Conformational preferences in a benzodiazepine series of potent nonpeptide fibrinogen receptor antagonists

Article abstract of DOI:10.1021/jm980166z

Previously, we reported the direct design of highly potent nonpeptide 3- oxo-1,4-benzodiazepine fibrinogen receptor antagonists from a constrained, RGD-containing cyclic semipeptide. The critical features incorporated into the design of these nonpeptides were the exocyclic amide at the 8-position which overlaid the Arg carbonyl, the phenyl ring which maintained an extended Gly conformation, and the diazepine ring which mimicked the γ-turn at Asp. In this paper, we investigate conformational preferences of the 8-substituted benzodiazepine analogues by examining structural modifications to both the exocyclic amide and the seven-membered diazepine ring and by studying the conformation of the benzodiazepine ring using molecular modeling, X-ray crystallography, and NMR. We found that the directionality of the amide at the 8-position had little effect on activity and the (E)-olefin analogue retained significant potency, indicating that the trans orientation of the amide, and not the carbonyl or NH groups, made the largest contribution to the observed activity. For the diazepine ring, with the exception of the closely analogous 3-oxo-2-benzazepine ring system described previously, all of the modifications led to a significant reduction in activity compared to the potent 3-oxo-1,4-benzodiazepine parent ring system, implicating this particular type of ring system as a desirable structural feature for high potency. Energy minimizations of a number of the modified analogues revealed that none could adopt the same low-energy conformation as the one shared by the active (S)-isomer of the 3-oxo-1,4-benzodiazepines and 3-oxo-2- benzazepines. The overall data suggest that the features contributing to the observed high potency in this series are the orientation of the 3-4 amide and the conformational constraint imposed by the seven-membered ring, both of which position the key acidic and basic groups in the proper spatial relationship.

Full text of DOI:10.1021/jm980166z

J . Med. Chem. 1999, 42, 545-559
545
Articles
Con for m a tion a l P r efer en ces in a Ben zod ia zep in e Ser ies of P oten t Non p ep tid e
F ibr in ogen Recep tor An ta gon ists
Richard M. Keenan,*,† J ames F. Callahan,*,† J ames M. Samanen,*,† William E. Bondinell,† Raul R. Calvo,†
Lichong Chen,† Charles DeBrosse,∇ Drake S. Eggleston,§ R. Curtis Haltiwanger,§ Shing Mei Hwang,|
Dalia R. J akas,† Thomas W. Ku,† William H. Miller,† Kenneth A. Newlander,† Andrew Nichols,‡
Michael F. Parker,† Linda S. Southhall,† Irene Uzinskas,† J anice A. Vasko-Moser,‡ J oseph W. Venslavsky,†
Angela S. Wong,| and William F. Huffman†
Research and Development Division, SmithKline Beecham Pharmaceuticals, P.O. Box 5089,
Collegeville, Pennsylvania 19426-0989
Received March 18, 1998
Previously, we reported the direct design of highly potent nonpeptide 3-oxo-1,4-benzodiazepine
fibrinogen receptor antagonists from a constrained, RGD-containing cyclic semipeptide. The
critical features incorporated into the design of these nonpeptides were the exocyclic amide at
the 8-position which overlaid the Arg carbonyl, the phenyl ring which maintained an extended
Gly conformation, and the diazepine ring which mimicked the γ-turn at Asp. In this paper, we
investigate conformational preferences of the 8-substituted benzodiazepine analogues by
examining structural modifications to both the exocyclic amide and the seven-membered
diazepine ring and by studying the conformation of the benzodiazepine ring using molecular
modeling, X-ray crystallography, and NMR. We found that the directionality of the amide at
the 8-position had little effect on activity and the (E)-olefin analogue retained significant
potency, indicating that the trans orientation of the amide, and not the carbonyl or NH groups,
made the largest contribution to the observed activity. For the diazepine ring, with the exception
of the closely analogous 3-oxo-2-benzazepine ring system described previously, all of the
modifications led to a significant reduction in activity compared to the potent 3-oxo-1,4-
benzodiazepine parent ring system, implicating this particular type of ring system as a desirable
structural feature for high potency. Energy minimizations of a number of the modified analogues
revealed that none could adopt the same low-energy conformation as the one shared by the
active (S)-isomer of the 3-oxo-1,4-benzodiazepines and 3-oxo-2-benzazepines. The overall data
suggest that the features contributing to the observed high potency in this series are the
orientation of the 3-4 amide and the conformational constraint imposed by the seven-membered
ring, both of which position the key acidic and basic groups in the proper spatial relationship.
In tr od u ction
fore, the development of fibrinogen receptor antagonists
for the treatment of thrombogenic disorders has been
The aggregation of platelets is a critical event in
hemostasis and arterial thrombosis. However, the un-
controlled deposition of platelets and the formation of
large aggregates can lead to vascular occlusion resulting
in serious ischemic disorders such as acute myocardial
infarction (AMI), unstable angina, or thrombotic stroke.
Platelets aggregate in response to a wide variety of
agonists including adenosine diphosphate (ADP), throm-
boxane A2, platelet activating factor (PAF), epinephrine,
serotonin, thrombin, vasopressin, and collagen.1 Inde-
pendent of the agonist, the final step in platelet ag-
gregation and subsequent thrombus formation is the
cross-linking of the dimeric plasma protein fibrinogen
to its receptor, the membrane-bound glycoprotein com-
plex GPIIb/IIIa, on adjacent activated platelets.2 There-
the focus of considerable research in recent years.3
The minimum binding epitope of fibrinogen to its
receptor, a member of the integrin superfamily of
adhesion receptors, is defined by the tripeptide se-
quence: Arg-Gly-Asp (RGD). Extensive early work on
modified peptides led to the identification of short linear
peptides containing the RGD sequence which effectively
block platelet aggregation. These studies pointed to the
basic group of the arginine and the acidic group of the
aspartic acid as the key elements for antagonist activ-
ity.4,5 Subsequent studies showed that incorporating the
RGD motif into constrained, cyclic peptides was a
successful approach to the discovery of high-affinity
fibrinogen receptor antagonists.6-13 Structural and con-
formational analyses of such constrained peptides led
to several related proposals for a biologically important
conformation about the RGD backbone, including a
“turn-extended-turn” conformation,14-17 a “cupped” pre-
sentation,18 and a Gly-Asp â-turn,19,20 which were used
† Department of Medicinal Chemistry.
| Department of Cellular Biochemistry.
‡ Department of Cardiovascular Pharmacology.
∇ Department of Analytical Chemistry.
§ Department of Physical and Structural Chemistry.
10.1021/jm980166z CCC: $18.00 © 1999 American Chemical Society
Published on Web 02/10/1999

546 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4
Keenan et al.
F igu r e 1. Comparison of the structures of fibrinogen receptor antagonists cyclic peptide 1, peptidomimetic 2, and nonpeptide 3.
Sch em e 1a
a
(a) SOCl2, NaN3, toluene, 80 °C; (b) 4-(N-Cbz-amidino)benzoyl chloride; (c) 10% Pd/C; (d) 3 N HCl; (e) SOCl2, NaBH4; (f) CBr4, Ph3P;
(g) Ph3P; (h) 3-(N-Boc-piperidin-4-yl)propionaldehyde, NaH; (i) 1 N NaOH; (j) TFA.
as templates for the successful design of potent non-
peptide RGD mimetics. The development of RGD mi-
metics as potent nonpeptide fibrinogen receptor antago-
nists therefore not only offers a promising new treatment
for thrombotic disorders but also represents a consider-
able achievement in peptidomimetic drug design.21
In our previous work, we used NMR, X-ray crystal-
lographic, and conformational data to identify a low-
energy turn-extended-turn conformation for the RGD
backbone of the highly potent cyclic peptide 1 (Figure
1).14,15 The conformational mimetic 2 containing the
seven-membered ring γ-turn mimetic about Asp re-
tained a good deal of the activity of the cyclic peptide 1,
supporting the biological importance of the proposed
model.22 Finally, the 3-oxo-1,4-benzodiazepine nonpep-
tide fibrinogen receptor antagonist 3 had equal potency
to the constrained cyclic peptide 1 used as the template
in its design.23 The critical features incorporated into
the structure of nonpeptide 3 were the diazepine ring
to mimic a γ-turn at Asp, the phenyl ring to maintain
an extended Gly conformation, and the exocyclic amide
at the 8-position which overlaid the Arg carbonyl. These
structural features closely mimicked the turn-extended-
turn Arg-Gly-Asp backbone found in 1 and were believed
to contribute to the potent activity observed for 3.
To date, we have described modifications to the Arg
mimetic region of the nonpeptide which have led to
surprising improvements in oral bioavailability and
duration of action in vivo,24 including a potent series of
7-substituted 3-oxo-1,4-benzodiazepines.25 Also, we have
reported that replacement of N-1 of the benzodiazepine
nucleus with CH2 afforded a series of 3-oxo-2-benza-
zepines with comparable activity, indicating that nei-
ther the nitrogen lone pair nor the N-H group at this
position is required for high affinity.26 In this paper, we
report on the synthesis and activity of 8-substituted
benzodiazepine analogues containing modifications to
the exocyclic amide and to the seven-membered diaz-
epine ring and also discuss the implications of the data
on our original design hypothesis.
Ch em istr y
The syntheses of the analogues in Table 1 are depicted
in Scheme 1. The carboxylic acid 2324a was subjected to
Curtius rearrangement via conversion to the acyl azide
to furnish the 8-amino derivative 24 in modest yield.
Coupling with the acid chloride derived from 4-(N-Cbz-
amidino)benzoic acid27 and global deprotection afforded
the retro-amide analogue 5. Reduction of carboxylic acid
23 gave the alcohol 25, which was converted to a
mixture of olefin isomers 26 and 27 via a Wittig
reaction. The isomers were separated and deprotected
to give the (E)- and (Z)-olefin analogues 8 and 9.
Hydrogenation of the olefin mixture and subsequent
deprotection provided the saturated compound 10.
The synthesis of the other compounds described in
this paper involved construction of the corresponding
6-7 ring system followed by attachment of the amidi-

Conformational Preferences in Benzodiazepine Series
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4 547
Sch em e 2a
a
(a) ClCOCOCl, tBuOH; (b) H2, Pd/C, MeOH; (c) (Boc)2O, DMAP; (d) 1 N NaOH; (e) methyl 4-phenethylaminocrotonate, Et3N, HOBt,
BOP reagent; (f) TFA, CH2Cl2; (g) MeOH, reflux; (h) SOCl2, 4-(Boc-amidino)aniline; (i) 20% HOAc, reflux; (j) Cbz-Cl; (k) methyl
4-phenethylaminocrotonate, Et3N.
Sch em e 3a
a
(a) Mg, ZnBr2, 5-bromo-1-pentene; (b) O3, MeOH/CH2Cl2, NaBH4; (c) Ac2O, pyridine; (d) HNO3/H2SO4; (e) 1 N NaOH; (f) CsHCO3,
benzyl bromide, DMF; (g) PCC, CH2Cl2; (h) Ph3PdCHCO2Et; (i) SnCl2, EtOH.
Sch em e 4a
nophenyl group to the aromatic ring via an amide linker
using standard coupling conditions as previously de-
scribed.27 Global deprotection then afforded the desired
product. The various ring systems were constructed
utilizing a variety of synthetic pathways outlined below.
The modified seven-membered ring compounds 12-
a
15 in Table 2 were made via an intramolecular conju-
gate addition of the appropriate aniline to an unsatur-
ated ester. Scheme 2 shows the syntheses of analogues
12 and 13, in which the amide carbonyl has been moved
to the 5-position on the benzodiazepine ring. For com-
pound 12, partial hydrolysis of dimethyl nitrotereph-
thalate afforded 28, which was converted to 29 by
reesterification as the tert-butyl ester and subsequent
reduction of the nitro group and protection as its Boc
derivative. Selective ester hydrolysis and coupling with
methyl 4-(phenethylamino)crotonate27 gave the cycliza-
tion precursor 30 which, after acid deprotection, under-
went an intramolecular conjugate addition to furnish
31. For the synthesis of compound 13, esterification and
protection of 5-nitroanthranilic acid yielded 32. Nitro
group reduction and protection gave the methyl ester
(a) DMAD, MeOH; (b) H2, Pd/C.
33, which was converted as before to 13 via the ring-
closed intermediate 34. Alkylation of the aminocrotonate
with the benzyl bromide 3527 furnished 36. Reduction
of the nitro group and cyclization as before provided the
bicyclic intermediate for the synthesis of the related
reduced amide analogue 14.
The precursor for the azepine analogue 15 was
synthesized as depicted in Scheme 3. Palladium-
catalyzed coupling of the bromide 37 with 4-pentenyl-
zinc iodide, ozonolysis of the olefin followed by reductive
workup, protection of the primary alcohol, and ring
nitration afforded 38. Treatment with base liberated the
alcohol, and the benzoic acid was reesterified to yield
the benzyl ester 39. Oxidation and chain elongation

