Synthesis, inhibitory activity towards human leukocyte elastase and molecular modelling studies of 1-carbamoyl-4-methyleneaminoxyazetidinones

Article abstract of DOI:10.1016/S0223-5234(00)00111-2

Some monocyclic β-lactam derivatives of type 3 (MAOAs) in which the leaving group (LG) on the C(4) is a methyleneaminoxy moiety, were synthesised and tested in vitro and in vivo for their inhibitory activity towards human leukocyte elastase (HLE). Some compounds showed an appreciable in vitro inhibitory activity against this enzyme. Effects on the anti-HLE activity due to the nature of the substituents R and R₁ present on their LG were observed and rationalised by means of molecular modelling techniques. The results of in vivo pharmacological tests indicated that MAOAs, while showing an inhibitory activity on the haemorrhage induced by HLE, did not exhibit any effects due to the R and R₁ substituents. (C) 2000 Editions scientifiques et medicales Elsevier SAS.

Full text of DOI:10.1016/S0223-5234(00)00111-2

Eur. J. Med. Chem. 35 (2000) 53−67
© 2000 Éditions scientifiques et médicales Elsevier SAS. All rights reserved
S0223523400001112/FLA
53
Original article
Synthesis, inhibitory activity towards human leukocyte elastase and molecular
modelling studies of 1-carbamoyl-4-methyleneaminoxyazetidinones
Bruno Macchiaa*, Daniela Gentilia, Marco Macchiaa, Francesca Mamonea, Adriano Martinellia,
Elisabetta Orlandinia, Armando Rosselloa, Giovanni Cercignanib, Rita Pierottib, Marcello Allegrettic,
Cinzia Astic, Gianfranco Casellic
aDipartimento di Scienze Farmaceutiche, Facoltà di Farmacia, Università degli Studi di Pisa, Via Bonanno, 6, 56100 Pisa, Italy
bLaboratorio di Biochimica, Dipartimento di Fisiologia e Biochimica, Università degli Studi di Pisa, Via S. Maria, 55, 56100 Pisa, Italy
cDompè S.p.A., Chemistry and Pharmacology Laboratories, Via Campo di Pile, 67100 L’Aquila, Italy
Received 6 May 1999; revised 7 July 1999; accepted 15 July 1999
Abstract – Some monocyclic â-lactam derivatives of type 3 (MAOAs) in which the leaving group (LG) on the C(4) is a methyleneaminoxy
moiety, were synthesised and tested in vitro and in vivo for their inhibitory activity towards human leukocyte elastase (HLE). Some
compounds showed an appreciable in vitro inhibitory activity against this enzyme. Effects on the anti-HLE activity due to the nature of the
substituents R and R1 present on their LG were observed and rationalised by means of molecular modelling techniques. The results of in vivo
pharmacological tests indicated that MAOAs, while showing an inhibitory activity on the haemorrhage induced by HLE, did not exhibit any
effects due to the R and R1 substituents. © 2000 Éditions scientifiques et médicales Elsevier SAS
elastase inhibitor / â-lactam monocyclic inhibitor / 1-carbamoyl-4-methyleneaminoxyazetidinone derivative / oxime derivative
1. Introduction
destruction of connective tissue, which is typical of
various diseases. An excessive proteolysis of elastin by
HLE has been considered to be responsible for several
chronic pulmonary diseases such as emphysema, acute
respiratory distress syndrome, cystic fibrosis and chronic
bronchitis [11–14]. Compounds capable of inhibiting
HLE may restore the balance between the free enzyme
and endogenous antiproteinases, which is found to be
altered in these pathologies, and therefore may have
considerable therapeutic potential. Several studies have
focused upon the development of potent, selective syn-
thetic HLE-inhibitors, whether of a proteic [15–20] or
non-proteic nature [21–33]. Among these last types of
inhibitors, some monocyclic â-lactams 1, substituted on
the C-4 by an arylethereal moiety, have proved to possess
HLE specificity and biopharmacological properties useful
for their clinical utilisation (figure 1) [28–30].
Human leukocyte elastase (HLE) is a serine proteinase
which is produced by neutrophils and is released from
their azurophilic granules upon physiological stimula-
tion [1]. This proteinase is one of the degradative en-
zymes involved in phagocytosis and plays a role in the
homeostatic regulation systems. Its extracellular activity
is normally regulated by circulating plasma and tissue-
localised natural protease inhibitors. The principal role of
intracellular HLE appears to be the degradation of foreign
proteins ingested by leukocytes during phagocytosis,
whereas the main target for extracellular elastase appears
to be certain structural proteins such as elastin [2],
laminin [3], fibronectin [4, 5], collagen and proteoglycan
[6–10].
A pathological derangement of one of the natural
controls of the activity of HLE may lead to its hyperex-
pession, which in turn may bring about an uncontrolled
For these compounds, as well as for other classes of
non-proteic inhibitors, the anti-elastase activity appears
to be linked to the presence, on their reactive nucleus, of
an appropriate leaving group (LG) (figure 2). The expul-
*Correspondence and reprints

54
Figure 1.
sion of this group leads to the formation of a new reactive
electrophilic species, in some cases composed of an
imine (E-I), capable of binding to the His-57 of the active
site of the enzyme, thus creating an enzyme–inhibitor
complex (E-I*) stabilised by the presence of two covalent
bonds between the two groups of the catalytic site, from
which the enzyme can no longer be regenerated [21–33].
This type of mechanism suggested to us the idea of
studying analogues of monocyclic â-lactams 1 in which
their LG (i.e. the C-4 aryloxy group) is replaced by a
substituent with different leaving-group characteristics,
such as the oximate moiety. Accordingly some methyl-
eneaminoxyazetidinones (MAOAs) of types 2 and 3 were
synthesised and tested for their inhibitory activity to-
wards HLE. While the type 2 derivatives proved to be
inactive towards this enzyme, the 3,3-diethyl derivatives
3 proved to possess a good anti-HLE inhibitory activity,
maintaining the same selectivity as the aryloxy type 1
compounds towards this enzyme (figure 3) [34].
These promising results obtained with a limited num-
ber of compounds suggested to us the idea of gaining
insight into the structure–activity relationships in this
class of â-lactams by studying the effects on the inhibi-
tory activity towards HLE of substitution on the methyl-
eneaminoxy molecular portion, which represents the LG
in this class of derivatives (figure 2). On the basis of the
observation that the MAOA 3b (R = Me), previously
studied, proved to possess a better inhibitory activity with
respect to the unsubstituted MAOA 3a (R = H), we
decided to evaluate first the effects on the anti-HLE
activity of the presence on the methyleneaminoxy group
of substituents R with the greatest steric hindrance, while
maintaining on the same molecular portion the phenyl
group, which is present as the second substituent in the
MAOAs 3. A substituent R bulkier than the methyl group
might interact with a suitable hydrophobic region of the
enzyme active site, thus increasing the forces involved in
the binding of the enzyme–inhibitor complex and conse-
quently the activity. On this basis, we synthesised the new
MAOAs 3c and 3d in which the substituent R is respec-
tively an ethyl and a phenyl group. Unfortunately, these
new MAOAs, tested in vitro for their inhibitory activity
towards HLE, appeared to be less active than the methyl-
substituted analogue 3b. Therefore, at this point, we
decided to verify the effects on the anti-HLE activity of
the introduction of a polar group, R1, in the para position
of the phenyl group which is present as the second
substituent on the methyleneaminoxy molecular portion
in the more active MAOAs 3a and 3b previously studied.
As R1 substituents for these new MAOAs, we chose
carboxyl (3g and 3h) and the O-ethyl-N-morpholino
group (3i and 3l) (negatively and positively charged
under physiological conditions, respectively) and a hy-
drogen bonding acceptor, like the fluorine atom of 3e and
3f (figure 3). This structural modification might improve
the aqueous solubility of the new compounds, increasing
their in vivo activity through an increase in their bioavail-
ability. Also for type 1 compounds, the presence of an
acidic or basic group, as in L680,833 and L694,458, at
present the leaders in this class of compounds, had
appeared to improve the biopharmacological properties.
2. Chemistry
Compounds 3c–f and 3h–l were prepared as outlined in
the scheme indicated in figure 4. Base-catalysed treat-
ment of 3,3-diethyl-4-acetoxyazetidin-2-one 11 [35] with
the benzophenone oxime 12d [36] or the E [37] oximes
12c [38], 12e [39], 12f [39] and 12h–l gave the corre-
sponding â-lactam intermediates 13c–l which, by reac-
tion with benzylisocyanate in the presence of triethyl-
amine (TEA), afforded the benzylurea derivatives

