via the acid chlorides (Scheme 3). Deprotection
of the carbonyl was achieved with mild acid hydrolysis.
The compounds 17–19 exist as mixtures of keto and
enol tautomers in deuterochloroform with the keto
forms existing as 1:1 diastereomeric mixtures.
Synthesis
Amines 8a and 8b were derived from norephedrine in a
reaction sequence affording a double inversion of stereo-
chemistry to retain the norephedrine configuration
(Scheme 1). The norephedrine amine was protected as
the phosphonamide. The alcohol was converted to the
mesylate, which allowed cyclization to aziridine 7 on
treatment with base. The aziridine was opened with the
appropriate phenoxide and the protecting phosphon-
amide was removed with acid treatment. The aziridine
ring opening reaction affords some of the alternative
regioisomeric products 9a and 9b. The protecting phos-
phonamide group in this sequence serves to activate both
aziridine ring formation and aziridine ring opening.
The butyrolactam carboxylic acid esters 20 and 21 were
converted to the desired amides by hydrolysis and EDC
coupling with the requisite amines (Scheme 4). The
valerolactam carboxylic acid 25 was similarly converted
to the amides 26a and 26b. Compounds 22a, 22b, 26a
and 26b were each isolated as a 1:1 diastereomeric mix-
ture that was not resolved.
The cyanoacetamide 5b, an analogue of 5, was prepared
by acylation of 8b with 2-cyanobutyryl chloride20 via
conditions described previously.15 The mixtures of 23a
and 24a and of 23b and 24b were separated into the
individual diastereomers by chromatography.
Amine 11 for incorporation into compound 15d was
prepared from the phthalimide protected d-alaninol
Scheme 1.

4146
G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
17c and 17d, respectively, impart high inhibitory
potency (Ki’s<50pM) when coupled to any of the left
hand carboxylate derived fragments contained in com-
pounds 1, 2, 3, 4 and 6.10ꢀ14,17 However, when the
amines are incorporated into the cyclopentanone car-
boxamides 17c and 17d, the Ki’s are 3–4 orders of mag-
nitude higher (Table 1). Compound 17a derived from
norephedrine has an added phenyl substituent relative
to 17d and shows a 2200-fold improvement in binding
potency, in accord with our model.
Scheme 2.
Bioefficacy
Cyclopentanone (17), cyclohexanone (18) and tetra-
hydrothiophenone (19) carboxamides were synthesized
to evaluate the relationship of ring size to binding
potency (Table 2). From the proton NMR analysis,
compounds 17–19 exist to varying degrees as a mixture
In order to validate the inhibitor binding model of this
work, cyclopentanone carboxamides 17a, 17c and 17d
were initially prepared. The chiral bromophenyl amine
and phenoxypropyl amine used for the preparations of
Scheme 3.
Scheme 4.

G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
Table 1. Ki values of SD and in vivo activity for three cyclopentanone carboxamides
4147
Compound
Structure
Ki (pM)
86ꢂ8
Sys act @3 ppm
15
Sys act @0.8 ppm
11
17a
17c
17d
23,000ꢂ300
0
0
0
190,000ꢂ2000
29
of keto and enol tautomers. Each of the keto tautomers
further exists as a mixture of diastereomers. It can be
presumed that the interconversion of the diastereomers
occurs rapidly via the enol tautomer relative to the
incubation time (1 min) of the enzyme assay. The tetra-
hydrothiophenone carboxamides 19, which are inter-
mediate in ring size to the cyclopentanone and
cyclohexanone, showed the highest binding potencies
with Ki’s ranging from 15 to 16 pM. Both the 2,5-chloro
and 2,5-difluoro substitution on the phenoxy group in
the analogue series synthesized around 4 and 5 had
previously been shown to impart high SD binding
potency relative to substitution at other positions or
replacement with more polar functionalities.11,12 There-
fore, synthesis of compounds in this work was limited to
the 2,5-dihalo substitution.
Table 2. Ki values of SD and in vivo activity for cyclopentanone, cyclohexanone and tetrahydrothiopenone carboxamides
Compound
X
Z
# of diast.
Ki (pM)
Sys. act @3 ppm
Sys. act @0.8 ppm
17aa
17b
18a
18b
19a
19b
Cl
F
Cl
F
Cl
F
CH2
CH2
CH2CH2
CH2CH2
S
2
2
2
2
2
2
86ꢂ8
71ꢂ5
15
37
0
11
0
130ꢂ8
71ꢂ1013
16ꢂ2
0
0
0
0
27
S
15ꢂ3
0
aData repeated from Table 1.
Table 3. Ki values of SD and in vivo activity for butyrolactam and valerolactam carboxamides
Compound
n
R (configuration)
X
# of diast.
Ki (pM)
Sys. act @3 ppm
0
Sys. act @0.8 ppm
0
22a
22b
26a
26b
23a
23b
24a
24b
1
1
2
2
1
1
1
1
H
H
H
Cl
F
Cl
F
Cl
F
Cl
F
2
2
2
2
1
1
1
1
2000ꢂ100
460ꢂ2055
3600ꢂ100
960ꢂ16017
1000ꢂ1030
2300ꢂ200
58ꢂ5
42
0
0
H
0
CH3 (4S,3R)
CH3 (4S,3R)
CH3 (4R,3S)
CH3 (4R,3S)
33
44
97
95
52
55
92
110ꢂ9

4148
G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
The cyclic lactam carboxamides 22, 23, 24 and 26 (Table
3) exhibited lower binding affinity to SD than com-
pounds 19, though notably Ki’s still ranged from 58 to
3600 pM. The five-member ring lactams 22 and 24
showed higher binding affinity than the six-membered
ring compounds 26. The addition of the methyl group
trans to the carboxamide improves binding affinity
(compare compounds 24a and 24b with 22a and 22b,
respectively). The diastereomers 23 and 24 could be
separated by chromatography, and there is no epimer-
ization to the cis diastereomers during the course of the
enzyme assay. Contamination of the more active isomer
in the less active isomer would appreciably overestimate
the binding efficacy of the latter though such con-
tamination was not seen within the limits of NMR
analysis. The more active diastereomers 24 bind at least
20-fold tighter than the less active. The assignment of
the relative stereochemistries of 23a/24a and 23b/24b
could not be positively identified though spectroscopic
means. The (3S,4R) configuration about the lactam ring
[vs the (4S,3R) configuration] correlates with the model
we have derived and is therefore assigned to the more
potent compounds 24a and 24b.
stants of compounds 5b, 19b and 24b for the Asn131Ala
mutant4 and wild-type SD to compare the relative
importance of displaying a nitrile, a ketone and a car-
boxamide towards the Asn131 carboxamide. The three
compounds of Table 4 are representative of three classes
of inhibitors discussed herein. The binding affinity dete-
rioration is greatest (160-fold) for 24b in going from
wild-type SD to the Asn131Ala mutant. In contrast, the
affinity of 5b for the Asn131Ala mutant is only 26-fold
less than for wild-type SD.
Discussion
The enzyme binding data support the inhibitor design
concept in that improved binding potency was achieved
by meshing the interaction of inhibitors with two SD
residues (Asn131 and Phe53) that are spatially located
in different regions of the binding pocket. Previously,
we described the design of 5 (and analogues) starting
from compounds such as 4. In order to accommodate
the additional phenyl group of 5, the t-butyl substituent
was truncated to an ethyl group as suggested by the
binding conformation of the latter delineated by X-ray
crystallography.12 The design involved two modifi-
cations of inhibitor structure to achieve higher binding
potency. Similarly, there are two important considera-
tions applied to the design of the inhibitors of this work.