548 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4
Keenan et al.
Sch em e 5a
Ta ble 1. Exocyclic Linker SAR
a
(a) Boc-glycine, BOP reagent; (b) 4 M HCl, dioxane; (c) Et3N,
DMSO; (d) H2, Pd/C.
Sch em e 6a
a
(a) KNO3, H2SO4; (b) H2NNH2; (c) 1,1′-carbonyldiimidazole;
(d) Lawesson’s reagent; (e) CH3I, K2CO3, BrCH2CO2CH3; (f) H2,
Pd/C.
afforded the unsaturated ester 40. Ring closure to 41
occurred in situ during reduction of the nitro group with
stannous chloride. For the synthesis of the ring-opened
analogue 16, intermediate 43 was made from reaction
of 3-aminobenzoic acid (42) with DMAD followed by
catalytic reduction (Scheme 4).
a
b
Reference 24a. Reference 24b.
inhibit ADP-induced platelet aggregation in human
platelet-rich plasma.29
The regioisomeric benzodiazepine analogues 17 and
18 in Table 2 were made from an intramolecular
nucleophilic aromatic substitution reaction. For 18,
reaction of 2-fluoro-5-nitrobenzaldehyde with tert-butyl
3-aminopropionate under reducing conditions afforded
the amine 44, which was acylated with Boc-glycine and
deprotected to give 45 (Scheme 5). Ring closure to 46
was effected by treatment with triethylamine overnight
in DMSO. Compound 17 was made in similar fashion,
but substituting tert-butylglycine in the first step.
The 2-azepine analogues 19 and 20 were made using
chemistry analogous to that developed for the synthesis
of the benzazepine series of GPIIb/IIIa antagonists.28
Here, methyl acrylate was employed in the palladium-
catalyzed addition in place of dimethyl itaconate, and
either glycine methyl ester or â-alanine methyl ester
replaced phenethylamine as the amine component. The
urea analogue 21 was synthesized via nonselective
alkylation of the thiourea derived from urea 48 which
was made in four steps from R,R′-dibromo-o-xylene
(Scheme 6).
A series of analogues was prepared to investigate the
role of the exocyclic amide at the 8-position of the
benzodiazepine nucleus (Table 1). We have previously
reported that N-methylation of the amide to yield 4 led
to a slight increase in in vitro activity, most notably in
the platelet aggregation assay.24a Interestingly, the
retro-amide analogue 5 showed comparable activity to
3 in both binding and inhibition of platelet aggregation
indicating the directionality of the exocyclic amide bond
was not important for activity in this series. We also
investigated olefin replacements, in this case in the
piperidinoethyl series,24b where the N-methylamide 7
exhibited similar binding affinity but had much im-
proved platelet aggregation inhibitory activity compared
to the secondary amide 6.30 The (E)-olefin 8 had 40-fold
better activity in both assays compared to the (Z)-olefin
9, and only 4-fold diminished binding affinity relative
to 6. The data from this set of analogues revealed that
the trans orientation of the amide was likely the
bioactive conformation and suggested the carbonyl and
NH had minimal contributions to the observed activity.
The incorporation of a seven-membered diazepine ring
to mimic the γ-turn observed around Gly-Asp in the
peptide model was another of the critical elements in
our original design hypothesis.23 A number of modifica-
tions to the seven-membered ring amide region were
investigated to probe the structural requirements for
activity (Table 2). Compounds 12 and 13, containing the
Resu lts a n d Discu ssion
For the compounds in this study, fibrinogen receptor
binding affinity was determined by assaying for inhibi-
tion of [3H]-1 binding to purified GPIIb/IIIa isolated
from human platelets and reconstituted in liposomes.
Also, the compounds were evaluated for their ability to

Conformational Preferences in Benzodiazepine Series
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4 549
Ta ble 2. Seven-Membered Ring SAR

550 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4
Ta ble 2 (Continued)
Keenan et al.
a
b
Reference 24a. Reference 26.
regioisomeric 5-oxo-1,4-benzodiazepine ring system,
were much less potent than the parent 3-oxo com-
pounds. In addition, reduction of the amide carbonyl (14)
or removal of the amide altogether (15) also led to a
significant reduction in activity compared to 3. Finally,
opening the seven-membered ring to the meta-disub-
stituted benzene compound 16 caused a similar reduc-
tion of activity, which is notable in light of the numerous
reports of highly potent fibrinogen receptor antagonists
which lack the constraints of a bicyclic ring system.31
Thus, an intact seven-membered ring containing a 3-4
amide contributes significantly to the observed fibrino-
gen receptor antagonist activity of the potent 3-oxo-1,4-
benzodiazepines such as 3. The contribution to activity
by the ring amide could be the result of a binding
interaction between the receptor and the amide carbonyl
group and/or the stabilization by the ring amide of a
highly potent conformation of the seven-membered ring.
Previously, we have shown that replacement of N-1
of the benzodiazepine nucleus with CH2 affords a 3-oxo-
2-benzazepine series with comparable activity (i.e., 11,
Table 2), indicating that neither the nitrogen lone pair
nor the N-H group at this position is required for high
affinity.26 We have also examined a number of related
analogues in which the chiral carbon atom in the seven-
membered ring has been replaced by a nitrogen, but the
key carbonyl at position 3 has been retained (Table 2).
All of the analogues 17-21 containing such modified
seven-membered rings were much weaker in activity
than the benzazepine 11. Although 17-21 lack a
phenethyl side chain on the seven-membered ring, this
substitution has been shown not to be important for
good activity in either the 3-oxo-2-benzazepine26 or
3-oxo-1,4-benzodiazepine24b systems. Thus, the chiral
center on the seven-membered ring appears to be crucial
for good activity and possibly is instrumental in placing
the carboxylate in the proper position for binding. Still,
these less active analogues all contain the carbonyl
group at position 3 found in the potent benzodiazepine
and benzazepine analogues, which would suggest either
that this carbonyl group is not a receptor-binding
element or that the conformation of the seven-mem-

Conformational Preferences in Benzodiazepine Series
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4 551
bered ring in these less active analogues positions the
carbonyl group in an unfavorable orientation. To shed
further light on this issue, we have performed molecular
modeling studies to examine the families of conforma-
tions available to each of the 6-7 ring systems described
in this paper.
Mod elin g a n d X-r a y Cr ysta llogr a p h y. Each ring
system was subjected to a Monte Carlo32 simulation
(1000 iterations) using an Amber33 force field in a water
matrix. All conformers within 50 kJ /mol of the global
minimum were saved and ranked by relative energy.
The substituent at N-4 was truncated to a methyl group,
and the arginine mimetic was replaced by an N-
methylamide for simplicity. The (S)-absolute stereo-
chemistry of the 3-substituent was fixed in AM1, and
all corresponding analogues maintained the same ab-
solute stereochemistry.34
The modeling of the parent 3-oxo-1,4-benzodiazepine
ring system (AM1) showed two low-energy families of
conformers, AM1-A and AM1-B, whose lowest energy
members differed by 10.6 kJ /mol. These consisted of the
two flip forms of the seven-membered ring (Table 3).
Replacement of N-1 with a methylene group to yield a
3-oxo-2-benzazepine ring system had no effect on the
observed ring conformers C-AM1 or the relative energy
of the two flip forms of the seven-membered ring,
consistent with the similar biological activity observed
for compounds containing these two closely related ring
systems.
We previously reported an X-ray crystal structure of
the racemic benzodiazepine 22a (Table 4).24a As seen
in Figure 2, the benzodiazepine ring of the (S)-enanti-
omer of 22a is superimposable upon AM1-A. We now
have obtained several X-ray crystal structures of various
3-oxo-1,4-benzodiazepine and 3-oxo-2-benzazepine com-
pounds (22b-f, Table 4), all of which display the same
AM1-A conformation about the seven-membered ring.
Since the crystal packing should be different due to the
different substituents in these analogues, the X-ray data
strongly implies that AM1-A must be an inherently low-
energy conformation for the 3-oxo-1,4-benzodiazepine
ring system.
The proton NMR spectrum of 22c in hexadeuterioac-
etone at 25 °C displays one set of multiplets for each
proton in the spectrum and does not suggest the
presence of multiple conformers. Lowering the temper-
ature to -60 °C induces a shift in the proton at N-1 but
has no effect on any of the other resonances, supporting
the lack of dynamic behavior in the benzodiazepine ring.
As with all other 3-oxo-1,4-benzodiazepine analogues,
the methylene protons at position 5 of 22c display a
characteristic doublet at ∼5.5 ppm that arises from the
different magnetic environments experienced by each
methylene proton. In the conformation AM1-A, one
proton (H-5a, Figure 3) resides above the plane of the
adjacent aromatic ring, while the other (H-5b, Figure
3) sits within the plane. In the proton spectrum, the
NOE enhancements between H-5a and H-2 and between
H-5b and H-6 (Figure 3) are consistent with the lowest
energy conformation of AM1 (AM1-A), wherein H-5a is
separated by only 2 Å from H-2.
observed by both proton NMR spectroscopy and X-ray
crystallography. Thus AM1-A may be the conformation
responsible for the pharmacological activity of these
3-oxo-1,4-benzodiazepine fibrinogen receptor antago-
nists.
In Table 3, we have compared the conformations
obtained from molecular modeling analysis of the other
6-7 ring systems to AM1-A, to probe for differences
which may explain the reduction in biological activity.
The analogous structure AM2, which contains a 4-5
amide in place of a 3-4 amide, also displays two families
of conformations which represent two flip forms of the
seven-membered ring. Both the vector of the acetic acid
substituent and the ring pucker of the seven-membered
ring in the lower energy conformers AM2-A are signifi-
cantly different when compared to those in AM1-A. The
analogue 12, which contains an AM-2 type 6,7 template,
displays considerably lower activity than the parent 3,
which contains an AM1 type template. These results
suggest that the lower energy conformers AM2-A place
the acetic acid group in an unfavorable orientation,
while the higher energy conformers AM2-B place the
ring amide in an unfavorable orientation.
The conformers of AM3, which are related to analogue
14, display a variety of low-energy conformations due
to the absence of the constraining 3-4 amide bond.
Interestingly, the conformer AM3-A positions the acetic
acid group in an equatorial orientation in the plane of
the aromatic ring similar to that in AM1-A, with minor
variations in the seven-membered ring conformation at
C-3 and C-4 and the obvious loss of the carbonyl oxygen.
The higher energy conformers AM3-B and AM3-C vary
either the position of the acetic acid side chain or the
conformation of the seven-membered ring. The lower
affinity of 14 which lacks the ring amide suggests that
the 3-4 amide may contribute a positive binding
interaction in addition to its stabilizing conformational
role. The potential importance of the amide’s interaction
is also seen in the strong preference for the (S)-absolute
stereochemistry at C-2 in potent (S)-3-oxo-1,4-benzodi-
azepine GPIIb/IIIa antagonists (3S vs 3R, Table 2).24a,25
The less potent (R)-isomers maintain an equatorial
acetic acid side chain but have a different relative
orientation of the seven-membered ring amide in inter-
acting with the receptor (Figure 4). The reduction in
activity for the acyclic compound 16, where the ap-
propriate pseudoequatorial conformation of the carboxyl-
ate-containing side chain should be accessible, also
points to a positive binding interaction for the ring
amide. However, in each of these examples, it is not
possible to rule out a negative interaction at the receptor
caused by the introduction of new functionality as the
reason for the dropoff in activity.
The conformers of AM4, which are related to analogue
19, contain a 2-3 amide instead of a 3-4 amide. The
vector of the acetic acid side chain in all of the conform-
ers AM4-A, AM4-B, and AM4-C position the acetic acid
side chain below the plane of the aromatic ring with a
vector significantly different than seen in AM1-A. This
is due to the fact that the side chain originates from an
amide nitrogen and not a methylene group, providing
a rationale for the low activity of 19. Similar results
were observed using the diazepine structure related to
17, another indication that interchange of carbon and
These molecular modeling studies of the 3-oxo-1,4-
benzodiazepine ring system have identified a low-energy
conformation AM1-A which is also the only conformation

552 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4
Keenan et al.
Ta ble 3. Conformations of the Different Seven-Membered Ring Systems
a
Relative energy in kJ /mol.

