Discovery of further pyrrolidine trans-lactams as inhibitors of human neutrophil elastase (HNE) with potential as development candidates and the crystal structure of HNE complexed with an inhibitor (GW475151)

Article abstract of DOI:10.1021/jm020881f

Described herein is a modern approach to the rapid preparation and evaluation of compounds as potential back-up drug candidates. GW311616A, 1, a derivative of pyrrolidine trans-lactams, has previously been described as a potent, orally active inhibitor of

Full text of DOI:10.1021/jm020881f

3878
J . Med. Chem. 2002, 45, 3878-3890
Discover y of F u r th er P yr r olid in e tr a n s-La cta m s a s In h ibitor s of Hu m a n
Neu tr op h il Ela sta se (HNE) w ith P oten tia l a s Develop m en t Ca n d id a tes a n d th e
Cr ysta l Str u ctu r e of HNE Com p lexed w ith a n In h ibitor (GW475151)
Simon J . F. Macdonald,*,† Michael D. Dowle,*,† Lee A. Harrison,† Geoffrey D. E. Clarke,† Graham G. A. Inglis,†
Martin R. J ohnson,† Pritom Shah,† Robin A. Smith,‡ Augustin Amour,§ Gill Fleetwood,| Davina C. Humphreys,‡
Christopher R. Molloy,‡ Mary Dixon, Rosalind E. Godward, Alan J . Wonacott,# Onkar M. P. Singh,#
Simon T. Hodgson,† and George W. Hardy†
Medicinal Chemistry 1, In Vitro Pharmacology, Systems Research, Respiratory Systems, Stevenage CEDD DMPK, and
Computational and Structural Sciences, GlaxoSmithKline Medicines Research Centre, Gunnels Wood Road,
Stevenage SG1 2NY, United Kingdom
Received March 18, 2002
Described herein is a modern approach to the rapid preparation and evaluation of compounds
as potential back-up drug candidates. GW311616A, 1, a derivative of pyrrolidine trans-lactams,
has previously been described as a potent, orally active inhibitor of human neutrophil elastase
(HNE) for the treatment of respiratory disease. These properties made it a suitable candidate
for development. Described here is the discovery of three further derivatives of pyrrolidine
trans-lactams, which fulfill the criteria required for back-up candidates 28, 29, and 32. These
include increased activity in inhibiting HNE in human whole blood (HWB) and comparable
pharmacokinetic properties, in particular clearance, in two species. To provide a rapid
assessment of clearance, cassette dosing in dog was used. Modern array techniques, including
the synthesis of mixtures, were used to synthesize compounds rapidly. Having selected three
potential compounds as back-up candidates, they were prepared as single enantiomers and
profiled in in vitro and in vivo assays and evaluated pharmacokinetically in rat and dog. These
compounds are highly potent and selective HNE inhibitors, with a prolonged pharmacodynamic
action. Pharmacokinetically, these compounds are comparable with 1 while they are more potent
in HWB. Compound 28, however, has a higher clearance. One of these compounds, 32, was
cocrystallized with HNE, and features of this structure are described and compared with the
cocrystal structure of 1 in porcine pancreatic elastase.
In tr od u ction
structure of one of these compounds in HNE is de-
scribed. We particularly highlight the value of the
cassette approach and the judicial use of modern
medicinal chemistry techniques in aiding the rapid
discovery of these back-up compounds. We also highlight
the HNE crystal structure, one of very few published.
1. Ba ck gr ou n d . A considerable body of work now
describes inhibitors of human neutrophil elastase (HNE),
a serine protease from the trypsin class, for the treat-
ment of respiratory diseases such as chronic bronchitis,
cystic fibrosis, and emphysema.1 We have introduced
pyrrolidine trans-lactams as inhibitors of HNE2 and
described some of their properties. In particular, we
have described their intracellular mechanism and the
medicinal chemistry program3 that led to the discovery4
of a drug candidate GW311616A 1 suitable for develop-
ment.5 The final phase of our program was to generate
potential back-up development compounds in case 1
failed in the drug development phase. We describe here
the goals for the back-up compounds and the medicinal
chemistry strategy and its application using a library,
pools, and array techniques. We report our development
and use of cassette dosing in dog for the rapid evaluation
of clearance. A summary of the biological data of the
three potential back-up compounds and the crystal
Although GW311616A was a suitable compound for
development, it did contain features that were regarded
as potential liabilities. From a structural viewpoint, we
were keen to remove the acrylamide functionality since
it was perceived as a potential toxicological risk because
it is a Michael acceptor. There were also concerns over
the long-term photostability of the double bond.6 Con-
sidering the biological profile of 1, the only available
measure of its potential efficacy in humans is its
potency, ex vivo in human whole blood (HWB) (0.67 µM).
This is less potent than the Merck compound7 L-694,458
(0.45 µM). We also noted that despite the excellent
bioavailability of GW311616A in rat (60%) and dog
(100%), its clearance in rat (79 mL/min/kg) and dog (28
mL/min/kg) approximated to liver blood flow (about 66
and 38 mL/min/kg, respectively).
* To whom correspondence should be addressed. Fax: +44 1438 76
36 15. E-mail: sjfm5947@gsk.com and mdd3463@gsk.com.
† Medicinal Chemistry 1.
These concerns directed our goals for the back-up
compounds. Thus, our first two goals were to improve
potency in HWB and maintain, if not reduce, the
clearance in rat and dog. To address these goals, we
could replace the piperidinocrotyl side chain and/or the
‡ In Vitro Pharmacology.
§ Systems Research.
| Respiratory Systems.
Stevenage CEDD DMPK.
Computational and Structural Sciences.
#
10.1021/jm020881f CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/03/2002

Pyrrolidine trans-Lactams as Inhibitors of HNE
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18 3879
pyrrolidine trans-lactam. We knew from previous struc-
ture-activity relationships (SAR) that most of the
enzyme inhibitory activity of 1 is derived from the highly
optimized trans-lactam core and that replacing this
feature would be difficult and time-consuming. We
therefore decided to replace the piperidinocrotyl side
chain and keep the trans-lactam core intact. We were
conscious, however, that the trans-lactam core might
define threshold levels of potency in HWB and clearance
rate whatever side chain was appended. To minimize
this risk, we rapidly explored a wide range of piperidino-
crotyl side chain replacements by using the latest
medicinal chemistry technologies (libraries, arrays, and
cassette dosing).
The results obtained (Figure 1) clearly show that the
estimates of clearance, half-life, and volume generated
using the cassette approach were in excellent agreement
with those generated from definitive pharmacokinetic
studies in which each compound was administered
alone. This experiment validated the use of this ap-
proach as a high throughput means of assessing clear-
ance of potential back-up compounds.
4. Ca ssette P r otocol Design . The data generated
from the validation study was used to optimize the
experimental design and maximize compound through-
put. Analysis of the data revealed a good inverse cor-
relation between the plasma concentration measured 1
h after administration of the cassette dose and the
plasma clearance (r2 ) 0.835). This correlation was
underpinned by the similarity in volume of distri-
bution of all of the analogues studied (3 L/kg). Given
these findings and the relatively limited range in the
physicochemical properties expected for the potential
back-ups, the plasma concentration at 1 h was used as
a measure of clearance. To correct for variability in the
system, to ensure the integrity of the data generated,
from each cassette, and to enable results to be easily
compared with the project’s lead, the racemate of 1,
compound 2, was included in each cassette as an
internal standard.
There were beneficial implications of this strategy.
Initially, the back-up compounds would provide struc-
tural novelty and thereby further patent protection. The
structural variety would also spread the risk in case of
toxicity (assuming the toxicity is associated with the side
chain). Second, we felt this synthetic program would be
relatively straightforward. Third, if we were successful
in improving potency in HWB, then this could reduce
the cost of the medicine. A final major benefit was that
if a back-up compound was required, then the trans-
lactam core intermediate available from the multi-
kilogram development synthesis of 1 could serve as a
versatile intermediate to provide large quantities of the
back-up compound quickly. This would minimize the
amount of new synthetic development work required
and reduce lead timesssubstantial benefits.
5. Med icin a l Ch em istr y Str a tegy. The medicinal
chemistry strategy was based principally upon the SAR
learned from the discovery of 1.3,4 Modification of the
piperidinocrotonyl side chain allows tuning of the phar-
macodynamics and kinetics. The carbonyl is required
for potency against HNE, the unsaturation gives a log
unit increase in activity in vivo over the saturated
analogue, and the amine, as its salt, provides solubility.
We therefore decided (Figure 2) to explore surrogates
2. Settin g Up th e Scr een s a n d Ca sca d e. We were
keen to measure HWB potency and clearance early in
the screening cascade in order to address our goals. We
anticipated that screening against HWB3 would be rela-
tively straightforward for the numbers of compounds
we would make from libraries and arrays. However,
evaluating clearance using traditional pharmacokinetic
assays has only a relatively low throughput. In general,
the stability of the compounds to in vitro hepatic
microsomes3 precluded the use of the usual in vitro
metabolic screens as predictors of plasma clearance. A
rapid in vivo assessment of clearance was therefore
required. The cassette dosing approach8 potentially
offered the high throughput required to rapidly identify
analogues with similar/superior pharmacokinetics to 1.
Sch em e 1a
We therefore defined our screening cascade with
narrowing filters as first in vitro assessment of inhibi-
tion potency against HNE and HWB followed by in vivo
clearance in the dog using cassette dosing. The most
promising compounds were then evaluated in in vivo
assays, and full pharmacokinetic profiles were obtained.
3. Ca ssette Dosin g Va lid a tion . Prior to its inclusion
in the screening cascade, the cassette dosing technique
was validated to ensure the absence of drug-drug
interactions, which, on occasion, can undermine the
predictive value of the estimated clearance rates. Five
analogues, whose plasma clearance had previously been
established in the dog, were coadministered together
with three new pyrrolidine trans-lactams. A 1.6 mg/kg
dose made up from 0.2 mg/kg of each of the eight
compounds, was administered intravenously to a male
beagle dog. (This dose level was selected to minimize
interaction potential while ensuring that detectable
plasma concentrations would be achieved.)
a
All structures racemic. Reagents: (a) Me2N(CH2)3NCNEt‚HCl
(EDC) (2 equiv), 1-hydroxybenzotriazole (HOBT) (2 equiv), 10,
MeCN, room temperature, 18 h, 39%. (b) Secondary amine (1.5
equiv), MeOH:CH2Cl2 (1:1) for 3 h and then BH4 on Amberlite
IRA-400 (2.5 mmol/g) (approx 2 equiv), room temperature, 24 h
and then polymer-bound benzaldehyde resin (0.27 mmol/g) (approx
1 equiv), 24 h.
-

