N(ω)-nitroarginine-containing dipeptide amides. Potent and highly selective inhibitors of neuronal nitric oxide synthase

Article abstract of DOI:10.1021/jm990111c

Selective inhibition of the isoforms of nitric oxide synthase (NOS) could be therapeutically useful in the treatment of certain disease states arising from the overproduction of nitric oxide (NO). Recently, we reported the dipeptide methyl ester, D-Phe-D-

Full text of DOI:10.1021/jm990111c

J . Med. Chem. 1999, 42, 3147-3153
3147
Nω-Nitr oa r gin in e-Con ta in in g Dip ep tid e Am id es. P oten t a n d High ly Selective
In h ibitor s of Neu r on a l Nitr ic Oxid e Syn th a se
Hui Huang,†,§ Pavel Martasek,‡,| Linda J . Roman,‡, Bettie Sue Siler Masters,‡,|, and Richard B. Silverman*,†
Department of Chemistry and Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University,
Evanston, Illinois 60208-3113, and Department of Biochemistry, The University of Texas Health Science Center,
San Antonio, Texas 78284-7760
Received March 11, 1999
Selective inhibition of the isoforms of nitric oxide synthase (NOS) could be therapeutically
useful in the treatment of certain disease states arising from the overproduction of nitric oxide
(NO). Recently, we reported the dipeptide methyl ester, D-Phe-D-ArgNO -OMe (19), as a modest
2
inhibitor of nNOS (Ki ) 2 µM), but with selectivity over iNOS as high as 1800-fold (Silverman,
R. B.; Huang, H.; Marletta, M. A.; Martasek, P. J . Med. Chem. 1997, 40, 2813-2817). Here a
library of 152 dipeptide amides containing nitroarginine and amino acids other than Phe are
synthesized and screened for activity. Excellent inhibitory potency and selectivity for nNOS
over eNOS and iNOS is achieved with the dipeptide amides containing a basic amine side
chain (20-24), which indicates a possible electrostatic (or hydrogen bonding) interaction at
the enzyme active site. The most potent nNOS inhibitor among these compounds is L-ArgNO
-
2
L-Dbu-NH2 (23) (Ki ) 130 nM), which also exhibits the highest selectivity over eNOS (>1500-
fold) with a 192-fold selectivity over iNOS. These compounds do not exhibit time-dependent
inhibition. The order and the chirality of the amino acids in the dipeptide amides have profound
influences on the inhibitory potency as well as on the isoform selectivity. These dipeptide amide
inhibitors open the door to the design of potent and highly selective inhibitors of nNOS.
In tr od u ction
Nitric oxide (NO) is an important biological messenger
fore, the source of the NOS used in various studies may
not be important.
Overproduction of NO has been implicated in a wide
variety of diseases. NO overproduction by nNOS has
been associated with strokes,5 migraine headaches,6 and
Alzheimer’s disease.7 Enhanced formation of NO fol-
lowing the induction of iNOS appears to be important
in the tolerance to and dependence on morphine,8
development of colitis,9 and tissue damage and inflam-
mation.10 Because the three isoforms of NOS have
unique roles in separate tissues, selective inhibition of
one isoform over the others is essential.11 Numerous
pharmaceutical companies have initiated programs to
identify potent and selective inhibitors of the isoforms
of NOS as potential treatments for some of these disease
states. In particular, nNOS inhibition has been targeted
for the treatment of strokes,12 and iNOS inhibition for
the treatment of septic shock13 and arthritis.14 It is very
important not to inhibit eNOS because of its important
role in maintaining blood flow.
Many of the NOS inhibitors are analogues of the
substrate L-arginine (Table 1) and include Nω-methyl-
L-arginine (1, L-NMA),15 Nω-nitro-L-arginine (2, L-NA),16
Nδ-(iminoethyl)-L-ornithine (3, L-NIO),17 N5-(imino-3-
butenyl)-L-ornithine (4, L-VNIO),18 Nω-alkyl-L-arginines
(5, 6),19 and S-alkyl-L-thiocitrullines (7, 8).20 Most of
these L-arginine analogues are inactivators of NOS, but
they have minimal selectivity among the isozymes
(Table 1), except for Nω-propyl-L-arginine (5)19,21 and N5-
(imino-3-butenyl)-L-ornithine (4),18 which are highly
selective inhibitors for nNOS over iNOS and eNOS.
There also are numerous non amino acid based NOS
inhibitors, for example (Table 2), aminoguanidine (9),22
S-ethylisothiourea (10),23 substituted N-phenylisothio-
involved in a variety of physiological processes.1
A
family of enzymes, the nitric oxide synthases (NOS, EC
1.14.13.39), catalyzes the oxidation of L-arginine to
L-citrulline and nitric oxide in a NADPH- and O2-
dependent process.2 There are three paradigms of NO
biological functions which correlate with three distinct
isoforms of NO synthase.3 The constitutive endothelial
isoform (eNOS) is involved in the regulation of smooth
muscle relaxation and blood pressure and in the inhibi-
tion of platelet aggregation. Another constitutive iso-
form is neuronal NOS (nNOS), which is important for
neurotransmission and long-term potentiation. NO
produced by the inducible isoform (iNOS) in activated
macrophage cells acts as a cytotoxic agent in normal
immune responses. All of the isoforms utilize NADPH,
FAD, FMN, tetrahydrobiopterin, and heme as cofactors.
The constitutive isoforms also require added Ca2+ and
calmodulin for activity, while the inducible isoform has
tightly bound Ca2+ and calmodulin. Cloning of the
isoforms of NOS has revealed that they share only
approximately 50% of primary sequence homology,
suggesting that they may differ from each other in
regulatory aspects.4 However, interspecies similarity for
each isoform is quite high (81-93% identical);4 there-
† Northwestern University.
‡ The University of Texas Health Science Center.
§ Carried out all of the chemical and biological work except for the
overexpression of nNOS and eNOS into E. coli and the purification of
eNOS used in these studies.
| Provided the purified recombinant bovine eNOS used in these
studies.
Provided the overexpressed nNOS used in these studies.
10.1021/jm990111c CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/17/1999