Conformational Preferences in Benzodiazepine Series
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4 553
Ta ble 4. Solved Benzodiazepine and Benzazepine X-ray
Structures
F igu r e 5. Comparison of lowest energy conformations of AM1
(blue), AM1′ (green), AM2 (yellow), AM3 (red), AM4 (purple),
and AM5 (light blue).
The low-energy conformations of the cyclic urea-
containing analogue 21 are represented by AM5-A and
AM5-B. Neither of these conformers is capable of
positioning the acetic acid side chain within the plane
of the aromatic ring, and therefore a favorable interac-
tion with the receptor cannot be achieved.
In summary, the data collected in the Monte Carlo
simulations allows for the rationalization of the biologi-
cal data for this series of related compounds. Analogues
containing the seven-membered ring conformation found
in AM1-A showed the optimal biological profile of the
ring systems described. The related 6-7 ring analogues
with diminished activity adopt conformations in which
either the acetic acid or the ring amide groups are
positioned in regions of space other than those of AM1-A
(Figure 5). Although not conclusive, these calculations
suggest the ring amide of the 3-oxo-1,4-benzodiazepine
and 3-oxo-2-benzazepine analogues may contribute a
positive binding interaction in addition to its stabilizing
conformational role.
F igu r e 2. Overlay of AM1-A, the low-energy conformation
calculated for the 3-oxo-1,4-benzodiazepine ring system, with
the conformation derived from the X-ray structure of 22a .
Con clu sion
The initial benzodiazepine analogues designed from
the peptide model contain a number of structural
features which map directly onto the peptide backbone
of Arg-Gly-Asp in a proposed bioactive conformation.23
Our subsequent studies have shown that the NH at
position 1 of the benzodiazepine ring which aligns with
the Gly-Asp amide bond NH is not necessary for good
fibrinogen receptor-binding affinity.26 In this paper, we
have shown that the carbonyl at the 8-position which
corresponds to the Arg-Gly amide bond carbonyl is also
not important for good activity. In addition, we have
demonstrated that the 3-oxo-1,4-benzodiazepine or 3-oxo-
2-benzazepine displays substantially better antiaggre-
gatory activity than several closely related ring systems.
Computational analysis reveals that these two ring
systems are able to adopt a specific conformation of the
seven-membered ring which may be associated with
enhanced activity. The overall data suggest that the
high potency in this series of compounds arises from the
orientation of the ring amide bond and the conforma-
tional constraint it imposes on the seven-membered
ring35 to position the key acidic and basic groups in the
preferred relative orientation.
F igu r e 3. Depiction of conformations AM1-A and AM1-B of
the 3-oxo-1,4-benzodiazepine ring system showing the observed
NOE relationships consistent with conformation AM1-A.
F igu r e 4. Overlay of the (R)- and (S)-enantiomers of the
lowest energy conformation AM1-A.
nitrogen at position 1 of the benzodiazepine ring has
little to no conformational impact in these ring systems.
The fact that extension of the acetic acid side chains as
in analogues 18 and 20 gives modest increases in
affinity suggests that greater flexibility of the propionate
side chain allows for a slightly better fit with the
receptor by overcoming the original vector from the
template.
Exp er im en ta l Section
Gen er a l. Melting points were measured with a Thomas-
Hoover melting point apparatus and are uncorrected. Proton
NMR spectra were obtained using Bruker AM-250 or AM-400

554 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4
Keenan et al.
spectrometers and are reported as ppm downfield from TMS
with multiplicity, number of protons, and coupling constant-
(s) in hertz (Hz) indicated parenthetically. Elemental analyses
were obtained using a Perkin-Elmer 240C elemental analyzer.
Chromatography refers to flash chromatography using Kie-
selgel 60, 230-400 mesh silica gel.
NMR Sp ectr oscop y of 22c. 1H NMR spectra were mea-
sured at 400.13 MHz using a Bruker Instruments AMX 400
for 10 mg/mL solutions of 22c in acetone-d6. Nuclear Over-
hauser enhancements (NOE) were determined by difference
spectroscopy, requiring collection of a series of spectra in which
the resonance frequencies of interest, including an off-
resonance control, were selectively irradiated with low (60-
db) RF power for 4 s prior to acquisition. For each spectrum
eight scans were averaged; then the process was repeated for
four cycles. Following Fourier transformation, difference spec-
tra were generating by subtracting the control spectrum from
each irradiated spectrum.
oxo-4-(2-p h en ylet h yl)-1H -1,4-b en zod ia zep in e-2-a cet a t e
(25). Carboxylic acid 23 (3.8 g, 10 mmol) was treated with
thionyl chloride (50 mL), and the resulting suspension was
heated at reflux under argon with stirring for 15 min, cooled
to room temperature, and concentrated. The residual oil was
redissolved in toluene (25 mL) and concentrated (2×), and the
residue was taken up in dry tetrahydrofuran (50 mL). The
solution was stirred at room temperature, and sodium boro-
hydride (2.0 g, 53 mmol) as added in one portion. After stirring
for 16 h, the reaction was carefully quenched at 0 °C with 1 N
hydrochloric acid, basified with 1 N sodium carbonate, ex-
tracted with ethyl acetate (150 mL), washed with brine, dried
(magnesium sulfate), and concentrated. Purification by flash
chromatography (1% methanol:chloroform) gave 25 (1.71 g,
47%) as a solid foam: 1H NMR (CDCl3) δ 2.62 (dd, 1H), 2.77
(m, 2H), 2.97 (dd, 1H), 3.27 (br s, 2H), 3.68 (m, 3H), 3.72 (s,
3H), 4.51 (s, 2H), 4.94 (t, 1H), 5.24 (d, J ) 16.5 Hz, 1H), 6.56
(s, 1H), 6.62 (d, J ) 7.4, 1H), 6.80 (d, J ) 7.7 Hz, 1H), 7.20 (m,
5H).
Variable-temperature spectra were measured in both sol-
vents with the AMX 400 instrument using a Eurotherm
temperature controller, calibrated against an external metha-
nol standard. Spectra were measured over a range of ca. 100
K with the lowest temperature (ca. -70 °C) limited by the
characteristics of the solvent.
Meth yl (R,S)-8-[4-[N-(ter t-Bu toxyca r bon yl)p ip er id in -
4-yl]-1(E)-bu ten yl]-2,3,4,5-tetr a h yd r o-3-oxo-4-(2-p h en yl-
eth yl)-1H-1,4-ben zod ia zep in e-2-a ceta te (26) a n d Meth yl
(R,S)-8-[4-[N-(ter t-Bu t oxyca r bon yl)p ip er id in -4-yl]-1(Z)-
bu t en yl]-2,3,4,5-t et r a h yd r o-3-oxo-4-(2-p h en ylet h yl)-1H-
1,4-ben zod ia zep in e-2-a ceta te (27). (a ) To a solution of
alcohol 25 (1.20 g, 3.3 mmol) and carbon tetrabromide (1.4 g,
4.2 mmol) in dry tetrahydrofuran (40 mL) stirred at 0 °C under
argon was added triphenylphosphine (1.1 g, 4.2 mmol) in one
portion. After 10 min, the reaction was allowed to warm to
room temperature and stirred for 3 h. The mixture was
concentrated, and the residue was purified by flash chroma-
tography (50% ethyl acetate:hexane) to yield the bromomethyl
intermediate (1.17 g, 83%). To this compound (1.15 g, 2.67
mmol) in dry tetrahydrofuran (30 mL) was added triphen-
ylphosphine (0.71 g, 2.7 mmol). The reaction was refluxed for
4 h, cooled to room temperature, concentrated, and triturated
with ether, filtered, and dried under vacuum to give the
triphenylphosphonium bromide Wittig reagent (1.89 g,
100%): 1H NMR (CDCl3) δ 2.01 (1H, s), 2.73 (2H, m), 2.90 (1H,
dd), 3.08 (1H, dd), 3.54 (1H, m), 3.65 (1H, m), 3.70 (3H, s),
3.74 (1H, d, J ) 16.7 Hz), 4.49 (2H, dt), 5.08 (1H, d, J ) 15.5
Hz), 5.17 (1H, t), 6.40 (1H, d), 6.58 (1H, d, J ) 11.4 Hz), 6.87
(1H, s), 7.18 (5H, m), 7.58-7.80 (15H, m).
(b) To a stirred solution of oxalyl chloride (2 mL, 22 mmol)
in dry dichloromethane (50 mL) at -78 °C under argon was
added dropwise a solution of dimethyl sulfoxide (3.4 mL, 44
mmol) in dichloromethane (10 mL). After 2 min, a solution of
3-[N-(tert-butoxycarbonyl)piperidin-4-yl]propanol (4.86 g, 20
mmol) in dichloromethane (10 mL) was added dropwise over
5 min. After stirring an additional 15 min at -78 °C,
triethylamine (14 mL) was added dropwise. After 5 min, the
reaction became a thick white slurry and was allowed to warm
to room temperature and stirred for 16 h. The reaction mixture
was washed with cold 1 N hydrochloric acid and with brine,
dried (magnesium sulfate), and concentrated. The residue was
purified by flash chromatography (30% ethyl acetate:hexane)
to yield 3-[N-(tert-butoxycarbonyl)piperidin-4-yl]propionalde-
hyde (4.18 g, 87%) as a clear oil which solidified in the
freezer: 1H NMR (CDCl3) δ 0.95-1.80 (7H, m), 1.48 (9H, s),
2.40-2.90 (4H, m), 4.14 (2H, br dt), 9.88 (1H, t).
Met h yl (R,S)-8-Am in o-1,2,4,5-t et r a h yd r o-3-oxo-4-(2-
ph en yleth yl)-3H-1,4-ben zodiazepin e-2-acetate (24). A mix-
ture of carboxylic acid 2324a (0.6 g, 1.3 mmol) and thionyl
chloride (3 mL) was heated to reflux under argon for 15 min.
The mixture was concentrated in vacuo, treated with meth-
ylene chloride (3 × 20 mL), and concentrated in vacuo to give
the acid chloride as a yellow solid. The acid chloride was
dissolved in dry acetone (6 mL) and added dropwise to a
solution of sodium azide (120 mg, 1.8 mmol) in water (3 mL)
stirred in an ice bath. The mixture was stirred for 1 h, diluted
with water (15 mL), and extracted with ethyl acetate. The
organic phase was dried with magnesium sulfate and concen-
trated in vacuo. The residue was dissolved in dry toluene (12
mL) and heated to 80 °C in an argon atmosphere for 2 h. The
mixture was concentrated in vacuo, and the residue was
stirred in 3 N HCl (6 mL) and tetrahydrofuran (10 mL) for 1
h. The mixture was concentrated in vacuo, and solid sodium
bicarbonate was added to pH 8. The mixture was extracted
with ethyl acetate, and the organic phase was dried with
magnesium sulfate and concentrated in vacuo to give 24 (0.3
g, 67%).
(R,S)-8-[(4-Am id in oben zoyl)a m in o]-1,2,4,5-tetr a h yd r o-
3-oxo-4-(2-p h en yleth yl)-3H-1,4-ben zod ia zep in e-2-a cetic
Acid (5). A mixture of 4-[N-(benzyloxycarbonyl)amidino]-
benzoic acid (0.23 g, 0.001 mol) and thionyl chloride (3 mL) in
methylene chloride (3 mL) was heated to reflux for 10 min,
concentrated in vacuo, treated with toluene, and concentrated
in vacuo several times to give 4-[N-(benzyloxycarbonyl)-
amidino]benzoyl chloride. The acid chloride was dissolved in
methylene chloride (3 mL) and added dropwise to a solution
of 24 (0.3 g, 0.9 mmol) and diisopropylethylamine (130 mg, 1
mmol) in dry methylene chloride (5 mL). The mixture was
stirred for at room temperature under argon for 5 h, diluted
with methylene chloride (20 mL), and extracted with water, 3
N HCl, 5% sodium bicarbonate, and brine. The organic phase
was dried with magnesium sulfate and concentrated in vacuo.
The residue was chromatographed on silica gel eluted with
2:98 methanol:methylene chloride to give the coupled product
(0.17 g, 32%). A solution of the coupled product (0.1 g, 0.15
mmol) in methanol (40 mL) containing 3 N HCl (8 drops) and
10% Pd/C (20 mg) was shaken in a hydrogen atmosphere (45
psi) for 30 min. The mixture was filtered and the filtrate
concentrated in vacuo. The crude amidine was dissolved in
methanol (15 mL), water (2 mL), and 1 N NaOH (1 mL) and
was stirred at room temperature under argon overnight. The
mixture was treated with 3 N HCl (1 mL) and concentrated
in vacuo. The residue was dissolved in 33:67 acetonitrile:water
and purified by HPLC to give 5 (26 mg, 33%): MS (EI) m/e
486 (M + H)+.
(c) To a solution of the Wittig reagent (1.0 g, 1.44 mmol)
and 3-[N-(tert-butoxycarbonyl)piperidin-4-yl]propionaldehyde
(0.42 g, 1.75 mmol) in dry tetrahydrofuran (30 mL) was added,
with stirring under argon, a 60% dispersion of sodium hydride
in mineral oil (58 mg, 1.44 mmol) in one portion. After stirring
for 16 h, the reaction was concentrated. The residue was
purified by flash chromatography (5% ethyl acetate:chloroform)
to give a mixture of the title compounds (0.717 g, 86%). Pure
olefin isomers in the ratio of approximately 1:2 Z:E were
obtained by HPLC chromatography [silica gel (Apex, 8 × 250
mm), 40% ethyl acetate:hexane]. (E)-Isomer 26: 1H NMR
(CDCl3) δ 1.11 (2H, m), 1.43 (4H, m), 1.45 (9H, s), 1.68 (2H, d,
J ) 12.5 Hz), 2.20 (2H, dd), 2.67 (2H, dt), 2.72 (1H, dd), 2.81
(1H, dd), 3.72 (3H, m), 3.74 (3H, s), 4.08 (2H, d, J ) 13.7 Hz),
Meth yl (R,S)-2,3,4,5-Tetr a h yd r o-8-(h yd r oxym eth yl)-3-