3880 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18
Macdonald et al.
F igu r e 1. Details of the cassette validation experiment.
for the double bond, which would provide either π
character and/or rigidity. We also selected diverse sets
of amines based upon analyses of size (CMR), basicity
(pKa), and lipophilicity (clogP). Finally, we decided in
the first instance to prepare racemates until potential
back-up candidates had been identified. At the time,
enantiomerically pure material was in short supply.
This also safeguarded against the risk that the absolute
stereochemistry of the enantiomer with the superior
biological profile switched from that observed for 1.9
6. 2° a n d 3° Ben zyla m in essP r ep a r a tion a n d
SAR. 6.1. P r ep a r a tion . An array of 18 secondary
amines (selected to represent diverse lipophilicity and
molecular weight) was prepared using supported re-

Pyrrolidine trans-Lactams as Inhibitors of HNE
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18 3881
Sch em e 3a
F igu r e 2. Cartoon summarizing the medicinal chemistry
strategy.
a
All compounds racemic. Reagents: (a) BrCH2COCO2H (1.1
equiv), EtOH, 4 Å sieves, reflux, 18 h, 68%. (b) K2CO3 (1.8 equiv),
H2O, MeOH, reflux, 4.5 h, 86%. (c) Compound 10, EDC (1.1 equiv),
HOBT (1.1 equiv), DMF, room temperature, 4 h, 76%. (d) MeSO2Cl
(1.7 equiv), Et3N (1.9 equiv), CH2Cl2, room temperature, 5 h, 38%.
(e) Mixture of secondary amines (0.67 equiv in total), Et3N (2
equiv), CH2Cl2, room temperature, 2 d, and then workup and 1 M
HCl in Et2O, EtOAc, and then trituration with Et2O.
Sch em e 2a
Using this process in a 96 well plate in a HiTOPS
block, with a set of 80 amines selected on the basis of
diverse lipophilicity and CMR, we prepared 73 tertiary
amines of general structure 14 in yields of 10-96%. The
seven amines that failed were all very hindered (e.g.,
diisopropylamine and dicyclohexylamine). Over 80% of
the amines were prepared in at least 70% yield. All of
the products were characterized by LC/MS and some
by LC/MS/NMR.
a
All structures racemic. Reagents: (a) Compound 10 (0.77
equiv), NaHCO3 (4 equiv), CH2Cl2, room temperature, 18 h, 42%.
(b) NaI (3 equiv), Me2CO, room temperature, 18 h, 97%. (c)
R1R2NH (0.83 equiv), Amberlyst A-21 (excess), MeCN, room
temperature, 2 h. (d) SCX bond elut purification.
6.2. SAR . We prepared large numbers of benzyl-
amines with varying substitution patterns principally
because they were among the most potent examples in
the HWB assay (33% of examples were <0.5 µM in
HWB). Most were also potent against isolated elastase
(>50% of examples were <50 nM). However, in this
series, the majority of these compounds were metaboli-
cally labile as determined by high turnover in hamster
liver microsomes and low plasma concentrations in the
dog cassette. We were unable to resolve these short-
comings in this series.
7. Th ia zoles. 7.1. Mixtu r es of Th ia zoles for Ca s-
sette Dosin g. After a reasonable number of compounds
were synthesized and screened, we found the most
discriminating assay to be the measurement of the
clearance in the dog cassette. Typically, discrete com-
pounds were synthesized, evaluated in the in vitro
assays, and then mixed for the cassette study. It
occurred to us that it might be more efficient to initially
prepare a mixture, prioritize which compounds were
cleared more slowly in a cassette study, and then
synthesize a much smaller subset as discrete compounds
for further evaluation.
In general, the plasma clearance of each component
compound of any cassette mixture was estimated by
comparison of its initial concentration with the concen-
tration at various timepoints, as judged by MS. The
implication of this is that the initial concentrations of
the component compounds do not have to be the same.
This meant that neither a uniform molar ratio of the
component compounds nor a precise determination of
the molar ratio was required.
2-(tert-Butylcarbonyloxy)thioacetamide and bromo-
pyruvic acid were condensed to give the thiazole 15 in
68% yield (Scheme 3). Hydrolysis of the pivalic ester
followed by coupling to the trans-lactam 10 proceeded
smoothly. The mesylate 16 was then prepared in 38%
yield in the usual fashion; it proved to be a stable,
crystalline intermediate. The next step was alkylation
of 16 with our mixture of amines. To avoid the possibil-
agents (Scheme 1). The aldehyde 11 was prepared from
3-formylbenzoic acid and 10. It was then treated with
the primary amine for 3 h in 1:1 methanol:dichloro-
methane to preform the intermediate imine prior to
addition of the borohydride resin since the aldehyde
would otherwise be cleanly reduced to the corresponding
alcohol. Following reduction of the imine,10 the excess
amine was removed by addition of benzaldehyde resin,
which selectively scavenges the excess primary amine
from the secondary amine product. The products were
then characterized by liquid chromatography (LC)/mass
spectrometry (MS)/NMR. Twelve out of the 18 reactions
proceeded in 80+% yields, five out of 18 in 50-80%
yields, and one reaction failed. The principal impurities
were the alcohol from aldehyde reduction and ring-
opened lactam products from the reaction with excess
amine.
An alternative strategy was adopted for the array of
tertiary amines 14 (Scheme 2). m-Chloromethylbenzoyl
chloride was coupled to 10 in 42% yield, and the iodide
13 was then efficiently prepared by a Finkelstein
reaction. The iodide reacts much more quickly than the
corresponding chloride with a secondary amine (1 h),
and very little quaternary ammonium salt is generated.
To scavenge the acid produced, excess amine was
initially used, but subsequent removal of residual amine
with scavenging isocyanate resin was capricious and
required vigorous shaking. Instead, an ion exchange
resin, Amberlyst A-21 (superior to Dowex MWA1 or
Dowex 66), proved suitable. From control experiments,
acetonitrile was preferred to DCM, tetrahydrofuran
(THF), or dimethylformamide (DMF) as solvent. Even-
tually, we opted to use excess iodide rather than excess
amine, as we had developed a rapid and simple purifi-
cation process using SPE SCX cartridges (benzene
sulfonic acid-derivatized silica).11

3882 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18
Macdonald et al.
F igu r e 3. Mass spectra (A and C) and HPLC chromatographs (B and D) for the synthetic mixture 17 (A and B) and the mixture
of discretes (C and D). Peaks corresponding to each component of the mixture are present in both sets of spectra. Note for the
synthetic mixture: the presence of unreacted phenylpiperazine and the corresponding relative decrease in peak height in the
HPLC for the phenylpiperazine analogue 17-5.
Ta ble 1. Plasma Concentrations at 1 h (Relative to the
Standard 2) of Synthetic Mixtures and Mixtures of Discretes
and Results from a Third Cassette, Which Included 17-2 and
ity of dialkylation of primary amines, we selected five
secondary amines each with differing molecular weights
to facilitate mass spectroscopic analysis (dimethylamine,
pyrrolidine, piperidine, 1-methylpiperazine, and 1-phen-
yl piperazine). A 50% excess of mesylate 16 (with respect
to the total number of moles of amine) was treated with
an equimolar ratio of dimethylamine, pyrrolidine, pi-
peridine, methylpiperazine, and phenylpiperazine in
dichloromethane with triethylamine as an acid scaven-
ger. At the end of the reaction, the solution was washed
with aqueous bicarbonate and water, dried, and con-
centrated. The residue now contained the products,
unreacted mesylate, and side products formed by de-
composition of the mesylate in solution (formation of
quaternary ammonium salts was very slow under these
conditions). This residue was then dissolved in ether and
treated with hydrogen chloride leading to precipitation
of the hydrochlorides. Trituration with ether then
allowed removal of the unreacted mesylate and nonbasic
side products. The LC/MS analysis (Figure 3, spectra
A and B) shows the presence of each compound in
mixture 17; however, the reaction with phenylpiper-
azine was only partially complete as indicated by the
presence of peaks for the phenylpiperazine adduct 17-5
and unreacted phenylpiperazine. Comparison of the
spectra A and B (Figure 3) for the synthetic mixture
with the spectra C and D (Figure 3) for the mixture of
discretes shows an excellent correlation. (The retention
time differences in the high-performance liquid chro-
17- 5
plasma concnsa
cassette 1 cassette 2
synthetic mixture of
amine
mixture
discretes
cassette 3b
0.4
17-1 Me2N
17-2 pyrrolidine
17-3 piperidine
17-4 1-methylpiperazine
17-5 1-phenylpiperazine
1.2
0.5
0.3
1.2
0.8
1.0
0.4
0.2
0.9
0.3
0.9
a
Plasma concentrations relative to the standard 2 at 1 h.
Cassette 3 included 17-2, 17-5, and eight other compounds
b
(structure and data not included).
matography (HPLC) spectra B and D are due to solvent
noise and interference.)
The results from evaluation of the synthetic mixture
17 in the cassette (cassette 1) are in good agreement
with those from an equimolar mixture of the same
compounds, which had been prepared and purified as
discretes (cassette 2) (Table 1). The exception is the
phenylpiperazine analogue 17-5 where the relative
plasma concentrations at 1 h for the synthetic mixture
is 0.8 as compared with 0.3 for the mixture of discretes.
From the analytical data (Figure 3), it is known that
there was less of this compound present in the synthetic
mixture. However, when the phenylpiperazine 17-5 was