55
Figure 2.
3c–14h, respectively. Treatment of 14h with trifluoro-
acetic acid and anisole afforded the corresponding free
acid 3h.
gen (see Experimental section). For these compounds, in
agreement with findings for aryl-alkyl or aryl-hydrogen
analogues of unsymmetrical oxime ethers which present a
syn relationship between the oximic oxygen and the alkyl
or the hydrogen iminic substituent [41], the protonation
of the oximic nitrogen determines an appreciable shift of
the signal of the nearest protons of the alkyl groups or the
signals of the hydrogen, linked to the adjacent iminic
carbon atom, to a higher field, depending on the quantity
of acid added to the oximic compound. The E geometry
around the C=N double bond of the oxime ether interme-
diates 13c–l and 14h and therefore of the MAOAs 3c–l
was assigned on the basis of the knowledge of the E
configuration of the starting oximes, bearing in mind that
these oximes have proved to be configurationally stable
under the different synthetic procedures used for the
preparation both of 13c–l and 14h and of the desired new
azetidinone derivatives 3c–f and 3h–l (see Experimental).
The E oxime 12h was synthesised by reaction of
4-acetylbenzoic acid 15 with diphenyldiazomethane and
then with hydroxylamine hydrochloride in the presence
of potassium carbonate. The E oximes 12i and 12l were
synthesised by Mitsunobu [40] reaction of phenols 17
and 18 with 4-(2-hydroxyethyl)morpholine and then by
treatment of the carbonyl derivative 19 and 20 with
hydroxylamine hydrochloride in basic media (figure 5).
The syn relationship around the double bond of the
methyleneaminoxy molecular portion, between the oxime
oxygen and the alkyl substituent linked to the iminic
carbon atom, of the newly synthesised oximes 12h–l, was
1
assigned by H-NMR study of the shielding induced by
an anisotropic solvent (C6D6) on 12h–l in the unproto-
nated and protonated forms of the oximic nitro-

56
Figure 3.
3. Biological results
activity increased from 3a (R = H, kinact/KI = 30 900
M–1s–1) to 3b (R = Me, kinact/KI = 44 000 M–1s–1). In
contrast, the activity decreased slightly going to 3c (R =
Et, kinact/KI = 8 000 M–1s–1) and to 3d (R = Ph, kinact/KI
= 3 400 M–1s–1), which represents the worst derivative in
the whole new series.
The newly synthesised MAOAs 3c–f and 3h–l, like
those previously described, 3a, 3b and 3g, were evaluated
for their ability to inhibit HLE-catalysed hydrolysis of the
substrate N-methoxysuccinyl-Ala-Ala-Pro-Val-p-nitro-
anilide. Table I shows the inhibitory activity of MAOAs
3 towards this enzyme as a concentration-independent
second-order rate constant kinact/KI. Table I also shows
the inhibitory activity observed in the same test for two
type 1 drugs used as references: compound 1a, already
used in our preceding paper [34], and compound
L680,833, one of the two leaders in this class of deriva-
tives. All the MAOAs 3 tested were found to possess an
appreciable inhibitory activity, with values of kinact/KI
ratio ranging from 3 400 M–1s–1 for 3d to 98 300 M–1s–1
for 3f.
Among the MAOAs 3e–l, substituted on the phenyl
group of the methyleneaminoxy molecular portion, 3f, 3h
and 3l, which contain a methyl on the iminic carbon of
the LG, showed, in all cases, better values of the kinact/KI
ratio than the corresponding analogues 3e, 3g and 3i, in
which a hydrogen is present in the place of the methyl on
the same carbon atom. This difference is particularly
marked in the para-fluorine substituted compounds
where the methyl-substituted 3f, the most interesting
compound in the whole series, presents a value of
kinact/KI = 98 300 M–1s–1, about sixteen times higher than
the unsubstituted 3e analogue.
In the MAOAs 3 unsubstituted on the phenyl group
linked to the methyleneaminoxy molecular portion, the

57
Figure 4. (a) Oxime, NaOH, Me2CO/H2O 1:1, r.t., 1h; (b) BnNCO, DMAP, TEA, CH2Cl2, r.t., 24h; (c) TFA/anisole 1:2, 0°C, 4h.
Under the same experimental conditions, the MAOAs
3 showed a similar inhibitory activity to those of the
reference drugs. All the MAOAs studied, 3a–l, showed
activity from 1.5–44 times higher than that of compound
1a. The most active MAOA 3f displayed an inhibitory
activity about 6 times lower than that of L-680,833.
All the MAOAs 3 and the reference drug L-680,833
were also evaluated in vivo for their HLE-inhibitory
activity by the method of elastase-induced lung haemor-
rhage in the mouse [34]. As confirmed by the data shown
in table I, only some of the MAOAs 3 studied showed an
appreciable inhibitory property on HLE-induced lung
Figure 5. (a) Diphenyldiazomethane, CHCl3/DMF, r.t.; (b) NH2OH HCl, K2CO3 1 M sol./THF, 0°C, r.t. 24h; (c) 4-(2-
hydroxyethylmorpholine), TPP, DEAD, dry THF, r.t., 24h; (d) NH2OH HCl, K2CO3, EtOH/H2O, r.t., 24h.