Conformational Preferences in Benzodiazepine Series
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4 555
4.90 (1H, t), 5.18 (1H, d, J ) 16.5 Hz), 6.14 (1H, dt, J trans
)
lin e (29). To a stirred suspension of the acid 28 (7.0 g, 31
mmol) in toluene (25 mL) was added oxalyl chloride (5 mL, 57
mmol) followed by 1 drop of dry DMF. After stirring overnight
(the solution became clear after ∼30 min), the reaction was
evaporated to dryness and reevaporated from fresh toluene
two times. To the resulting acid chloride were added CHCl3
(10 mL) and t-BuOH (1 mL) followed by pyridine (0.6 mL).
After stirring for 6 h the reaction was evaporated, taken up
in ethyl acetate, washed with 1 N NaHCO3 and brine, dried
over MgSO4, and evaporated. Purification by flash chroma-
tography eluting with 10% ethyl acetate in hexane gave the
intermediate ester as an oil (7.28 g, 83%): 1H NMR (CDCl3) δ
1.6 (9H, s), 3.95 (3H, s), 7.83 (1H, d, J ) 8 Hz), 8.3 (1H, dd),
8.5 (1H, d, J ) 2 Hz). A solution of the ester (7.28 g, 26 mmol)
in MeOH (100 mL) was hydrogenated over 5% Pd/C (0.5 g) at
50 psi H2 in a Parr shaker for 4 h. After filtration of the
catalyst through a pad of Celite and evaporation of the filtrate,
the remaining residue was taken up in CH2Cl2 (100 mL) to
which was added di-tert-butyl dicarbonate (6.2 g, 28.4 mmol)
followed by DMAP (0.6 g, 5 mmol) with stirring. The reaction
was refluxed for 16 h, then evaporated to dryness, taken up
in ethyl acetate, washed with 1 N HCl and brine, dried over
MgSO4, and evaporated. Purification by flash chromatography
on silica gel eluted with 10% ethyl acetate in n-hexane gave
product 29 (3.68 g, 40%): 1H NMR (CDCl3) δ 1.55 (9H, s), 1.6
(9H, s), 3.93 (3H, s), 7.6 (1H, dd), 8.05 (1H, d, J ) 8 Hz), 9.1
(1H, d, J ) 2 Hz), 10.28 (1H, s).
(Z)-Meth yl 4-[N-[4,N-Bis(ter t-bu toxyca r bon yl)-2-a m i-
n oben zoyl]-N-(2-p h en yleth yl)a m in o]-2-bu ten oa te (30). To
a stirred solution of ester 29 (3.68 g, 10.5 mmol) in dioxane
(50 mL) was added aqueous 1 N NaOH (12 mL). The reaction
was stirred for 3 h, acidified with aqueous 1 N HCl (12 mL),
extracted with ethyl acetate, washed with brine, dried over
MgSO4, and evaporated. To the remaining solid were added
DMF (75 mL), (Z)-methyl 4-(phenethylamino)-2-butenoate27
(3.7 g, 16.9 mmol), Et3N (7.3 mL, 52 mmol), and HOBt (2.8 g,
20.7 mmol) followed by BOP reagent (5.6 g, 12.7 mmol). After
stirring overnight the reaction was evaporated to dryness.
Purification by flash chromatography eluting with 99:1 CHCl3,
MeOH gave product 30 as a solid foam (3.59 g, 62%).
15.8 Hz), 6.24 (1H, d, J trans ) 15.8 Hz), 6.67 (1H, s), 6.72 (1H,
dd, J ) 7.8 Hz), 6.80 (1H, d, J ) 7.8 Hz), 7.20 (5H, m); MS
(ES) m/e 576.2 (M + H)+. (Z)-Isomer 27: 1H NMR (CDCl3) δ
1.08 (2H, m), 1.41 (4H, m), 1.45 (9H, s), 1.62 (2H, d, J ) 12.9
Hz), 2.31 (2H, dd), 2.64 (2H, dt), 2.72 (1H, dd), 2.82 (2H, dt),
3.03 (1H, dd), 3.73 (3H, m), 3.75 (3H, s), 4.06 (2H, d, J ) 13.2
Hz), 4.93 (1H, t), 5.22 (1H, d, J ) 16.5 Hz), 5.59 (1H, dt, J cis
)
11.6 Hz), 6.27 (1H, d, J cis ) 11.7 Hz), 6.55 (1H, s), 6.63 (1H, d,
J ) 7.7 Hz), 6.82 (1H, d, J ) 7.7 Hz), 7.20 (5H, m); MS (ES)
m/e 576.2 (M + H)+.
(R,S)-2,3,4,5-Tet r a h yd r o-3-oxo-4-(2-p h en ylet h yl)-8-[4-
(p ip er id in -4-yl)-1(E)-bu ten yl]-1H-1,4-ben zod ia zep in e-2-
a cetic Acid (8). Compound 26 (273 mg, 0.47 mmol) was
dissolved in dioxane (50 mL), and 1.0 N NaOH (0.9 mL, 0.9
mmol) was added. The solution was stirred at room temper-
ature overnight, acidified with 1.0 N HCl (0.9 mL, 0.9 mmol),
diluted with CHCl3 (50 mL), washed with brine, dried (MgSO4),
and evaporated. The residue which remained was treated with
neat TFA (20 mL), stirred at room temperature for 30 min,
evaporated to dryness, reevaporated from toluene several
times, triturated with petroleum ether/ether, and then dried
under vacuum. The solid which remained was purified by
reversed-phase flash chromatography (Hamilton PRP-1 (10 ×
250 mm) eluted with 40% CH3CN, H2O, 0.1% TFA) to give the
title compound: 1H NMR (CD3OD) δ 1.38 (2H, m), 1.45 (2H,
m), 1.61 (1H, br s), 1.95 (2H, d, J ) 13.6 Hz), 2.22 (2H, dd),
2.61 (1H, m), 2.77 (2H, m), 2.91 (3H, m), 3.35 (2H, m), 3.67
(2H, m), 3.79 (1H, d, J ) 16.8 Hz), 5.02 (1H, dd), 5.37 (1H, d,
J ) 16.5 Hz), 6.15 (1H, dt), 6.26 (1H, d, J ) 15.9 Hz), 6.58
(1H, s), 6.60 (1H, d, J ) 7.8 Hz), 6.81 (1H, d, J ) 7.8 Hz), 7.17
(5H, m); MS (ES) m/e 462.2 (M + H)+. Anal. (C28H35N3O3‚1.5CF3-
CO2H) Calcd: C, 58.86; H, 5.82; N, 6.64. Found: C, 58.68; H,
5.84; N, 6.32.
(R,S)-2,3,4,5-Tet r a h yd r o-3-oxo-4-(2-p h en ylet h yl)-8-[4-
(p ip er id in -4-yl)-1(Z)-bu ten yl]-1H-1,4-ben zod ia zep in e-2-
a cetic Acid (9). Compound 27 (185 mg, 0.32 mmol) was
hydrolyzed and deprotected as in the preparation of compound
8 to give the title compound: 1H NMR (CD3OD) δ 1.30 (2H,
m), 1.41 (2H, dd), 1.58 (1H, br s), 1.82 (2H, d, J ) 13.1 Hz),
2.34 (2H, m), 2.60 (1H, m), 2.76 (2H, m), 2.88 (3H, m), 3.29
(3H, m), 3.69 (2H, m), 3.81 (1H, d, J ) 16.9 Hz), 5.04 (1H, m),
5.40 (1H, d, J ) 16.6 Hz), 5.58 (1H, dt), 6.31 (1H, d, J ) 11.6
Hz), 6.48 (1H, d, J ) 7.9 Hz), 6.50 (1H, s), 6.85 (1H, d, J ) 7.6
Hz), 7.18 (5H, m); MS (ES) m/e 462.2 (M + H)+. Anal.
(C28H35N3O3‚1.5CF3CO2H‚0.5H2O) Calcd: C, 57.83; H, 5.73; N,
6.27. Found: C, 58.03; H, 5.89; N,6.55.
(R,S)-2,3,4,5-Tet r a h yd r o-3-oxo-4-(2-p h en ylet h yl)-8-[4-
(piper idin -4-yl)bu tyl]-1H-1,4-ben zodiazepin e-2-acetic Acid
(10). A mixture of the (E/Z)-isomers 26 and 27 and 10%
palladium on carbon in acetic acid was shaken under hydrogen
(50 psi) for 4 h to give the saturated compound, which was
deprotected and hydrolyzed as in the preparation of compound
8 to give the title compound: 1H NMR (CD3OD) δ 1.32 (6H,
m), 1.57 (3H, m), 1.87 (2H, d, J ) 14.0 Hz), 2.49 (2H, t), 2.61
(1H, dd), 2.75 (2H, m), 2.90 (3H, m), 3.30 (2H, m), 3.68 (2H,
m), 3.80 (1H, d, J ) 16.7 Hz), 5.02 (1H, dd), 5.37 (1H, d, J )
16.5 Hz), 6.41 (1H, d, J ) 8.0 Hz), 6.42 (1H, s), 6.80 (1H, d, J
) 8.0 Hz), 7.17 (5H, m); MS (ES) m/e 464.2 (M + H)+. Anal.
(C28H37N3O3‚CF3CO2H‚1.25H2O) Calcd: C, 59.91; H, 6.51; N,
6.73. Found: C, 60.04; H, 6.80; N, 7.00.
5-Ca r boxy-2-(m eth oxyca r bon yl)n itr oben zen e (28). To
a stirred solution of dimethyl nitroterephthalate (12 g, 50
mmol) in dioxane (100 mL) was slowly added at room tem-
perature aqueous 1 N NaOH (50 mL) dropwise over 30 min.
After stirring overnight at room temperature the reaction was
diluted with water, washed with ether, acidified with aqueous
1 N HCl (50 mL), and then extracted several times with ethyl
acetate. After the ethyl acetate phase was dried over MgSO4
and evaporated, the resulting residue was purified by flash
chromatography eluting with 98:2:0.1 CHCl3, MeOH, HOAc
to give product 28 (8.07 g, 72%): 1H NMR (CDCl3) δ 3.96 (3H,
s), 7.80 (1H, d), 8.36 (1H, dd), 8.57 (1H, d).
Meth yl (R,S)-8-Ca r boxy-1,2,4,5-tetr a h yd r o-5-oxo-4-(2-
p h en yleth yl)-5H-1,4-ben zod ia zep in e-2-a ceta te (31). To 30
(3.59 g) was added 90% TFA in CH2Cl2 (50 mL). After stirring
at room temperature for 45 min the reaction was evaporated
to dryness and then reevaporated from anhydrous MeOH two
times. The residue which remained was taken up in MeOH
(100 mL) and refluxed for 2 days under Ar. Evaporation to
dryness gave 31 as a slightly yellow solid (2.5 g, 100%): 1H
NMR (CDCl3) δ 2.57 (2H, m), 2.88 (2H, m), 3.41 (1H, dd), 3.50
(2H, m), 3.66 (3H, s), 3.92 (2H, m), 6.18 (1H, s), 6.97 (1H, d),
7.10-7.38 (5H, m), 7.44 (1H, s), 7.64 (1H, d); MS (M + H)+
383.3.
)
(R,S)-8-[[(4-Am id in op h en yl)a m in o]ca r b on yl]-1,2,4,5-
t et r a h yd r o-5-oxo-4-(2-p h en ylet h yl)-5H -1,4-b en zod ia z-
ep in e-2-a cetic Acid (12). To a stirred solution of 4-amidi-
noaniline dihydrochloride (2.1 g, 10 mmol) in aqueous 1 N
NaOH (20 mL), H2O (10 mL), and THF (10 mL) was added
di-tert-butoxycarbonyl dicarbonate (2.4 g, 11 mmol). After
stirring for 5 h the reaction was extracted with ethyl acetate
three times, dried over Na2SO4, and evaporated. Purification
by flash chromatography on silica gel eluted with 96:4 CHCl3,
MeOH gave 4-(Boc-amidino)aniline as a white solid (1.31 g,
56%): 1H NMR (CDCl3) δ 1.44 (9H, s), 5.74 (2H, s), 6.52 (2H,
d), 7.70 (2H, d); TLC Rf 0.38 (5% MeOH, CHCl3). To the acid
31 (0.52 g, 1.4 mmol) was added thionyl chloride (5 mL, 69
mmol). After refluxing for 15 min under Ar the reaction was
evaporated and then reevaporated from fresh toluene two
times. To the resulting acid chloride in CH2Cl2 (30 mL) cooled
to 0 °C, with stirring under Ar, was added pyridine (0.33 mL,
4 mmol) followed by 4-(Boc-amidino)aniline (0.48 g, 2 mmol).
The reaction was stirred for 4 h at room temperature, diluted
with chloroform, washed with aqueous 1 N NaHCO3, dried
over Na2SO4, and evaporated. Purification by flash chroma-
tography eluting with 98:2 CHCl3, MeOH gave the coupled
5,N-Bis(ter t-bu toxyca r bon yl)-2-(m eth oxyca r bon yl)a n i-