Pyrrolidine trans-Lactams as Inhibitors of HNE
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18 3883
Sch em e 4a
from ethyl acetimidate and serine methyl ester (Scheme
6) followed by oxidation. The resulting oxazole 23
sublimes readily. Bromination15 of the methyl group
gave mixtures of starting material, R,R-dibromo mate-
rial, and the monobromo product 24 in low to moderate
yields. Originally, we had wished to couple the corre-
sponding bromo-acid to the lactam 10; unfortunately,
attempted hydrolysis of 24 under basic or acidic condi-
tions gave complex mixtures. However, in an array
format, reaction of 24 with secondary amines16 followed
by basic hydrolysis, conversion to the acid chloride, and
coupling with 10 proceeded to give the target amines
in fair to good yields.17 The pyrrolidine derivative shown
here was prepared using this sequence in 45% yield.
a
All compounds racemic. Reagents: (a) MeOH, concentrated
HCl, room temperature, 20 h, 70%. (b) NBS (1.0 equiv), (BzO)2
(0.37 equiv), hν, CCl4, reflux, 5 h, 26%. (c) KOH (10 equiv), H2O,
room temperature, 2 h, 68%. (d) Compound 10 (0.7 equiv), EDC
(1.2 equiv), MeCN, room temperature, 20 h, 50%. (e) Cyclopropyl-
amine (6 equiv) (see ref 12), NaI (1.5 equiv), K2CO3 (6 equiv),
CH2Cl2, room temperature, 1-3 d, 63%.
8.4. P yr a zin es, Azetid in es, a n d Oxa zoles: SAR.
The most striking feature of the SAR of the pyrazines,
azetidines, and oxazoles was that only one example
emerged from each array as a potential candidate
having the desired profile: 20, 22, and 25. While the
pyrazines as a class were extraordinarily potent vs
HWB, with 80% having IC50 values <0.5 µM, only the
cyclopropylamine derivative 20 emerged with acceptable
clearance and efficacy in the in vivo hamster assay.
Most of the azetidines and oxazoles failed on potency
grounds either against isolated HNE or in HWB. Only
the neopentyl-azetidine 22 and pyrrolidine-oxazole 25
have the desired profile.
included as a component of a mixture of discretes in a
third cassette experiment (cassette 3), the relative
plasma concentration was 0.9. It is apparent that the
original value for the mixture of discretes was in-
accurate rather than the value from the synthetic
mixture (cf. the values for the pyrrolidine derivative
17-2 across all three cassettes). This illustrates the point
made above that precise equimolar mixtures are not
necessary using this process. Having validated this
protocol, we subsequently prepared two further series
of novel thiazoles as synthetic mixtures (without pre-
paring them as discretes) for evaluation in the cassette.
7.2. SAR fr om th e Th ia zoles. Over two-thirds of the
thiazoles prepared inhibited HNE with IC50 values of
<20 nM, and half had IC50 values <0.5 µM in HWB.
However, the vast majority of the more potent thiazoles
in this series proved to have high turnover in hamster
liver microsomes and/or low plasma concentrations in
the dog cassette experiments. The few compounds
evaluated in vivo in hamster were mediocre.
8.5. Syn th eses of En a n tiom er ica lly P u r e Ca n d i-
d a tes. From the biological evaluation of the above
series, the racemates 20, 22, and 25 were selected for
further study. Our first task was to prepare large
quantities of these compounds in enantiomerically pure
form.
8.5.1. P r ep a r a tion of th e En a n tiom er ica lly P u r e
P yr a zin e 28 a n d Azetid in e 29. To reduce the cost of
synthesis, we wanted to streamline the processing of
the costly enantiomerically pure trans-lactam moiety 27
(Scheme 7).5 We therefore chose to couple the trans-
lactam to a cyclopropylamine-derived pyrazine acid. In
a three step sequence, the pyrazine acid 18 was first
brominated to give a mixture of mono and R,R-dibromo
acids. After workup, this crude mixture was treated
with cyclopropylamine. This amine only reacts with the
monobromide. The desired cyclopropylaminopyrazine
acid was then extracted into dilute hydrochloric acid,
which was then basified with sodium hydroxide and
reacted with di-tert-butyl dicarbonate to furnish the tert-
butyl carbamate 26 in 13% yield for the three steps.
Conditions for coupling this acid 26 to the trans-lactam
27, which worked both on a large and a small scale, were
eventually found with the use of O-(7-azabenzotriazol-
1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophos-
phate (HATU) in acetonitrile. This proceeded in 89%
yield on a 20 g scale. Deprotection of the carbamate and
hydrochloride formation were achieved with hydrogen
chloride in dioxan. Recrystallization from water-2-
propanol gave 28 in greater than 99% purity as judged
by HPLC. The enantiomerically pure neopentyl-azeti-
dine 29 was prepared using the enantiomerically pure
trans-lactam 27 via the same route as that described
(Scheme 5).
8. P yr a zin es, Azetid in es, a n d Oxa zoles. 8.1. Ar -
r a y P r ep a r a tion s: P yr a zin es. Synthesis of an array
of pyrazine analogues (which have a predicted clogD of
over 2 log units lower than the corresponding benzene
derivatives) was straightforward (Scheme 4). The com-
mercially available pyrazine acid 18 was esterified and
brominated under photolytic conditions to give the mono
bromide in 26% yield together with 22% of the R,R-
dibromide. After ester hydrolysis, the bromo-acid 19 was
coupled to 10 and the array was prepared by treatment
with either the secondary amines12 and sodium iodide
in dichloromethane or the amine hydrochlorides, sodium
iodide, and potassium carbonate in dichloromethane.
Exemplified here is the cyclopropyl derivative 20 pre-
pared in 63% yield.
8.2. Ar r a y P r ep a r a tion s: Azetid in es. The prepa-
ration of the azetidine array commenced from 3-azeti-
dinecarboxylic acid (Scheme 5). Protection of the azeti-
dine N as its benzylcarbamate allowed efficient coupling
with 10 in excellent yield. Removal of the benzyl
carbamate is best achieved by hydrogenolysis in acetic
acid using platinum oxide as catalyst. Hydrogenolysis
with palladium hydroxide as catalyst gives complex
mixtures. Reductive amination with the amine 21 as
its acetate with a range of aldehydes, followed by
formation of the hydrochlorides, gave the products in
moderate to good yield13 (exemplified for the neopentyl
derivative 22).
8.5.2. P r ep a r a tion of th e En a n tiom er ica lly P u r e
Oxa zole 32. The synthesis used for the racemic pyr-
rolidine-oxazole 25 (Scheme 6) involved a low-yielding
8.3. Ar r a y P r ep a r a tion s: Oxa zoles. The oxazole
ring system was prepared using standard procedures14