First, a carbonyl functionality was displayed towards
the carboxamide of the side chain of Asn131 to afford a
hydrogen bonding capability. The carbonyl was posi-
tioned as part of a ring to afford a better steric fit into
the binding pocket and to pre-organize the carbonyl
into an orientation that allows for a hydrogen bond. In
the case of the lactam inhibitors, the ring locks the
cisoid geometry of the carboxamide NH and carbonyl
oxygen necessary for the potential formation of two
hydrogen bonds. The second consideration involves the
proper choice of the carboxamide N-substituent for
interaction with Phe53. The SD inhibitory potencies of
17c and 17d are lower due to an unfavorable contact
A stringent in vivo assay that approximates farm prac-
tices was employed to evaluate rice blast disease con-
trol.14 The ratings in Tables 1–3 represent the per cent
disease control in a systemic assay in which compound
applied to the soil root zone of a rice seedling must
translocate to the leaf surface where inoculum is applied.
The data from the assay are best interpreted as yes–no
results in which disease control above 90% at a given
dose represents compound effectiveness while control
below 90% is deemed unacceptable for the purpose of
disease management (i.e., at the level of commercially
viable products). Compounds 24a and 24b are the only
ones that afford effective control by this assay.
The inhibitor modifications described are targeted at
improving the interaction with the side-chain carbox-
amide of Asn131. Table 4 shows the inhibition con-
Table 4. Ki values (pM) of wild-type SD and Asn131Ala SD for three carboxamides
Compound
Structure
(A) Ki (Asn131Ala)
740ꢂ1028
(B) Ki (wild-type)
(A)/(B) relative
5b
ꢂ4
26
19b
24b
750ꢂ18015
ꢂ2
50
18,000ꢂ200
110ꢂ9
160

G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
4149
with the side chain of Phe53, which does not allow a
favorable alignment of the cyclopentanone carbonyl
with Asn131. The phenyl group of Phe53 stacks with the
p-face of the unsubstituted phenyl group of the nor-
ephedrine derived 17a and 17b and enters into a favor-
able CH–p interaction21 with the substituted phenoxy
group. These interactions relieve the steric congestion
around the cyclopentanone five-membered ring and
thereby allow for alignment of the hydrogen bond giv-
ing rise to the strong binding potency of the inhibitor.
ability of the nitrile of 5a, to the improved capability of
the carbonyl compounds 17–19, to the further improve-
ment of lactams 22, 24 and 26, there is a progression
towards increasing hydrophilicity, which has been cor-
related with improved plant translocation dynamics.22
Though the ketones 17–19 display some of the lower
Ki’s, they do not exhibit commensurate in vivo activity
(see Table 2). They exist as a mixture of keto-enol tau-
tomers, which serve as potential handles for metabo-
lism. That compounds 24a and24b controlled disease in
the systemic assay likely reflects their greater hydro-
philicity and bioavailability together with the measure
of higher binding potency to SD.
Inhibitor binding involves a number of factors including
the contacts to individual protein residues and the
energies associated with desolvation of both the inhi-
bitor and the surface of the enzyme binding pocket. The
progression of the compounds in Table 4 from 5b to 19b
to 24b might be thought to disfavor binding due to the
concomitant increase in hydrophilicity, which favors
partitioning into the aqueous environment. Though the
ketone moiety of 19b can enter into only one hydrogen
bond with the enzyme while the butyrolactam of 24b
can enter into two hydrogen bonds, the binding con-
stant of the latter is higher due to the entropic penalty
associated with desolvation of the more polar lactam
moiety. Since the enzyme contacts with the five-mem-
bered rings of 19b and 24b differ from one another and
from the ethyl substituted cyanoacetamide fragment of
5b, evaluation of inhibitor binding constants with the
Asn131Ala mutant addresses the role of the Asn131
side-chain carboxamide for recognition of the inhibi-
tors. From this analysis, the better hydrogen bonding of
19b and 24b in comparison to 5b are considered to make
up for a loss of hydrophobicity. Though compound 24b
binds less well to SD than 19b and 5b, the 160-fold
deterioration in binding to Asn131Ala in comparison to
wild-type SD indicates that the specific interaction with
the side chain carboxamide is strongest among the three
compounds. The 50-fold deterioration in binding to
Asn131Ala in comparison to wild-type SD indicates
that 19b has a stronger contact with Asn131 than 5b
where the corresponding loss is 26-fold. Hence, the
value of comparing inhibitor binding to Asn131Ala and
wild-type SD is to dissect apart some of the factors that
contribute to the binding phenomenon. Ordinarily,
analysis of site-directed mutants serves to delineate cat-
alytic mechanisms of enzyme reactions, as has certainly
been exemplified for building a mechanistic hypothesis
for SD catalysis.4,8 Here, inhibitor binding studies on
the Asn131Ala mutant serve to substantiate the design
thesis: compounds envisioned to provide improved
hydrogen bonding capability to the Asn131 carbox-
amide indeed do so. In the analyses, it is assumed that
the desolvation energy associated with the binding of
the inhibitors is factored away by the comparison
between binding to Asn131Ala and wild-type SD.
In summary, the cyclopentanone, tetrahydro-
thiophenone and butyrolactams constitute novel
pharmacophores for binding to SD leading to novel
fungicides. The design of the potent inhibitors goes a
step beyond ordinary structure-based design methods
by realizing the potential strength of hydrogen bonding
to the Asn131 carboxamide from functional studies of
the enzyme. Purposely, for obtaining more polar inhib-
itors with improved in vivo expression of fungicide
activity, the more polar substituents of inhibitors were
designed to mesh with the Asn131 side chain carbox-
amide. Proof of the designed interaction with the car-
boxamide of Asn131 comes from analysis of inhibitor
binding constants of the Asn131Ala mutant of SD in
comparison to the wild-type SD. This study of SD
inhibitors underscores the value of structural studies
and site-directed mutagenesis as tools for linking struc-
ture to function to practice.
Experimental
Methods and materials
Homogeneous wild-type scytalone dehydratase was
purified as described.23 The homogeneous Asn131Ala
mutant of scytalone dehydratase was expressed in
Escherichia coli and purified as described.4 All other
materials were from Aldrich Chemical (Milwaukee, WI,
USA), Sigma Chemical (St. Louis, MO, USA), Hamp-
ton (Laguna Hills, CA, USA), Fluka (Buchs, Switzer-
land), ICN Biomedicals, Inc. (Costa Mesa, CA, USA),
or Lancaster Synthesis Inc. (Windham, NH, USA).
Chromatographic separations were performed with E.
Merck silica gel (230–400 mesh) under nitrogen pressure
(flash chromatography). Except where indicated, anhy-
drous solvents and reagents were used as received. All
1
melting points are uncorrected. H NMR spectra where
noted, were recorded at 300 MHz. 19F NMR spectra
were recorded at 282 MHz. 13C NMR spectra were
recorded at 126 MHz. 31P NMR spectra were recorded
at 212 MHz. Optical rotation were measured with a
Perkin-Elmer model 241 polarimeter. Mass spectra were
obtained on a Micromass Platform 2 using Atmospheric
Pressure Chemical Ionization detection at 3.5 kV posi-
tive ion and 2.5 kV negative ion. Accurate mass ele-
mental composition measurements were performed on a
Finnigan MAT900XL mass spectrometer using Elec-
tron Multiplier detection at 1.80kV positive ion and
Most often in drug design, improvement in binding
potency is not the only criterion for superior in vivo
performance as bioavailability, metabolism and mem-
brane permeability can degenerate the expression of
desired in vivo activity. The interest in finding new
classes of chemistry stems from the desire to overcome
these hurdles. In going from weak hydrogen bond cap-

4150
G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
1.90kV negative ion. Inhibition constants were deter-
mined for both wild-type SD and Asn131Ala via meth-
ods described previously with a minimum of three
inhibitor concentrations and duplicate activity mea-
surements.10,15 Protocols for the in vivo assay for
determination of disease control have been described.14
layers were dried (MgSO4) and concentrated, and the
residue was chromatographed on silica gel. Elution with
ether removed a higher Rf component and with ethyl
acetate a lower Rf component. Each of the isolated
components were dissolved in ether and treated with
excess 1N HCl in ether to precipitate solids. The higher
Rf component (4.4 g) was identified as (2S,1R)-2-(2,5-
dichlorophenoxy)-1-phenylpropylamine (9a) and was
characterized as its hydrochloride salt: mp 225–227 ꢁC.