556 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4
Keenan et al.
product as a solid (585 mg, 71%): 1H NMR (CDCl3) δ 1.52 (9H,
s), 2.37 (1H, dd), 2.52 (1H, dd), 2.93 (2H, t), 3.17 (1H, dd), 3.30
(1H, d), 3.48 (1H, dt), 3.67 (3H, s), 3.90 (1H, br s), 4.03 (1H,
dt), 4.67 (1H, s), 6.90-7.85 (12H, m).
(0.96 g) as an oil which was used in the next reaction. The
amine was dissolved in anhydrous methanol (50 mL) and
refluxed under Ar for 72 h. After stripping off the solvent the
crude material was purified by flash chromatography eluted
with 60% ethyl acetate in n-hexane to give the cyclized product
(0.87 g, 91%): 1H NMR (CDCl3) δ 1.59 (9H, s), 2.39 (1H, dd),
2.51 (1H, dd), 2.68-3.50 (6H, m), 3.56 (1H, br s), 3.73 (3H, s),
3.95 (2H, br s), 4.47 (1H, br s), 7.15-7.27 (6H, m), 7.42 (1H,
s), 7.51 (1H, d, J ) 7.4 Hz); MS (ES) m/e 425.0 (M + H)+.
To the cyclized product was added 90% TFA in CH2Cl2 (30
mL). After stirring at room temperature for 45 min the reaction
was evaporated to dryness and then reevaporated from
anhydrous toluene two times. The free acid was converted to
its acid chloride and coupled with (4-Boc-amidino)aniline, and
the product deprotected and hydrolzed as in the preparation
of 12. Evaporation left a residue which was purified by prep-
HPLC on a PRP-1 column eluted with 25% CH3CN, 0.1% TFA/
0.1% TFA, H2O to give product 14 as a white solid: MS (M +
H)+ ) 472.2. Anal. (C27H29N5O3‚2.5CF3CO2H‚1.0H2O).
Meth yl 3-Nitr o-4-(4-a cetoxybu tyl)ben zoa te (38). A mix-
ture of dry Mg (900 mg, 37 mmol) and ZnBr2 (5.63 g, 25 mmol)
in THF (25 mL) was treated with 5-bromopent-1-ene (Aldrich;
4.45 mL, 37.6 mmol), and the subsequent suspension was
heated at 56 °C (oil bath) for 18 h. The mixture was cooled to
room temperature, methyl 4-bromobenzoate (37) (Eastman; 7.1
g, 33 mmol) was added along with tetrakis(triphenylphos-
phine)palladium(0) (1.0 g, 0.865 mmol), and the mixture was
stirred at room temperature for 24 h. The mixture was
quenched with 3 N HCl (aqueous) and extracted with ethyl
acetate. The combined organic extracts were dried over
anhydrous MgSO4 and evaporated at reduced pressure. The
residue was purified by flash chromatography (5% ethyl
acetate in hexane) to give partially purified methyl 4-pent-4-
enylbenzoate which was carried on to the next step without
further purification.
The partially purified methyl 4-pent-4-enylbenzoate was
dissolved in a mixture of MeOH/CH2Cl2, cooled to -78 °C, and
treated with excess O3. After the excess O3 was purged, solid
NaBH4 (2.50 g, 66 mmol) was added and the solution slowly
brought to room temperature overnight. The reaction mixture
was evaporated at reduced pressure and the residue taken into
ethyl acetate. It was then washed with 1 N HCl (aqueous),
dried over anhydrous MgSO4, and evaporated at reduced
pressure. The residue was purified by flash chromatography
(30-40% ethyl acetate in hexane) to give 2.13 g of a partially
purified methyl 4-(4-hydroxybutyl)benzoate which was carried
on to the next step without further purification.
The partially purified methyl 4-(4-hydroxybutyl)benzoate
was dissolved in dry pyridine (25 mL) and treated with acetic
anhydride (1.93 mL, 20.4 mmol) at room temperature for 24
h. The solvent was evaporated at reduced pressure, and the
residue dissolved in ethyl acetate. This solution was washed
with 1 N HCl (aqueous), dried over anhydrous MgSO4, and
evaporated at reduced pressure. The residue was purified by
flash chromatography (15% ethyl acetate in hexane) to give
partially purified methyl 4-(4-acetoxybutyl)benzoate which was
carried on to the next step without further purification.
The partially purified methyl 4-(4-acetoxybutyl)benzoate
was dissolved in acetic anhydride (5 mL) and the solution
cooled to -12 °C. A mixture of fuming HNO3 (4.1 mL) and
concentrated H2SO4 (3.4 mL) was added dropwise over 20 min
and the reaction stirred an additional 40 min at -12 °C. The
reaction mixture was then poured onto ice and then extracted
with Et2O. The Et2O extracts were washed carefully with 5%
Na2CO3 (aqueous), dried over anhydrous MgSO4, and evapo-
rated at reduced pressure. The residue was first purified by
flash chromatography (15-25% ethyl acetate in hexane) to give
1.12 g (11.5% overall) of 38: 1H NMR (CDCl3) δ 1.57-1.90
(m, 4H), 2.03 (s, 3H), 2.77-3.10 (m, 2H), 3.93 (s, 3H), 3.97-
4.20 (m, 2H), 7.43 (d, 1H, J ) 7.5 Hz), 8.13 (d of d, 1H, J )
7.5, 1.5 Hz), 8.47 (d, 1H, J ) 1.5 Hz); MS (ES) m/e 296 (M +
H)+.
To the above (585 mg, 1.0 mmol) was added 90% TFA in
CH2Cl2 (20 mL). After stirring for 45 min the reation was
evaporated to dryness. The remaining residue was then taken
up in 20% HOAc in water and refluxed under Ar for 24 h.
Evaporation left a residue which was purified by prep-HPLC
on a PRP-1 column eluted with 26% CH3CN, 0.1% TFA/0.1%
TFA, H2O to give product 12 as a white solid: 1H NMR
(DMSO-d6, 2%TFA) δ 2.52 (4H, m), 2.89 (2H, m), 3.4-3.57 (3H,
m), 3.92 (2H, m), 7.27 (6H, m), 7.37 (1H, s), 7.72 (1H, d), 7.83
(2H, d), 8.0 (2H, d), 8.97 (2H, s), 9.22 (2H, s); MS (M + H)+
)
486.2. Anal. (C27H27N5O4‚2CF3CO2H‚0.5CH3CO2H‚2.5H2O).
N-(ter t-Bu toxycar bon yl)-4-n itr oan th r an ilic Acid Meth -
yl Ester (32). 5-Nitroanthranilic acid (10 g, 55 mmol) was
treated with diazomethane from Diazald (34.5 g, 110 mmol)
and then protected as its Boc derivative as in the preparation
of 29 to give product 32 (6.61 g, 43%): 1H NMR (CDCl3) δ 1.58
(9H, s), 4.0 (3H, s), 7.8 (1H, dd), 8.2 (1H, d, J ) 8 Hz), 9.38
(1H, d, J ) 2 Hz), 10.4 (1H, br s).
N-(ter t-Bu toxyca r bon yl)-4-(ca r boben zyloxya m in o)-2-
a m in oben zoic Acid Meth yl Ester (33). The nitro compound
32 (6.6 g, 23.4 mmol) was hydrogenated as in the preparation
of 29 and then treated with Cbz-Cl (4.0 mL, 28 mmol) to give
product 33 (8.24 g, 88%): 1H NMR (CDCl3) δ 1.5 (9H, s), 3.87
(3H, s), 5.2 (2H, s), 7.35 (5H, s), 7.44 (1H, dd), 7.57 (1H, s),
7.97 (1H, d, J ) 8 Hz), 8.42 (1H, d, J ) 2 Hz), 10.46 (1H, s).
Meth yl (R,S)-8-(Car boben zyloxyam in o)-1,2,4,5-tetr ah y-
d r o-5-oxo-4-(2-p h en yleth yl)-5H-1,4-ben zod ia zep in e-2-a c-
eta te (34). The methyl ester 33 (7.91 g, 19.8 mmol) was
saponified and coupled with 4-phenethylaminocrotonate as in
the preparation of 30, then deprotected with TFA, and cyclized
as in the preparation of 31 to give 3.73 g (76%) of product 34:
1H NMR (CDCl3) δ 2.40 (1H, dd), 2.53 (1H, dd), 2.94 (2H, t),
3.13 (1H, dd), 3.35 (1H, dd), 3.38 (1H, dt), 3.70 (3H, s), 3.90
(1H, m), 4.10 (1H, dt), 5.16 (2H, s), 6.67 (1H, dd), 7.00 (1H, s),
7.10-7.40 (10H, m), 7.71 (1H, d, J ) 8.5); MS (M + H)+
488.3.
)
(R,S)-8-[[(4-Am id in op h en yl)ca r b on yl]a m in o]-1,2,4,5-
t et r a h yd r o-5-oxo-4-(2-p h en ylet h yl)-5H -1,4-b en zod ia z-
ep in e-2-a cetic Acid (13). The Cbz-aniline derivative 34 (0.5
g, 1 mmol) was hydrogenolyzed, coupled to 4-Cbz-amidinoben-
zoyl chloride, and deprotected as in the preparation of 5. Ester
hydrolysis as in the preparation of 10 gave product 13 (218
mg, 48%): 1H NMR (DMSO-d6, 2%TFA) δ 2.53 (4H, m), 2.89
(2H, m), 3.53-3.36 (3H, m), 3.9 (2H, m), 7.08 (1H, dd), 7.26
(5H, m), 7.44 (1H, d), 7.6 (1H, d), 7.95 (2H, d), 8.14 (2H, d),
9.2 (2H, s), 9.43 (2H, s); MS (M + H)+ ) 486.2. Anal.
(C27H27N5O4‚3CF3CO2H‚2CH3CO2H‚3H2O).
(E)-Meth yl N-[(2-Nitr o-4-(ter t-bu toxyca r bon yl)p h en -
yl)m eth yl]-4-[(2-p h en yleth yl)a m in o]-2-bu ten oa te (36). To
a solution of 2-nitro-4-(tert-butoxycarbonyl)benzyl bromide
(35)27 (4.06 g, 12.8 mmol) in THF (50 mL) was added with
stirring at room temperature (E)-methyl 4-(phenethylamino)-
2-butenoate (3.4 g, 15.5 mmol) followed by triethylamine (2.2
mL, 15.7 mmol). After stirring for 16 h the reaction was diluted
with ethyl acetate (50 mL), washed with 1 N Na2CO3 (100 mL)
and brine (100 mL), dried (Na2SO4), and evaporated to dryness.
Purification by flash chromatography on silica gel eluted with
15% ethyl acetate in n-hexane gave 36 as an oil (5.16 g, 88%):
1H NMR (CDCl3) δ 1.60 (9H, s), 2.74 (4H, br s), 3.30 (2H, dd),
3.74 (3H, s), 4.00 (2H, s), 6.00 (1H, d, J ) 15 Hz), 7.02 (1H,
dt), 7.25 (5H, m), 7.67 (1H, d, J ) 8 Hz), 8.15 (1H, dd), 8.45
(1H, d, J ) 2 Hz).
(R,S)-8-[[(4-Am id in op h en yl)a m in o]ca r b on yl]-1,2,4,5-
tetr a h yd r o-4-(2-p h en yleth yl)-1,4-ben zod ia zep in e-2-a ce-
tic Acid (14). To a solution of the nitro compound 36 (1.03 g,
2.26 mmol) in 1:1 acetic acid, ethanol (50 mL) was added with
stirring at room temperature iron powder (0.8 g, 14.3 mmol).
After stirring for 16 h the reaction was evaporated at reduced
pressure taken up in ethyl acetate, washed with 1 N Na2CO3
and brine, dried (Na2SO4), and evaporated to give the amine
Ben zyl 3-Nitr o-4-(4-h yd r oxybu tyl)ben zoa te (39). The
ester 38 (1.04 g, 3.51 mmol) in dioxane (36 mL) was treated

