Synthesis of a novel cyclic prodrug of RGD peptidomimetic to improve its cell membrane permeation

Article abstract of DOI:10.1016/S0045-2068(02)00013-5

The objective of this work was to synthesize cyclic prodrug 2 derived from the parent RGD peptidomimetic 1 and to evaluate its chemical and enzymatic stabilities and antithrombic activity. Cyclic prodrug 2 was formed to improve the cell membrane permeation of RGD peptidomimetic 1 by transiently masking the unfavorable physicochemical properties of compound 1. Cyclic prodrug 2 was synthesized by linking the amino and carboxylic acid groups of parent 1 via the (acyloxy)alkoxy promoiety. The prodrug-to-drug conversion of cyclic prodrug 2 was evaluated in isolated esterase and human plasma in the absence and presence of the esterase inhibitor paraoxon. The rate of hydrolysis of cyclic prodrug 2 was significantly faster in plasma (t_1/2 = 33.5 ± 0.6 min) than in PBS (t_1/2 = 314 ± 11 min). Cyclic prodrug 2 was converted by esterase to the parent compound 1 and this conversion was inhibited by an esterase inhibitor, paraoxon. The IC₅₀ (4 μM) of cyclic prodrug 2 was higher than the IC₅₀ (1.9 μM) of parent drug 1. The antithrombic activity of cyclic prodrug 2 depends on the incubation time in platelet-rich plasma; the activity increases with incubation time, suggesting that the prodrug-to-drug conversion is time-dependent and mediated by esterase. Cyclic prodrug 2 was more stable under acidic and neutral conditions than under basic conditions, suggesting that handling and formulation of this prodrug should be undertaken under acidic conditions.

Full text of DOI:10.1016/S0045-2068(02)00013-5

BIOORGANIC
CHEMISTRY
Bioorganic Chemistry 30 (2002) 285–301
www.academicpress.com
Synthesis of a novel cyclic prodrug of RGD
peptidomimetic to improve its cell
membrane permeation
Xiaoping Song, Christine R. Xu,
Henry T. He, and Teruna J. Siahaan*
Department of Pharmaceutical Chemistry, The University of Kansas, 2095 Constant Avenue,
Lawrence, KS66047, USA
Received 13 November 2001
Abstract
The objective of this work was to synthesize cyclic prodrug 2 derived from the
parent RGD peptidomimetic 1 and to evaluate its chemical and enzymatic stabilities
and antithrombic activity. Cyclic prodrug 2 was formed to improve the cell mem-
brane permeation of RGD peptidomimetic 1 by transiently masking the unfavorable
physicochemical properties of compound 1. Cyclic prodrug 2 was synthesized by
linking the amino and carboxylic acid groups of parent 1 via the (acyloxy)alkoxy
promoiety. The prodrug-to-drug conversion of cyclic prodrug 2 was evaluated in
isolated esterase and human plasma in the absence and presence of the esterase in-
hibitor paraoxon. The rate of hydrolysis of cyclic prodrug 2 was significantly faster
in plasma (t1=2 ¼ 33:5 Æ 0:6 min) than in PBS (t1=2 ¼ 314 Æ 11 min). Cyclic prodrug 2
was converted by esterase to the parent compound 1 and this conversion was in-
hibited by an esterase inhibitor, paraoxon. The IC50 (4 lM) of cyclic prodrug 2 was
higher than the IC50 (1.9 lM) of parent drug 1. The antithrombic activity of cyclic
prodrug 2 depends on the incubation time in platelet-rich plasma; the activity in-
creases with incubation time, suggesting that the prodrug-to-drug conversion is time-
dependent and mediated by esterase. Cyclic prodrug 2 was more stable under acidic
* Corresponding author. Fax: 1-785-864-5736.
E-mail address: siahaan@ku.edu (T.J. Siahaan).
0045-2068/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved.
PII: S0045-2068(02)00013-5

286
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
and neutral conditions than under basic conditions, suggesting that handling and
formulation of this prodrug should be undertaken under acidic conditions. Ó 2002
Elsevier Science (USA). All rights reserved.
Keywords: Cyclic prodrug; RGD peptidomimetic; (acyloxy)alkoxy linker; Antithrombic
agents
1. Introduction
Thrombotic disease is one of the major causes of death in the United States; in
addition, more than one-and-one-half million people are hospitalized with myo-
cardial infarctions each year [1,2]. The basic cause of this group of diseases is the
formation of a thrombus that attaches to blood vessels, causing disruption of normal
blood flow as well as tissue damage. There is a need to develop inhibitors for
thrombosis that can be delivered orally because oral drug delivery is convenient and
cost-effective [3]. In the process of thrombosis, activated platelets aggregate to form
blood clots via multivalent interactions between GPIIb/IIIa receptors on the surface
of platelets and fibrinogen (Fg) [4–7]. Binding between Fg and GPIIb/IIIa receptors
is mediated by the Arg–Gly–Asp (RGD) sequence on fibrinogen a- and c-chains
[8–12]. Therefore, RGD peptides and peptidomimetics have been developed as
antithrombic agents to inhibit platelet aggregation in vitro and in vivo [2,13–17].
Potent RGD peptidomimetics contain a positively charged guanidino or amino
group and a negatively charged carboxylic acid group, separated by a certain dis-
tance for their binding selectivity to GPIIb/IIIa [16,18–20]. Unfortunately, these
charges and other physicochemical properties (i.e., hydrogen-bonding potential,
conformation) hinder the permeation of RGD peptidomimetics through cell
Scheme 1.

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
287
membranes of the intestinal mucosa for oral bioavailability (Scheme 1); for this
reason, RGD peptidomimetics have been developed only for intravenous (IV) ap-
plications [2]. Many attempts have been made to develop orally active RGD pep-
tidomimetics; however, some of these failed in clinical trials. Thus, a new method to
improve cell membrane permeation of RGD peptidomimetics is necessary to increase
their oral bioavailability. One approach to increase membrane permeation of RGD
peptidomimetics is in transiently altering the physicochemical properties (e.g.,
charge, hydrogen-bonding potential, conformation, hydrodynamic radii, and lipo-
philicity) of the RGD peptidomimetics. This can be done by linking the amino and
carboxylic acid groups via the acyloxyalkoxy promoiety to form cyclic prodrugs
(Scheme 1) [21–23]. Previously, we have shown that the formation of cyclic prodrugs
improves the permeation of peptides via the Caco-2 cell monolayer, a model for the
intestinal mucosa [23–25]. The cyclic prodrugs permeate through the cell membranes
more significantly than do the parent linear peptides [23,24,26,27]. The improvement
in membrane permeation of the cyclic prodrugs is due to: (a) increased conforma-
tional stability and intramolecular hydrogen-bonding potential, (b) an increase in
hydrophobicity produced by masking the charges in the N- and C-termini, (c)
Scheme 2.