58
Table I. In vitro and in vivo inhibition of HLE by 3a–l.
Compound
R
R1
HLEa kinact/KI M-1s-1 (± SD)b
Lung haem. % inhib. of HLEc (± SD)
3a
3b
3c
3d
3e
3f
H
Me
Et
Ph
H
Me
H
H
H
H
H
F
F
30 900 (3 100)
44 000 (3 900)
8 000 (700)
3 400 (300)
6 000 (800)
98 300 (9 000)
6 100 (700)
13 500 (1 400)
39 (12)
43 (17)
*
34 (15)
19 (15)
31 (18)
14 (5)
*
3g
3h
COOH
COOH
Me
3i H2C2O4
H
17 900 (1 900)
21 (9)
3l H2C2O4
Me
35 200 (3 300)
*
1a
2 200 (300)
56**
L680,833
590 000 (80 000)
100 (0.02)
aSubstrate: N-methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide. bSD: standard deviation. cPercentage inhibition in elastase-induced lung
haemorrhage in mice when the compound was dosed orally at 10 mg/kg 5 h before challenge with HLE. *Inactive at the dose screened.
**Value reported in ref. [35] for HLE-induced lung haemorrhage in hamsters when dosed at 30 mg/kg.
haemorrhage at the dose screened (10 mg/kg). Com-
pounds 3a, 3b, 3d and 3f at this dose showed a percent-
age of inhibitory activity ranging from 31% of 3f,
para-fluorine phenyl-substituted, to 43% of its analogue
3b, which is devoid of this type of substituent on the
phenyl ring.
placed in the catalytic site, taking into account the
experimental data [43] available for compound L674,197
(figure 6), which is structurally similar to it. Subse-
quently, the complex thus obtained was minimised. At
this point, all MAOAs 3a–l were inserted in turn into the
catalytic site in place of L680,833 and every complex
was minimised.
The HLE–inhibitor complexes were modelled in a
situation in which no bonds were present between the
inhibitor and the enzyme.
4. Molecular models
With the aim of investigating the role of each molecu-
lar moiety of the MAOAs 3 in determining interaction
with the enzyme, a molecular model of HLE was con-
structed by means of molecular mechanics and molecular
dynamics calculations. For this purpose, we started from
the crystallographic structure of HLE complexed with the
peptidic inhibitor MSACK (Metoxysuccinyl-Ala-Ala-
Pro-Ala-Chloromethylketone) [42]. The MSACK was
then substituted with compound L680,833, which was
Figure 7 shows an overall view of HLE complexed
with the MAOA 3b (R = Me). Compound 3b is inserted
into the very hydrophobic pocket composed of the
residues of the sites S1 (Val190, Phe192, Ala213, Val216
and Phe228), S1≠ (Phe41, Cys42 and Gly43) and S2
(His57, Leu99 and Phe215). The innermost part of the
pocket is occupied by the two ethyl groups linked to the
C-3 of the â-lactam nucleus, which are enclosed by the

59
The structures of the interaction complexes of the other
MAOAs 3 and that of L680,833 were found to be quite
similar to that of 3b, shown in figure 7; the main
differences were in the region of the iminoethereal
portion and of the phenyl ring attached to it.
Figure 8 shows some details of the interaction between
HLE and compound 3b. The â-lactam carbonyl group of
3b is H-bonded with Ser195 and Gly193; the iminoethe-
real portion of 3b does not seem to interact strongly with
any particular residue of the enzyme; however the closest
residue to this polar group is His57, a residue considered
very important for the catalytic action of HLE. Therefore,
the distance between one of the two nitrogens of the
imidazolic ring of His57 and the oxygen of the imino-
ethereal group of compound 3b (3.5 Å) could allow a
polar interaction and therefore could improve the inter-
action of 3b with HLE.
Figure 6.
residues of the S1 site. The two phenyl rings occupy a
part of the S2 site and point out of the pocket towards the
solution.
Figure 7. Compound 3b inserted into the catalytic site of HLE. The ribbon of HLE is coloured magenta. The atoms of the residues
which form the sites S1, S2 and S1≠ are shown as CPK spheres coloured in yellow (residues of S1 site), brown (residues of S2 site)
and cyan (residues of S1≠ site). The solvent accessible surfaces of these sites are shown in grey.

60
a weaker ability, consequently, to interact. This fact could
explain the lower activity of compounds 3c and 3d with
respect to not only MAOA 3b, but also 3a.
Figure 10A indicates the possibility of a hydrogen
bond between the fluorine atom in the para position of
the phenyl ring linked to the iminic carbon of compound
3f and the amidic group of the side chain of Asn61
situated on the external surface of the enzyme, with a
distance F-H of about 2.5 Å. Figure 10B shows that in
analogue 3e, the different arrangement of the iminic
phenyl ring in the enzyme pocket due to the lack of the
methyl group on the iminic carbon makes the distance
F-H greater (3.5 Å); for this reason the F-H interaction is
weaker. The results obtained for the para-F substituted
compounds 3e and 3f might therefore explain the syner-
gic effect of this substitution and the introduction of a
methyl on the iminic carbon which makes the inhibitory
activity of compound 3f particularly high.
5. Conclusions
In a previous paper [34] it was observed that the
substitution of the LG of monocyclic â-lactams 1, the
C-4 aryloxy group, with a methyleneaminoxy molecular
portion leads to a new class of HLE inhibitors, the
MAOAs 3.
In the present work, the effects induced on the anti-
elastase properties of the MAOAs 3 by the R and R1
substituents present on the LG have been more accurately
investigated. Among the compounds studied, 3a, 3b, 3f
and 3l were found to possess an appreciable in vitro
inhibitory activity against HLE, with kinetic constant
values (kinact/KI) ranging from 30 900–98 300 M–1s–1.
The results made it possible to rationalise how the
substitution on the iminic carbon present on the LG of
MAOAs 3 can modulate the in vitro inhibitory activity
towards the HLE of such compounds, showing, in par-
ticular, that the optimal substitution was the one in which
R is a methyl group. Moreover it was observed that the
inhibitory activity was markedly improved by a para-
fluorine substitution on the phenyl ring of the LG
(R1 = F).
Molecular modelling calculations made it possible to
understand the reason for the higher activity possessed by
compounds in which the R substituent on the LG is a
methyl group. In fact the R substituent occupies a little
lipophilic pocket near the active site capable of accom-
modating groups not larger than a methyl. These theo-
retical studies also indicated that a fluorine substituent in
the para position on the phenyl ring of the methylene-
aminoxy LG (R1) is able to improve the interaction with
HLE, forming a hydrogen bond with the residue Asn61.
Figure 8. A detail of the interaction of compound 3b with
HLE; only the residues His57, Gly193 and Ser195 of HLE are
shown.
Figure 9 shows another detail of the interaction of
compound 3b with HLE. It can be seen that the methyl
group of 3b occupies a little pocket in the S2 site of HLE,
thus allowing a hydrophobic interaction of this group
with the side chains of residues Leu99 and Phe215. This
interaction is also found in the other MAOAs possessing
a methyl group on the iminic carbon (3f, 3h and 3l) and
could supply an explanation for the higher inhibitory
activity of these compounds with respect to the corre-
sponding ones without the methyl group on the iminic
carbon (cfr. compound 3b vs. 3a, 3f vs. 3e, 3h vs. 3g and
3l vs. 3i).
Molecular modelling calculations also indicate that the
size of this pocket is too small for a bulkier group like an
ethyl or a phenyl, and that the presence of such bulky
groups, as in compounds 3c and 3d, induces a modifica-
tion of the arrangement of the molecule in the pocket
mainly at the level of the phenyl ring linked to the iminic
carbon. In this arrangement the phenyl rings of these
compounds were found to be farther from the S2 site with