Conformational Preferences in Benzodiazepine Series
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4 557
with 18 mL of 1 M NaOH (aqueous) at room temperature for
3 days. The reaction mixture was acidified with 1 N HCl
(aqueous) and extracted with ethyl acetate. The combined
organic extracts were dried over anhydrous MgSO4 and
evaporated at reduced pressure. The partially hydrolyzed
material was resubmitted to the above conditions at room
temperature for 3 days to give crude 3-nitro-4-(4-hydroxybu-
tyl)benzoic acid which was used after workup without further
purification.
The crude 3-nitro-4-(4-hydroxybutyl)benzoic acid was dis-
solved in aqueous MeOH and neutralized to pH ) 7.6 with
solid CsHCO3. The solution was then evaporated under
vacuum to dryness and evaporated from MeOH/toluene to
remove traces of H2O. The salt was then dissolved in anhy-
drous DMF (30 mL) and treated with benzyl bromide (1 mL,
8.41 mmol). The subsequent suspension was heated at 57 °C
(oil bath) for 18 h. The reaction mixture was evaporated under
vacuum and the residue dissolved in ethyl acetate. The
solution was washed with 1 N HCl (aqueous), dried over
anhydrous MgSO4, and evaporated at reduced pressure. The
residue was purified by flash chromatography (40% ethyl
acetate in hexane) to give 624 mg (54%) of 39: 1H NMR
(CDCl3) δ 1.43-1.95 (m, 5H), 2.93 (t, 2H, J ) 7.5 Hz), 3.67 (t,
2H, J ) 6 Hz), 5.37 (s, 2H), 7.37 (s, 5H), 7.40 (d, 1H, J ) 7.5
Hz), 8.13 (d of d, 1H, J ) 7.5, 1.5 Hz), 8.74 (d, 1H, J ) 1.5
Hz).
(E)-Ben zyl 3-Nit r o-4-(5-ca r b et h oxyp en t -4-en yl)b en -
zoa te (40). The alcohol 39 (271 mg, 0.823 mmol) in CH2Cl2
(10 mL) was treated with pyridinium chlorochromate (355 mg,
1.65 mmol) at room temperature for 4 h. The reaction mixture
was diluted with Et2O, filtered through a pad of Florisil, and
evaporated at reduced pressure. The residue was dissolved in
ethyl acetate and refiltered through a pad of Florisil. The crude
aldehyde in CH2Cl2 (10 mL) was treated with (carbethoxy-
methylene)triphenylphosphorane (344 mg, 0.988 mmol) and
stirred at room temperature for 3 days. The reaction mixture
was evaporated at reduced pressure, and the residue was
purified by flash chromatography (20% ethyl acetate in hex-
ane) to give 195 mg (60%) of 40: 1H NMR (CDCl3) δ 1.27 (t,
3H, J ) 7.5 Hz), 1.63-2.03 (m, 2H), 2.13-2.47 (m, 2H), 2.80-
3.10 (m, 2H), 4.17 (q, 2H, J ) 7.5 Hz), 5.37 (s, 2H), 5.83 (br d,
1H, J ) 16.5 Hz), 6.93 (d of t, 1H, J ) 16.5, 7.5 Hz), 7.40 (s,
5H), 7.40 (d, 1H, J ) 9 Hz), 8.15 (d of d, 1H, J ) 9, 1.5 Hz),
8.50 (d, 1H, J ) 1.5 Hz).
2-(Ca r beth oxym eth yl)-8-(ca r boben zyloxy)-tetr a h yd r o-
1-ben za zep in e (41). The nitro compound 40 (304 mg, 0.765
mmol) in dry EtOH (10 mL) was treated with SnCl2 (725 mg,
3.83 mmol), and the mixture was heated at reflux for 18 h.
The reaction mixture was evaporated and the residue taken
into ethyl acetate and 5% Na2CO3 (aqueous). The resulting
precipitate was filtered off and discarded and the organic layer
separated, dried over anhydrous MgSO4, and evaporated at
reduced pressure. The residue was purified by flash chroma-
tography (15-50% ethyl acetate in hexane) to give 128 mg of
41 contaminated with about 20% of the corresponding diethyl
ester. An additional amount (67 mg) of the uncyclized aniline
was also isolated. Compound 41: 1H NMR (CDCl3) δ 1.20 (t,
3H, J ) 6 Hz), 1.43-2.07 (m, 4H), 2.33-2.57 (m, 2H), 2.67-
2.93 (m, 2H), 3.17-3.50 (m, 1H), 4.13 (q, 2H, J ) 6 Hz), 4.30
(br s, 1H), 5.32 (s, 2H), 7.00-7.67 (m, 8H).
deprotected and hydrolyzed as in the preparation of 5 to give
15: MS (ES) m/e 367 (M + H)+.
(R,S)-Dim eth yl N-(3-Ca r boxyp h en yl)a sp a r ta te (43). To
a solution of 3-aminobenzoic acid (42) (1.04 g, 7.5 mmol) in
methanol (40 mL) was added dimethyl acetylenedicarboxylate
(1.25 g, 8.75 mmol). The solution was heated at reflux for 1 h,
cooled to room temperature, and concentrated. The resulting
oil was chromatographed (5:95 methanol:dichloromethane-
0.5% acetic acid) to give a yellow solid (2.0 g, 95%). The solid
was dissolved in methanol (100 mL) containing 10% palladium
on carbon (0.40 g), and the solution was shaken in a hydrogen
atmosphere (35 psi) for 3 h. The mixture was filtered, and the
filtrate was concentrated to give 43 as a pale-yellow solid: 1H
NMR (400 MHz, CDCl3) δ 7.55 (d, 1H), 7.45 (s, 1H), 7.3 (t,
1H), 4.55 (t, 1H), 3.8 (s, 3H), 3.7 (s, 3H), 2.95 (d, 2H).
(R,S)-N-[3-[[(4-Am idin oph en yl)am in o]car bon yl]ph en yl]-
a sp a r tic Acid Dih yd r och lor id e (16). The acid 43 (0.5 g, 1.78
mmol) was refluxed for 5 min in thionyl chloride. The solution
was concentrated to an oil which was treated with dichlo-
romethane (25 mL) and concentrated three times. The result-
ing oil was dissolved in dichloromethane (5 mL) and the
solution added dropwise to a stirred solution of 4-(N-Cbz-
amidino)aniline (0.5 g, 1.85 mmol) and diisopropylethylamine
(0.24 g, 1.9 mmol) in dichloromethane (50 mL). After 16 h,
the solution was treated with diisopropylethylamine (0.24 g,
1.9 mmol), extracted with water (2 × 25 mL), and dried
(sodium sulfate). The organic phase was concentrated, and the
resulting brown solid was purified by preparative TLC. The
resultant pale-yellow solid was deprotected and the ester
hydrolyzed as in the preparation of 5. Purification by prepara-
tive HPLC gave a white solid which was dissolved in water
(20 mL) and 6 N hydrochloric acid (1.0 mL). Lyophilization
gave the title compound (0.04 g, 31%) as a colorless solid: 1H
NMR (400 MHz, DMSO-d6) δ 9.3 (s, 2H), 8.9 (s, 2H), 8.0 (d,
2H), 7.7 (d, 2H), 7.3-7.1 (m, 3H), 6.85 (d, 1H), 4.4 (t, 1H), 2.8
(dd, 1H), 2.7 (dd, 1H); MS (ES) m/e 371.2 (M + H)+.
ter t-Bu tyl N-[(2-F lu or o-5-n itr op h en yl)m eth yl]-3-a m i-
n op r op ion a te (44). 3-Aminopropionic acid tert-butyl ester
(9.7 g, 53.3 mmol) and 2-fluoro-5-nitrobenzaldehyde (9.0 g, 53.3
mmol) were added to a suspension of sodium acetate (6.5 g,
78.3 mmol) in methanol (150 mL), followed by portionwise
addition of sodium cyanoborohydride (6.5 g, 0.1 mol). After 2
h at room temperature, the reaction mixture was quenched
with ice and diluted with sodium bicarbonate solution. The
mixture was extracted with EtOAc, and the combined organic
extracts were dried over MgSO4. Filtration and evaporation
of the solvent in vacuo yielded the title compound as a yellow
oil (15.7 g, 99%): MS m/e 299 (M + H)+.
ter t-Bu t yl N-[(2-F lu or o-5-n it r op h en yl)m et h yl]-N-(2-
a m in oa cetyl)-3-a m in op r op ion a te (45). A solution of com-
pound 44 (17.3 g, 58.1 mmol) in CH2Cl2 (250 mL) was stirred
at room temperature under an argon atmosphere. Triethyl-
amine (18 mL, 128 mmol) and BOP reagent (28.2 g, 63.9 mmol)
were added, followed by Boc-glycine (11.2 g, 63.9 mmol). The
reaction mixture was stirred overnight at room temperature
and poured into ice water (300 mL). The mixture was extracted
with ethyl acetate. The combined organic extracts were washed
with 1 M KHSO4, water, 5% NaHCO3, and brine and dried
over MgSO4. Filtration of the mixture and evaporation of the
filtrate in vacuo yielded the intermediate Boc-protected amine
which was stirred in CH2Cl2 (250 mL) and 4 M HCl/dioxane
(25 mL) at 0 °C for 3 h. Evaporation of the solvents in vacuo
yielded the free amine 45 (20.0 g, 100%): MS m/e 356 (M +
H)+.
(R,S)-2-(Ca r boxym eth yl)-8-[[(4-a m id in op h en yl)a m in o]-
ca r boxy]-tetr a h yd r o-1-ben za zep in e (15). A suspension of
benzazepine 41 (128 mg, 0.348 mmol) contaminated with about
20% of the diethyl ester and 5% Pd/C (20 mg) in MeOH was
treated with H2 at 50 psi (Parr apparatus) at room tempera-
ture for 3 h. The reaction mixture was filtered through a plug
of Celite and evaporated at reduced pressure to give the free
acid. The acid was converted to its acid chloride and reacted
with 4-(N-Cbz-amidino)aniline (141 mg, 0.522 mmol) as in the
preparation of 12 to give the coupled product: 1H NMR (CDCl3)
δ 1.20 (t, 3H, J ) 7.5 Hz), 1.30-2.03 (m, 4H), 2.30-2.50 (m,
2H), 2.60-2.90 (m, 2H), 3.10-3.50 (m, 1H), 4.10 (q, 2H, J )
7.5 Hz), 4.30 (br s, 1H), 5.17 (s, 2H), 6.93-7.90 (m, 14H), 8.40
(s, 1H); MS (ES) m/e 529 (M + H)+. The coupled product was
ter t-Bu t yl 7-Am in o-3-oxo-1,2,3,5-t et r a h yd r o-4H -1,4-
ben zod ia zep in e-2-p r op a n oa te (46). To a solution of the
amine 45 (12.0 g, 30.7 mmol) in DMSO (400 mL) were added
triethylamine (15 mL) and water (5 mL), and the reaction was
stirred overnight. The reaction mixture was poured into water
(500 mL) and extracted with ethyl acetate. The combined
organic phases were washed with aqueous brine and dried over
sodium sulfate. Filtration and concentration of the organic
extracts in vacuo yielded the cyclized nitro compound (1.8 g,
18%).