3148 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Huang et al.
Ta ble 1. Isoform Selectivity of Some Amino Acid Based NOS
Inhibitors (L-Arginine Analogues)
nNOS. Since L-nitroarginine (2) is reported to have
about a 250-fold selectivity in favor of nNOS over iNOS,
and L-nitroarginine-containing dipeptides have been
reported to be inhibitors of the release of endothelium-
derived relaxing factor (now known to be nitric oxide),28
we synthesized and evaluated a library of nitroarginine-
containing dipeptides to see if the incorporation of
nitroarginine into a dipeptide could increase the inhibi-
tory potency and selectivity of L-nitroarginine. We
selectivitya
compdref
R
eNOS/nNOS
eNOS/iNOS
1 L-NMA15
2 L-NA16
NHCH3
NHNO2
CH3
CH2CHdCH2
NHCH2CH2CH3
NHCH2CHdCH2
SCH3
0.6
2.5
0.3
120
149
15
0.5
0.01
0.1
0.2
0.05
1.5
recently reported that D-Phe-D-ArgNO methyl ester (19)
2
3 L-NIO17
4 L-VNIO18
5 L-NPA19,21
6 L-AA19
was an 1800-fold more selective inhibitor of nNOS than
of iNOS; unfortunately, its selectivity for nNOS over
eNOS was minimal,29 as is the case with nitroarginine
itself. We now report the first family of molecules that
exhibits high selectivity for inhibition of nNOS over both
iNOS and eNOS.
7 L-SMTC20
8 L-SETC20
10
50
0.4
1.5
SCH2CH3
a
Defined as the ratio of the Ki or IC50 values.
Ch em istr y
Ta ble 2. Isoform Selectivity of Some Non Amino Acid Based
NOS Inhibitors
All of the dipeptide amides were synthesized by
automated solid-phase peptide synthesis. Dipeptide
amides were prepared as a result of the use of Rink
resin as the solid support. All four isomers of each
dipeptide combination (L-L; L-D; D-L; D-D) were
synthesized with Nω-nitroarginine at either the N-
terminus or the C-terminus except for isoleucine and
cysteine, which were used only as the L-isomers. In
addition to the standard 19 natural amino acids (ex-
cluding Arg, which was replaced by ArgNO ), D- and
2
L-ornithine (Orn), L-2,4-diaminobutyric acid (Dbu), and
L-2,3-diaminopropionic acid (Dpr) also were used in the
construction of the dipeptide amides with Nω-nitroargi-
nine. Each of the dipeptide amides was >80% pure after
cleavage from the resin, as determined by NMR analy-
sis, and was used directly for IC50 activity screening.
Ten compounds (20-29), which showed either excellent
potency, high selectivity, or both during the screening
process, were selected for further purification and
characterization. Ki values of these compounds were
determined after they were purified by HPLC. The
analytical data for these compounds are shown in Table
3.
selectivitya
compdref
eNOS/nNOS
eNOS/iNOS
922
1023
1124
1225
1326
1427
1523
1614b
1713a
1814a
10
1.2
29
13
5-10
90
38
25
155
130
500
2
0.3
Because of the limited synthetic scale by automated
solid-phase peptide synthesis and the relatively long
purification time by HPLC, the most promising dipep-
40
5-10
900
180
5000
3
tide amide, L-ArgNO -L-Dbu-NH2 (23), was synthesized
2
on a larger scale by a solution-phase synthesis. In the
initial attempts to make this dipeptide amide from the
corresponding dipeptide methyl ester, the nitro group
on the ArgNO was removed during the aminolysis
314
2
a
reaction in a pressure bottle. This problem was over-
come by first synthesizing L-Dbu amide for use in the
peptide coupling step (Scheme 1). The best route to the
amide was via treatment of the corresponding p-nitro-
phenyl ester of Dbu with dry ammonia. The inhibition
constants for all three isozymes measured with the
compound obtained by solution-phase synthesis were
the same as those obtained with the compound from the
solid-phase synthesis.
The ratio of the IC50 or Ki values.
urea (11),24 2-amino-5,6-dihydro-6-methyl-4H-1,3-thi-
azine (12),25 7-nitroindazole (13),26 (+)-cis-5-pentyl-4-
(trifluoromethyl)pyrrolidin-2-imine (14),27 bis(isothio-
ureas) (15),23 N-(3-(aminomethyl)benzyl)acetamidine
(16),14b N-(3-(aminomethyl)phenyl)acetamidine (17),13a
and 7-(2-ethylbutyl)hexahydro-1H-azepin-2-imine (18).14a
As shown in Table 2, the selectivity has been signifi-
cantly improved by these non amino acid based inhibi-
tors, especially iNOS selective ones. In some cases, the
selectivity is as high as 5000- (16), 900- (14), and 500-
fold (9) in favor of iNOS over eNOS. However, such high
selectivity has not been observed for the inhibition of
nNOS over eNOS.
Resu lts a n d Discu ssion
Initially, the 152 dipeptide amides prepared were
screened by measuring the percentage of remaining
enzyme activity in the presence of 100 µM inhibitors
when compared to an untreated enzyme control. A rapid
onset of inhibition was observed, and the dipeptide
Our interest has been in the selective inhibition of