½aꢃD =+36.4 (c 1.02, EtOH); H NMR (DMSO-d6) d
1.2 (d, 3H), 3.7 (m, 1H), 5.9 (d, 1H), 6.8 (m, 1H), 7.0(m,
1H), 7.25–7.4 (m, 6H), 8.3–8.4 (s, br, 3H); 13C NMR
(DMSO-d6) d 15.8, 57.2, 75.5, 116.7, 122.3, 122.4, 128.1,
128.8, 129.3, 131.0, 132.3, 133.4, 152.8; IR (KBr) 3446
(br), 3017, 2983, 2913, 1582, 1502, 1477, 1403, 1260,
1057, 700 cmꢀ1. HRMS calcd for C15H16Cl2NO
[(M+ꢀHCl)+1]: 296.0609. Found: 296.0622. The lower
Rf component corresponded to 2.6 g of additional
desired product 8a.
Synthesis of ((2R,3R)-3-methyl-2-phenylaziridinyl)bis
(methylethoxy)phosphino-1-one (7). A solution of 68 g
(215 mmol) of [((2S,1R)-2-hydroxy-1-methyl-2-phenyl-
ethyl)amino]bis(methylethoxy)phosphino-1-one24 and
33 mL (240mmol) triethylamine in 30mL ether was
cooled in ice water. Methanesulfonyl chloride (18.3 mL,
240mmol) was added dropwise. The mixture was stirred
at room temperature overnight before being diluted
with ether and washed with water and brine. Drying
(MgSO4) and removal of solvent gave 78 g of an
unstable solid, which was used without further purifica-
25
1
ꢁ
1
tion: mp 121–123 C; H NMR (CDCl3) d 1.15 (d, 3H),
1.3–1.4 (m, 12H), 2.7 (m, 1H), 2.9 (s, 3H), 3.6–3.7 (m,
1H), 3.8 (s, br, 1H), 4.6 (m, 2H), 5.65 (d, 1H), 7.4 (m,
5H); 13C NMR (CDCl3) d 16.6, 23.8, 38.8, 52.3, 86.7,
126.6, 127.5, 128.7, 136.2. The solid was dissolved in dry
THF cooled in ice water while 30.6 g (280 mmol) of
potassium t-butoxide was added all at once. After
warming to room temperature and stirring for 2 h, the
solvent was removed, and the residue was diluted in
ether. The ether was washed with water and brine.
Drying (MgSO4) and removal of solvent gave 58 g of
Synthesis of (1S,2R)-1-(2,5-difluorophenoxy)-1-phenyl-
prop-2-ylamine hydrochloride (8b). Potassium t-butoxide
(1.7 g, 15 mmol) was added to a solution of 4.4 g (32
mmol) of 2,5-difluorophenol in 50mL DMF at room
temperature. Aziridine 7 (5.0g, 17 mmol) was added,
and the solution was heated at 130 ꢁC for 21 h. The
solution was quenched with aqueous NH4Cl and
extracted with EtOAc. The EtOAc was washed three
times with water and once with brine. Drying (MgSO4)
and removal of solvent gave a 9 g of a viscous oil. The
oil was heated at reflux in 50mL of 20% aqueous HCl
for 2 h. After cooling and diluting with water, the solu-
tion was extracted with ether. Insoluble solids were fil-
tered, washed with hexanes and dried in vacuo to give
25
product as an oil: ½aꢃD =ꢀ60.3 (c 1.02, EtOH); 1H
NMR (CDCl3) d 1.2–1.3 (m, 9H), 1.3–1.4 (d, 3H), 1.6
(d, 3H), 2.6 (m, 1H), 3.3–3.4 (dd, 1H), 4.6–4.8 (m, 2H),
7.3 (m, 5H); 13C NMR (CDCl3) d 15.7, 25.7, 46.1, 46.9,
73.5, 127.8, 129.5, 130.4, 139.6; 31P NMR (CDCl3) d
9.5; IR (neat) 2979, 2934, 1460, 1419, 1385, 1375, 1263,
1241, 1179, 1142, 1109, 1077, 1063, 1001, 923, 756, 701
25
1.5 g of 8b, mp 233–235 ꢁC; ½aꢃD =ꢀ3.7 (c 1.02, EtOH);
1H NMR (DMSO-d6) d 1.2 (d, 3H), 3.7 (m, 1H), 5.9 (d,
1H), 6.8 (m, 1H), 7.0(m, 1H), 7.25–7.4 (m, 6H), 8.3–8.4
(s, br, 3H); 19F NMR (CDCl3) d ꢀ116, ꢀ138; 13C NMR
(DMSO-d6) d 12.3, 52.6, 80.2, 105.4, 108.6, 117.3, 126.8,
129.0, 129.6, 136.0, 145.7, 149.3 (d, J=243 Hz), 158.2
(d, J=241 Hz); IR (KBr) 3426 (br), 2864, 1511, 1207,
cmꢀ1
;
298.1572. Found: 298.1581.
HRMS calcd for C15H25NO3P (M++1):
Synthesis of (1S,2R)-1-(2,5-dichlorophenoxy)-1-phenyl-
prop-2-ylamine hydrochloride (8a). Potassium t-butoxide
(7.85 g, 70mmol) was added to a solution of 25.3 g (155
mmol) of 2,5-dichlorophenol in 150mL DMF at room
temperature. Aziridine 7 (23 g, 78 mmol) was added,
and the solution was heated at 130 ꢁC for 23 h. The
solution was diluted with ether and washed three times
with water and once with brine. Drying (MgSO4) and
removal of solvent gave a 35 g of a viscous oil. The oil
was heated at reflux in 200 mL of 20% aqueous HCl for
2 h. After cooling and diluting with water, the solution
was extracted with ether. Insoluble solids were filtered,
washed with ether and dried in vacuo to give 12.9 g of
1156, 842, 751, 700 cmꢀ1
;
HRMS calcd for
C15H16F2N2O (M++1): 264.1199. Found: 264.1199.