558 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4
Keenan et al.
A solution of the nitro compound (2.2 g, 6.6 mmol) and
platinum oxide catalyst (0.6 g) in ethyl acetate (100 mL) was
shaken on a Parr shaker under hydrogen (40 psi) for 1 h. The
catalyst was filtered from the solution, and the solvent was
evaporated in vacuo to yield 46 (2.0 g, 98%): MS m/e 306 (M
+ H)+.
mL), filtered, and dried at reduced pressure to give a mixture
of the 7- and 8-nitro compounds (1.38 g, 68%).
(c) The mixture of the above compounds (470 mg, 1.7 mmol)
was taken up in 1:1 methanol:dimethylformamide (20 mL).
Argon was bubbled through the system, 10% palladium on
carbon (80 mg) was added, and the mixture was shaken under
hydrogen (40 psi) for 1.5 h. The solution was filtered through
Celite and concentrated. The residue was chromatographed
(silica gel, 4% methanol:dichloromethane) to separate the
7-amino regioisomer from the 8-isomer 49 (110 mg, 42%): 1H
NMR (360 MHz, DMSO-d6) δ 3.6-3.65 (s, 3H), 4.0-4.1 (m,
4H), 4.25-4.3 (s, 2H), 5.7 (s, 2H), 6.35-6.45 (dd, 1H), 6.45-
6.5 (d, 1H), 6.6-6.65 (t, 1H), 6.85-6.9 (d, 1H); MS (ES) m/e
250.2 (M + H)+.
8-[(4-Am id in oben zoyl)a m in o]-2,3,4,5-tetr a h yd r o-3-oxo-
1H-2,4-ben zod ia zep in e-2-a cetic Acid (21). The amine 49
was coupled with 4-[(benzyloxycarbonyl)amidino]benzoic acid
and deprotected as in the preparation of 5 to furnish the title
compound: MS (ES) m/e 382.0 (M + H)+; 1H NMR (DMSO-d6,
360 MHz) δ 2.5 (s, 3H), 3.9-4.0 (d, 2H), 4.2-4.3 (d, 2H), 4.4-
4.5 (d, 2H), 6.5-6.6 (m, 1H), 7.2-7.3 (d, 1H), 7.6-8.1 (m, 10H).
Anal. (C21H26N5O7.5) Calcd: C, 53.84; H, 5.59; N, 14.95.
Found: C, 54.04; H, 5.92; N, 14.81.
7-[[(4-Am id in op h en yl)ca r b on yl]a m in o]-3-oxo-1,2,3,5-
tetr ah ydr o-4H-1,4-ben zodiazepin e-4-pr opan oic Acid (18).
A mixture of compound 46 (2.0 g, 6.6 mmol), 4-[N-(benzyloxy-
carbonyl)amidino]benzoic acid (2 g, 6.6 mmol), BOP reagent
(2.9 g, 6.6 mmol), and triethylamine (1.6 mL, 13.2 mmol) in
DMF (20 mL) was stirred overnight under an argon atmo-
sphere. The solution was poured into a mixture of ice water
(150 mL) and extracted with EtOAc. The combined extracts
were washed with 5% NaHCO3 solution, dried over sodium
sulfate, filtered, and concentrated in vacuo to a yellow solid.
This solid was chromatographed (silica gel, 3% methanol/CH2-
Cl2) to yield the coupled compound (150 mg, 4%); 10% pal-
ladium on carbon (150 mg, activated) was added to a solution
of the compound (150 mg, 0.26 mmol) in glacial acetic acid:
ethyl acetate (1:1, 30 mL) and concentrated hydrochloric acid
(0.5 mL). The mixture was hydrogenated in a Parr shaker (45
psi) for 24 h. The reaction mixture was filtered, the filtrate
was evaporated, and a portion (0.36 g) of the resulting residue
was purified on HPLC (YMC ODS-AQ, 50 × 250 mm, 85 mL/
min, 15% CH3CN/H2O-0.1% TFA, UV detection at 220 nm)
to yield the title compound (15 mg): MS m/e 362 (M + H)+.
Anal. (C20H21N5O4‚1.1TFA‚0.25H2O) Calcd: C, 47.46; H, 4.23;
N, 12.47. Found: C, 47.76; H, 4.05; N, 12.17.
Ack n ow led gm en t. We thank the Department of
Physical and Structural Chemistry for mass spectral
support and Edith A. Reich of the Department of
Physical and Structural Chemistry for elemental analy-
ses.
Dip h th a lim id oyl-o-xylen e (47). A mixture of R,R′-di-
bromo-o-xylene (26 g, 98 mmol) and potassium phthalimide
(54.7 g, 294 mmol) in dimethylformamide (500 mL) was
refluxed for 18 h. The solution was cooled to room temperature,
poured into water, and filtered and the filter cake was dried
at reduced pressure to give the title compound (35 g, 90%).
Su p p or tin g In for m a tion Ava ila ble: X-ray crystallogra-
phy data for compounds 22b-f (Table 4). This material is
available free of charge via the Internet at http://pubs.acs.org.
Refer en ces
2,3,4,5-Te t r a h yd r o-7-n it r o-3-oxo-1H -2,4-b e n zod ia z-
ep in e (48). To a cold solution of potassium nitrate (6.2 g, 61.2
mmol) in concentrated sulfuric acid (350 mL) was added
portionwise compound 47 (22 g, 55.6 mmol). The resulting
solution was warmed to room temperature and stirred for 18
h. The mixture was carefully poured into ice water, and the
mixture was filtered. The filter cake was washed with water
and dried at reduced pressure to give the nitrated compound.
This compound was added portionwise to a solution of hydra-
zine (26 mL, 535 mmol) in ethanol (1000 mL), and the solution
was refluxed for 1 h. Additional hydrazine (26 mL, 535 mmol)
was added followed by ethanol (500 mL). The solution was
stirred at room temperature for 18 h and filtered, and the
filtrate was concentrated. The residue was triturated with
chloroform (600 mL) for 1 h. After filtration and concentration,
the residue was dissolved in tetrahydrofuran (500 mL), and
1,1′-carbonyldiimidazole (9.2 g, 56.8 mmol) dissolved in tet-
rahydrofuran (250 mL) was added dropwise. The resulting
mixture was stirred at room temperature for 6 days, filtered,
and concentrated to yield the title compound (7.7 g, 71% yield).
Meth yl 8-Am in o-2,3,4,5-tetr a h yd r o-3-oxo-1H-2,4-ben -
zod ia zep in e-2-a ceta te (49). (a ) A heterogeneous solution
consisting of compound 48 (5.5 g, 26.6 mmol) and Lawesson’s
reagent (7.0 g, 17.3 mmol) in toluene (100 mL) was heated to
80 °C under an argon atmosphere for 1.5 h. The solution was
cooled and filtered. The solid was triturated with 7:3 dichlo-
romethane:methanol for 1 h, filtered, and concentrated to give
the thiourea (4.3 g, 73%).
(b) To a solution of the thiourea (2.2 g, 9.9 mmol) in
dimethylformamide (50 mL) was added dropwise iodomethane
(0.614 mL, 9.9 mmol) dissolved in dimethylformamide (5 mL).
The solution was stirred for 1 h, and potassium carbonate (3.0
g, 21.8 mmol) was added followed by methyl bromoacetate
(0.934 mL, 9.9 mmol). The solution was stirred for 18 h,
filtered, and concentrated to give a residue which was parti-
tioned between ethyl acetate and water. The organic layer was
concentrated, and the residue was taken up in water/dioxane
(1:1) (50 mL) and refluxed for 18 h. The solution was
concentrated, and the residue was triturated with water (30
(1) Coller, B. S.; Anderson, K.; Weisman, H. F. New antiplatelet
agents: platelet GPIIb/IIIa antagonists. Thromb. Haemost. 1995,
74, 302-308.
(2) Kieffer, N.; Phillips, D. R. Platelet Membrane Glycoproteins:
Functions in Cellular Adhesion and Aggregation. Annu. Rev.
Cell. Biol. 1990, 6, 329-357.
(3) See: (a) Lefkovits, J .; Plow, E. F.; Topol, E. J . Platelet Glyco-
protein IIb/IIIa Receptors in Cardiovascular Medicine. N. Engl.
J . Med. 1995, 332, 1553-1559. (b) Samanen, J . Annu. Rep. Med.
Chem. 1996, 31, 91-100.
(4) D’Souza, S. E.; Ginsberg, M. H.; Plow, E. F. Arginyl-glycyl-
aspartic acid (RGD): A Cell Adhesion Motif. Trends Biochem.
Sci. 1991, 16, 246-250.
(5) Nichols, A. J .; Ruffolo, R. R., J r.; Huffman, W. H.; Poste, G.;
Samanen, J . Trends Pharmacol. Sci. 1992, 13, 413.
(6) Samanen, J .; Ali, F.; Romoff, T.; Calvo, R.; Sorenson, E.; Vasko,
J .; Storer, B.; Berry, D.; Bennett, D.; Stroksacker, M.; Powers,
D.; Stadel, J .; Nichols, A. Development of a Small RGD Peptide
Fibrinogen Receptor Antagonist with Potent Antiaggregatory
Activity in Vitro. J . Med. Chem. 1991, 34, 3114-3125.
(7) Nutt, R. F.; Brady, S. F.; Colton, C. D.; Sisko, J . T.; Ciccarone,
T.; Levey, M. R.; Duggan, M. E.; Imagire, I. S.; Gould, R. J .;
Anderson, P. S.; Veber, D. F. Development of Novel, Highly
Selective Fibrinogen Receptor Antagonists as Potentially Useful
Antithrombotic Agents. In Peptides: Chemistry and Biology.
Proceedings of the Twelfth American Peptide Symposium; Smith,
J . A., River, J . E., Eds.; ESCOM: Leiden, The Netherlands, 1992;
pp 914-916.
(8) Barker, P. L.; Bullens, S.; Bunting, S.; Burdick, D. J .; Chan, K.
S.; Deisher, T.; Eigenbrot, C.; Gadek, T. R.; Gantzos, R.; Lipari,
M. T.; Muir, C. D.; Napier, M. A.; Pitti, R. M.; Padua, A.; Quan,
C.; Stanley, M.; Struble, M.; Tom, J . Y. K.; Burnier, J . P. Cyclic
RGD Analogues as Antiplatelet Antithrombotics. J . Med. Chem.
1992, 35, 2040-2048.
(9) Scarborough, R. M.; Naughton, M. A.; Teng, W.; Rose, J . W.;
Phillips, D. R.; Nannizzi, L.; Arfsten, A.; Campbell, A. M.; Charo,
I. F. Design of Potent and Specific Integrin Antagonists. J . Biol.
Chem. 1993, 268, 1066-1073.
(10) Ali, F. E.; Bennett, D. B.; Raul, C. R.; Elliott, J . D.; Hwand, S.
M.; Ku, T. W.; Lago, M. A.; Nichols, A. J .; Romoff, T. T.; Shah,
D. H.; Vasko, J . A.; Wong, A. S.; Yellin, T. O.; Yuan, C. K.;
Samanen, J . M. Conformationally Constrained Peptides and
Semipeptides Derived from RGD as Potent Inhibitors of the
Platelet Fibrinogen Receptor and Platelet Aggregation. J . Med.
Chem. 1994, 37, 769-780.