61
Figure 9. The methyl substituent on the iminic carbon of compound 3b occupies a little liphophilic pocket of HLE formed by residues
Leu99 and Phe215 (represented as CPK spheres coloured brown) of the S2 site.
The in vivo results indicated the absence of any
relationship between the kind of R and R1 substituent on
the LG and the lung haemorrhage inhibition induced by
HLE in mice. In any case, some of the MAOAs, 3a, 3b,
3d and 3f, showed a certain inhibitory activity towards
haemorrhage induced by HLE.
for all compounds, with the only exceptions of 9 H2C2O4
and 10 H2C2O4 (DMSO-d6). For configurational 1H-
NMR studies C6D6 and CF3CO2D were used as solvents.
The proton magnetic resonance assignments were estab-
lished on the basis of the expected chemical shifts and the
multiplicity of the signals. Analytical TLC was carried
out on 0.25 mm layer silica gel plates containing a
fluorescent indicator, and spots were detected under UV
light (254 nm). Flash chromatography or preparative
medium pressure liquid chromatography (MPLC) were
carried out through glass columns containing 40–63 µm
silica gel (Macherey-Nagel Silica Gel 60) or reversed
phase of octadecyl silica gel (C18 Polygosil 60–4063
Macherey-Nagel). The MPLCs were performed using a
chromatographic apparatus consisting of a Buchi 681
pump, a Knauer differential refractometer detector, and a
Philips PM 8220 pen recorder. Solvents and reagents
were obtained from commercial vendors in the appropri-
6. Experimental protocols
6.1. Chemistry
Melting points were determined on a Kofler hot-stage
apparatus and are uncorrected. Infrared (IR) spectra for
comparison of compounds were taken as paraffin oil
mulls or as liquid film on a Mattson 1000 FTIR spec-
trometer. Nuclear magnetic resonance (1H-NMR) spectra
were recorded on a Varian CFT-20 (80 MHz) or on a
Bruker AC-200 (200 MHz) in a ca 2% solution of CDCl3

62
A.
B.
Figure 10. A. Interaction of the para-fluorine substituent present on compound 3f with the residue Asn61 situated on the external
surface of HLE; B. The same interaction, but for compound 3e.
ate grade and were used without further purification
unless otherwise indicated. Elemental analyses were
carried out by our analytical laboratory and were consis-
tent with theoretical values to within ± 0.4%.
dried and evaporated to give an oily residue which, after
purification by flash chromatography eluting with a 2:1
hexane-AcOEt mixture, yielded the appropriate pure
azetidinone 13c–f and 13h. 13c (52% yield), m.p.
72–73 °C, 1H-NMR (CDCl3) δ 0.96–1.24 (m, 9H),
1.67–2.04 (m, 4H), 2.80 (q, 2H, J = 9.6 Hz), 5.52 (s, 1H),
6.51 (br, 1H), 7.38–7.56 (m, 5H). Anal. C16H22N2O2 (C,
H, N). 13d (47% yield), m.p. 69–70 °C, 1H-NMR
(CDCl3) δ 0.90 and 1.13 (2t, 6H, J = 7 Hz), 1.47–1.97 (m,
4H), 5.63 (s, 1H), 6.52 (br, 1H), 7.29–7.62 (m, 10H).
Anal. C20H22N2O2 (C, H, N). 13e (52% yield), m.p.
75–78 °C, 1H-NMR (CDCl3) δ 1.01 and 1.03 (2t, 6H, J =
7.2 Hz), 1.55–1.90 (m, 4H), 5.45 (s, 1H), 6.50 (br, 1H),
6.95–7.70 (m, 4H), 8.09 (s, 1H). Anal. C14H17FN2O2 (C,
6.1.1. General procedure for the preparation of the
(E)-4-(arylideneaminoxy)-2-azetidinones 13c–f and 13h
1 N NaOH (6.5 mL) was added to a stirred solution of
one of the appropriate E oximes 12c–f or 12h (6.5 mmol)
in acetone (6.5 mL). After stirring for 10 min at room
temperature, a solution of 11 (5.8 mmol) in acetone (3.5
mL) was added, and the resulting mixture was stirred at
the same temperature for 40 min. The acetone was then
removed in vacuo and the aqueous layer was extracted
with ether. The organic phase was washed with brine,