More desired product could be recovered by basifying
the aqueous layer with Na2CO3 extracting additional
ether. The combined ether layers were dried (MgSO4)
and concentrated, and the residue was chromato-
graphed on silica gel. Elution with ether removed a
higher Rf component and with ethyl acetate a lower Rf
component. Each of the isolated components were dis-
solved in ether and treated with excess 1N HCl in ether
to precipitate solids. The higher Rf component (0.5 g)
was identified as (2S,1R)-2-(2,5-difluorophenoxy)-1-
phenylpropylamine (9b) and was characterized as its
25
product 8a: mp 209–211 ꢁC; ½aꢃD =ꢀ45.7 (c 1.03,
EtOH); 1H NMR (DMSO-d6) d 1.4 (d, 3H), 3.7 (m,
1H), 5.75 (d, 1H), 6.7 (d, 1H), 6.85 (d, 1H), 7.25 (d, 1H),
7.35 (m, 5H), 8.5 (s, br, 3H); 13C NMR (DMSO-d6) d
11.8, 51.3, 79.0, 115.5, 121.5, 122.2, 126.3, 128.6, 128.8,
131.2, 132.0, 135.4, 152.6; IR (KBr) 3452 (br), 3085,
25
hydrochloride salt: mp 228–231 ꢁC; ½aꢃD =ꢀ10.0 (c 0.94,
EtOH) 1H NMR (DMSO-d6) d 1.2 (d, 3H), 3.7 (m, 1H),
5.9 (d, 1H), 6.8 (m, 1H), 7.0(m, 1H), 7.25–7.4 (m, 6H),
8.3–8.4 (s, br, 3H); 19F NMR (DMSO-d6) d ꢀ116,
ꢀ137; 13C NMR (DMSO-d6) d 15.8, 57.2, 75.6, 105.5,
108.2, 116.7, 128.2, 128.7, 128.9, 133.6, 145.0, 149.4 (d,
J=240Hz), 158.1 (d, J=241 Hz); IR (KBr) 3340(br),
3040, 3000, 2904, 1625, 1596, 1515, 1459, 1434, 1285,
1206, 1161, 1146, 1101, 1031, 981, 852, 811, 773, 702
3037, 2999, 2904, 1581, 1497, 1474, 1404, 1255 cmꢀ1
;
HRMS calcd for C15H16Cl2NO (M++1): 296.0609.
Found: 296.0609. More desired product could be
recovered by basifying the aqueous layer with Na2CO3
extracting with additional ether. The combined ether

G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
4151
25
was isolated as a solid: mp 82–90 ꢁC; ½aꢃD =+35.9 (c
cmꢀ1. HRMS calcd for C15H16F2NO [(M+ꢀHCl)+1):
264.1200. Found: 264.1189. The lower Rf component
corresponded to 2.6 g of additional desired product 8b.
1
1.01, EtOH); H NMR (CDCl3) d 1.2–1.3 (m, 3H), 1.8–
2.4 (m, 6H), 2.9–3.0(m, 1H), 4.4 (m, 1H), 5.3 (d of d,
1H), 6.4–6.6 (2m, 1H), 7.0–7.1 (m 2H), 7.2–7.4 (m, 5H);
19F NMR (CDCl3) d ꢀ117, ꢀ140; 13C NMR (CDCl3) d
13.8 and 13.3, 20.4, 25.8, 38.8, 50.7 and 50.5, 54.2, 83.2,
104.8 and 105.0, 107.3 and 107.5, 126.2, 128.2, 128.7,
137.2, 146.8, 149.3 (d, J=240Hz), 158.4 (d, J=239 Hz),
166.4, 216.0; IR (KBr) 3290, 2990, 2880, 1740, 1630,
1580, 1480, 1455, 1400, 1270, 1250, 1135, 1100, 1060,
1035 cmꢀ1; HRMS calcd for C21H22F2NO3 (M++1):
374.1568. Found: 374.1557.
Synthesis of 6,9-dioxa-3-thiaspiro[4.4]nonanecarboxylic
acid (13c). The carboalkoxyketone 12c25 (12 g, 75
mmol), ethylene glycol (4.7 mL, 86 mmol) and 0.5 g
p-toluenesulfonic acid in 300 mL benzene were heated at
reflux overnight with azeotropic removal of water. After
dilution with ether, the mixture was washed with water
and brine. Drying (MgSO4) and removal of solvent gave
an oil that was chromatographed on silica gel (3:1 hex-
anes–ether) to give 5 g of an oil. The oil was dissolved in
75 mL MeOH along with 2.0g of 50% NaOH. The
solution was heated at reflux overnight. Solvent was
removed and the residue was partitioned between aqu-
eous Na2CO3 and ether. The ether was separated, and
the aqueous phase was extracted again with ether. The
aqueous phase was acidified with concd HCl and
extracted twice with EtOAc. The EtOAc extracts were
washed with brine, dried (MgSO4) and concentrated to
give 3.7 g of 13c as an oil: 1H NMR (CDCl3) d 2.95–3.1
(m, 4H), 3.2–3.3 (m, 1H), 4.0–4.8 (m, 4H); 13C NMR
(CDCl3) d 25.6, 29.2, 36.6, 38.2, 65.1, 175.4; IR (KBr)
3500 (br), 3056, 2984, 2956, 2897, 2600 (br), 1714, 1421,
1327, 1286, 1229, 1207, 1155, 1126, 1060, 1035, 1005,
950, 911 cmꢀ1; HRMS calcd for C7H11O4S (M++1):
191.0378. Found: 191.03852.
N-[(1R)-1-(4-Bromophenyl)ethyl](2-oxocyclopentyl)car-
boxamide (17c). Starting with 200 mg of 13a and 250
mg of 4-bromo-a-methylbenzylamine hydrochloride,
140mg of 17c was isolated as an solid: mp 88–107 ꢁC;
25
½aꢃD =+109.2 (c 1.03, EtOH); 1H NMR (CDCl3) d 1.45
(m, 3H), 1.8 (m, 1H), 2.0–2.2 (m, 4H), 2.9 and 3.0 (2t,
1H), 5.0–5.1 (m, 1H), 7.0 (m, br, 1H), 7.1–7.2 (2d, 2H),
7.45 (2d, 2H); 13C NMR (CDCl3) d; 20.4, 21.9, 22.4,
25.4, 25.7, 35.4, 35.7, 38.9, 47.6, 48.6, 51.8, 54.1, 64.3,
66.0, 71.7, 121.0, 127.9, 131.7, 142.5, 165.8, 216.6; IR
(KBr) 3325, 2978, 1738, 1646, 1539, 1491, 1137, 1009,
823 cmꢀ1. HRMS calcd for C14H17BrNO2 (M++1):
310.0443. Found: 310.0433.
N-[(1R)-2-(2,5-Dichlorophenoxy)-isopropyl](2-oxocyclo-
pentyl)carboxamide (17d). Starting with 200 mg of 13a
and 270mg of 11,15 120mg of 17d was isolated as a
General procedure for the synthesis of 17–19
25
solid: mp 92–98 ꢁC; ½aꢃD =+53.2 (c 0.99, EtOH); H
NMR (CDCl3) d 1.35 (m, 3H), 1.7–1.9 (m, 1H), 2.0–2.2
(m, 1H), 2.2–2.5 (m, 4H), 2.95 (q, 1H), 3.9–4.1 (m, 2H),
4.2 (m, 1H), 6.8–7.0(m, 3H), 7.25 (d, 1H); 13C NMR
(CDCl3) d 17.4, 20.4, 25.9, 38.9, 44.7, 54.2, 71.7, 114.2,
121.7, 130.7, 133.1, 154.7, 154.5, 166.5, 216.2; IR (KBr)
3285, 2979, 2880, 1740, 1646, 1586, 1554, 1484, 1460,
1
A few drops of DMF were added to a solution of oxalyl
chloride (1.2 equiv) and acid 13 (1.1 equiv) in CH2CH2.
The mixture was heated at reflux overnight. A solution
of amine 8 (1.0equiv) and triethylamine (2.1 equiv) was
added, and the resulting mixture was stirred at room
temperature overnight. Solvent was removed, and the
residue was taken up in EtOAc and washed with water
and brine. Drying (MgSO4) and removal of solvent gave
an oil, which was dissolved in 1:1 THF–1 N aqueous
HCl with stirring overnight. The solution was diluted
with EtOAc and washed with water and brine. Drying
(MgSO4) and removal of solvent gave an oil, which was
chromatographed on silica gel (35% hexanes in ether) to
give product.