Conformational Preferences in Benzodiazepine Series
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 4 559
(11) Cheng, S.; Craig, W. S.; Mullen, D.; Tschopp, J . F.; Dixon, D.;
Pierschbacher, M. D. Design and Synthesis of Novel Cyclic RGD-
Containing Peptides as Highly Potent and Selective Integrin
Design of a Potent Non-Peptide Fibrinogen Receptor Antagonist
Based on the Structure and Conformation of a Highly Con-
strained Cyclic RGD Peptide. J . Am. Chem. Soc. 1993, 115,
8861-8862.
R
IIbâIII Antagonists. J . Med. Chem. 1994, 37, 1-8.
(12) J ackson, S.; DeGrado, W.; Dwivedi, A.; Parthasarathy, A.;
Higley, A.; Krywko, J .; Rockwell, A.; Markwalder, J .; Wells, G.;
Wexler, R.; Mousa, S.; Harlow, R. Template-Constrained Cyclic
Peptides: Design of High-Affinity Ligands for GPIIb/IIIa. J . Am.
Chem. Soc. 1994, 116, 3220-3230.
(13) Muller, G.; Gurrath, M.; Kessler, H. Pharmacophore refinement
of gpIIb/IIIa antagonists based on comparative studies of anti-
adhesive cyclic and acyclic RGD peptides. J . Comput. Aided. Mol.
Des. 1994, 8, 709-730.
(14) Kopple, K. D.; Baures, P. W.; Bean, J . W.; D’Ambrosio, C. A.;
Hughes, J . L.; Peishoff, C. E.; Eggleston, D. S. Conformations
of Arg-Gly-Asp Containing Heterodetic Cyclic Peptides: Solution
and Crystal Studies. J . Am. Chem. Soc. 1992, 114, 9615-9623.
(15) Peishoff, C. E.; Ali, F. E.; Bean, J . W.; Calvo, R.; D′Ambrosio, C.
A.; Eggleston, D. S.; Hwang, S. M.; Kline, T. P.; Koster, P. F.;
Nichols, A.; Powers, D.; Romoff, T.; Samanen, J . M.; Stadel, J .;
Vasko, J . A.; Kopple, K. D. Investigation of Conformational
Specificity at GPIIb/IIIa: Evaluation of Conformationally Con-
strained RGD Peptides. J . Med. Chem. 1992, 35, 3962-3969.
(16) (a) Bach, A. C., II; Eyermann, C. J .; Gross, J . D.; Bower, M. J .;
Harlow, R. L.; Weber, P. C.; DeGrado, W. F. Structural Studies
of Family of High Affinity Ligands for GP IIB-IIIa. J . Am. Chem.
Soc. 1994, 116, 3207-3219. (b) Wityak, J .; Sielecki, T. M.; Pinto,
D. J .; Emmet, G.; Sze, J . Y.; Liu, J .; Tobin, E.; Wang, S.; J iang,
B.; Ma, P.; Mousa, S. A.; Wexler, R. R.; Olson, R. Discovery of
Potent Isoxazoline Glycoprotein IIb/IIIa Receptor Antagonists.
J . Med. Chem. 1997, 40, 50-60.
(24) (a) Bondinell, W. E.; Keenan, R. M.; Miller, W. H.; Ali, F. E.;
Allen, A. C.; DeBrosse, C. W.; Eggleston, D. S.; Erhard, K. F.;
Haltiwanger, R. C.; Lee, C. P.; Nichols, A. J .; Ross, S. T.;
Samanen, J . M.; Valocik, R. E.; Vasko-Moser, J . A.; Venslavsky,
J . W.; Wong, A. S.; Yuan, C. K. Design of a Potent and Orally
Active Nonpeptide Platelet Fibrinogen Receptor (GPIIB/IIIA)
Antagonist. Bioorg. Med. Chem. 1994, 2, 897-908. (b) Bondinell,
W. E.; et al. Manuscript in preparation.
(25) Samanen, J . M.; Ali, F. E.; Barton, L. S.; Bondinell, W. E.;
Burgess, J . L.; Callahan, J . F.; Calvo, R. R.; Chen, W.; Chen, L.;
Erhard, K.; Feuerstein, G.; Heys, R.; Hwang, S.-M.; J akas, D.
R.; Keenan, R. M.; Koster, P. F.; Ku, T. W.; Kwon, C.; Lee, C.-
P.; Miller, W. H.; Newlander, K. A.; Nichols, A.; Parker, M.;
Peishoff, C. E.; Rhodes, G.; Ross, S.; Shu, A.; Simpson, R.;
Takata, D.; Vasko-Moser, J . A.; Valocik, R. E.; Yellin, T. O.;
Uzinskas, I.; Venslavsky, J . W.; Wong, A.; Yuan, C.-K.; Huffman,
W. F. Potent, Selective, Orally Active 3-Oxo-1,4-benzodiazepine
GPIIb/IIIa Integrin Antagonists. J . Med. Chem. 1996, 39, 4867-
4870.
(26) Miller, W. H.; Bondinell, W. E.; Callahan, J . F.; Eggleston, D.
S.; Huffman, W. F.; Hwang, S. M.; J akas, D. R.; Keenan, R. M.;
Koster, P. F.; Ku, T. W.; Kwon, C. K.; Nichols, A. J .; Samanen,
J . M.; Takata, D. T.; Uzinskas, I. N.; Valocik, R. E.; Vasko-Moser,
J . A.; Wong, A. S. Structure-Activity Relationships in 3-Oxo-
1,4-benzodiazepine-2-acetic Acid GPIIb/IIIa Antagonists. The
2-Benzazepine Series. Bioorg. Med. Chem. Lett. 1996, 6, 2481-
2486.
(17) Eldred, C. P.; Evans, B.; Hindley, S.; J udkins, B. D.; Kelly, H.
A.; Kitchin, J .; Lumley, B. D.; Porter, B.; Ross, B. C.; Smith, K.
J .; Taylor, N. R.; Wheatcroft, J . R. Orally Active Non-Peptide
Fibrinogen Receptor (GPIIb/IIIa) Antagonists: Identification of
4-[4-[-4(Aminoimino-methyl)Phenyl]-1-Piperazinyl-1-Piperidineace-
tic acid as a Long-Acting, Broad-Spectrum Antithrombotic Agent.
J . Med. Chem. 1994, 37, 3882-3885.
(27) Bondinell, W. E.; Callahan, J . F.; Huffman, W. F.; Keenan, R.
M.; Ku, T. W.; Newlander, K. A.; Samanen, J . M.; Uzinskas, I.
N. WO 9414776, J uly 7, 1994.
(28) Miller, W. H.; Newlander, K. A.; Eggleston, D. S.; Haltiwanger,
R. C.; Synthesis of a 2-Benzazepine Analogue of a Potent,
Nonpeptide GPIIb/IIIa Antagonist. Tetrahedron Lett. 1995, 36,
373-376.
(18) (a) McDowell, R. S.; Gadek, T. R.; Barker, P. L.; Burdick, D. J .;
Chan, K. S.; Quan, C. L.; Skelton, N.; Struble, M.; Thorsett, E.
D.; Tischler, M.; Tom, J . Y. K.; Webb, T. R.; Burnler, J . P. From
Peptide to Non-Peptide. 1. The Elucidation of a Bioactive
Conformation of the Arginine-Glycine-Aspartic Acid Recognition
Sequence. J . Am. Chem. Soc. 1994, 116, 5069-5077. (b) Mc-
Dowell, R. S.; Blackburn, B. K.; Gadek, T. R.; McGee, L. R.;
Rawson, T.; Reynolds, M. E.; Robarge, K. D.; Somers, T. C.;
Thorsett, E. D.; Tischler, M.; Webb, R. R.; Venuti, M. C. From
Peptide to Non-Peptide. 2. The de Novo Design of Potent, Non-
Peptidal Inhibitors of Platelet Aggregation Based on a Benzo-
diazepinedione Scaffold. J . Am. Chem. Soc. 1994, 116, 5077-
5083.
(19) Fisher, M. J .; Gunn, B.; Harms, C. S.; Kline, A. D.; Mullaney, J .
T.; Nunes, A.; Scarborough, R. M.; Arfsen, A. E.; Skelton, M. A.;
Um, S. L.; Utterback, B. G.; J acubowski, J . A. Non-Peptide RGD
Surrogates Which Mimic a Gly-Asp â-turn: Potent Antagonists
of Platelet Glycoprotein IIb-IIIa. J . Med. Chem. 1997, 40, 2085-
2101.
(20) Hirschmann, R.; Sprengeler, P. A.; Kawasaki, T.; Leahy, J . W.;
Shakespeare, W. C.; Smith, A. B., III. The First Design and
Synthesis of a Steroidal Peptidomimetic. The Potential Value
of Peptidomimetics in Elucidating the Bioactive Conformation
of Peptide Ligands. J . Am. Chem. Soc. 1992, 114, 9699-9701.
(21) For recent reviews of the progress in the development of
nonpeptide IIb/IIIa antagonists, see ref 3b and: Ojima, I.;
Chakravarty, S.; Dong, Q. Antithrombotic agents: from RGD
to peptide mimetics. Bioorg. Med. Chem. 1995, 3, 337-360.
(22) Callahan, J . F.; Bean, J . W.; Burgess, J . L.; Eggleston, D. E.;
Hwang, S. M.; Kopple, K. D.; Koster, P. F.; Nichols, A.; Peishoff,
C. E.; Samanen, J . M.; Vasko, J . A.; Wong, A.; Huffman, W. F.
Design and Synthesis of a C7 Mimetic for the Predicted γ-turn
Conformation Found in Several Constrained RGD Antagonists.
J . Med. Chem., 1992, 35, 3970-3972.
(29) For a full description of the in vitro assays, see ref 24a.
(30) The intestinal permeability of these compounds was the subject
of a recent review; See: Samanen, J . M.; Lee, C.-P.; Smith, P.
L.; Bondinell, W.; Calvo, R. R.; J akas, D. R.; Newlander, K.;
Parker, M.; Uzinskas, I.; Yellin, T. O.; Nichols, A. J . The Use of
Rabbit Intestinal Permeability as an in vitro Assay in the Search
for Orally Active GPIIb/IIIa Antagonists. Adv. Drug Deliv. Rev.
1996, 23, 133-142.
(31) For prototypical acyclic nonpeptide fibrinogen receptor antago-
nists, see: (a) Hartman, G. D.; Egbertson, M. S.; Halczenko, W.;
Laswell, W. L.; Duggan, M. E.; Smith, R. L.; Naylor, A. M.;
Manno, P. D.; Lynch, R. J .; Zhang, G.; et al. Non-peptide
fibrinogen receptor antagonists. 1. Discovery and design of
exosite inhibitors. J . Med. Chem. 1992, 35, 4640-4642. (b)
Zablocki, J . A.; Rico, J . G.; Garland, R. B.; Rogers, T. E.;
Williams, K.; Schretzman, L. A.; Rao, S. A.; Bovy, P. R.; Tjoeng,
F. S.; Lindmark, R. J .; Toth, M. V.; Zupec, M. E.; McMackins,
D. E.; Adams, S. P.; Miyano, M.; Markos, C. S.; Milton, M. N.;
Paulson, S.; Herin, M.; et al. Potent in vitro and in vivo inhibitors
of platelet aggregation based upon the Arg-Gly-Asp sequence of
fibrinogen. (Aminobenzamidino)succinyl (ABAS) series of orally
active fibrinogen receptor antagonists. J . Med. Chem. 1995, 38,
2378-2394.
(32) Monte Carlo simulations were done using MacroModel version
6.0 (Department of Chemistry, Columbia University, New York).
(33) Minimizations within BatchMin V6.0 (Department of Chemistry,
Columbia University, New York) were performed using AM-
BER*. All force-field equations are identical with those of
authentic AMBER from P. Kollman.
(34) The biological activity for the benzodiazepine series has been
shown to reside in the (S)-configuration at the acetic acid
stereocenter, as in L-aspartic acid; see refs 24a and 25.
(35) Blackburn, B. K.; Lee, A.; Baier, M.; Kohl, B.; Olivero, A. G.;
Matamoros, R.; Robarge, K. D.; McDowell, R. S. From peptide
to non-peptide. 3. Atropisomeric GPIIbIIIa antagonists contain-
ing the 3,4-dihydro-1H-1,4-benzodiazepine-2,5-dione nucleus. J .
Med. Chem. 1997, 40, 717-729.
(23) Ku, T. W.; Ali, F. E.; Barton, L. S.; Bean, J . W.; Bondinell, W.
E.; Burgess, J . L.; Callahan, J . F.; Calvo, R. R.; Chen, L.;
Eggleston, D. S.; Gleason, J . G.; Huffman, W. F.; Hwang, S.-M.;
J akas, D. R.; Karash, C. B.; Keenan, R. M.; Kopple, K. D.; Miller,
W. H.; Newlander, K. A.; Nichols, A.; Parker, M. F.; Peishoff, C.
E.; Samanen, J . M.; Uzinskas, I.; Venslavsky, J . W. Direct
J M980166Z