63
H, N). 13f (60% yield), m.p. 77–79 °C, 1H-NMR
(CDCl3) δ 1.02 and 1.11 (2t, 6H, J = 7.2 Hz), 1.53–1.98
(m, 4H), 2.26 (s, 3H), 5.49 (s, 1H), 6.90–7.85 (m, 4H).
Anal. C15H191FN2O2 (C, H, N). 13h (43% yield), m.p.
125–126 °C, H-NMR (CDCl3) δ 1.09 and 1.11 (2t, 6H,
J = 7 Hz), 1.62–1.91 (m, 4H), 2.27 (s, 3H), 5.51 (s, 1H),
6.35 (br, 1H), 7.09 (s, 1H), 7.22–7.42 (m, 10H), 7.68 and
8.11 (2d, 4H, J = 8.8 Hz). Anal. C29H30N2O4 (C, H, N).
4.46 (d, 2H, J = 5.6 Hz), 5.90 (s, 1H), 6.84 (br, 1H),
7.20–7.75 (m, 10H). Anal. C24H29N3O3 (C, H, N). 3d
(47% yield), m.p. 70–71 °C, H-NMR (CDCl3) δ 1.02
1
and 1.06 (2t, 6H, J = 7 Hz), 1.53–1.85 (m, 4H), 4.44 (d,
2H, J = 5.6 Hz), 5.94 (s, 1H), 6.81 (br, 1H), 7.24–7.51 (m,
15H). Anal. C28H29N3O3 (C, H, N). 3e (53% yield), m.p.
1
122–124 °C, H-NMR (CDCl3) δ 1.01 and 1.03 (2t, 6H,
J = 7.2 Hz), 1.55–1.99 (m, 4H), 4.47 (d, 2H, J = 5.6 Hz),
5.86 (s, 1H), 6.80–7.84 (m, 10H), 8.09 (s, 1H). Anal.
C22H24FN3O3 (C, H, N). 3f (55% yield), m.p. 83–85 °C,
1H-NMR (CDCl3) δ 1.01 and 1.03 (2t, 6H, J = 7.2 Hz),
1.52–1.98 (m, 4H), 2.25 (s, 3H), 4.44 (d, 2H, J = 5.6 Hz),
5.90 (s, 1H), 6.75–7.75 (m, 10H). Anal. C23H26FN3O3
(C, H, N). 3i (55% yield), m.p. 150–151 °C (3i.H2C2O4),
1H-NMR (CDCl3) δ 1.04 and 1.07 (2t, 6H, J = 7 Hz),
1.71–1.92 (m, 4H), 2.45–2.70 (m, 4H), 2.79 (t, 2H, J =
5.6 Hz), 3.50–3.70 (m, 4H), 4.12 (t, 2H, J = 5.6 Hz), 4.51
(d, 2H, J = 5.6 Hz), 5.86 (s, 1H), 6.85 (d, 2H, J = 8.8 Hz),
7.20–7.32 (m, 5H), 7.48 (d, 2H, J = 8.8 Hz), 8.07 (s, 1H).
Anal. C28H36N4O5 (C, H, N). 3l (62% yield), m.p.
6.1.2. General procedure for the preparation of the
(E)-4-(arylideneaminoxy)-2-azetidinones 13i and 13l
1 N NaOH (12.7 mL) was added to a stirred solution of
12i or 12l (3.7 mmol) in acetone (3.5 mL). After stirring
for 10 min at room temperature, a solution of 11
(3.3 mmol) in acetone (3.5 mL) was added, and the
resulting mixture was stirred at the same temperature for
40 min. The acetone was then removed in vacuo and the
aqueous layer was extracted with ether. The organic
phase was washed with brine, dried and evaporated to
give an oily residue which, after purification by MPLC on
C18 reversed phase using a 6:4 mixture of MeOH-H2O
with 8% of an aqueous solution of NH4OH (30%) and 4%
of 2-propanol, yielded the appropriate pure azetidinone
13i or 13l as oils. 13i (28% yield),1H-NMR (CDCl3) δ
1.06 and 1.08 (2t, 6H, J = 7 Hz), 1.64–1.86 (m, 4H),
2.45–2.70 (m, 4H), 2.85 (t, 2H, J = 5.6 Hz), 3.60–3.90
(m, 4H), 4.18 (t, 2H, J = 5.6 Hz), 5.50 (s, 1H), 6.43 (br,
1H), 6.93 (d, 2H, J = 8.8 Hz), 7.53 (d, 2H, J = 8.8 Hz),
8.12 (s, 1H). Anal. C20H29N3O4 (C, H, N). 13l (31%
yield),1H-NMR (CDCl3) δ 1.07 and 1.10 (2t, 6H, J = 7
Hz), 1.62–1.81 (m, 4H), 2.23 (s, 3H), 2.45–2.60 (m, 4H),
2.80 (t, 2H, J = 5.6 Hz), 3.55–3.73 (m, 4H), 4.18 (t, 2H,
J = 5.6 Hz), 5.47 (s, 1H), 6.45 (br, 1H), 6.85 (d, 2H, J =
8.8 Hz), 7.51 (d, 2H, J = 8.8 Hz). Anal. C21H31N3O4 (C,
H, N).
1
155–157 °C (3l.H2C2O4), H-NMR (CDCl3) δ 1.05 and
1.08 (2t, 6H, J = 7 Hz), 1.64–1.80 (m, 4H), 2.30 (s, 3H),
2.50–2.70 (m, 4H), 2.75 (t, 2H, J = 5.6 Hz), 3.65–3.80
(m, 4H), 4.17 (t, 2H, J = 5.6 Hz), 4.52 (d, 2H, J = 5.6 Hz),
5.95 (s, 1H), 6.85 (d, 2H, J = 8.8 Hz), 7.25–7.32 (m, 5H),
7.56 (d, 2H, J = 8.8 Hz). Anal. C29H38N4O5 (C, H, N).
1
14h (83% yield), m.p. 112–113 °C, H-NMR (CDCl3) δ
1.01 and 1.04 (2t, 6H, J = 7 Hz), 1.61–1.89 (m, 4H), 2.29
(s, 3H), 4.51 (d, 2H, J = 5.6 Hz), 5.51 (s, 1H), 6.87 (br,
1H), 7.09 (s, 1H), 7.22–7.42 (m, 5H), 7.68 and 8.11 (2d,
4H, J = 8.8 Hz). Anal. C37H37N3O5 (C, H, N).
6.1.4. (E)-4-(arylideneaminoxy)-1-N-
(benzylaminocarbamoyl)-2-azetidinone 3h
Trifluoroacetic acid (2 mL) was added dropwise to a
stirred and cooled solution of 14h (0.30 g, 0.5 mmol) in
anisole (4 mL) and the resulting mixture was stirred at
0 °C for 4 h and extracted with an ice-cooled 10%
aqueous NaHCO3 solution (3 × 25 mL). The aqueous
layer, after being washed with CHCl3 (2 × 30 mL), was
acidified to pH 3.5 at 0 °C by the addition of 10%
aqueous H3PO4 and then extracted with CHCl3 (3 ×
30 mL). Evaporation of the washed (brine) organic ex-
tracts yielded a white solid residue, which, after crystal-
lisation with AcOEt-hexane, gave pure 3h: 0.12 g as a
white solid (72% yield), m.p. 144–145 °C, 1H-NMR
(CDCl3) δ 1.09 and 1.10 (2t, 6H, J = 7 Hz), 1.74–2.03 (m,
4H), 2.35 (s, 3H), 4.53 (d, 2H, J = 5.6 Hz), 6.05 (s, 1H),
6.98 (br, 1H), 7.33–7.41 (m, 5H), 7.72 and 8.04 (2d, 4H,
J = 8 Hz). Anal. C24H27N3O5 (C, H, N).
6.1.3. General procedure for the preparation of the
(E)-4-(arylideneaminoxy)-1-N-(benzylaminocarbamoyl)-
2-azetidinones 3c–f, 3i–l and 14h
A stirred solution of the appropriate azetidinone 13
(1 mmol), Et3N (0.14 mL, 1 mmol) and a catalytic
amount of 4-(dimethylamino)pyridine (DMAP, 2 crys-
tals) in anhydrous CH2Cl2 (5 mL), was treated under
nitrogen with benzylisocyanate (0.37 mL, 3 mmol), and
the mixture was stirred at room temperature for 24 h. The
solvent was evaporated and the residue was purified by
crystallisation with i-PrOH (in the case of 3c–f and 14h)
or by MPLC on C18 reversed phase using a 6:4 mixture of
MeOH-H2O with 8% of an aqueous solution of NH4OH
(30%) and 4% of 2-propanol (in the case of 3i–l). 3c
1
(62% yield), m.p. 94 °C, H-NMR (CDCl3) δ 0.94–1.25
(m, 9H), 1.64–1.86 (m, 4H), 2.77 (q, 2H, J = 7.2 Hz),