1403, 1266, 1248, 1135, 1096, 1060, 1035, 897, 837 cmꢀ1
;
HRMS calcd for C15H18Cl2NO3 (M++1): 330.0664.
Found: 330.0649.
N-[(2S,1R)-2-(2,5-Dichlorophenoxy)-1-methyl-2-phenyle-
thyl](2-oxocyclohexyl)carboxamide (18a). Starting with
250mg of 13b27 and 370mg of 8a, 330mg of 17a was
25
isolated as a solid: mp 101–114 ꢁC; ½ꢀꢃD =ꢀ19.4 (c 1.04,
1
EtOH); H NMR (CDCl3) d 1.2 (m, 3H), 1.6–2.4 (m,
N-[(2S,1R)-2-(2,5-Dichlorophenoxy)-1-methyl-2-phenyl-
ethyl](2-oxocyclopentyl)carboxamide (17a). Starting
with 230mg of 13a26 and 370mg of 8a, 380mg of 17a
8H), 3.1–3.2 (m, 0.5H), 4.4–4.5 (m, 1H), 5.4 (m, 1H), 5.8
(d, 0.5H),6.6 (m, 1H), 6.85 (t 1H), 7.1–7.4 (m, 6.5H),
14.0(s, 0.5H); 13C NMR (CDCl3) d 12.4 and 12.3, 19,
20, 23.1, 26.2, 28, 30.2, 33, 41.1, 49.0, 49.5, 55.0, 63.4,
81.5 and 82.0, 82.5, 19.2, 114.9 and 114.6 and 114.2,
120.8 and 121.0, 124.9, 127.2, 127.8, 129.6, 132.2, 136.0,
163.2, 169.7 and 170.9, 210; IR (KBr) 3403, 2937, 1630,
1605, 1588, 1520, 1484, 1395, 1260, 1220, 1090, 1055,
1000 cmꢀ1; HRMS calcd for C22H24Cl2NO3 (M++1):
420.1133. Found: 420.1122.
25
1
was isolated as an oil: ½aꢃD =ꢀ5.7 (c 0.98, EtOH); H
NMR (CDCl3) d 1.2–1.3 (m, 3H), 1.7–2.4 (m, 6H), 2.9–
3.0(m, 1H), 4.4 (m, 1H), 5.4 (d of d, 1H), 6.6 (2s, 1H),
6.8 (m 1H), 7.0–7.1 (m, 1H), 7.2–7.4 (m, 6H); 13C NMR
(CDCl3) d 13.7 and 13.4, 20.3, 25.9, 38.7, 50.5 and 50.8,
82.7, 115.2 and 115.5, 121.7, 122.3, 126.0, 128.1, 128.7,
130.6, 132.8, 136.8, 154.0, 170.7, 215.8; IR (KBr) 3280,
2980, 2880, 1740, 1630, 1605, 1585, 1480, 1460, 1403, 1266,
1250, 1135, 1100, 1060, 1035, 900 cmꢀ1; HRMS calcd for
C21H22Cl2NO3 (M++1): 406.0977. Found: 406.0992.
N-[(2S,1R)-2-(2,5-Difluorophenoxy)-1-methyl-2-phenyl-
ethyl](2-oxocyclohexyl)carboxamide (18b). Starting with
250mg of 13b and 330mg of 8b, 200 mg of 17b was
25
N-[(2S,1R)-2-(2,5-Difluorophenoxy)-1-methyl-2-phenyl-
ethyl](2-oxocyclopentyl)carboxamide (17b). Starting
with 230mg of 13a and 330mg of 8b, 240mg of 17b
isolated as a solid: mp 86–92 ꢁC; ½aꢃD =+7.8 (c 1.01,
1
EtOH); H NMR (CDCl3) d 1.2 (m, 3H), 1.7–2.4 (m,
8H), 3.1–3.2 (m, 0.5H), 4.4 (m, 1H), 5.4 (m, 1H), 5.8 (d,

4152
G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
0.5H), 6.6–6.8 (m, 1H), 7.0 (m 1H), 7.2–7.4 (m, 5.5H),
14.0(s, 0.5H); 19F NMR (CDCl3) d ꢀ117, ꢀ140; 13C
NMR (CDCl3) d 13.4, 13.7, 21.9. 22.5, 24.1, 27.2, 28,
29.3, 31.3, 33, 36, 42.0, 49.8, 50.3, 50.5, 55.9, 83.1, 83.3,
83.8, 96.9, 104, 107, 116, 126.0, 126.2, 128.2, 128.7,
137.2, 146.8, 150(d, J=240Hz), 159 (d, J=240Hz),
168.6, 170.8, 171.9, 210.0; IR (KBr) 3395, 2940, 1630,
1605, 1583, 1520, 1478, 1396, 1261, 1223, 1094, 1056,
999 cmꢀ1; HRMS calcd for C22H24F2NO3 (M++1):
388.1724. Found: 388.1713.
140, 1375, 1261, 1094, 1060, 995 cmꢀ1; HRMS calcd for
C20H21 Cl2N2O3 (M++1): 407.0929. Found: 407.0941.
Synthesis of N - [(2S,1R) - 2- (2,5 - difluorophenoxy) - 1 -
methyl-2-phenylethyl](2-oxopyrrolidin-3-yl)carboxamide
(22b). A solution of 180mg (1.4 mmol) pyrrolidinone
carboxylate 20, 0.39 mL (2.8 mmol) triethylamine, 420
mg (1.4 mmol) 8b, and 270mg (1.4 mmol) EDC in 10
mL CH2CH2 was stirred at room temperature over-
night. The solution was washed with water and brine,
dried (MgSO4) and concentrated. The residue was
chromatographed on silica gel (EtOAc) to give product
N-[(2S,1R)-2-(2,5-Dichlorophenoxy)-1-methyl-2-phenyl-
ethyl](4-oxo(3–2,3,5-trihydrothienyl))carboxamide (19a).
Starting with 260mg of 13c and 370mg of 8a, 110mg
25
as a 1:1 mixture of diastereomers as an oil: ½aꢃD =+29.7
1
(c 1.14, EtOH); H NMR (CDCl3) d 1.2–1.3 (2d, 3H),
of 19a was isolated as
a
solid: mp 90–98 ꢁC;
2.3–2.7 (m, 2H), 3.2 (m, 1H), 3.4 (m, 2H), 4.4 (m, 1H),
5.3 (dd, 1H), 5.7 (d, br 1H), 6.5 (m, 2H), 7.0(m, 1H),
7.2–7.4 (m, 5H), 7.8 (2d, 1H); 19F NMR (CDCl3) d
ꢀ117, ꢀ139; 13C NMR (CDCl3) d 12.5 and 12.7, 22.2.
and 22.4, 39.2, 45.0, 49.5 and 49.7, 82.3, 103.8 and
104.0, 106.1 and 106.3, 115.1 and 115.4, 126.0, 127.9,
128.4, 137.1, 147.4 (d, J=207), 158.2 (d, J=242 Hz),
167.2, 175.3; IR (CH2CH2) 3430, 3300 (br), 3060,
2990, 2965, 2680, 1702, 1675, 1625, 1510, 1453, 1431,
1322, 1207, 1155, 1100, 1002, 800 cmꢀ1; HRMS calcd
for C20H21F2N2O3 (M++1): 375.1520. Found:
375.1539.