3884 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18
Macdonald et al.
Sch em e 5a
a
Reagents: (a) CBZCl (1.05 equiv), 0.5 M K2CO3, dioxan, room temperature, 18 h, 92% crude. (b) Compound 10, EDC (1.2 equiv), 5:1
CH2Cl2:MeCN, room temperature, 7 h, 97% crude. (c) 5% PtO2, H2, AcOH, room temperature, 8 h, 98% crude. (d) t-BuCHO (2 equiv),
Na(AcO)3BH (2 equiv), trace AcOH, CH2Cl2, room temperature, 18 h, 73%. (e) 1 M HCl in Et2O, CH2Cl2 86%.
Sch em e 6a
Sch em e 8a
a
All compounds racemic. Reagents: (a) Et3N (1 equiv), CH2Cl2,
reflux, 5.5 h, 77%. (b) DBU (0.25 equiv), CCl4:pyridine:MeCN (1.2:
1.8:2.3), room temperature, 18 h, 65%. (c) NBS (1.1 equiv),
(PhCO2)2 (0.05 equiv), CCl4, hν (200 W tungsten lamp), reflux, 1.5
h, 15-34%. (d) Pyrrolidine (1.1 equiv), K2CO3 (1.5 equiv), MeCN,
room temperature, 6-48 h, see 17 for other amines. (e) NaOH (3
equiv), H2O, dioxan, room temperature, 5 h. (f) (COCl)2 (5 equiv),
DMF (catalytic), CH2Cl2, room temperature, 1.5 h. (g) Compound
10 (0.8 equiv), NaHCO3 (4 equiv), CH2Cl2, room temperature, 4
h, 45% (see ref 17 for yields of other amines); purify then 1 M HCl
in Et2O (1.1 equiv), CH2Cl2, quantitative.
a
Reagents: (a) ClCHNMe2 + Cl- (0.5 equiv), CHCl3, -5 to 20
°C, 3 h, and then 10% AcOH in H2O, room temperature, 3 h, 73%.
(b) BrCH2CN (as solvent), Rh2(OAc)4 (1.5 mol %), 75 °C, 8 h, 38%.
(c) Pyrrolidine (3.0 equiv), MeCN, room temperature, 45 min, 57%.
(d) K2CO3 (1.0 equiv), EtOH, H2O, reflux, 4 h. (e) Compound 27
(1 equiv), EDC (2 equiv), HOBT (1 equiv), MeCN, room temper-
ature, 20 h. (f) 1 M HCl in Et2O (1.1 equiv), room temperature, 15
min and then recrystallized from Me2CO, 65% for two steps.
Ta ble 2. Kinetic Parameters Measured at pH 7.4a
Sch em e 7a
c
kon
koff
Ki
t1/2
inhibitor
k
obs/[I] (s-1 M-1
)
(× 10-6 s-1
)
(nM)b
(h)
1 GW311616
28 GW447631 28 735
29 GW469002
32 GW475151
757
6630
251 668
6306
2.07
3.53
10.9
4.85
0.31
93
0.014 54.5
720
3483
1.73
0.16
17.6
39.7
30 505
a
HNE (3.4 nM) was assayed with 1 mM of the chromogenic
substrate MeOSuc-AAPV-pNA (Sigma, M4765) at 30 °C in 0.1 M
Hepes (pH 7.4), 0.3 M NaCl, 10% (v/v) DMSO, and 0.03% (v/v)
TritonX-100 containing 0-1 µM inhibitor. See refs 21 and 22 for
b
further details. Ki ) koff/kon, koff and kon are, respectively, the
c
dissociation and association rate constants.22 t1/2 ) 0.693/koff
represents the half-life time of the acyl-enzyme.
ponification of the ester, and coupling with 27 using
EDC in the presence of hydroxybenzotriazole. After the
hydrochloride was formed, recrystallization from ace-
tone gave analytically pure 32 in 65% yield.
a
Reagents: (a) Br2 (1.1 equiv), AcOH, 80 °C, 1 h. (b) Cyclopro-
pylamine (2.0 equiv), Et3N (1.0 equiv), room temperature, 20 h.
(c) Boc2O (1.2 equiv), 1,4-dioxan, room temperature, 44 h, 13% for
three steps. (d) Compound 27 (1 equiv) (see ref 5), HATU (1.1
equiv), DIPEA (2.0 equiv), MeCN, room temperature, 2 h, 89%.
(e) 4 M HCl, 1,4-dioxan, room temperature, 2 h, and then
9. Su m m a r y of P oten cy a n d In Vitr o a n d In Vivo
Da ta . The previous section highlights the use of recent
chemical technologies in our medicinal chemistry pro-
gram. Emerging from the large number of derivatives
that were prepared were three that satisfied our criteria
for back-up candidates. These were the cyclopropyl-
pyrazine 28 (GW447631A),20 the neopentylazetidine
29 (GW469002A),21 and the pyrrolidinyloxazole 32
(GW475151A),21 which represent a diverse set of side
chains. (For further details on the biological activity of
many of these derivatives, see refs 20 and 21.)
i
recrystallize from 5% H2O in PrOH (82%).
bromination, which was not amenable on a large scale.
An alternative multigram process featuring a rhodium-
catalyzed insertion reaction was thus developed.18 Ethyl
diazoacetate was formylated with the Vilsmeier reagent
to afford, after hydrolysis, the aldehyde 30 in 73% yield
(Scheme 8). Treatment of 30 with catalytic rhodium
acetate generated the carbene, which was reacted with
bromoacetonitrile19 to afford the oxazole 31 in 38% yield
after chromatography. This oxazole was then readily
converted into 32 by treatment with pyrrolidine, sa-
Table 2 compares the kinetic parameters determined
for the three potential back-up compounds with those
for 1 (GW311616A). In each case, the acylated enzyme

Pyrrolidine trans-Lactams as Inhibitors of HNE
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18 3885
Ta ble 3. Biological Profiles of the Back-Up Compoundsa
parameter
1 GW311616
28 GW447631
29 GW469002
32 GW475151
HNE IC50 (µM)b,c
0.022
0.011
0.037
0.010
Selectivity Assays IC50 (µM)
chymotrypsin,b,d
trypsinb,e
4.2
>100
>100
0.54
>100
>100
31
>100
>100
3
>100
>100
cathepsin Gb,d
Whole Blood IC50 (µM)c
human
dog
0.67
0.35
0.15
0.26
0.19
0.17
0.25
0.39
Hamster In Vivoc
(% Elastase Inhibition at 2.2 mg/kg po)
bronchoalveolar lavage
bone marrow
81
48
90
69
83
56
93
65
Dog In Vivoc
(% Elastase Inhibition in Peripheral Blood at 2 mg/kg po)
6 hours
3 days
100
97
99
92
99f
92f
100
100
a
b
For comparison of values for 1 with literature compounds, see ref 4. Values represent a mean of three experiments. All values are
within 30% of the mean. These values are after 15 min of preincubation. c See Experimental Section for practical details. See ref 2 for
d
experimental details. e See ref 22 for experimental details. f Dose level was 1.1 mg/kg po.
dissociates very slowly, so that the inhibitors behave
essentially irreversibly. Compound 28 is particularly
potent, having a picomolar Ki value. As with 1, selectiv-
ity against other serine proteases has been maintained
(Table 3), and these new compounds are even more
potent than the original lead in the HWB assay,
demonstrating inhibition of intracellular elastase. Po-
tent inhibition was also shown in the analogous dog
whole blood assay.
This potency was maintained in vivo after oral dosing
in both hamsters and dogs (Table 3). In the hamster,
inhibition of intracellular elastase in neutrophils recov-
ered by bronchoalveolar lavage is thought to reflect
inhibition within cells in the circulation. These cells
have then been recruited to the lung with IL-8. Inhibi-
tion in the bone marrow, on the other hand, is believed
F igu r e 4. Effect of orally dosed 32 (GW475151) on intracel-
lular elastase in peripheral dog blood. Elastase and myleo-
to give rise to prolonged duration of action of the
compounds, since any neutrophils released into the
peroxidase activities were determined in lysed white cells from
circulation would already have their complement of
elastase inhibited. (The level of inhibition in the bone
marrow is lower than that in bronchoalveolar lavage in
each case.) This hypothesis is supported by the dog data
(Figure 4) and suggests a common mechanism between
the species. At the lowest dose of compound (0.22 mg/
kg), elastase inhibition in circulating neutrophils rapidly
reaches greater than 50% but has returned to control
values by 24 h. At the 2 mg/kg dose level, elastase
activity is virtually eliminated for at least 5 days. At
the intermediate dose, inhibition is between the two.
Each of the three back-up compounds was potent in
vivo.
10. P h a r m a cok in etics (See Ta ble 4). All four
compounds 1, 28, 29, and 32 exhibit high plasma
clearance in rat, which equates to or exceeds liver blood
flow (69-225 mL/min/kg). These clearances combined
with the moderate to high volumes of distribution (2.6-
7.5 L/kg) result in short plasma half-lives of less than
or equal to 1 h. Despite the high clearance observed with
these compounds, oral bioavailability ranging from 24
to 60% is evident, which suggests that clearance is
predominantly extrahepatic. While the pharmaco-
kinetics of the compounds studied are broadly similar,
the plasma clearance of 28 is approximately 3-fold
higher than the other inhibitors and the oral exposure
peripheral blood (2 dogs/dose level). Ratios were compared with
the mean of three predose samples for each dog.
is notably lower. As an oral therapy, it is therefore
considered to be pharmacokinetically inferior to the
other compounds.
The pharmacokinetic profile in the dog broadly re-
flects that seen in the rat with clearance values that
approach/exceed liver blood flow (23-50 mL/min/kg),
moderate volumes of distribution (1.9-3.5 L/kg), and
short plasma half-lives of typically less than 2 h. The
oral bioavailability again suggests that clearance is
predominantly extrahepatic, and the data (two dogs per
compound) shows that essentially complete oral bio-
availability is achieved with all four compounds. How-
ever, as evident in the rat, the oral exposure of 28 is
lower than that of the other three compounds and its
plasma clearance is higher. Therefore, on the basis of
its pharmacokinetic profile in both the rat and the dog,
this compound is considered to be inferior to the other
elastase inhibitors studied.
11. Cocr yst a l St r u ct u r e of H NE w it h 32
(GW475151). The X-ray crystal structure of 32 in a
complex with HNE at pH 4.0 has been determined at
2.0 Å resolution. The electron density that covers the
whole of the inhibitor molecule provides a clear basis

3886 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18
Macdonald et al.
Ta ble 4. Pharmacokinetic Parameters of 1 (GW311616A),a 28 (GW447631A),a 29 (GW469002A),a and 32 (GW475151A)a from iv and
po Dosing in Rat at 2 mg/kg and Dog at 0.66 mg/kg
1 (GW311616A)
rat
dogb
28 (GW447631A)
ratc
dog
29 (GW469002A)
rat dog
32 (GW475151A)
rat dog
param
iv
po
iv
po
iv
po
iv
po
205
0.25
220 147
iv
po
iv
po
iv
po
iv
po
Cmax (ng/mL)
Tmax (h)
AUC∞ (ng/mL h) 424
232
152
0.7
135
142
130
0.9
147
138
0.05 0.5
0.03 0.25
148
0.3
225
5.0
0.03 0.75
0.03 0.25 0.03 0.5
263 405 430
0.8
64
0.4
482
0.8
69
217
1.2
488 417 392
1.8
23
93
1.3
286
1.0
39
308
1.4
T1/2 (h)
1.1
79
7.5
1.5
28
3.5
1.6
0.4
50
1.9
0.4
2.1
0.4
85
2.6
Clp (mL/min/kg)
Vd area (L/kg)
F (%)
4.9
3.5
3.4
60
∼100
43
50-120
45
90
24
∼100
All of the data were gathered on the hydrochlorides. The 1 dog data were normalized from a 2 to 0.66 mg/kg dose. c The 29 rat data
a
b
were normalized from a 2.4 to 2 mg/kg dose.
F igu r e 5. Crystal structure of HNE complexed with GW475151. The ring-opened ligand covalently linked to Ser195 is shown in
thick bonds with the electron density of the omit map in a green/blue chicken wire net representation. Key residues of HNE are
drawn in thin bonds with residues numbered as for chymotrypsin. S1 is the primary enzyme specificity pocket. The alternative
positions for the side chain of Phe215 are denoted “a” and “b”.
for interpretation of ligand interactions within the active
site (Figure 5). The ring-opened trans-lactam is co-
valently bonded to Ser195; the S1 pocket is occupied by
the isopropyl group while the ester carbonyl of the
inhibitor points toward the main chain nitrogen atoms
of Gly193 and Ser195 (the oxyanion hole), but distances
are too long for good hydrogen bonding. His57 is
displaced from its catalytic position and is replaced by
two water molecules; this will contribute to the stability
of the complex. The pyrrolidine group makes hydro-
phobic contacts with both Phe215 and Leu99B. All
residues of the active site and binding pocket match
closely for the two independent molecules in the asym-
metric unit of the crystal apart from the side chain of
Phe192. This group makes contact with the methyl-
sulfonamide group of the inhibitor and has alternate
conformations (both shown in Figure 5), which arise
from differences in intermolecular contacts within the
crystal. As compared with the structure of HNE in
complex with a chloromethyl ketone inhibitor (1HNE),23
the loop containing residues Val99, Asn99A, Leu99B,
and Leu100 is displaced by up to 1 Å in the main chain.
This probably occurs as a consequence of the hydro-
phobic contact between the pyrrolidine group of the
ligand and the side chain of Leu99B, which moves ∼1.5
Å.
A comparison with the binding of 1 to porcine pan-
creatic enzyme (PPE)5 indicates significant differences
(Figure 6). Although the methylsulfonamide groups are
essentially superimposed, the ring-opened lactam is