288
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
improvement of enzymatic stability against exopeptidase and endopeptidases, and
(d) reduction of hydrodynamic radii [25,28,29]. After crossing the cell membranes,
the prodrug can be converted to the parent drug by esterase in the bloodstream.
Herein, we describe the synthesis of cyclic prodrug 2 from the parent RGD
peptidomimetic 1 (Scheme 2) to improve its membrane permeation properties. The
parent peptidomimetic 1 is the derivative of a known potent RGD peptidomimetic 3
[30]; the amidine group in compound 3 was replaced with an amino group in
compound 1. The antithrombic activity of cyclic prodrug 2 was compared to that of
the parent compound 1, RGD peptidomimetics 3 and 4, and RGD peptide 5. The
synthesis of cyclic prodrug 2 was accomplished by linking the amino and carboxylic
acid groups via the acyloxyalkoxy promoiety. Because RGD peptidomimetic 1 has
aromatic and piperidine rings, cyclic prodrug 2 has high rigidity, which contributes
to the physicochemical and biopharmaceutical properties of the prodrug. The en-
zymatic conversion of cyclic prodrug 2 was evaluated to determine the ability of the
prodrug to release the RGD peptidomimetic 1.
2. Results and discussion
2.1. Synthesis of cyclic prodrug 2
The synthesis of cyclic prodrug 2 was accomplished by forming an amide bond
between the amino functional group of the piperidine ring and the carboxylic acid of
the alanine in linear precursor 6, as shown in Scheme 2. The linear precursor 6 was
synthesized from intermediate 18, which can be formed via two different paths
(Scheme 4). The first path is conjugation of the two key intermediates 7 and 8.
Unfortunately, intermediate 7 contains an (acyloxy)alkoxy promoiety linked to the
piperidine acetic acid group and has a side product 15. Intermediate 8 is the 4-
aminomethylbenzoylalanine peptide. The second path to make intermediate 18 is the
formation of intermediate 21 from compound 8. Intermediate 21 was reacted with
compound 14 to give intermediate 18 without a side product.
The synthesis of intermediate 7 (Scheme 3) was completed by conjugating com-
pounds 10 and 14. The synthesis of compound 10 was initiated using commercially
available 1-chloromethyl chloroformate. Treatment of 1-chloromethyl chlorofor-
mate with p-nitrophenol in the presence of N-methylmorpholine (NMM) produced
compound 9 in 94% yield. The chloride atom in 9 was replaced with iodine by re-
acting it with sodium iodide in acetone to produce compound 10 in 93% yield.
Synthesis of compound 14 was started from the commercially available 4-hydrox-
ypiperidine [30]. The secondary amine of 4-hydroxypiperidine was protected with a
benzyloxycarbonyl (Cbz) group; this was carried out by reacting 4-hydroxypiperi-
dine with benzylchloroformate to obtain compound 11 in 96% yield. The hydroxyl
group in compound 11 was alkylated with 2-bromoacetate-t-butyl ester to yield 90%
of compound 12. The t-butyl ester-protecting group in 12 was removed with triflu-
oroacetic acid (TFA) in CH2Cl2 to produce acid 13 in 100% yield. Acid 13 was
converted to cesium salt by treating it with Cs2CO3 in methanol to give a quanti-
tative yield of salt 14. Addition of cesium salt 14 into iodo intermediate 10 in

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
289
Scheme 3.
dimethylformamide (DMF) gave the key intermediate 7 in 42% yield; the low yield
was due to the formation of side product 15 in 58% yield. Previously, we have shown
that this is one of the most challenging steps in the synthesis of this type of cyclic
prodrug; the yield of this intermediate depends on the structure of the acid [21]. The
mixture of compounds 7 and 15 was used directly in the next reaction due to the
difficulty in separating them.

290
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
The key intermediate 8 was produced by coupling the Boc-p-aminomethyl benzoic
acid (16) and the alanine benzyl ester as shown in Scheme 4. Protection of the amino
group in p-aminomethylbenzoic acid with the t-Boc group gave compound 16 in 85%
yield. Activation of the carboxylic acid group of 16 was accomplished using EDC
and HOBT; the activated acid was subsequently reacted with alanine benzyl ester to
give compound 17 in 99% yield. Deprotection of the Boc group with TFA/CH2Cl2
produced the desired intermediate 8 in a quantitative yield.
The final process in the synthesis of cyclic prodrug 2 was the formation of key
intermediate 6, which can be formed by hydrogenation of compound 18. Early in our
study, compound 18 was made by conjugation of intermediate 8 with a mixture of
compounds 7 and 15 in DMF in the presence of HOBT and NMM. This reaction
gave a side product 19, which did not contain the (acyloxy)alkoxy promoiety.
Compound 18 (78%) can be separated with difficulty from the side product 19 using
silica gel column chromatography. In a separate attempt, the mixture was also
subjected to the next reaction without purification; the Cbz- and OBzl-protecting
groups in compounds 18 and 19 were removed simultaneously by hydrogenation to
give compounds 6 and 4. Compound 6 was purified by semi-preparative HPLC on a
C-18 reversed-phase column. Compound 4 was also isolated and was tested for
antithrombic activity. Because this synthetic route is extremely time-consuming and
the yield is low, we developed a new route to prepare 18 in fewer steps with higher
yield and easier purification (Scheme 4).
When the TFA salt 8 was initially converted to free amine 20, it reacted readily
with 1-chloromethyl chloroformate in the presence of a base (TEA) to produce
compound 21 in 95% yield. Unlike the formation of compound 10 in Scheme 3,
metathesis of chloride to iodide atom in 21 was unsuccessful. However, the coupling
reaction between 21 and cesium salt 14 proceeded smoothly to give the key inter-
mediate 18 in 76% yield after silica gel column chromatography. Because no side
products were formed, the subsequent hydrogenation compound 6 could be used
Scheme 4.

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
291
without purification. This optimal method can provide large amounts of cyclic
prodrug for in vitro and in vivo studies. Cyclization of linear precursor 6 was ac-
complished under high-dilution conditions in DMF using HBTU as an activating
reagent in the presence of N,N-diisopropylethyl amine (DIEA); the desired cyclic
prodrug 2 was isolated in 20% yield after HPLC purification. The low yield obtained
in this cyclization step may be due to (a) the difficulty in the formation of a rigid
macrocyclic ring, (b) the possibility of forming oligomers, even under high-dilution
conditions, or (c) the low recovery during purification by preparative HPLC.
The proton NMR spectrum of cyclic prodrug 2 showed two conformers; these
conformers were derived from the cis- and trans-isomerization at the peptide bond
between the (acyloxy)alkoxy linker and the aminobenzoic acid group as shown in
Scheme 2. A similar isomerization phenomenon was also found in other cyclic
prodrugs that contain the (acyloxy)alkoxy linker [29,31].
2.2. Synthesis of compounds 1 and 3
Compound 1: The synthesis of parent compound 1 is shown in Scheme 5; it is
started from compounds 12 (Scheme 3) and 17 (Scheme 4). The Cbz-protecting
Scheme 5.