25
1
½aꢃD =ꢀ41.4 (c 1.02, EtOH); H NMR (CDCl3) d 1.2–
1.3 (m, 3H), 3.2–3.8 (m, 4.75H), 4.35–4.55 (m, 1H), 5.3–
5.4 (dd, 1H), 5.6 (d, 0.25H), 6.6 (m, 1H), 6.8–6.9 (m
1H), 7.0–7.1 (m, 0.75H), 7.2–7.4 (m, 6H), 12.8 (s,
0.25H); 13C NMR (CDCl3) d 13.1 and 13.3, 27, 30.7,
35.9, 42.4, 49.7 and 49.9 and 50.7; 54.8 and 56.4, 82.4
and 83.6, 100.1, 115.2 and 115.4 and 115.8, 121.6 and
121.7 and 122.1, 125.6 and 125.8, 128.1 and 128.2, 128.6
and 128.8, 130.5, 136.5, 165; IR (KBr) 3410, 2940, 1720
1630, 1605, 1587, 1505, 1480, 1395, 1260, 1220, 1090,
1055, 1000 cmꢀ1; HRMS calcd for C20H20Cl2NO3S
(M++1): 424.0541. Found: 424.0527.
Synthesis of N - [(2S,1R) - 2- (2,5 - dichlorophenoxy) - 1 -
methyl-2-phenylethyl]((4S,3R)-4-methyl-2-oxopyrrolidin-
3-yl)carboxamide (23a)and N-[(2S,1R)-2-(2,5-dichloro-
phenoxy)-1-methyl-2-phenylethyl]((4R,3S)-4-methyl-2-
oxopyrrolidin-3-yl)carboxamide (24a). A solution of 200
mg (1.4 mmol) methyl pyrrolidinone carboxylate 21,29
0.6 mL (1.4 mmol) triethylamine, 470 mg (1.4 mmol) 8a,
and 220mg (1.4 mmol) EDC in 10mL CH 2CH2 was
stirred at room temperature overnight. Solvent was
removed, and the residue was dissolved in EtOAc. The
solution was washed with water and brine, dried
(MgSO4) and concentrated. The residue was chromato-
graphed on silica gel (1:1 EtOAc–CH2CH2) to separate
two materials. The higher Rf component (23a) was iso-
N-[(2S,1R)-2-(2,5-Difluorophenoxy)-1-methyl-2-phenyl-
ethyl](4-oxo(3–2,3,5-trihydrothienyl))carboxamide (19b).
Starting with 260mg of 13c and 330mg of 8b, 100 mg
25
of 19b was isolated as an oil: ½aꢃD =ꢀ17.5 (c 1.04,
1
EtOH); H NMR (CDCl3) d 1.2–1.3 (m, 3H), 3.2–3.8
(m, 4.75H), 4.4 (m, 1H), 5.3 (m, 1H), 5.6 (d, 1H), 6.4–
6.6 (m, 2H), 7.1–7.2 (m 1.75H), 12.8 (s, 0.25H); 19F
NMR (CDCl3) d ꢀ117, ꢀ140; 13C NMR (CDCl3) d
13.4, 13.5, 16.5, 28.6, 30.8, 36.1, 38.8, 42.3, 49.7, 50.1,
50.6, 50.9, 55.0, 83.2, 84.1, 98, 100.3, 104.4, 104.6, 107.1,
107.3, 116.3, 126.1, 128.3, 128.8, 136.9, 159.3, 157.4,
156.4, 146, 148, 150, 164.7, 171.4, 109.2; 3390, 2941,
1720 1630, 1605, 1590, 1520, 1484, 1395, 1260, 1220,
25
1090, 1050, 1000 cmꢀ1
;
HRMS calcd for
lated as an oil: ½aꢃD =+9.0( c 1.03, EtOH); H NMR
(CDCl3) d 1.2 (d, 3H), 1.3 (d, 3H), 2.7–2.9 (m, 3H), 3.4–
3.5 (m, 1H), 4.4 (m, 1H), 5.4 (m, 1H), 5.6 (s, 1H), 6.6 (s,
1H), 6.8 (d, 1H), 7.2–7.4 (m, 6H), 7.85 (d, 1H). 1H
NMR (CDCl3) d 1.2–1.3 (2d, 3H), 2.3–2.7 (m, 2H), 3.2
(m, 1H), 3.4 (m, 2H), 4.4 (m, 1H), 5.3 (dd, 1H), 5.7 (d,
br 1H), 6.5 (m, 2H), 7.0(m, 1H), 7.2–7.4 (m, 5H), 7.8
(2d, 1H); 13C NMR (CDCl3) d 13.8, 20.5, 33.6, 48.4,
51.5, 53.6, 83.5, 116.2, 122.4, 122.7, 127.1, 129.0, 129.6,
131.5, 133.7 137.9, 154.9, 168.3, 176.0; IR (KBr) 3424
(br), 3276 (br), 1700, 1664, 1582, 1537, 1477, 1452, 1402,
1382, 1261, 1094, 1058, 996, 703 cmꢀ1; HRMS calcd for
C21H23Cl2N2O3 (M++1): 421.1086. Found: 421.1084.
The lower Rf component (24a) was isolated as an oil:
1
C20H19F2NO3S (M++1): 392.1131. Found: 392.1126.
Synthesis of N - [(2S,1R) - 2- (2,5 - dichlorophenoxy) - 1 -
methyl-2-phenylethyl](2-oxopyrrolidin-3-yl)carboxamide
(22a). A solution of 180mg (1.4 mmol) pyrrolidinone
carboxylate 20,28 0.39 mL (2.8 mmol) triethylamine, 465
mg (1.4 mmol) 8a, and 270mg (1.4 mmol) ethyl-3(3-
dimethylamino)propyl carbodiimide (EDC) in 10mL
CH2CH2 was stirred at room temperature overnight.
The solution was washed with water and brine, dried
(MgSO4) and concentrated. The residue was chromato-
graphed on silica gel (EtOAc) to give product as a 1:1
25
mixture of diastereomers as an oil: ½aꢃD =ꢀ13.8 (c 0.99,
25
1
EtOH); H NMR (CDCl3) d 1.2–1.3 (2d, 3H), 2.3–2.7
½aꢃD =ꢀ23.4 (c 1.02, EtOH); 1H NMR (CDCl3) d 1.2 (d,
(m, 2H), 3.2 (m, 1H), 3.4 (m, 2H), 4.4 (m, 1H), 5.4 (dd,
1H), 5.6–5.7 (d, br 1H), 6.6 (s, 1H), 6.8 (m, 1H), 7.2–7.4
(m, 6H), 7.7–7.9 (2d, 1H); 13C NMR (CDCl3) d 13.4,
13.5, 28.7, 36, 38.8, 50.1, 50.8, 51.0, 55, 82.7, 83.8, 115.3,
115.6, 121.8, 122.0, 125.8, 126.0, 128.3, 128.8, 129.0,
130.7, 154.0, 164.8, 167.2, 171.5, 209.1; IR (CH2CH2)
3420 (br), 3280 (br), 1700, 1662, 1580, 1540, 1480, 1450,
3H), 1.3 (d, 3H), 2.65 (d, 1H), 2.9 (m, 2H), 3.4–3.6 (m,
1H), 4.4 (m, 1H), 5.4 (s, 1H), 5.6 (s, 1H), 6.6 (s, 1H), 6.8
(d, 1H), 7.2–7.4 (m, 6H), 7.75 (d, 1H); 13C NMR
(CDCl3) d 13.1, 19.3, 32.4, 47.3, 50.4, 52.5, 82.5, 115.2,
121.3, 121.8, 125.9, 127.9, 128.5, 130.4, 132.6 136.8,
153.9, 167.1, 174.9; IR (KBr) 3424 (br), 3284 (br), 1700,
1658, 1582, 1539, 1477, 1452, 1402, 1261, 1094, 1060,

G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
4153
996, 702 cmꢀ1; HRMS calcd for C21H23Cl2N2O3
(M++1): 421.1086. Found: 421.1090.