Pyrrolidine trans-Lactams as Inhibitors of HNE
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18 3887
F igu r e 6. Stereoview of the superposition of the active site residues of the HNE-GW475151 complex (atom colors) with that of
the porcine pancreatic elastase-GW311616 complex (green). The difference in the orientation of the ring-opened lactam leads to
a marked shift in the pendant atoms with the loss of the hydrogen bond from the carbonyl to the main chain NH of residue Val
216. Labeled residues are those of the HNE complex.
× 100 mL). The combined extracts were dried (MgSO4) and
rotated anticlockwise relative to 1, bringing the inhibitor
filtered, and the solvent was removed in vacuo to give a brown
chain down and out in the pocket. As a consequence,
oil, which was purified by flash column chromatography on
there is no hydrogen bond from the carbonyl of the
silica (Merck 9385), with 100:8:1 dichloromethane/methanol/
ligand to the main chain NH of residue 216; instead,
acetic acid as the eluent. The required fractions were evapo-
rated to dryness in vacuo to give a tan-colored solid, which
this group points out into solvent.
12. Con clu sion . Described here is the strategy and
discovery of three back-up candidates to 1, which are
suitable for development. We successfully improved
HWB potency and with 29 and 32 maintained accept-
able clearance rates in rat and dog. Through the use of
modern array techniques, we were able to prepare
rapidly a substantial number of compounds for biologi-
cal evaluation. Evaluation of clearance using the dog
cassette protocol allowed 136 compounds to be screened
in 7 months of which 94 were screened in 2 months,
demonstrating its value.
was stirred vigorously in 5:1 cyclohexane/diethyl ether until
finely divided. The solid was filtered off and dried in vacuo to
give the title compound as an orange/brown solid (16.65 g).
TLC (100:8:1 dichloromethane/methanol/acetic acid) Rf ) 0.31.
MS MH+ (found) 294, MH+ (calcd) 294.
Cyclopr opyl-[5-(6S-isopr opyl-4-m eth an esu lfon yl-5-oxo-
h exa h yd r o-(3a R,6a S)-p yr r olo[3,2-b]p yr r ole-1-ca r bon yl)-
p yr a zin -2-ylm eth yl]ca r ba m ic Acid ter t-Bu tyl Ester . Com-
pound 275 (11.36 g), 26 (13.53 g), and O-(7-azabenzotriazol-
1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate (19.3
g) were stirred in acetonitrile (260 mL) at room temperature,
and N,N-diisopropylethylamine (16 mL) was added. After the
mixture was stirred for 2 h, the solvent was removed in vacuo,
and the residue was diluted with dichloromethane (250 mL)
and washed with 1 M sodium carbonate solution (250 mL).
The aqueous phase was extracted with dichloromethane (3 ×
150 mL). The combined organics were washed with 1 M sodium
carbonate solution (50 mL), dried (MgSO4), filtered, and the
solvent was evaporated in vacuo to leave a yellow-brown solid.
The solid was purified by flash column chromatography (Merck
9385 silica) using 50% ethyl acetate/cyclohexane as eluent, and
the required fractions were evaporated to dryness in vacuo to
give the title compound as a white foam (21.55 g); [R]D20 +69.5°
(c ) 0.8, MeCN). MS MH+ (found) 522, MH+ (calcd) 522.
Exp er im en ta l Section
For descriptions of equipment and techniques, see refs 2 and
5. For the preparative details of array and library compounds,
see refs 20 and 21. For descriptions of assay conditions, refer
to the published procedures cited in the references.
P r ep a r a tion of 28. 5-[(ter t-Bu toxyca r bon yl-cyclop r o-
p yl-a m in o)m eth yl]p yr a zin e-2-ca r boxylic Acid 26. Bro-
mine was added to a stirred suspension of 2-methylpyrazin-
5-carboxylic acid (60 g) in acetic acid (300 mL). The reaction
mixture was then heated to 80 °C for 1 h. The solvent was
evaporated in vacuo, and the residue was partitioned between
ethyl acetate (250 mL) and 2 M aqueous HCl (250 mL). The
aqueous phase was extracted with further ethyl acetate (5 ×
250 mL), the combined organics were washed with 2 M HCl
(100 mL) and saturated brine solution (100 mL), dried
(MgSO4), and filtered, and the solvent was evaporated in vacuo
to leave a brown solid. The solid was stirred in acetonitrile
(900 mL), and triethylamine (60 mL) and cyclopropylamine
(30 mL) were added. After the mixture was stirred for 20 h at
room temperature, further cyclopropylamine (30 mL) was
added and the mixture was stirred for a further 15 min. The
volatiles were evaporated in vacuo, and the residue was
partitioned between ethyl acetate (200 mL) and 2 M aqueous
HCl (300 mL). The organic phase was extracted with further
2 M HCl (4 × 200 mL), and the combined aqueous extracts
were washed with ethyl acetate (50 mL), cooled in an ice bath,
and basified with 10 M aqueous sodium hydroxide (120 mL).
The solution was washed with ethyl acetate (3 × 200 mL) and
diethyl ether (200 mL), and remaining organic volatiles were
removed in vacuo to give a brown aqueous solution. To the
solution was added 1,4-dioxane (500 mL) and di-tert-butyl
dicarbonate (71 g), and the mixture was stirred at room
temperature for 20 h. Further di-tert-butyl dicarbonate (10 g)
was added, and stirring was continued for a further 24 h. Citric
acid (85 g) was added to the stirred mixture before it was
extracted with ethyl acetate (2 × 200 mL + 3 × 150 mL + 2
(3S,3a S,6a R)-4-(5-Cyclop r op yla m in om eth yl-p yr a zin e-
2-ca r bon yl)-3-isop r op yl-1-m eth a n esu lfon yl-h exa h yd r o-
p yr r olo[3,2-b]p yr r ol-2-on e Hyd r och lor id e 28. A solution
of cyclopropyl-[5-(6S-isopropyl-4-methanesulfonyl-5-oxo-hexa-
hydro-(3aR,6aS)-pyrrolo[3,2-b]pyrrole-1-carbonyl)pyrazin-2-yl-
methyl]carbamic acid tert-butyl ester (21.53 g) and 4.0 M HCl
in 1,4-dioxan (200 mL) was stirred at room temperature for 2
h. The solvent was removed in vacuo to give an off-white solid.
The solid was recrystallized from hot 5% water/2-propanol (2.3
L) to give 28 (15.54 g) as a white solid; mp 183-185 °C; TLC
(Silica, eluent 200:8:1 dichloromethane:ethanol:0.880 am-
20
monia) Rf ) 0.21; [R]D +51.3° (c ) 0.9, 1:1 H2O/MeCN). CD:
λmax 250.2 nm (∆E -1.34 M-1 cm-1), λmax 285.4 nm (∆E +0.99
M-1 cm-1), (MeCN/H2O). 1H NMR (DMSO-d6, 400 MHz): δ
9.80 (bs, 2H), 9.04 (s, 1H), 8.88 (s, 1H), 4.56 (s, 2H), 4.03-
3.83 (m, 3H), 3.62 (t, J ) 11 Hz, 1H), 3.32 (s, 3H), 3.13-3.07
(m, 1H), 2.94 (m, 1H), 2.80 (m, 1H), 2.33 (m, 1H), 2.09 (m,
1H), 1.22 (d, J ) 7 Hz, 3H), 1.02 (d, J ) 7 Hz, 3H), 0.93 (m,
2H), 0.76 (m, 2H). MS MH+ (found) 422.19, MH+ (calcd) 422.19.
Anal. Found C, 47.4; H, 6.4; N, 14.3; S, 6.5; Cl, 7.8; water,
4.9%. C19H27Cl N5O4S‚HCl‚1.3H2O requires C, 47.4; H, 6.4; N,
14.6; S, 6.7; Cl, 7.4; water, 4.9%.
P r ep a r a tion of 29. 1-Ben zyloxyca r bon yl Azetid in -3-oic
Acid . To azetidine-3-carboxylic acid (0.50 g), potassium car-
bonate (0.82 g), dioxan (5 mL), and water (10 mL) was added
benzyl chloroformate (0.74 mL) at room temperature, and the
mixture was stirred for 6 h under nitrogen. Piperazine (5