292
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
group in compound 12 was removed by hydrogenation to give amine 22 in a
quantitative yield. Similarly, the benzyl ester-protecting group in compound 17 was
also removed by hydrogenation to give acid 23 in a quantitative yield. Acid 23 was
activated with HBTU and NMM followed by reaction with amine 22 to give com-
pound 24 in 57% yield. Boc- and t-Bu-protecting groups in compound 24 were re-
moved with TFA in dichloromethane to give the desired parent compound 1 in 100%
yield.
Compound 3: The synthesis of compound 3 was accomplished using the method
described by Alig et al. [30].
2.3. Biological study
Antithrombic activity: The antithrombic activities of compounds 1–5 were eval-
uated using an ADP-activated platelet aggregation assay and their IC50 results are
given in Scheme 2 [32]. RGDF peptide 5, with an IC50 of 12 lM, was used as a
positive control. RGD peptidomimetic 1 has an IC50 of 1.9 lM, about tenfold higher
than that of parent peptidomimetic 3 (IC50 of 0.14 lM). Replacement of the amid-
inine group with an amino group in 1 lowers the selectivity of RGD peptidomimetics
for GPIIb/IIIa. This may be due to the preference of the GPIIb/IIIa receptor for an
amidinine over an amino group. The side product 4 was also tested in the platelet
aggregation assay; this compound has an IC50 of 300 lM. Although the charge
separation in 4 is the same as that in 1, compound 4 is less active than compound 1;
this result confirms the hypothesis that not only the charge separation but also the
structure of RGD peptidomimetic is important. To prove that cyclic prodrug 2 can
be converted to parent drug 1, the anti-platelet aggregation activity of cyclic prodrug
2 was investigated. The inhibitory activity of cyclic prodrug 2 was evaluated under
different conditions. Cyclic prodrug 2 has an IC50 of 10 lM when platelet aggrega-
tion is induced immediately after the addition of the prodrug; however, it has an
optimal IC50 of 4 lM when incubated for 10–20 min PRP, prior to induction of
aggregation. This result suggests that the effective conversion of prodrug to drug by
esterase in PRP takes about 10 min.
Chemical and enzymatic stability: The chemical stability of cyclic prodrug 2 was
evaluated at pH 4, 7, and 10 in PBS (Table 1). Prodrug 2 is more stable at pH 4.0
(t1=2 ¼ 985 Æ 37 min) than at pH 7.0 (t1=2 ¼ 314 Æ 11 min) or pH 10.0 (t1=2 ¼ 1:46 Æ
0:07 min). As we observed previously, the (acyloxy)alkoxy promoiety is chemically
Table 1
Half-life (t1=2, min) for the stability in aqueous buffer of various pH values (37 °C, l ¼ 0:15)
and enzymatic stability of cyclic prodrug 2
Enzymatic stability
No paraoxon
With paraoxon
Isolated esterase
Plasma
16:5 Æ 0:9
33:5 Æ 0:6
567 Æ 14
75:1 Æ 2:1
Chemical stability
PBS
pH ¼ 4
pH ¼ 7
pH ¼ 10
985 Æ 37
314 Æ 11
1:46 Æ 0:07

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
293
unstable at basic pH and stable at acidic pH [21,24,33]. This result suggests that
cyclic prodrug 2 will be stable at the pH of the stomach.
The ability of esterase to hydrolyze the ester bond of the promoiety is necessary
for releasing the RGD peptidomimetic from the cyclic prodrug into the bloodstream.
The target for RGD peptidomimetics is the platelets in the bloodstream. The con-
version of prodrug 2 to parent drug 1 by esterase was confirmed by comparing the
stability of the prodrug in platelet-rich plasma or isolated esterase in the presence
and absence of the esterase inhibitor paraoxon [24,26]. As shown in Table 1, the
prodrug can be converted to the parent drug by isolated porcine and plasma ester-
ases as determined by HPLC. In general, the half-lives of the prodrugs in isolated
esterase (Table 1) were shorter than those in PBS at pH 7.0, suggesting that esterase
can hydrolyze the prodrugs. The rate of hydrolysis of cyclic prodrug 2 was signifi-
cantly faster in human plasma in the absence of paraoxon (33:5 Æ 0:6 min) than in
the presence of paraoxon (t1=2 ¼ 75:1 Æ 2:1 min) (Table 1). This enzymatic reaction
was also conducted in isolated esterase in the presence and the absence of esterase
inhibitor. The rate of conversion of prodrug 2 in isolated esterase solution showed a
half-life of 16:5 Æ 0:9 min in the absence of paraoxon and 567 Æ 14 min in the
presence of paraoxon. These results suggest that prodrug 2 is converted to parent
compound 1 by esterase. The production of parent compound 1 in the enzymatic
reaction was observed by HPLC. Paraoxon was less effective in inhibiting the pro-
drug-to-drug conversion in plasma than in isolated esterase; this may be due to the
presence of many different types of esterases in human plasma. Thus, paraoxon may
not effectively inhibit all of the different esterases in the plasma.
In another study, the intrinsic membrane permeation of cyclic prodrug 2 was
compared to that of the parent compound 1 using Caco-2 cell monolayers, an in
vitro model for the intestinal mucosa. The cyclic prodrug 2 (Papp ¼ 10:3 Æ 0:74 Â
10À7 cm=s) can permeate the cell membrane about 13 times better than the parent
RGD peptidomimetic 1 (Papp ¼ 0:77 Æ 0:10 Â 10À7 cm=s). These results are congru-
ent with previous studies on cyclic prodrugs of RGD peptidomimetics using
coumaric acid linkers [34,35]. The improvement in the membrane permeation of
cyclic prodrug 2 is due to the alteration of its physicochemical properties (i.e.,
charges, hydrogen-bonding potential, and conformation) compared to those of
parent RGD peptidomimetic 1. Although the in vitro results suggest that cyclic
prodrug formation may improve the oral bioavailability of RGD peptidomimetics,
in vivo studies of the prodrugs will be essential for determining the utility of this
method.
3. Conclusions
We have developed a route to synthesize a new rigid cyclic peptide prodrug 2 by
using an (acyloxy)alkoxy promoiety to link the amino and carboxylic acid groups of
the parent peptidomimetic 1. Cyclic prodrug 2 is susceptible to esterase metabolism
(slow step), leading to a cascade of chemical reactions and resulting in the release of
the linear parent peptidomimetic 1. The antithrombic activity of cyclic prodrug 2 is
due to prodrug-to-drug conversion by esterase in the plasma. The chemical stability