2852, 1675, 1582, 1526, 1489, 1477, 1451, 1402, 1356,
1324, 1286, 1200, 1134, 1094, 1060, 997, 899,842, 805
cmꢀ1; HRMS calcd for C21H23Cl2N2O3 (M++1):
421.1086. Found: 421.1087.
Synthesis of N - [(2S,1R) - 2- (2,5 - difluorophenoxy) - 1 -
methyl-2-phenylethyl]((4S,3R)-4-methyl-2-oxopyrrolidin-
3-yl)carboxamide (23b) and N-[(2S,1R)-2-(2,5-difluoro-
phenoxy)-1-methyl-2-phenylethyl]((4R,3S)-4-methyl-2-
oxopyrrolidin-3-yl)carboxamide (24b). A solution of 200
mg (1.4 mmol) methyl pyrrolidinone carboxylate 21, 0 .6
mL (1.4 mmol) triethylamine, 420mg (1.4 mmol) 8b,
and 220mg (1.4 mmol) EDC in 10mL CH 2CH2 was
stirred at room temperature overnight. Solvent was
removed, and the residue was dissolved in EtOAc. The
solution was washed with water and brine, dried
(MgSO4) and concentrated. The residue was chromato-
graphed on silica gel (1:1 EtOAc–CH2CH2) to separate
two materials. The higher Rf component (23b) was iso-
Synthesis of N - [(2S,1R) - 2- (2,5 - difluorophenoxy) - 1-
methyl - 2- phenylethyl](2- oxo(3 - piperidyl))carboxamide
(26b). A solution of 140mg (1.0mmol) valerolactam
carboxylate 25, 0.28 mL (2.0 mmol) triethylamine, 300
mg (1.0mmol) 8b, and 190mg (1.0mmol) EDC in 10
mL CH2CH2 was stirred at room temperature over-
night. The solution was washed with water and brine,
dried (MgSO4) and concentrated. The residue was
chromatographed on silica gel (EtOAc) to give product
as a 1:1 mixture of diastereomers as a solid: mp 52–
25
ꢁ
1
58 C; ½aꢃD =+29.4 (c 1.02, EtOH); H NMR (CDCl3)
d 1.25 (m, 3H), 1.6–2.0(m, 3H), 2.3–2.4 (m, 1H), 3.1–3.4
(m, 3H), 4.4 (m, 1H), 5.3 (dd, 1H), 5.9 (m, 1H), 6.5 (m,
2H), 7.0(m, 1H), 7.2–7.4 (m, 5H), 7.9 (2d, 1H); 19F
NMR (CDCl3) d ꢀ117, ꢀ139; 13C NMR (CDCl3) d 13.4
and 14.0, 20.4 and 20.7, 22.6, 42.5, 46.4, 50.5, 83.1 and
83.5, 104.7 and 105.1, 107.3, 126.3, 128.0, 128.6, 137.4,
146, 149 (d, J=240Hz), 158.3 (d, J=242 Hz), 167.3 and
167.5, 170.7; IR (CH2CH2) 3400, 3300 (br), 3060, 2983,
2943, 2876, 1675, 1626, 1511, 1452, 1431, 1384, 1356,
25
lated as a solid: mp 58–61 ꢁC; ½aꢃD =+55.5 (c 1.07,
1
EtOH); H NMR (CDCl3) d 1.2 (d, 3H), 1.3 (d, 3H),
2.7–3.0(m, 3H), 3.4 (t, 1H), 4.4 (m, 1H), 5.3 (d, 1H),
5.65 (s, br, 1H), 6.4–6.6 (m, 2H), 6.9–7.0(m, 1H), 7.2–
7.4 (m, 5H), 7.85 (d, 1H); 19F NMR (CDCl3)
dꢀ117,ꢀ140; 13C NMR (CDCl3) d 13.7, 19.6, 32.6, 47.4,
50.5, 52.7, 83.3, 104.8, 107.3, 116.2, 126.3, 128.1, 128.6,
137.4, 137.4, 146.8, 149.1 (d, J=241 Hz), 158.3 (d,
J=245 Hz), 167.3, 175.1; IR (CH2CH2) 3431, 3304 (br),
2985, 1702, 1668, 1532, 1511, 1453, 1432, 1321, 1207,
1155, 1100, 1002, 839, 802 cmꢀ1; HRMS calcd for
C21H22F2N2O3 (M++1): 389.1676. Found: 389.1671.
The lower Rf component (24b) was isolated as a solid:
1323, 1286, 1207, 1154, 1100, 1003, 839, 801 cmꢀ1
;
HRMS calcd for C21H22F2N2O3 (M++1): 389.1676.
Found: 389.1678.
Synthesis of 2- cyano - N - [(1R,2S) - 2 - (2,5 - difluorophe-
noxy)-1-methyl-2-phenylethyl]-butanamide (5b). A few
drops of DMF were added to a 10mL CH Cl2 solution
25
mp 56–58 ꢁC; ½ꢀꢃD =ꢀ3.1 (c 1.02, EtOH); 1H NMR
(CDCl3) d 1.2 (d, 3H), 1.3 (d, 3H), 2.7 (m, 1H), 2.9
(m, 2H), 3.45 (m, 1H), 4.4 (m, 1H), 5.4 (s, 1H), 5.6
(s, 1H), 6.4–6.6 (m, 2H), 7.0(m, 1H), 7.2–7.4 (m, 5H),
7.75 (d, 1H); 19F NMR (CDCl3) d ꢀ117, ꢀ140; 13C
NMR (CDCl3) d 13.4, 19.6, 32.4, 47.4, 50.7, 83.2, 104.7,
107.3, 116.3, 126.2, 128.1, 128.7, 137.3, 143, 149
(d, J=240Hz), 158 (d, J=240Hz), 167.2, 175.1;
IR (CH2CH2) 3431, 3306 (br), 3059, 2985, 2961,
2677, 1702, 1675, 1627, 1511, 1453, 1431, 1322,
1207, 1155, 1100, 1002, 839, 801 cmꢀ1; HRMS calcd
for C21H22F2N2O3 (M++1): 389.1676. Found:
389.1676.
2
20
of 115 mg (1.0mmol) of 2-cyanobutanoic acid
and
0.086 mL (1.0 mmol) oxalyl chloride, and the mixture
was stirred at room temperature overnight. A solution
of 280mg (0.84 mmol) 8b and 0.26 mL (1.9 mmol) trie-
thylamine in 10mL CH Cl2 was added. After stirring
2
overnight, 1.0g of Amberlite IRA-68 resin was added,
and the mixture was filtered through a plug of silica gel
rinsing through with EtOAc. The filtrate was stripped,
and the residue was chromatographed (55% Et2O–hex-
anes) to give 150mg product as an oil that slowly soli-
25
dified: mp 84–86 ꢁC; ½aꢃD =+26.6 (c 1.03, EtOH); H
NMR (CDCl3) d 1.0and 1.1 (2d, J=7 Hz, 3H), 1.25 (d,
J=7 Hz, 3H), 1.7–2.1 (m, 2H), 3.3–3.4 (m, 1H), 4.4 (m,
1H), 5.5 (m, 1H), 6.4 (m, 2H), 6.7–6.8 (m, 1H), 7.0(m,
1H), 7.3–7.4 (m, 5H); 19F NMR (CDCl3) d ꢀ117, ꢀ139;
13C NMR (CDCl3) d 11.0and 11.1, 13.6 and 13.9, 23.9,
40.1, 51.1 and 51.3, 83.3, 105.3 and 105.1, 108.1 and
108.0, 116.6 and 116.4, 117.9, 126.3, 128.6, 129.3 and
129.0, 136, 146.5, 147.6 (d, J=240Hz), 158.5 (d,
J=244), 154.2; IR (KBr) 3321, 2977, 2937, 2248, 1659,
1627, 1551, 1512, 1455, 1532, 1320, 1206, 1160, 1099,
794, 736, 707 cmꢀ1; HRMS calcd for C20H20F2N2O2
(M++1): 359.1572. Found: 359.1571.