3888 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18
Macdonald et al.
drops) was then added. After 0.5 h, the dioxan was removed
in vacuo and the residue was diluted with 2 M hydrochloric
acid (25 mL). After it was extracted with ethyl acetate (35 mL),
the organic layer was washed with brine (10 mL) and dried
(MgSO4). Solvent removal in vacuo gave the title compound
P r ep a r a tion of 32. 2-P yr r olid in -1-ylm eth yl-oxa zole-4-
ca r boxylic Acid Eth yl Ester . To a stirred solution of
2-(bromomethyl)oxazole-4-carboxylic acid ethyl ester (43.9 g)
in acetonitrile (300 mL) was added pyrrolidine (15.7 mL). After
it was stirred for 10 min, more pyrrolidine (7.8 mL) was added.
After a further 30 min, the solvent was removed in vacuo to
leave an orange oil. The oil was partitioned between 1 M
sodium carbonate (400 mL) and dichloromethane (500 mL),
and the phases were separated. The organic phase was washed
with water (100 mL), dried (MgSO4), filtered, and the solvent
was removed in vacuo to give the title compound as an orange
oil (24.0 g). MS MH+ (found) 225, MH+ (calcd) 225.
+
+
as a colorless oil (1.07 g). MS MNH4 (found) 253, MNH4
(calcd) 253.
(3S,3a S,6a R)-3-(6-Isop r op yl-4-m eth a n esu lfon yl-5-oxo-
h exa h yd r o-p yr r olo[3,2-b]p yr r ole-1-ca r bon yl)a zetid in e-
1-ca r boxylic Acid Ben zyl Ester . A mixture of 275 (0.90 g),
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide (0.91 g), and
acetonitrile (12 mL) was stirred for 5 min under nitrogen after
which 1-benzyloxycarbonyl azetidin-3-oic acid (0.90 g) in
acetonitrile (30 mL) was added. After 18 h, the solvent was
removed in vacuo and the residue was treated with 2 M
hydrochloric acid (30 mL). The mixture was extracted with
ethyl acetate (90 mL). The organic layer was washed with 2
M hydrochloric acid (30 mL), saturated sodium bicarbonate
solution (2 × 30 mL), and brine (30 mL), dried (MgSO4), and
concentrated in vacuo. The residue was purified by flash
chromatography on silica (Merck 9385) eluting with 10% and
then 20% acetonitrile in dichloromethane to give the title
compound as a colorless foam (0.80 g). MS MH+ (found) 464,
MH+ (calcd) 464.
2-P yr r olid in -1-ylm eth yl-oxa zole-4-ca r boxylic Acid /2-
P yr r olid in -1-ylm eth yl-oxa zole-4-ca r boxylic Acid P ota s-
siu m Sa lt. Potassium carbonate (14.8 g) was added to a
solution of 2-pyrrolidin-1-ylmethyl-oxazole-4-carboxylic acid
ethyl ester (24.0 g) in ethanol (150 mL) and water (150 mL).
The reaction mixture was stirred at reflux for 4 h. The solvent
was removed in vacuo. The orange/brown residue was azeo-
troped with toluene (×3) and then dried in vacuo. The solid
obtained was stirred vigorously with ether (100 mL) and
filtered off before it was dried in vacuo to give a mixture of
the title compound together with potassium bicarbonate as a
brown solid (34.5 g). This material was used without further
purification. MS MH+ (found) 197, MH+ (calcd) 197.
(3S,3a S,6a R)-4-(Azet id in e-3-ca r b on yl)-3-isop r op yl-1-
m e t h a n e su lfon yl-h e xa h yd r o-p yr r olo[3,2-b ]p yr r ol-2-
on e Aceta te. Platinum(IV) oxide (1.0 g) and acetic acid (50
mL) were prehydrogenated for 0.5 h and then (3S,3aS,6aR)-
3-(6-isopropyl-4-methanesulfonyl-5-oxohexahydro-pyrrolo[3,2-
b]pyrrole-1-carbonyl)azetidine-1-carboxylic acid benzyl ester
(0.80 g) in acetic acid (50 mL) was added. After it was
vigorously stirred under hydrogen for 19 h, the reaction was
filtered through a 3 M Empore C18 extraction disk cartridge
and concentrated in vacuo. The residual gum was dried by
azeotropic distillation with toluene (3 × 50 mL), and the
residue was triturated with ether (2 × 20 mL) to give the title
compound as a white powder (0.63 g). MS MH+ (found) 330,
MH+ (calcd) 330.
(3S,3a S,6a R)-3-Isop r op yl-1-m eth a n esu lfon yl-4-(2-p yr -
r olidin -1-ylm eth yl-oxazole-4-car bon yl)h exah ydr o-pyr r olo-
[3,2-b ]p yr r ol-2-on e H yd r och lor id e 32. 2-Pyrrolidin-1-
ylmethyl-oxazole-4-carboxylic acid/2-pyrrolidin-1-ylmethyl-
oxazole-4-carboxylic acid potassium salt (32.2 g) was added
rapidly to a stirred solution of 1-hydroxybenzotriazole (13.0
g) in acetonitrile (350 mL). A solution of 27 (21.7 g)5 and 1-(3-
dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (37.0
g) in acetonitrile (70 mL) was then added, and the reaction
mixture was stirred for 20 h. The solvent was removed in
vacuo, and the residue was partitioned between dichloro-
methane (900 mL) and 1.0 M sodium carbonate solution (600
mL). The aqueous phase was separated and extracted with
dichloromethane (150 mL). The combined organics were
washed with brine (250 mL), dried (MgSO4), and concentrated
in vacuo to leave a brown solid. The solid was purified by flash
column chromatography (Merck 9385 silica; eluent di-
chloromethane:ethanol:ammonia 150:8:1 to 135:8:1) to give a
cream solid (29.3 g). The solid was dissolved in dichloro-
methane (150 mL) and treated with 1.0 M hydrogen chloride
in ether (75 mL). The solvent was removed in vacuo to leave
a solid, which was redissolved in dichloromethane (150 mL)
and again treated with 1.0 M hydrogen chloride in ether (75
mL). The solvent was removed in vacuo to leave a solid, which
was recrystallized from acetone to give 32 (26.3 g) as a white
solid; mp 156-158 °C; TLC (Silica; dichloromethane:ethanol:
ammonia 100:8:1; double elution) Rf ) 0.66. 1H NMR (400
MHz; DMSO-d6): δ 8.78 (s, 1H), 4.68 (s, 2H), 4.13 (ddd, J )
11, 11, 7 Hz, 1H), 4.08 (dd, J ) 11, 10 Hz, 1H), 3.80 (ddd, J )
12, 10.5, 5.5 Hz, 1H), 3.60 (m, 2H), 3.55 (dd, J ) 12, 10.5 Hz,
1H), 3.31 (s, 3H), 3.20 (m, 2H), 3.03 (dd, J ) 12, 2.5 Hz, 1H),
2.88 (md, J ) 2.5 Hz, 1H), 2.34 (m, 1H), 2.12 (m, 1H), 1.96 (m,
4H), 1.19 (d, J ) 7 Hz, 3H), 0.98 (d, J ) 7 Hz, 3H). Contains
0.16 Mol % acetone. IR (KBr diffuse reflectance): νmax 3633,
3474, 3149, 3102, 2956, 2882, 2668, 2576, 2475, 1747, 1709,
1639, 1634, 1567, 1442, 1380, 1347, 1161, 1146, 967, 810, 547
cm-1. MS MH+ (found) 425.186372, MH+ (calcd) 425.185867
(error 1.2 ppm). Anal. Found: C, 48.65; H, 6.39; N, 11.41; S,
6.19; Cl, 7.13%. C19H28N4O5S‚HCl‚0.75H2O‚0.2Me2CO requires
C, 48.43; H, 6.57; N, 11.53; S, 6.60; Cl, 7.29%.
(3S,3a S,6R)-4-[1-(2,2-Dim eth yl-p r op yl)a zetid in e-3-ca r -
bon yl]-3-isopr opyl-1-m eth an esu fon yl-h exah ydr o-pyr r olo-
[3,2-b]p yr r ol-2-on e 29. To (3S,3aS,6aR)-4-(azetidine-3-car-
bonyl)-3-isopropyl-1-methanesulfonyl-hexahydro-pyrrolo[3,2-
b]pyrrol-2-one acetate (0.625 g) in dichloromethane (90 mL)
were added acetic acid (9 drops, ca. 100 µL), sodium tri-
acetoxyborohydride (0.680 g), and pivalaldehyde (0.24 mL).
After 18 h, the mixture was concentrated in vacuo and the
residue was treated with ethyl acetate (90 mL). This was
extracted with 1 M hydrochloric acid (3 × 30 mL), and the
combined extracts were taken to pH 8 with solid sodium
bicarbonate. The mixture was extracted with dichloromethane
(3 × 30 mL), and the combined extracts were dried (MgSO4)
and concentrated in vacuo. The residue was purified by flash
chromatography on silica gel (Merck 9385) eluting with 200:
8:1 dichloromethane:ethanol:ammonia to afford a gum (0.466
g). The gum was dissolved in dichloromethane (20 mL), treated
with 1 M hydrogen chloride in ether (5 mL), and concentrated
in vacuo. Recrystallization from 7% water in 2-propanol (75
mL) gave 29 as a white powder (0.298 g); mp 194-195 °C;
TLC SiO2 (100:8:1 dichloromethane:ethanol:ammonia): Rf
1
0.56. H NMR (CD3OD, 57 °C, 400 MHz): δ 4.40 (t, 2H), 4.28
(bs, 2H), 3.88 (m, 1H), 3.83-3.71 (m, 3H), 3.59 (sextet, 1H),
3.36 (t, 1H), 3.23 (s, 3H), 3.10 (s, 2H), 3.03-2.96 (m, 1H), 2.95-
2.90 (m, 1H), 2.52 (quintet, 1H), 2.12 (quintet, 1H), 1.27 (d,
3H), 1.04 (s, 9H), 1.02 (d, 3H). IR (KBr diffuse reflectance):
3479, 3415, 2966, 2608, 1738, 1661 cm-1. MS MH+ (found)
400.226259, MH+ (calcd) 400.227004 (1.9 ppm). Anal. Found:
C, 50:80; H, 7.76; N, 9.37; S, 6.97%; C19H33N3O4S‚0.8H2O HC1
requires C, 50.67; H, 7.97; N, 9.33; S, 7.12%. HPLC (Inertsil
ODS2-IK5, 40 °C, 215 nm, 1 mL/min, solvent A H2O + 0.1%
H3PO4, solvent B 95% MeCN/H2O + 0.1% H3PO4, gradient 0%
B for 2 min, to 100% B in 40 min, 100% B for 10 min) retention
time 14.247 min. HPLC (Chiralcel OD-H, 25 °C, 215 nm, 1
mL/min, isocratic 60% EtOH in heptane) retention time 7.796
min.
Cr ysta llogr a p h y. A 4-fold excess of 32, GW475151, was
added to HNE (Calbiochem) in 10 mM sodium acetate, pH 5.0,
at a protein concentration of 10 mg/mL. Crystals of the complex
grew in hanging drops containing equal volumes of protein
and a precipitant solution containing 1.1-1.2 M (NH4)2SO4
(AS), 100 mM citrate, pH 3.8-4.0, at room temperature. After
they were harvested into 1.2 M AS and gradual addition of
1.0 M AS, pH 4.0, with 20% glycerol, crystals were flash-frozen
in a nitrogen gas stream cooled to 100 K and stored in liquid