294
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
of the cyclic prodrug 2 is higher at acidic pH than at neutral and basic pH. Finally,
prodrug 2 can permeate the cell membrane better than parent compound 1.
4. Experimental methods
4.1. General methods
1H-NMR spectra were recorded on 500 and 400 MHz Bruker NMR instruments.
Chemical shifts are expressed in parts per million (d), relative to the internal standard
tetramethylsilane (TMS) or residual proton from solvent. Abbreviations are as
follows: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; and br, broad.
High-resolution mass spectra (HRMS) were obtained using a VG Analytical ZAB
double-focusing spectrometer. Elemental analyses were performed by Quantitative
Technologies, New Jersey. All starting materials were purchased from Aldrich
Chemical, Sigma Chemical, Fluka Chemical, or Bachem Bioscience and used as
received.
HPLC purification and analysis were conducted using a Rainin HPXL gradient
system with a Dynamax UV detector. The desired product was purified by semi-
ꢀ
preparative reversed-phase HPLC using a C-18 column (12 lm, 300 A, 25cm Â
21:4mm i.d., flowrate 10 mL/min) and eluting with a gradient of solvent A (0.1%
TFA/H2O:5% MeCN) and solvent B (100% MeCN). The gradient method used for
the preparative HPLC started from 0% to 35% of solvent B for 23 min, followed by an
increase to 100% of solvent B in 10 min. The final elution was at 100% solvent B for
2 min, followed by gradient change to 100% solvent A in 2 min. The desired peptide
ꢀ
was analyzed by analytical reversed-phase HPLC using a C-18 column (5 lm, 300 A,
25cm  4:6mm i.d., flowrate 1 mL/min) eluting with a gradient of solvents A and B.
The total HPLC analysis run time was 18 min. The solvent gradient used for the
analytical HPLC was 0–50% of solvent B over 12 min. The composition of solvent B
was increased to 100% in 2 min, followed by elution at 100% of solvent B for 2 min.
Finally, the solvent was changed to 0% solvent B over 2 min for the final equilibration.
1-Chloromethyl-p-nitrophenyl carbonate (9). Compound 9 was prepared and
obtained (94%) in the same way, as reported in the literature [21].
1-Iodomethyl-p-nitrophenyl carbonate (10). Compound 10 was synthesized (93%)
from compound 9 according to reported procedures [21].
N-Benzyloxycarbamate-4-hydroxypiperidine (11). Triethylamine (TEA) (5.06 g,
50 mmol) and benzyl chloroformate (8.53 g, 50 mmol) were added to a solution of
4-hydroxypiperidine (5.06 g, 50 mmol) in CH2Cl2 at 0 °C. The reaction mixture
was then stirred at room temperature for 16 h. The resulting suspension was filtered
and the residue after solvent evaporation was taken up in EtOAc (50 mL), which
was washed with H2O (2 Â 20 mL) and 10% citric acid (40 mL) and dried over
anhydrous Na2SO4. The solvent was removed to give the desired compound 11
(11.32 g, 96%) as a pale yellow oil. This compound was used without further puri-
fication. 1H-NMR (CDCl3; d): 1.48 (2H, m), 1.85 (2H, m), 3.15 (2H, m), 3.89 (3H, m),
5.12 (2H, s), 7.35 (5H, m). The NMR spectrum of 11 was identical to that reported in
the literature [30].

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
295
2-(N-Benzyloxycarbamate-4-piperidinyloxy)-t-butyl acetate (12). Tert-butyl bro-
moacetate (8.8 g, 45 mmol) and tetra-n-butylammonium hydrogen sulfate (0.339 g,
1 mmol) dissolved in H2O (5 mL) were added to toluene solution (50 mL) of com-
pound 11 (7.058 g, 30 mmol). A solution of NaOH (0.625 M, 25 mL) was added
dropwise to the reaction mixture followed by vigorous stirring for 20 h. The organic
layer was then separated and dried over anhydrous Na2SO4. The solvent was re-
1
moved to give compound 12 (9.46 g, 90%) as light yellow oil. H-NMR (CDCl3; d):
1.48 (9H, s), 1.61 (2H, m), 1.85 (2H, m), 3.15 (2H, m), 3.58 (1H, m), 3.83 (2H, m), 4.0
(2H, s), 5.16 (2H, s), 7.34 (5H, m).
2-(N-Benzyloxycarbamate-4-piperidinyloxy)-acetic acid (13). TFA (25 mL) was
added to a stirred solution of compound 12 (9 g, 25.8 mmol) in CH2Cl2 at 0 °C. The
reaction mixture was stirred at room temperature for 1 h. Volatile compounds in the
reaction mixture were removed by rotary evaporation under vacuum. The residue
was triturated and washed with Et2O/petroleum ether (1:1) to give a white solid. The
residual solvent was removed under vacuum to give compound 13 as a white solid
(7.56 g, 100%); this compound was used in the next step without further purification.
2-(N-Benzyloxycarbamate-4-piperidinyloxy)-acetyl-oxymethyl-p-nitro-phenyl car-
bonate (7). The cesium salt of compound 13 (2.93 g, 10 mmol) was prepared by re-
acting it with Cs2CO3 (1.7 g, 5.2 mmol) in MeOH (35 mL). After stirring for 1 h, the
solvent was removed under reduced pressure to afford a white powder of cesium salt
14. Cesium salt 14 (4.25 g, 10 mmol) in DMF (150 mL) was added slowly to an ice-
cold stirred solution of iodomethyl-p-nitrophenyl carbonate 10 (3.23 g, 10 mmol) in
DMF (50 mL) over a period of 2 h. After stirring for 29 h at room temperature,
DMF was removed under reduced pressure to give an oily residue. This residue was
dissolved in ethyl acetate (150 mL), followed by washing with 10% NaHCO3
(2 Â 40 mL), H2O (2 Â 50 mL), and brine (40 mL). The ethyl acetate layer was sep-
arated and dried over anhydrous Na2SO4; the solvent was then removed to give the
desired compound 7 (42%) and the side product 15 (58% yield) as determined by
proton NMR. Due to the difficulty of separating these compounds, this mixture was
used in the next step. Compound 7 1H-NMR (CDCl3; d): 1.65 (2H, m), 1.90 (2H, m),
3.23 (2H, m), 3.68 (1H, m), 3.84 (2H, m), 4.24 (2H, s), 5.13 (2H, s), 5.96 (2H, s), 7.36
(7H, m), 8.29 (2H, d, J ¼ 9.0 Hz). MS (FAB) m/z: 489 (Mþ þ 1).
N-Boc-4-aminomethyl-benzoic acid (16). 4-Aminomethyl benzoic acid (5.36 g,
20 mmol) and di-t-butyl dicarbonate were dissolved in a mixture of water (67 mL)
and dioxane (100 mL). NaOH (1.0 N) was added dropwise to this solution at 0 °C to
adjust the pH to 9.0. The reaction mixture was stirred at room temperature for 1.5 h
and during this period some white precipitate was formed. The reaction mixture was
concentrated to a volume of 60 mL, followed by addition of EtOAc (70 mL). The
mixture was cooled in an ice bath, followed by addition of 15% citric acid to adjust
the pH to 3.0. The aqueous layer was extracted with EtOAc (4 Â 40 mL) and the
combined organic layer was dried over anhydrous Na2SO4. A white solid after
solvent evaporation was purified by recrystallization in EtOAc to yield pure com-
pound 16 (5.02 g, 85%). 1H-NMR (DMSO-d6; d): 1.38 (9H, s), 4.16 (2H, d,
J ¼ 6.0 Hz), 7.31 (2H, d, J ¼ 8.0 Hz), 7.46 (1H, t, J ¼ 6.0 Hz), 7.86 (2H, d, J ¼ 8.0 Hz).
4-(N-Boc-aminomethyl)-benzoyl-Ala-OBzl (17). Acid 16 (3.765 g, 15 mmol),
HOBT (2.027 g, 15 mmol), and EDC (2.876 g, 15 mmol) were dissolved in CH2Cl2