64
6.1.5. Benzhydryl (E)-acetophenonoxime-
4-carboxylate 12h
(m, 4H), 4.18 (t, 2H, J = 5.6 Hz), 6.87 and 7.88 (2d, 4H,
J = 8.8 Hz). Anal. C14H19NO3 (C, H, N).
A solution of diphenyldiazomethane (6.5 g, 33 mmol)
in CHCl3 (10 mL) was added portionwise to an unstirred
solution of 4-acetylbenzoic acid 15 (5.5 g, 33 mmol) in
DMF (25 mL). The resulting mixture was poured into a
saturated aqueous NaHCO3 solution (100 mL) at 0 °C
and then extracted with CHCl3 (2 × 100 mL). The organic
phase was dried and evaporated to give an oily residue,
which after crystallisation with AcOEt-hexane, yielded
4.5 g of pure benzhydryl 4-acetylbenzoate 16 (41%
6.1.7. General procedure for the preparation of 4-[2≠-
ethyloxy-N-morpholino]-(E)-benzaldoxime 12i and 4-
[2≠-ethyloxy-N-morpholino]-(E)-acetophenonoxime 12l
Solid K2CO3 (2.80 mmol) and then a solution of
NH2OH.HCl (5.60 mmol) in H2O (9.5 mL) were added to
a stirred and cooled (0 °C) solution of the appropriate
phenylcarbonyl derivative 19, 20 (4.25 mmol) in EtOH
(25 mL). The resulting mixture was stirred at room
temperature for 24 h and then neutralised with 21%
H3PO4. The EtOH was removed under reduced pressure
and the resulting mixture was extracted with Et2O
(50 mL). Evaporation of the dried and filtered extracts
gave a solid residue which was crystallised by a 2:1
hexane-AcOEt mixture. 12i, 0.93 g (57% yield), m.p.
1
yield), m.p. 106–107 °C, H-NMR (CDCl3) δ 2.62 (s,
3H), 7.10 (s, 1H), 7.22–7.35 (m, 10 H), 7.97 and 8.19 (2d,
4H, J = 8.8 Hz). Anal. C22H18O3 (C, H). 16 (4.2 g,
13 mmol) was dissolved in THF (85 mL) and the result-
ing solution, after being cooled at 0 °C, was treated with
an aqueous solution of 1 M K2CO3 (9 mL) and then,
portionwise, with an aqueous solution of 0.55 M hy-
droxylamine hydrochloride (31 mL). The resulting mix-
ture was stirred at room temperature for 24 h and then
acidified at 0 °C (pH 5) with aqueous 5% HCl. After
evaporation of THF, the aqueous layer was diluted with
water (45 mL) and extracted with Et2O (2 × 100 mL).
The organic phase was dried and evaporated to give a
solid residue which, after crystallisation with AcOEt-
hexane, yielded pure 12h, 3.2 g (71% yield), m.p. 148 °C,
1H-NMR (CDCl3) δ 2.30 (s, 3H), 7.09 (s, 1H), 7.23–7.36
(m, 10 H), 7.68 and 8.11 (2d, 4H, J = 8.8 Hz). Anal.
C22H19NO3 (C, H, N).
1
114–115 °C, H-NMR (CDCl3) δ 2.55–2.75 (m, 4H),
2.88 (t, 2H, J = 5.6 Hz), 3.70–3.90 (m, 4H), 4.23 (t, 2H,
J = 5.6 Hz), 6.90 and 7.46 (2d, 4H, J = 8.8 Hz), 8.08 (s,
1H). Anal. C13H18N2O3 (C, H, N). 12l, 0.62 g (56%
1
yield), m.p. 124–125 °C, H-NMR (CDCl3) δ 2.23 (s,
3H), 2.50–2.70 (m, 4H), 2.82 (t, 2H, J = 5.6 Hz),
3.58–3.80 (m, 4H), 4.18 (t, 2H, J = 5.6 Hz), 6.85 and 7.50
(2d, 4H, J = 8.8 Hz). Anal. C14H20N2O3 (C, H, N).
6.1.8. Configurational 1H-NMR studies on oximes 12h–l
1H-NMR studies of the shielding effect induced by an
anisotropic solvent on 12h–l were carried out with 5%
(w/w) in C6D6 (0.75 mL), adding increasing amounts (5,
10, 15 and 25 µL) of CF3CO2D. The spectra were
recorded for the free bases and after each addition of acid.
The chemical shifts of the protons of the methyl group or
the proton linked to the iminic carbon atom in deutero-
benzene solutions were respectively 2.30 ppm (Me pro-
tons, 12h), 8.05 ppm (H, 12i) and 2.23 ppm (Me protons,
12l) as free bases. After the addition of the above-
mentioned amounts of CF3CO2D, the chemical shift of
the same protons of 12h–l changed to 2.28, 2.26, 2.24 and
2.21 ppm for 12h, 7.83, 7.65, 7.63 and 7.61 ppm for 12i
and 1.99, 1.87, 1.77 and 1.74 ppm for 12l for their
protonated forms.
6.1.6. General procedure for the preparation of 4-[2≠-
ethyloxy-N-morpholino]-benzaldehyde 19 and 4-[2≠-
ethyloxy-N-morpholino]-acetophenone 20
A solution of the appropriate 4-hydroxyphenylcarbonyl
derivative 17, 18 (12 mmol), 4-(2-hydroxyethylmor-
pholine) (1.57 g, 12 mmol), triphenylphosphine (3.15 g,
12 mmol) and diethylazodicarboxylate (21.0 g, 12 mmol)
in dry THF (70 mL) was stirred at room temperature for
24 h under N2 atmosphere. The THF was removed under
reduced pressure and the crude residue was dissolved
with AcOEt (100 mL) and extracted with 10% HCl (2 ×
50 mL). The aqueous phase was alkalinised to pH 9 with
solid K2CO3, and extracted with AcOEt (2 × 100 mL).
Evaporation of the dried and filtered extracts gave a crude
residue which was purified by MPLC on silica gel using
a 1:1 mixture of hexane-AcOEt with 2% of triethyamine
to afford the desired compounds as oils. 19 [44]: 1.7 g
6.1.9. Configurational stability test for the
asymmetric oximes 12c–f and 12h–l
The above-mentioned oximes 12c–f and 12h–l
(2 mmol) in acetone (6.9 mL) were treated with a 1 N
NaOH solution (6.9 mL) and the resulting solutions were
stirred for 1 h at room temperature. The acetone was then
removed in vacuo and the aqueous layers were then
neutralised with H3PO4 solution (10%) and extracted
with ether. The organic phases were washed with brine,
dried and evaporated to give quantitatively the unchanged
1
(60% yield), H-NMR (CDCl3) δ 2.65–2.70 (m, 4H),
2.88 (t, 2H J = 5.6 Hz), 3.55–3.70 (m, 4H), 4.23 (t, 2H,
J = 5.6 Hz), 7.05 and 7.89 (2d, 4H, J = 9 Hz), 9.97 (s,
1H). Anal. C13H17NO3 (C, H, N). 20 [45]: 1.6 g (53%
yield), 1H-NMR (CDCl3) δ 2.41–2.90 (m, 9H), 3.75–3.80