Pyrrolidine trans-Lactams as Inhibitors of HNE
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18 3889
nitrogen for subsequent data collection. Crystallographic data
collection was carried out on Station 9.6 at the Daresbury-
SRS using a QUANTUM4 CCD detector system (Area Detector
Systems Corporation, San Diego, CA.) and evaluated by the
DENZO/SCALEPACK program package.24 The structure was
determined by molecular replacement using AmoRe25 with
coordinates of the starting model from the deposited structure
1HNE23 with the ligand and water molecules removed. The
resulting structure, which has two molecules in the asym-
metric unit of P4322, was refined using cycles of REFMAC26,27
alternating with rebuilding the structure, modeling of the
ligand, and addition of solvent molecules, which was carried
out with the aid of the QUANTA program (Accelrys Inc., San
Diego, CA). Coordinates of the HNE-GW475151 structure have
been deposited in the Protein Data Bank, ID 1h1b.
Assa ys. HWB Ela sta se In h ibition Assa y. Triplicate
aliquots of fresh, heparinized HWB (200 µL) were added to
appropriately diluted samples (10 µL) of compounds under test.
Control samples (six replicates) contained water in place of
compound. Samples were mixed thoroughly by pipette and
were then incubated for 30 min at 37 °C. Cold red cell lysis
buffer (750 µL of 155 mM ammonium chloride, 0.12 mM
ethylenediaminetetraacetic acid (EDTA), 10 mM potassium
bicarbonate, pH 7.4) was then added. Tubes were capped,
inverted several times, and maintained at 4 °C for 15 min,
inverting every 5 min. After the tubes were centrifuged at 250g
for 10 min at 4 °C, the resulting pelleted cells were washed.
The wash was with saline (300 µL), followed by centrifugation
at 100g for 10 min at 4 °C. Pellets were washed twice more,
before resuspension of the final cell pellet in buffer (200 µL of
100 mM Tris, 300 mM NaCl, 1% (w/v) HTAB, pH 8.6). Samples
were stored at -20 °C. After the samples were freeze-thawed
four times, elastase activity was determined by a colorimetric
assay in 50 mM Tris, 150 mM NaCl, 0.6 mM MeO-Succ-Ala-
Ala-Pro-Val-pNA at pH 8.6, measuring the rate of increase in
absorbance at 405 nm.
In Vitr o Assa ys of HNE. Assay contents: 50 mM Tris/
HCl (pH 8.6), 150 mM NaCl, and 11.8 nM purified HNE.
Suitable concentrations of compound under test were diluted
with water from a 10 mM stock solution in DMSO. Values
above were final concentrations after the addition of substrate
solution. The mixture above was incubated for 15 min at 30°C
at which time the remaining elastase activity was measured
for 10 min in a BioTek 340i plate-reader, after the addition of
0.6 mM MeO-succinyl-alanyl-alanyl-prolyl-valyl-p-nitroanilide.
The rate of increase in absorbance at 405 nm was proportional
to elastase activity. Enzyme activity was plotted against
concentration of inhibitor and an IC50 was determined using
curve-fitting software.
HNE Kin etics. HNE (3.4 nM) was assayed with 1 mM of
the chromogenic substrate MeOSuc-AAPV-pNA (Sigma, M4765)
at 30 °C in 0.1 M Hepes (pH 7.4), 0.3 M NaCl, 10% (v/v) DMSO
and 0.03% (v/v) TritonX-100 containing 0-1 µM inhibitor. See
ref 22 for further details.
Selectivity Assa ys. The trypsin, chymotrypsin, and cath-
epsin G assays were carried out essentially as described in
ref 2 (chymotrypsin and cathepsin G) and ref 22 (trypsin).
cation. Elastase and myeloperoxidase assays were then per-
formed on these samples to assess the efficacy of the com-
pounds and to normalize for neutrophil recovery.
In Vivo Activity of In h ibitor s of HNE Or a lly in Dog.
Elastase and myeloperoxidase activities were determined in
lysed white cells from peripheral blood (two dogs/dose level).
Ratios were compared with the mean of three predose samples
for each dog.
Dog Ca ssette: Va lid a tion . The test set of eight compounds
was administered intravenously via the femoral vein at a dose
level of 0.2 mg/kg body weight per compound to a female beagle
dog. Blood was sampled from the jugular vein over the time
period 5 min to 24 h postdose. Plasma samples were prepared
by solid phase extraction. The extracts were analyzed by HPLC
and detected by single ion monitoring on a Finnigan TSQ mass
spectrometer. The analytical column was a Capital Inertsil
ODS3 silica 5 µm, 50 mm × 2.1 mm. The mobile phase was
0.1% formic acid in water:methanol (70:30). The flow rate was
0.4 mL/min.
Dog Ca ssette: Sta n d a r d Assa y. The set of 10 novel
compounds plus the internal standard 2 was administered
intravenously via the femoral vein at a dose level of 0.2 mg/
kg body weight per compound to a female beagle dog. Blood
was sampled from the jugular vein over the time period 5 min
to 2 h postdose and analyzed as above.
Ack n ow led gm en t. We thank Professor Helquist for
his comments on the rhodium-catalyzed reaction.
Refer en ces
(1) For recent reviews, leading references, and patents, see (a) Metz,
W. A.; Peet, N. P. Inhibitors of Human Neutrophil Elastase as
a Potential Treatment for Inflammatory Diseases. Expert Opin.
Ther. Pat. 1999, 9, 851-868. (b) Skiles, J . W.; J eng, A. Y.
Therapeutic Promises of Leukocyte Elastase and Macrophage
Metalloelastase Inhibitors for the Treatment of Pulmonary
Emphysema. Expert Opin. Ther. Pat. 1999, 9, 869-895. (c)
Leung, D.; Abbenante, G.; Fairlie, D. P. Protease Inhibitors:
Current Status and Future Prospects. J . Med. Chem. 2000, 43,
305-341. (d) Veale, C. A.; Bernstein, P. R.; Bohnert, C. M.;
Brown, F. J .; Bryant, C.; Damewood, J . R., J r.; Earley, R.;
Feeney, S. W.; Edwards, P. D.; Gomes, B.; Hulsizer, J . M.;
Kosmider, B. J .; Krell, R. D.; Moore, G.; Salcedo, T. W.; Shaw,
A.; Silberstein, D. S.; Steelman, G. B.; Stein, M.; Strimpler, A.;
Thomas, R. M.; Vacek, E. P.; Williams, J . C.; Wolanin, D. J .;
Woolson, S. Orally Active Trifluoromethyl Ketone Inhibitors of
Human Leukocyte Elastase. J . Med. Chem. 1997, 40, 3173-
3181. (e) Ohmoto, K.; Yamamoto, T.; Okuma, M.; Horiuchi, T.;
Imanishi, H.; Odagaki, Y.; Kawabata, K.; Sekioka, T.; Hirota,
Y.; Matsuoka, S.; Nakai, H.; Toda, M.; Cheronis, J . C.; Spruce,
L. W.; Gyorkos, A.; Wieczorek, M. Development of Orally Active
Nonpeptidic Inhibitors of Human Neutrophil Elastase. J . Med.
Chem. 2001, 44, 1268-1285.
(2) Macdonald, S. J . F.; Belton, D. B.; Buckley, D. M.; Spooner, J .
E.; Anson, M. S.; Harrison, L. A.; Mills, K.; Upton, R. J .; Dowle,
M. D.; Smith, R. A.; Molloy, C.; Risley, C. Syntheses of trans-
5-Oxohexahydropyrrolo[3,2-b]pyrroles and trans-5-Oxohexa-
hydrofuro[3,2-b]pyrroles (Pyrrolidine trans-Lactams and trans-
Lactones): New Pharmacophores for Elastase Inhibition. J . Med.
Chem. 1998, 41, 3919-3922.
(3) Macdonald, S. J . F.; Dowle, M. D.; Harrison, L. A.; Spooner, J .
E.; Shah, P.; J ohnson, M. R.; Inglis, G. G. A.; Clarke, G. D. E.;
Belton, D. J .; Smith, R. A.; Molloy, C. R.; Dixon, M.; Murkitt,
G.; Godward, R. E.; Skarzynski, T.; Singh, O. M. P.; Kumar, K.
A.; Hodgson, S. T.; McDonald, E.; Hardy, G. W.; Finch, H.;
Fleetwood, G.; Humphreys, D. C. Intracellular Inhibition of
Human Neutrophil Elastase by Orally Active Pyrrolidine-trans-
lactams. Bioorg. Med. Chem. Lett. 2001, 11, 243-246.
(4) Macdonald, S. J . F.; Dowle, M. D.; Harrison, L. A.; Shah, P.;
J ohnson, M. R.; Inglis, G. G. A.; Clarke, G. D. E.; Smith, R. A.;
Humphreys, D.; Molloy, C. R.; Amour, A.; Dixon, M.; Murkitt,
G.; Godward, R. E.; Padfield, T.; Skarzynski, T.; Singh, O. M.
P.; Kumar, K. A.; Fleetwood, G.; Hodgson, S. T.; Hardy, G. W.;
Finch, H. The Discovery of a Potent, Intracellular, Orally
Bioavailable, Long Duration Inhibitor of Human Neutrophil
Elastase-GW311616A a Development Candidate. Bioorg. Med.
Chem. Lett. 2001, 11, 895-898.
In Vivo Activity of In h ibitor s of HNE: Or a l In Vivo
Mod el Usin g IL-8-In d u ced Lu n g In filtr a tes for th e As-
sessm en t of In tr a cellu la r Ela sta se In h ibition in Ha m -
ster s. Adult hamsters (100-150 g) were randomized into
groups (n ) 4) and fasted overnight. Under gaseous anaes-
thetic (3% isofluorane), animals were dosed orally with 1 mL/
100 g water as vehicle or containing predissolved compounds.
Either at the same time, or subsequently under anaesthetic,
animals were dosed intratracheally with 1 µg of recombinant
human IL-8 in 100 µL of sterile saline. Six hours after IL-8
dosing, animals were sacrificed using intraperitoneal pento-
barbitone. The lungs were lavaged with 2 × 2.5 mL sterile
saline, and femurs were removed by dissection. Intracellular
elastase was prepared from neutrophils collected by lavage and
from femoral bone marrow. This was achieved by sonication
of the neutrophils and centrifugation to yield intracellular
granules. These were disrupted by freeze/thawing and soni-
(5) Macdonald, S. J . F.; Clarke, G. D. E.; Dowle, M. D.; Harrison,
L. A.; Hodgson, S. T.; Inglis, G. G. A.; J ohnson, M. R.; Shah, P.;
Upton, R. J .; Walls, S. B. A Flexible, Practical, and Stereo-