296
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
(120 mL) at 0 °C. NMM (1.52 g, 15 mmol) and H-Ala-OBzl (2.685 g, 15 mmol) were
added to the reaction mixture, followed by stirring at 0 °C for 4 h and at ambient
temperature for 24 h. A residue remaining after solvent evaporation was dissolved
in EtOAc (150 mL); it was successively washed with 10% aqueous citric acid
(2 Â 40 mL), H2O (60 mL), saturated NaHCO3 (2 Â 40 mL), H2O (60 mL), and
brine (50 mL), and dried over anhydrous Na2SO4. Solvent was removed under
reduced pressure to give a white solid, which was purified by recrystallization in
1
EtOAc to yield compound 17 (6.1 g, 99%). H-NMR (CDCl3; d): 1.45 (9H, s), 1.53
(3H, d, J ¼ 7.3 Hz), 4.36 (2H, d, J ¼ 6.1 Hz), 4.82 (1H, m), 5.21 (2H, dd,
J1 ¼ 5:0 Hz, J2 ¼ 13:0 Hz), 6.8 (1H, br), 7.38 (7H, m), 7.76 (2H, d, J ¼ 8.0 Hz). MS
(FAB) m/z: 413 (Mþ þ 1). HRMS: Calcd for C23H28N2O5, 413.2076. Found:
413.2079.
TFA ꢀ 4-Aminomethyl-benzoyl-Ala-OBzl (8). Compound 17 (4.12 g, 10 mmol) in
CH2Cl2 (20 mL) was added to TFA (20 mL) at 0 °C. After stirring at room tem-
perature for 1 h, the reaction mixture was concentrated by a rotary evaporator under
reduced pressure. The remaining residue was triturated with Et2O/petroleum ether
(1:1) to give a white solid. The solid was isolated by decantation and the residual
ether was removed under vacuum to give compound 8 (4.26 g, 100%). This com-
1
pound was used in the next step without further purification. H-NMR (DMSO-
d6; d): 1.42 (3H, d, J ¼ 7.3 Hz), 4.11 (2H, t, J ¼ 5.4 Hz), 4.52 (1H, quintet, J ¼ 7.2 Hz),
5.13 (2H, dd, J1 ¼ 2:6 Hz, J2 ¼ 12:7 Hz), 7.29–7.38 (5H, m), 7.54 (2H, d, J ¼ 8.2 Hz),
7.92 (2H, d, J ¼ 8.2 Hz), 8.28 (3H, br), 8.87 (2H, d, J ¼ 6.9 Hz). MS (FAB) m/z: 313
(Mþ þ 1). HRMS: Calcd for C18H21N2O3, 313.1552. Found: 313.1541.
4-Aminomethyl-benzoyl-Ala-OBzl (20). Compound 8 (2.30 g, 5.40 mmol) was
dissolved in a mixture of CH2Cl2 (100 mL) and NaOH (2 M, 50 mL). The organic
phase was separated, washed thoroughly with NaOH (2 M, 50 mL), and dried over
anhydrous MgSO4. Solvent was removed to yield a white solid, which was dried
1
under vacuum to give compound 20 (1.60 g, 95%). H-NMR (CD3OD; d): 1.53 (3H,
d, J ¼ 7.3 Hz), 3.86 (2H, s), 4.61–4.69 (1H, m), 5.21 (2H, dd, J1 ¼ 6:0 Hz,
J2 ¼ 12:4 Hz), 7.40–7.38 (5H, m), 7.45 (2H, d, J ¼ 8.2 Hz), 7.84 (2H, d, J ¼ 8.2 Hz).
1-Chloromethyleneoxycarbonyl-4-aminomethyl-benzoyl-Ala-OBzl (21). A solution
of 1-chloromethyl chloroformate (0.79 g, 6.12 mmol) in dry CH2Cl2 (20 mL) was
added dropwise to compound 20 (1.60 g, 5.13 mmol) in dry CH2Cl2 (40 mL) and
TEA (0.62 g, 6.12 mmol) in an ice bath. After stirring overnight at room temperature,
the white residue remaining after solvent evaporation was taken up in EtOAc
(150 mL); this was followed by washing with 5% KHSO4 (2 Â 50 mL), H2O
(2 Â 50 mL), and brine (2 Â 50 mL). The organic layer was separated, dried over
anhydrous MgSO4, and concentrated until a precipitate formed; petroleum ether
(150 mL) was added to complete the precipitation. The precipitate was filtered,
washed with petroleum ether, and dried under vacuum to give compound 21 (2.0 g,
1
95%). H-NMR (CDCl3; d): 1.55 (3H, d, J ¼ 7.2 Hz), 4.47 (2H, d, J ¼ 6.1 Hz), 4.78–
4.89 (1H, m), 5.24 (2H, dd, J1 ¼ 5:0 Hz, J2 ¼ 12:3 Hz), 5.50 (1H, s), 5.80 (2H, s), 6.79
(1H, d, J ¼ 7.0 Hz), 7.34–7.42 (7H, m), 7.77 (2H, d, J ¼ 8.1 Hz). MS (FAB) m/z: 405
(Mþ þ 1).
2-(N-Benzyloxycarbamate-4-oxypiperidinyl)-acetyl-methyleneoxycarbonyl-4-ami-
nomethyl-benzoyl-Ala-OBzl (18). Method 1: A solution of compound 8 (2.03 g,