1
Synthesis
of
N-[(2S,1R)-2-(2,5-dichlorophenoxy)-1-
methyl - 2- phenylethyl](2- oxo(3 - piperidyl))carboxamide
(26a). A solution of 140mg (1.0mmol) valerolactam
carboxylate 25,30 0.28 mL (2.0 mmol) triethylamine, 330
mg (1.0mmol) 8a, and 190mg (1.0mmol) EDC in 10
mL CH2CH2 was stirred at room temperature over-
night. The solution was washed with water and brine,
dried (MgSO4) and concentrated. The residue was
chromatographed on silica gel (EtOAc) to give product
as a 1:1 mixture of diastereomers as a solid: mp 54–
25
64 ꢁC; ½aꢃD =ꢀ6.8 (c 1.17, EtOH); H NMR (CDCl3) d
1.25 (m, 3H), 1.6–2.0(m, 3H), 2.3 (m, 1H), 3.2 (m, 2H),
3.4 (m, 1H), 4.4 (m, 1H), 5.4 (dd, 1H), 6.0(d, br 1H),
6.6 (d, 1H), 6.8 (m, 1H), 7.2–7.4 (m, 6H), 7.9 (2d, 1H);
13C NMR (CDCl3) d 15.0and 14.5; 22.0and 21.7, 43.8
and 43.7, 47.7 and 47.6, 52.3 and 52.1, 83.9 and 83.8,
116.6, 122.8 and 122.7, 123.2, 127.5 and 127.3, 127.5,
129.9 and 129.3, 138.4, 155.5 and 155.4, 168.8 and
168.6, 171.8; IR (CH2CH2) 3400 (br), 3052, 2984, 2929,
1
Acknowledgements
Recognition must be given to J. Michael Bonman and
G. Shawn Smith for the in vivo rice blast disease control
data described in this work.

4154
G. S. Basarab et al. / Bioorg. Med. Chem. 10 (2002) 4143–4154
15. Jordan, D. B.; Basarab, G. S.; Steffens, J. J.; Schwartz,
References and Notes
R. S.; Doughty, J. G. Biochemistry 2000, 39, 8593.
16. Lundqvist, T.; Rice, J.; Hodge, C. N.; Basarab, G. S.;
Pierce, J.; Lindqvist, Y. Structure 1994, 2, 937.
17. Wawrzak, Z.; Sandalova, T.; Steffens, J. J.; Basarab, G. S.;
Lundqvist, T.; Lindqvist, Y.; Jordan, D. B. Proteins: Struct.,
Funct. Gen. 1999, 35, 425.
1. Howard, R. H.; Valent, B. Annu. Rev. Microbiol. 1996, 50, 491.
2. Bechinger, C.; Giebel, K.-F.; Schnell, M.; Leiderer, P.;
Deising, H. B.; Bastmeyer, M. Science 1999, 285, 1896.
3. Chumley, F. G.; Valent, B. Mol. Plant-Microbe Interact.
1990, 3, 135.
4. Basarab, G. S.; Steffens, J. J.; Wawrzak, Z.; Schwartz, R. S.;
Lundqvist, T.; Jordan, D. B. Biochemistry 1999, 38, 6012.
5. Jordan, D. B.; Basarab, G. S. Bioorg. Med. Chem. Lett.
2000, 10, 23.
18. Gomez, J.; Freire, E. J. Mol. Biol. 1995, 252, 337.
19. Basarab, G. S. Unpublished results.
20. Winters, G.; Di Mola, N.; Oppici, E.; Nathansohn, G.
Farmaco Ed. Sci. 1975, 30, 620.
6. Jordan, D. B.; Zheng, Y.-J.; Locket, B. A.; Basarab, G.
S Biochemistry 2000, 39, 2276.
21. Nishio, M.; Umezawa, Y.; Hirota, M.; Takeuchi, Y. Tetra-
hedron 1995, 31, 8665.
7. Basarab, G. S.; Jordan, D. B.; Zheng, Y.-J. Org. Lett. 2000,
2, 1541.
22. Bromilow, R. H.; Chamberlain, K. Monogr.— Br. Plant
Growth Regul. Group 1989, 18, 113.
8. Zheng, Y.-J.; Basarab, G. S.; Jordan, D. B. Biochemistry
2002, 41, 820.
9. Nakasako, M.; Motoyhama, T.; Kurahashi, Y.; Yama-
guchi, I. Biochemistry 1998, 37, 9931.
10. Chen, J. M.; Xu, S. L.; Wawrzak, Z.; Basarab, G. S.;
Jordan, D. B. Biochemistry 1998, 37, 17735.
11. Jordan, D. B.; Lessen, T.; Wawrzak, Z.; Bisaha, J. J.;
Gehret, T. C.; Hansen, S. L.; Schwartz, R. S.; Basarab, G. S.
Bioorg. Med. Chem. Lett. 1999, 9, 1607.
12. Basarab, G. S.; Jordan, D. B.; Gehret, T. C.; Schwartz,
R. S.; Wawrzak, Z. Biorg. Med. Chem. Lett. 1999, 9, 1613.
13. Jennings, L. D.; Wawrzak, Z.; Amorose, D.; Schwartz, R.;
Jordan, D. B. Bioorg. Med. Chem. Lett. 1999, 9, 2509.
14. Jennings, L. D.; Rayner, D. R.; Jordan, D. B.; Okonya,
J. F.; Basarab, G. S.; Amorose, D. E.; Anaclerio, B. M.; Lee,
J. K.; Schwartz, R. S.; Whitmore, K. A. Bioorg. Med. Chem.
2000, 8, 897.
23. Lundqvist, T.; Weber, P. C.; Hodge, C. N.; Braswell,
E. H.; Rice, J.; Pierce, J. J. Mol. Biol. 1993, 232, 999.
24. Schmidpeter, A.; Weinmaier, J. Chem. Ber. 1978, 111, 2086.
25. Chou, T. S.; Chang, R. C. J. Chem. Soc., Chem. Commun.
1992, 549.
26. Mondon, A.; Oelrich, E.; Schickfluss, R. Chem. Ber. 1972,
105, 2036.
27. Wright, J. L.; Caprathe, B. W.; Downing, D. M.; Glase,
S. A.; Heffner, T. G.; Jaen, J. C.; Johnson, S. J.; Kesten, S. R.;
MacKenzie, R. G. J. Med. Chem. 1994, 37, 3523.
28. Hamashima, Y.; Ishikura, K.; Kubota, T.; Minami, K.
DE 3424780 A1, Chem. Abstr. 1985, 103, 22372.
29. Hasegawa, M.; Nakayama, A.; Yokohama, S.; Hosokami,
T.; Kurebayashi, Y.; Ikeda, T.; Shimoto, Y.; Ide, S.; Honda,
Y.; Suzuki, N. Chem. Pharm. Bull. 1995, 43, 1125.
30. Narayanan, K.; Cook, J. M. J. Org. Chem. 1991, 56,
5733.