3890 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18
Macdonald et al.
selective Synthesis of Enantiomerically Pure trans-5-Oxohexa-
hydropyrrolo[3,2-b]pyrroles (Pyrrolidine-trans-lactams), a New
Class of Serine Protease Inhibitors, Using Acyliminium Meth-
odology. J . Org. Chem. 1999, 64, 5166-5175.
(18) A third process for synthesising this oxazole has recently been
published by our development chemistry colleagues. Cardwell,
K. S.; Hermitage, S. A.; Sjolin, A. A New Method for the
Formation of 2,4-Disubstituted Oxazoles: Internal Transfer of
Oxidation State Through a Molecular Framework. Tetrahedron
Lett. 2000, 41, 4239-4242.
(19) (a) Connell, R. D.; Tebbe, M.; Gangloff, A. R.; Helquist, P.;
Aakermark, B. Rhodium-Catalyzed Heterocycloaddition Route
to 1,3-Oxazoles as Building Blocks in Natural Products Synthe-
sis. Tetrahedron 1993, 49, 5445-5459. (b) Connell, R. D.; Tebbe,
M.; Helquist, P.; Aakermark, B. Direct Preparation of 4-Carbo-
ethoxy-1,3-oxazoles. Tetrahedron Lett. 1991, 32, 17-20. (c)
Gangloff, A. R.; Aakermark, B.; Helquist, P. Generation and Use
of a Zinc Derivative of a Functionalized 1,3-Oxazole. Solution
of the Virginiamycin/Madumycin Oxazole Problem. J . Org.
Chem. 1992, 57, 4797-4799.
(20) Clarke, G. D. E.; Dowle, M. D.; Finch, H.; Harrison, L. A.; Inglis,
G. G. A.; J ohnson, M. R.; Macdonald, S. J . F.; Shah, P.; Smith,
R. A. Pyrrolopyrrolone Derivatives as Inhibitors of Neutrophil
Elastase, and Their Preparation and Use. PCT Int. Appl. 1999,
WO9912933.
(21) Dowle, M. D.; Finch, H.; Harrison, L. A.; Inglis, G. G. A.;
J ohnson, M. R.; Macdonald, S. J . F. Pyrrolopyrrolone Derivatives
as Inhibitors of Neutrophil Elastase. PCT Int. Appl. 1999,
WO9912931.
(22) Values were determined by the kinetic methodology described
in Weir, M. P.; Bethell, S. S.; Cleasby, A.; Campbell, C. J .;
Dennis, R. J .; Dix, C. J .; Finch, H.; J hoti, H.; Mooney, C. J .; Patel,
S.; Tang, C. M.; Ward, M.; Wonacott, A. J .; Wharton, C. W. Novel
Natural Product 5,5-trans-Lactone Inhibitors of Human R-Throm-
bin: Mechanism of Action and Structural Studies. Biochemistry
1998, 37, 6645-6657.
(23) Navia, M. A.; McKeever, B. M.; Springer, J . P.; Lin, T.-Y.;
Williams, H. R.; Fluder, E. M.; Dorn, C. P.; Hoogsteen, K.
Structure of Human Neutrophil Elastase in Complex with a
Peptide Chloromethyl Ketone Inhibitor at 1.84A Resolution.
Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 7-11.
(6) In fact, chemical experiments suggested that the acrylamide was
a very poor Michael acceptor and the E-geometry of the double
bond was photostable.
(7) Finke, P. E.; Shah, S. K.; Fletcher, D. S.; Ashe, B. M.; Brause,
K. A.; Chandler, G. O.; Dellea, P. S.; Hand, K. M.; Maycock, A.
L.; Osinga, D. G.; Underwood, D. J .; Weston, H.; Davies, P.;
Doherty, J . B. Orally Active â-Lactam Inhibitors of Human
Leukocyte Elastase. 3. Stereospecific Synthesis and Structure-
Activity Relationships for 3,3-Dialkylazetidin-2-ones. J . Med.
Chem. 1995, 38, 2449-2462.
(8) Frick, L. W.; Adkison, K. K.; Wells-Knecht, K. J .; Woollard, P.;
Higton, D. M. Cassette Dosing: Rapid In Vivo Assessment of
Pharmacokinetics. Pharm. Sci. Technol. Today 1998, 1, 12-18.
(9) Such a switch had been observed in an earlier series in the
program.
(10) Yoon, N. M.; Kim, E. G.; Son, H. S.; Choi, J . Borohydride
Exchange Resin, a New Reducing Agent for Reductive Amina-
tion. Synth. Commun. 1993, 23, 1595-1599.
(11) Application of the reaction mixture to the SCX SPE cartridge
was followed by eluting unreacted iodide with EtOH (3 volumes),
followed by elution of product with 5% NH3 in EtOH. Any
quaternary ammonium salt is retained on the column. With this
process, 24 compounds were purified in 2.5 h.
(12) The amines used were morpholine, methylpiperazine, di-
cyclohexylamine, piperidine, pyrrolidine, cyclopropylamine,
(iPr)2CHNH2, dimethylamine, phenylpiperazine, and dibutyl-
amine in 24-93% yields.
(13) The aldehydes used and yields of hydrochlorides were cyclopro-
pylaldehyde (37%), 2,6-dichlorobenzaldehyde (12%), nPrCHO
(37%), PhCHO (68%), and pivalaldehyde (73%).
(14) Videnov, G.; Kaiser, D.; Kempster, C.; J ung, G. Synthesis of
Naturally Occurring, Conformationally Restricted Oxazole- and
Thiazole-Containing Di- and Tripeptide Mimetics. Angew. Chem.
Int. Ed. Engl. 1996, 35, 1503-1506.
(15) (a) Fujita, E. Organic Synthesis Utilizing Thiazolidine and the
Related Heterocycles. Heterocycles 1984, 21, 41-60. (b) Nagao,
Y.; Yamada, S.; Fujita, E. Synthetic Studies on Virginiamycin
M2: Functionalization at the 2-Methyl Group of 4-tert-Butoxy-
carbonyl-2-methyl-1,3-oxazole. Tetrahedron Lett. 1983, 24, 2291-
2294.
(16) The amines and yields were piperidine (100%), dimethylamine
(100%), pyrrolidine (88%), morpholine (81%), 1-phenylpiperazine
(98%), dibutylamine (86%), methylcyclopropylamine (59%), di-
isopropylamine (36%), and dicyclohexhylamine (26%).
(17) The yields over two steps (hydrolysis and coupling) for the amine
derivatives were piperidine (52%), dimethylamine (51%), pyr-
rolidine (45%), morpholine (93%), 1-phenylpiperazine (47%),
dibutylamine (78%), methylcyclopropylamine (21%), diisopro-
pylamine (6%), and dicyclohexhylamine (47%).
(24) Otwinowski, Z.; Minor, W. Processing of X-ray Diffraction Data
Collected in Oscillation Mode. In Methods in Enzymology; Carter,
C. W., J r., Sweet, R. M., Eds.; New York: Academic Press, 1997;
Vol. 276, pp 307-326.
(25) Navaza, J . AmoRe - an Automated Package for Molecular
Replacement. Acta Crystallogr. 1994, A50, 157-163.
(26) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J . Refinement of
Macromolecular Structures by the Maximum-Likelihood Meth-
odol. Acta Crystallogr. D 1997, 53, 905-921; CCP4 (Collabora-
tive Computational Project 4).
(27) The CCP4 Suite: Programs for Protein Crystallography. Acta
Crystallogr. D 1994, 50, 760-763.
J M020881F