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
297
4.76 mmol) and NMM (0.97 g, 9.6 mmol) in DMF (50 mL) was added to a stirred
solution of compound 7 (0.976 g, 2.0 mmol), side product 15 (1.11 g, 2.76 mmol), and
HOBT (0.643 g, 4.76 mmol) in DMF (25 mL). The reaction mixture was stirred at
room temperature for 3 h and then DMF was removed under reduced pressure to
give an oily residue. The oily residue was dissolved in EtOAc (150 mL), followed by
washing with 10% NaHCO3 (2 Â 50 mL), H2O (4 Â 50 mL), and brine (40 mL). The
organic layer was dried over anhydrous Na2SO4 and the solvent was removed under
reduced pressure to yield a pale yellow oily mixture of compound 18 and by-product
19. The pure compound 18 (1.035 g, 78%) was obtained as a colorless solid after
silica gel column purification using EtOAc:hexane:isopropanol (4:2:1) as eluent. 1H-
NMR (CDCl3; d): 1.53 (3H, d, J ¼ 7.3 Hz), 1.61 (2H, m), 1.85 (2H, m), 3.22 (2H, m),
3.60 (1H, m), 3.73 (2H, m), 4.18 (2H, s), 4.44 (2H, d, J ¼ 6.1 Hz), 4.84 (1H, m), 5.12
(2H, s), 5.21 (2H, dd, J1 ¼ 5:0 Hz, J2 ¼ 13:0 Hz), 5.82 (2H, s), 6.8 (1H, br), 7.36
(13H, m), 7.76 (2H, d, J ¼ 8.0 Hz). MS (FAB) m/z: 662 (Mþ þ 1). HRMS: Calcd for
C35H40N3O10, 662.2713. Found: 662.2736. Method 2: Compound 21 (3.0 g,
7.43 mmol) in DMF (40 mL) was added dropwise to the stirred cesium salt 14 (3.77 g,
8.87 mmol) in DMF (60 mL). After stirring overnight at room temperature, the white
residue left after solvent removal was taken up in EtOAc (200 mL), washed with 5%
KHSO4 (2 Â 50 mL), H2O (2 Â 50 mL), and brine (2 Â 50 mL), and dried over an-
hydrous MgSO4. A white solid remaining after solvent evaporation was purified on a
silica gel column using EtOAc:hexane (2:1 to 3:1) as eluent to give compound 18
(3.7 g, 76%).
TFA ꢀ 2-(4-oxypiperidinyl)-acetyl-methyleneoxycarbonyl-4-aminomethyl-benzoyl-
Ala-OH (6). Compound 18 (0.661 g, 1.0 mmol) was dissolved in MeOH (40 mL) and
TFA (2 mL) and to this solution was added 10% Pd/C (60 mg). The reaction mixture
was stirred at room temperature under a H2 atmosphere using a balloon on the top
of a round-bottomed flask. After stirring for 5 h, the reaction mixture was filtered to
remove the catalyst (Pd/C); the solvent was removed under reduced pressure to give
an oily residue. This residue was washed with Et2O (3 Â 20 mL) to afford the TFA
salt of compound 6 (0.552 g) in a quantitative yield. This compound was further
purified using semi-preparative HPLC. 1H-NMR (DMSO-d6; d): 1.38 (3H, d,
J ¼ 7.0 Hz), 1.66 (2H, m), 1.92 (2H, m), 2.95 (2H, m), 3.15 (2H, m), 3.65 (1H, m),
4.25 (2H, s), 4.26 (2H, d, J ¼ 6.0 Hz), 4.38 (1H, m, J ¼ 7.0 Hz), 7.32 (2H, d,
J ¼ 8.0 Hz), 7.82 (2H, d, J ¼ 8.0 Hz), 8.20 (1H, t, J ¼ 7.0 Hz), 8.61 (1H, d, J ¼ 6.0 Hz).
MS (FAB) m/z: 438 (Mþ þ 1). HRMS: Calcd for C20H28N3O8, 438.1876. Found:
438.1891.
Cyclic prodrug (2). Compound 6 (0.25 g, 0.45 mmol) was dissolved in DMF
(500 mL), followed by addition of HBTU (0.853 g, 2.25 mmol). The mixture was
stirred at room temperature for 30 min under N2, followed by addition of DIEA
(0.581 g, 4.5 mmol). After the reaction mixture was stirred for 16 h, the DMF was
removed under reduced pressure to give a light brown residue. Cyclic peptide 2 was
purified from this residue by semi-preparative HPLC with a C-18 column to give
20% yield of pure product. The HPLC run used a stepped gradient of 0–35% solvent
B for 23 min, 35–100% solvent B for 8 min, and 100% solvent B for 2 min. Cyclic
peptide 2 was analyzed by analytical reversed-phase HPLC using a C-18 column
ꢀ
(5.0 lm, 300 A, 25cm  4:6mm i.d., flowrate 1 mL/min) to give a retention time of

298
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
10 min. The cyclic prodrug can also be purified by recrystallization from CHCl3/
1
hexane. H-NMR (DMSO-d6; d) for major conformer: 1.24 (3H, d, J ¼ 6.2 Hz), 1.44
(2H, br), 1.74 (2H, br), 2.77 (2H, br), 2.80 (2H, br), 3.73 (1H, d, J ¼ 17 Hz), 3.83 (1H,
d, J ¼ 17 Hz), 4.04 (1H, dd, J1 ¼ 6:3 Hz, J2 ¼ 14 Hz), 4.12 (1H, dd, J1 ¼ 5:6 Hz,
J2 ¼ 14 Hz), 4.35 (1H, br), 4.97 (1H, m, J ¼ 7.3 Hz), 5.52 (1H, d, J ¼ 6.0 Hz), 5.63
(1H, d, J ¼ 6.0 Hz), 7.35 (2H, d, J ¼ 8.0 Hz), 7.85 (2H, d, J ¼ 8.0 Hz), 8.12 (1H, t,
1
J ¼ 6.2 Hz), 8.90 (1H, d, J ¼ 9.0 Hz). H-NMR (DMSO-d6; d) for minor conformer:
0.55 (2H, br), 0.87 (1H, bm), 1.22 (3H, d, J ¼ 6.0 Hz), 1.35 (1H, br), 2.35 (1H, br),
2.46 (1H, br), 2.67 (1H, br), 2.74 (1H, br), 3.84 (1H, d, J ¼ 6.0 Hz), 3.57 (1H, d,
J ¼ 6.0 Hz), 4.04 (1H, dd, J1 ¼ 6:3 Hz, J2 ¼ 14 Hz), 4.12 (1H, dd, J1 ¼ 5:6 Hz,
J2 ¼ 14 Hz), 4.24 (1H, br), 4.97 (1H, m, J ¼ 7.3 Hz), 5.55 (1H, d, J ¼ 6.0 Hz), 5.70
(1H, d, J ¼ 6.0 Hz), 7.25 (2H, d, J ¼ 8.0 Hz), 7.85 (2H, d, J ¼ 8.0 Hz), 7.91 (1H, t,
J ¼ 6.2 Hz), 8.95(1H, d, J ¼ 9.0 Hz). MS (FAB) m/z: 420 (Mþ þ 1). HRMS: Calcd for
C20H25N3O7, 420.1771. Found: 420.1797. Anal. Calcd for C20H25N3O7 ꢀ 0:8H2O: C,
55.37; H, 6.18; N, 9.69%. Found: C, 55.42; H, 6.07; N, 9.62%.
(4-Piperidinyloxy)-t-butyl acetate (22). To a solution of compound 12 (3.53 g,
10.0 mmol) in MeOH (50 mL) was added 10% Pd/C (0.53 g) in one portion. The
reaction mixture was stirred at room temperature for 4.5 h under a H2 atmosphere
using a balloon on the top of a round-bottomed flask. The reaction mixture was
filtered to remove the catalyst (Pd/C); the solvent was removed under reduced
1
pressure to give compound 22 (2.15 g) in quantitative yield as an oil. H-NMR
(DMSO-d6; d): 1.48 (2H, m, J ¼ 6.0 Hz), 1.49 (9H, s), 1.81 (2H, m, J ¼ 6.0 Hz), 2.5
(2H, br), 2.9 (2H, m, J ¼ 6.0 Hz), 3.45 (1H, m, J ¼ 6.0 Hz), 3.95 (2H, s). MS (FAB) m/
z: 216 (Mþ þ 1).
4-(N-Boc-aminomethyl)-benzoyl-Ala-OH (23). To a solution of compound 17
(1.257 g, 3.05 mmol) in MeOH (30 mL) was added 10% Pd/C (0.19 g) in one portion.
The reaction mixture was stirred at room temperature for 4.5 h under a balloon H2
atmosphere. The reaction mixture was filtered to remove the catalyst (Pd/C); the
solvent was removed under reduced pressure to give compound 23 (1.0 g, 100%) as a
1
colorless oil. H-NMR (DMSO-d6; d): 1.37 (3H, d, J ¼ 7.0 Hz), 1.42 (9H, s), 4.16
(2H, d, J ¼ 6.0 Hz), 4.40 (1H, m, J ¼ 7.3 Hz), 7.32 (2H, J ¼ 8.0 Hz), 7.45 (1H, t,
J ¼ 7.5 Hz), 7.81 (2H, d, J ¼ 8.0 Hz), 8.65 (1H, d, J ¼ 7.0 Hz). MS (FAB) m/z: 323
(Mþ þ 1).
2-[N-(Boc-aminomethyl-4-benzoyl-Ala)-4-piperidinyloxy]-t-butyl acetate (24).
Compound 22 (0.645 g, 3.0 mmol) and compound 23 (0.97 g, 3.0 mmol) were dis-
solved in DMF (70 mL) and the reaction mixture was kept in an ice bath. HBTU
(1.326 g, 3.5 mmol) was added to the stirred solution in one portion, followed by the
addition of NMM (0.405 g, 4.0 mmol). The reaction mixture was stirred at 0 °C for
2 h and then at room temperature for 24 h. DMF was removed under reduced
pressure to give an oily residue that was dissolved in EtOAc (100 mL), followed by
washing with 10% aqueous citric acid (2 Â 20 mL), H2O (4 Â 30 mL), 10% aqueous
NaHCO3 (2 Â 20 mL), and brine (30 mL). The organic layer was dried over anhy-
drous Na2SO4 and the solvent was removed under reduced pressure to yield a col-
orless oil that was purified on a silica-gel column using EtOAc:hexane (2:1) as eluent.
The pure compound 24 (0.89 g, 57%) was obtained as a white solid. 1H-NMR
(CDCl3; d): 1.42 (3H, d, J ¼ 7.0 Hz), 1.48 (9H, s), 1.50 (9H, s), 1.75 (2H, m), 1.90 (2H,