65
starting oximes 12c–f and 12h–l (1H-NMR analysis) as
residues. These residues were then solubilised with an-
hydrous CH2Cl2 (10mL) and treated under stirring with
Et3N (2 mmol) and DMAP (2 crystals) for 24 h at room
temperature. The mixture was then neutralised with
H3PO4 solution (10%) and the organic phase was washed
with brine, dried and evaporated to quantitatively give the
unchanged starting oximes 12c–f and 12h–l as residues
(1H-NMR).
6.2. HLE inhibition
The activity of HLE (EC 3.4.21.37, Calbiochem) was
assayed spectrophotometrically using N-methoxysuc-
cinyl-Ala-Ala-Pro-Val-p-nitroanilide (MAAPVNA) as
the substrate [46] in 0.05 M sodium phosphate buffer, pH
7.8; the hydrolytic release of p-nitroaniline was moni-
tored on the basis of the increase in absorbance at 390 nm
(e for MAAPVNA = 500 M–1 cm–1; e for p-nitroaniline =
13 000 M–1 cm–1) with a Beckman DU-7 UV/visible
spectrophotometer thermostated at 37 °C [35]. Both sub-
strate and inhibitor stock solutions were made in DMSO.
A typical inhibition assay (carried out in a final volume of
500 µL of the buffer containing 4% DMSO) was run as
follows: HLE (about 5 milliunits, corresponding to an
enzyme concentration of about 10 nM; one milliunit
releases one nmol of p-nitroaniline per min from
MAAPVNA) was added to a reaction mixture containing
MAAPVNA (0.2–0.4 mM) and the increase in absor-
bance was recorded for 3 min (time course A); one of the
inhibitors dissolved in DMSO was subsequently added to
a final concentration ranging from 0.05–4.0 µM and the
absorbance change was further recorded up to 30 min
(time course B). Control assays, in which the inhibitor
was omitted, ran linearly in time up to an absorbance of
1.0–1.1 AU, before the slope started to change due to
substrate consumption; assays in which the inhibitor was
added showed a swift time-dependent fall in slope, and
the absorbance levelled off between 0.4–1.0 AU at
30 min, depending on the inhibitor concentration and
potency.
Figure 11. A typical time course of the absorbance change in
the HLE inhibition assay, using compound 3b as the inhibitor,
is shown to illustrate the kinetic analysis method used herein;
arrows and labels mark the addition of components to the
reaction mixture. Black dots are absorbance readings every
10 s; the dotted straight line extrapolates the time course of the
absorbance in the absence of added inhibitor. Initial conditions:
[S]0 = 0.35 mM (A0 = 0.181); [HLE]0 = 3.3 nM; [3b] = 0.2 µM.
The reaction mixture (final volume 500 µL) was made in
0.05 M sodium phosphate buffer, pH 7.8, with 4% DMSO;
T = 37 °C. Values of the changing slope in phase B, taken over
60 s intervals, were divided by the slope in phase A. The natural
log of these ratios were plotted against time to get the kobs
value. The reciprocal of several kobs values, obtained at diffe-
rent inhibitor concentrations and fixed substrate concentration,
were eventually plotted against 1/[I] to get the inhibition
apparent second-order rate constant; this was finally multiplied
by
͑
1 +
S
͓ ͔
KM
͒
to get the kinact /KI value.
ր
presence of both inhibitor and substrate strictly obeyed
pseudo first-order kinetics up to a residual activity lower
than 5% (i.e. over 3 natural log units). 1/kobs values
obtained at the same substrate concentration and at
varying concentrations of a single inhibitor were plotted
vs. 1/[I]. The reciprocal of the slope of this plot yields
The first-order decay in HLE activity was evaluated by
taking the slope every 60 s over time course B and
dividing it by the slope in time course A (figure 11). The
natural log of this ratio was plotted as a function of time,
the negative slope of this plot yielding a kobs value
(apparent first-order rate constant). Pseudo first-order
conditions were checked by evaluating total HLE molar
concentration from the rate in time course A, substrate
concentration and the enzyme–substrate kinetic param-
eters at 37 °C (kcat = 20 s–1, KM = 0.12 mM, our data, not
shown). The inactivation time course of the enzyme in the
kinact
,
from which the concentration-
KI 1 + S
͓ ͔
KM
͑
ր
independent second-order rate constant kinact/KI is ob-
tained. Similar values of this parameter were obtained for
the same inhibitor at different substrate concentrations, in
agreement with a competitive mechanism of inhibition.

66
6.3. Pharmacology
formed with a tether constraint on the heavy atoms
gradually lowered till complete relaxation.
6.3.1. HLE-induced lung haemorrhage in mice
At this point the molecule of MSACK was removed
from the active site and substituted with the molecule of
L680,833, again bound to Ser195 in accordance with the
Michaelis complex structure. The molecule was manually
adapted in such a manner as to satisfy the findings from
Finke et al. [43] for an analogue of L680,833 (L674,197),
through X-ray crystallography; however, they do not
supply the experimental co-ordinates. In particular, suit-
able constraints were introduced in order to take into
account the interaction of the carbonylic oxygen of the
â-lactamic cycle with the amidic hydrogens of Gly193
and Ser195 [43]. A run of MD at a temperature of 900°K
was then carried out with quite a high tether constraint on
the backbone of the enzyme (with the exception of
residues 190–200, the nearest ones to the active site) in
order to prevent the unfolding of the protein. After a 10 ps
equilibration, a 100 ps run was carried out selecting one
conformation every 10 ps. Each of these conformations
was then minimised, gradually cancelling all constraints.
The minimised conformations thus obtained were found
to be similar to each other in the region of the active site
and the one in which the conformation of L680,833 was
energetically favoured was selected.
Mice, anaesthetised with a mixture of ketamine and
xylazine i.p., were instilled intratracheally with 50 µL of
saline (vehicle) or human leukocyte elastase obtained
from purulent sputum (Elastin Products Company inc.)
(3 200 U/mL dissolved in saline, corresponding to a
solution of 16 µg/50 µL). Test compounds were orally
administered at 10 mg/kg as a suspension of DMSO in
Methocel, 3 h before enzyme challenge. 1.5 h after
elastase instillation, the animals were sacrificed by CO2
asphyxiation and the trachea was exposed. The lungs
were washed using a 600 µL saline-filled syringe con-
nected to a tracheal cannula by gently expanding the
lungs and then withdrawing the saline. This operation
was repeated 3 times, yielding a final volume of about
1.5 mL of bronchoalveolar lavage fluid (BAL) for each
animal. The BAL fluids were diluted with Na2CO3 and
sonicated to ensure cell disruption. The blood content was
determined spectrophotometrically at 414 nm. The
amount of blood in each BAL sample was calculated by
a standard curve obtained by using sonicated whole
mouse blood supplemented with Na2CO3 2% w/v. The
haemorrhage was expressed as µL equivalents of blood in
1 mL of BAL fluid. The protective effect of the com-
pounds was calculated as a percentage of inhibition of
lung haemorrhage in treated versus control animals. The
basal value (vehicle) was subtracted from each treatment
group.
Subsequently the bond between the inhibitor and
Ser195 was broken and the structure was again mini-
mised through MD and MM in a way similar to that used
for the Michaelis complex.
At this point the molecule of L680,833 was substituted
in sequence with MAOAs 3a–l using the same strategy of
minimisation by means of MD and MM.
6.4. Computational analysis
All the molecular mechanics (MM) and dynamics
(MD) calculations were made by using the program
Discover with the CVFF forcefield [47]. A distance-
dependent dielectric constant equal to 4 was used in order
to simulate an aqueous environment, no explicit water
molecules were considered. All the ionisable groups of
HLE were considered in their most stable form at a pH of
7.4, thus obtaining a structure with an excess of 11
positive-charged groups. The minimisation methods were
steepest descents and conjugate gradient, and the MD
runs were made with a 1 fs time step. All the structures
were modelled and visualised through the software In-
sight II [47]. The starting-point for the modelling of the
inhibitor active site was the X-ray structure of HLE
complexed with the peptidic inhibitor MSACK
(Metoxysuccinyl-Ala-Ala-Pro-Ala-Chloromethylketone)
[42] whose co-ordinates were obtained from the Protein
Data Bank [48, 49]. In the experimental structure, the
hydrogens were added, the bond between MSACK and
Ser195 was formed, and then a minimisation was per-
Acknowledgements
Part of this work was supported by the contract
‘Programma Nazionale di Ricerca sui Farmaci (seconda
Fase), Tema 4°, granted by the Italian Ministry of
University and Scientific and Technological Research.
References
[1] Travis J., Dumbin A., Potempa J., Watorek W., Kurdowska A., Ann.
NY Acad. Sci. 624 (1991) 81–86.
[2] Janoff A., Scherer J., J. Exp. Med. 128 (1968) 1137–1155.
[3] Dal Nogare A.R., Toews G.B., Pierce A.K., Am. Rev. Respir. Dis.
135 (1987) 671–675.
[4] McDonald J.A., Kelly D.G., J. Biol. Chem. 255 (1987) 8848–8858.
[5] McDonald J.A., Baum B.J., Rosemberg D.M., Kelman J.A., Brin
S.C., Crystal R.G., Lab. Invest. 40 (1979) 350–357.
[6] Brown G.M., Donaldson K., Thorax 43 (1988) 132–139.