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
299
m), 3.48 (1H, m), 3.70 (2H, m), 3.78 (2H, m), 4.02 (2H, s), 4.35 (2H, br), 4.90 (1H,
br), 5.08 (1H, m, J ¼ 4.8 Hz), 7.35 (2H, d, J ¼ 8.0 Hz), 7.42 (1H, t, J ¼ 6.5 Hz), 7.80
(2H, J ¼ 8.0 Hz). MS (FAB) m/z: 520 (Mþ þ 1).
TFA ꢀ 2[N-(Aminomethyl-4-benzoyl-Ala)-4-piperidinyloxy] acetic acid (1). TFA
(5 mL) was added to a stirred solution of compound 24 (0.86 g, 1.65 mmol) in
CH2Cl2 (8 mL). The reaction mixture was stirred at room temperature for 1.5 h.
Volatile compounds in the reaction mixture were removed by rotary evaporation
under vacuum. The residue was washed with Et2O/petroleum ether (3:1) to afford a
1
white solid and dried under vacuum to yield 1 (7.87 g, 100%). H-NMR (DMSO-
d6; d): 1.27 (3H, d, J ¼ 6.0 Hz), 1.40 (2H, m, J ¼ 6.8 Hz), 1.81 (2H, m, J ¼ 6.8 Hz),
3.00 (1H, m, J ¼ 6.8 Hz), 3.30 (1H, m, J ¼ 6.8 Hz), 3.60 (1H, m, J ¼ 7.0 Hz), 3.75 (1H,
m, J ¼ 7.0 Hz), 4.11 (2H, m, J ¼ 8.0 Hz), 4.94 (1H, m, J ¼ 6.9 Hz), 7.52 (2H, d,
J ¼ 7.8 Hz), 7.92 (1H, d, J ¼ 7.8), 7.94 (1H, d, J ¼ 6.0 Hz), 8.22 (3H, br), 8.71 (1H, d,
J ¼ 7.5 Hz). MS (FAB) m/z: 364 (Mþ þ 1).
5. Platelet aggregation
Fresh platelet-rich plasma (PRP; 10 mL) was diluted with 10 mL TyrodeÕs buffer
and centrifuged at 1000 rpm for 10 min. The PRP suspension was decanted and used
for the next step. PRP (0.5 mL) in a siliconized cuvette was stirred at 1000 rpm and
pre-warmed to 37 °C for 2 min. Fibrinogen solution (10 lg/mL, Sigma) was added to
the cuvette and 1 min later adenosine diphosphate (ADP) (10 lM) was added. The
change in light transmission was monitored with a Chrono-log Model 490 optical
aggregometer. For the inhibition studies, a specific amount of peptide was added,
prior to addition of ADP [32]. Each inhibitor was tested at five different concen-
trations for the determination of IC50.
The effects of the incubation time of cyclic prodrug 2 with platelet-rich plasma
were investigated as follows. Cyclic prodrug 2 was added to PRP and incubated at
37 °C for various times before aggregation was induced by adding ADP. The extent
of platelet aggregation is determined by a change in light transmission through the
PRP on the aggregometer.
6. Enzymatic stability
The cyclic prodrug (final concentration 24 lM) was incubated for 6 h with 1 mL of
90% human plasma in a temperature-controlled (37:0 Æ 0:5°C) shaking water bath
(60 rpm). At various time points, aliquots (20 lL) were removed and the esterase
activity was immediately quenched by adding 150 lL freshly prepared 6 N guani-
dinium hydrochloride in acidified HBSS (HBSS containing 0.01% (v/v) phosphoric
acid). Aliquots (150 lL) of that acidic mixture at pH 3 were transferred to an Ul-
trafree-MC 5000 NMWL filter unit (Millipore, Bedford, MA) and centrifuged at
7500 rpm (5000g) for 60 min (4 °C). Aliquots (50 lL) of the filtrates were diluted with
mobile phase and injected on the HPLC column.

300
X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
Acknowledgments
Financial support for this research was provided by a grant from NIH (HL-
59931). We thank Dr. Todd Williams, Mr. Robert Drake, and Ms. Homigol Biesada
for technical support at the Mass Spectroscopy Laboratory. We thank Nancy
Harmony for her help in preparing the manuscript.
References
[1] R.J. Shebuski, in: J.A. Bristol (Ed.), Annual Reports in Medicinal Chemistry, vol. 26,
Academic Press, San Diego, CA, 1991, pp. 93–101.
[2] R.M. Scarborough, D.D. Gretler, J. Med. Chem. 43 (2000) 3453–3473.
[3] N.S. Nicholson, N.A. Abood, S.G. Panzer-Knodle, L.G. Frederick, J.D. Page, A.K.
Salyers, O.D. Suleymanov, J.A. Szalony, B.B. Taite, R.J. Anders, Med. Res. Rev. 21
(2001) 211–226.
[4] D.C. Levin, J.T. Fallon, Circulation 66 (1982) 316–320.
[5] R. Lottenberg, C.S. Kitchens, Thrombosis and Platelets in Myocardial Ischemia, F.A.
Davis Co., Philadelphia, PA, 1987.
[6] J.G. White, Thrombosis and Platelets in Myocardial Ischemia, F.A. Davis Co.,
Philadelphia, PA, 1987.
[7] E. Falk, Circulation 71 (1985) 699–708.
[8] R.B. Pytela, M.D. Pierschbacher, M.H. Ginsberg, E.F. Plow, E. Ruoslahti, Science 231
(1986) 1559–1562.
[9] E.F. Plow, M.D. Pierschbacher, E. Ruoslahti, G.A. Marguerie, M.H. Ginsberg, Proc.
Natl. Acad. Sci. USA 82 (1985) 8057–8061.
[10] D.B. Taylor, J. Biol. Chem. 267 (1992) 11729–11733.
[11] J.J. Cook, M. Trybulec, E.C. Lasz, S. Khan, S. Niewiaroski, Biochem. Biophys. Acta
1119 (1992) 312–321.
[12] J.W. Weisel, C. Nagaswami, G. Vilarie, J.S. Bennett, J. Biol. Chem. 267 (1992) 16637–
16643.
[13] J. Samanen, F. Ali, T. Romoff, R. Calvo, E. Sorenson, J. Vasko, B. Storer, D. Berry, D.
Bennett, M. Strohsacker, D. Powers, J. Stadel, A. Nichols, J. Med. Chem. 34 (1991) 3114–
3125.
[14] B.K. Blackburn, T.R. Gadek, in: J.A. Bristol (Ed.), Annual Reports in Medicinal
Chemistry, vol. 28, Academic Press, San Diego, CA, 1993, pp. 79–88.
[15] N.S. Nicholson, S.G. Panzer-Knodle, A.K. Salyers, B.B. Taite, L.W. King, M. Miyano,
R.J. Gorczynski, M.H. Williams, M.E. Zupec, F.S. Tjoeng, S.P. Adams, L.P. Feigen, J.
Pharmacol. Exp. Ther. 256 (1990) 876–882.
[16] P.L. Baker, S. Bullens, S. Bunting, D. Burdick, K.S. Chan, T. Deisher, C. Eigenbrot, T.R.
Gadek, R. Gantzos, M.T. Lipari, C.D. Muir, M.A. Napier, R.M. Pitti, A. Padua, C.
Quan, M. Stanley, M. Struble, J.Y.K. Tom, J.P. Burnier, J. Med. Chem. 35 (1992) 2040–
2048.
[17] M.S. Egbertson, T.-C. Chang, M.E. Duggan, R.J. Gould, W. Halczenko, G.D. Hartman,
W.L. Laswell, J.J. Lynch Jr., P.D. Manno, A.M. Naylor, J.D. Prugh, D.R. Tamjit, G.R.
Sitko, R.S. Smith, L.M. Turchi, G. Zhang, J. Med. Chem. 37 (1994) 2537–2551.
[18] J.H. Beer, K.T. Springer, B.S. Coller, Blood 79 (1992) 117–128.
[19] J.A. Zablocki, M. Miyano, S.N. Rao, S. Panzer-Knodle, N. Nicholson, L. Feigen, J. Med.
Chem. 35 (1992) 4914–4917.

X. Song et al. / Bioorganic Chemistry 30 (2002) 285–301
301
[20] J.F. Callahan, J.W. Bean, J.L. Burgess, D.S. Eggleston, S.M. Hwang, K.D. Kopple, P.F.
Koster, A. Nichols, C.E. Peishoff, J.M. Samanen, J.A. Vasko, A. Wong, W.F. Huffman,
J. Med. Chem. 35 (1992) 3970–3972.
[21] S. Gangwar, G.M. Pauletti, T.J. Siahaan, V.J. Stella, R.T. Borchardt, J. Org. Chem. 62
(1997) 1356–1362.
[22] A. Bak, T.J. Siahaan, O.S. Gudmundsson, S. Gangwar, G.J. Friis, R.T. Borchardt, J.
Pept. Res. 53 (1999) 393–402.
[23] A. Bak, O.S. Gudmundsson, G.J. Friis, T.J. Siahaan, R.T. Borchardt, Pharm. Res. 16
(1999) 24–29.
[24] G.M. Pauletti, S. Gangwar, F.W. Okumu, T.J. Siahaan, V.J. Stella, R.T. Borchardt,
Pharm. Res. 13 (1996) 1613–1621.
[25] O.S. Gudmundsson, S.D.S. Jois, D.G.V. Velde, T.J. Siahaan, B. Wang, R.T. Borchardt,
J. Pept. Res. 53 (1999) 383–392.
[26] G.M. Pauletti, S. Gangwar, B. Wang, R.T. Borchardt, Pharm. Res. 14 (1997) 11–17.
[27] O.S. Gudmundsson, K. Nimkar, S. Gangwar, T.J. Siahaan, R.T. Borchardt, Pharm. Res.
16 (1999) 16–23.
[28] G.M. Pauletti, S. Gangwar, T.J. Siahaan, J. Aube, R.T. Borchardt, Adv. Drug Del. Rev.
27 (1997) 235–256.
[29] S. Gangwar, S.D.S. Jois, T.J. Siahaan, D. Vander Velde, V.J. Stella, R.T. Borchardt,
Pharm. Res. 13 (1996) 1655–1660.
ꢁ
€
€
[30] L. Alig, A. Edenhofer, P. Hadvary, M. Hurzeler, D. Knopp, M. Muller, P. Steiner, A.
Trzeciak, T. Weller, J. Med. Chem. 35 (1992) 4393–4407.
[31] O.S. Gudmundsson, D.G.V. Velde, S.D.S. Jois, A. Bak, T.J. Siahaan, R.T. Borchardt, J.
Pept. Res. 53 (1999) 403–413.
[32] S.D.S. Jois, U.S.F. Tambunan, S. Chakrabarti, T.J. Siahaan, J. Biomol. Struct. Dyn. 14
(1996) 1–12.
[33] U. Gogate, A.J. Repta, J. Alexander, Int. J. Pharmaceutics 40 (1987) 235–248.
[34] B. Wang, W. Wang, G. Camenisch, J. Elmo, H. Zhang, R.T. Borchardt, Chem. Pharm.
Bull. 47 (1999) 90–95.
[35] G.P. Camenisch, W. Wang, B. Wang, R.T. Borchardt, Pharm. Res. 15 (1998) 1174–1181.