67
[7] Starkey P.M., Acta Biol. Med. Ger. 36 (1977) 1549–1554.
[28] Fireston R.A., Barker P.L., Pisano J.M., Ashe B.M., Dahlgren M.E.,
Tetrahedron 46 (1990) 2255–2262.
[8] Gadek J.E., Fells G.A., Wright D.G., Crystal R.G., Biochem.
Biophys. Res. Commun. 98 (1980) 1815–1822.
[29] Doherty J.B., Shah S.K., Finche P.E., Dorn C.P., Hagman K.W.,
Hale J.J. et al., Proc. Natl. Acad. Sci. USA 90 (1993) 8727–8731.
[9] Mainard C.L., Dixit S.N., Kang A.H., J. Biol. Chem. 225 (1980)
5435–5441.
[30] Finche P.E., Dorn C.P., Kissinger A.L., Shah S.K., Ball R.G., Davies
P. et al., 22nd National Meeting of the American Chemical Society,
Chicago, 1995, MEDI 083.
[10] Janoff A., Blondin J., Proc. Soc. Exp. Biol. Med. 135 (1970)
302–306.
[31] Alpegiani M., Bissolino P., Corigli R., Rizzo V., Perrone E., Bioorg.
Med. Chem. Lett. 5 (1995) 687–690.
[11] Snider G.L., Ciccolella D.E., Morris S.M., Ann. NY Acad. Sci. 624
(1991) 45–59.
[32] Chabin R., Green B.G., Gale P., Maycock A.L., Weston H., Dorn
C.P. et al., Biochemistry 32 (1993) 8970–8980.
[12] Merritt T.A., Cochrane C.G., Holcomb K., J. Clin. Invest. 72 (1983)
656–666.
[33] Groutas W.C., Epp J.B., Venkataraman R., Kuang R., Truong T.M.,
McClenahan J.J., Prakash O., Bioorg. Med. Chem. 4 (1996)
1393–1400.
[13] Jackson A.H., Hill S.L., Afford S.C., J. Resp. Dis. 65 (1984)
114–124.
[14] Stockley R.A., Hill S.L., Burnett D., Ann. NYAcad. Sci. 624 (1991)
257–266.
[34] Balsamo A., Asti C., Belfiore M.S., Brandolini L., Cercignani G.,
Gentili D. et al., Eur. J. Med. Chem. 32 (1997) 889–894.
[15] Burkhart J.P., Koehl J.R., Mehdi S., Durhan S.L., Janusz M.J.,
Huber E.W. et al., J. Med. Chem. 38 (1995) 223–233.
[35] Shah S.K., Dorn C.P., Finke P.E., Hale J.J., Hagmann W.K., Brause
K.A. et al., J. Med. Chem. 35 (1992) 3745–3754.
[16] Durhan S.L., Hare C.M., Angelastro M.R., Burkhart J.P., Koehl,
J.R., Marquat A.L. et al., J. Pharmacol. Exp. Ther. 270 (1994)
185–191.
[36] Meyer V., Chem. Ber. 16 (1883) 822–823.
[37] Karabatsos G.J., Taller R.A., Tetrahedron 24 (1968) 3347–3360.
[38] Campbell K.N., McKenna T.F., J. Org. Chem. 4 (1939) 198–205.
[17] Edwards P.D., Anisik D.W., Bryant C.A., Ewing B., Gomes B.,
Lewis J.J. et al., J. Med. Chem. 40 (1997) 1876–1885.
[39] Furniss B.S., Hannaford A.J., Smith P.W.G., Tatchell A.R., Vogel’s
Textbook of Practical Organic Chemistry, 5th ed., Longman Scien-
tific & Technical, John Wiley & Sons Inc., New York, 1989, p. 1259.
[18] Edwards P.D., Anisik D.W., Davis M.W., J. Med. Chem. 38 (1995)
76–85.
[40] Mitsunobu O., Synthesis (1981) 1–28.
[19] Edwards P.D., Meyer E.F., Vijayalakshmi J., Tuthill P.A., Andisik
D.A., Gomes B., Strimpler A., J. Am. Chem. Soc. 114 (1992)
1854–1863.
[41] Fox B.L., Reboulet J.E., Rondeau R.E., Rosemberg M.H., J. Org.
Chem. 35 (1970) 4234–4236.
[42] Navia M.A., McKeever B.M., Springer J.P., Lin Y.T., Williams H.R.,
Fluder E.M., Dorn C.P., Hoogsteen K., Proc. Natl. Acad. Sci. USA
86 (1989) 7–11.
[20] Williams J.C., Falcone R.C., Knee C., Stein R.L., Strimpler A.M.,
Reaves B., Giles R.E., Krell R.D., Am. Rev. Respir. Dis. 144 (1991)
875–883.
[43] Finche P.E., Shah S.K., Fletcher D.S., Ashe B.M., Brause K.A.,
Chandler G.O. et al., J. Med. Chem. 38 (1995) 2449–2562.
[21] Bernstein P.R., Gomes B.C., Kosmider B.J., Vavek E.P., Williams
J.C., J. Med. Chem. 38 (1995) 212–215.
[44] Cossey H.D., Shape C.J., Stephens F.F., J. Chem. Soc. 35 (1963)
4322–4330.
[22] Veale C.A., Bernstein P.R., Byrant C., Ceccarelli C., Damewood
J.R., Earley R. et al., J. Med. Chem. 38 (1995) 98–105.
[45] Najer H., Chabrier P., Guidicelli R., Bull. Soc. Chim. Fr. (1956)
1672–1674.
[23] Veale C.A., Damewood J.R., Steelman G.B., Byrant C., Gomes
B.C., Williams J.C., J. Med. Chem. 38 (1995) 86–97.
[46] Nakajima K., Powers J.C., Ashe B.M., Zimmerman M., J. Biol.
Chem. 254 (1979) 4027–4032.
[24] Bernstein P.R., Andisik D., Bradley P.K., Byrant C., Ceccarelli C.,
Damewood J.R. et al., J. Med. Chem. 37 (1994) 3313–3326.
[47] Insight II Version 2.3, Discover Version 2.9.5 Biosym Technologies,
San Diego, CA.
[25] Groutas W.C., Kuang R., Venkataraman R., Biochem. Biophys. Res.
Commun. 198 (1994) 341–349.
[48] Sussman J.L., Lin D., Jiang J., Manning N.O., Prilusky J., Ritter O.,
Abola E.E., Acta Cryst. D54 (1998) 1078–1084.
[26] Player M.R., Sowell J.W., Patil G.S., Kam C.M., Powers J.C.,
Bioorg. Med. Chem. Lett. 4 (1994) 949–954.
[49] Abola E.E., Sussman J.L., Prilusky J., Manning N.O., in: Carter Jr.
C.W., Sweet R.M. (Eds.), Methods in Enzymology, Academic Press,
San Diego, 1997, pp. 556–571.
[27] Finche P.E., Shah S.K., Ashe B.M., Ban R.G., Blacklock T.J.,
Bonney R.J. et al., J. Med. Chem. 35 (1992) 3731–3744.