Nω-Nitroarginine-Containing Dipeptide Amides
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3149
Ta ble 3. Analytical Data for the Selected Nω-Nitroarginine-Containing Dipeptide Amides
compd
1H NMR chemical shift (δ, D2O)
HRMS (calcd)
anal. (C, H, N)
NO2
20 L-ArgNO -L-Arg -NH2 4.32 (t, 1H), 4.05 (t, 1H), 3.28 (t, 4H), 1.56-1.98 (m, 8H)
420.2033 (420.2062) C12H25N11O6‚1.2TFA‚
1.5H2O
2
21 L-ArgNO -L-Lys-NH2
4.30 (t, 1H), 4.05 (t, 1H), 3.29 (t, 2H), 2.98 (t, 2H),
1.87-2.01 (m, 2H), 1.58-1.86 (m, 6H), 1.35-1.50 (m, 2H)
4.34 (t, 1H), 4.08 (t, 1H), 3.29 (t, 2H), 3.00 (t, 2H),
1.89-1.98 (m, 2H), 1.62-1.89 (m, 6H)
4.47 (dd, 1H), 4.09 (t, 1H), 3.30 (t, 2H), 3.08 (m, 2H),
2.01-2.25 (m, 2H), 1.89-2.01 (m, 2H), 1.60-1.75 (m, 2H)
4.12 (t, 1H), 3.49 (dd, 1H), 3.20-3.35 (m, 4H),
1.91-2.03 (m, 2H), 1.62-1.75 (m, 2H)
4.77 (t, 1H), 4.06 (t, 1H), 3.29 (t, 2H), 2.87 (dd, 1H),
2.69 (dd, 1H), 1.85-1.96 (m, 2H), 1.56-1.72 (m, 2H)
4.45 (t, 1H), 4.10 (t, 1H), 3.85 (dd, 2H), 3.30 (t, 2H),
1.89-2.00 (m, 2H), 1.63-1.76 (m, 2H)
4.34 (t, 1H), 4.08 (t, 1H), 3.29 (t, 2H), 3.00 (t, 2H),
1.89-1.98 (m, 2H), 1.62-1.89 (m, 6H)
347.2185 (347.2150) C12H26N8O4‚2TFA‚
2
0.8H2O
22 L-ArgNO -L-Orn-NH2
333.1979 (333.1993) C11H24N8O4‚2TFA‚H2O
2
23 L-ArgNO -L-Dbu-NH2
319.1819 (319.1837) C10H22N8O4‚2TFA‚
1.2H2O
305.1692 (305.1680) C9H20N8O4‚2TFA‚H2O
2
24 L-ArgNO -L-Dpr-NH2
2
25 L-ArgNO -D-Asn-NH2
333.1607 (333.1629) C10H20N8O5‚TFA‚H2O
306.1551 (306.1520) C9H19N7O5‚TFA‚0.6H2O
333.1985 (333.1993) C11H24N8O4‚2TFA‚H2O
347.2186 (347.2150) C12H26N8O4‚2TFA‚H2O
347.2146 (347.2150) C12H26N8O4‚2TFA‚H2O
2
26 D-ArgNO -L-Ser-NH2
2
27 L-ArgNO -D-Orn-NH2
2
28 L-Lys-D-ArgNO -NH2
4.31 (t, 1H), 4.07 (t, 1H), 3.30 (t, 2H), 2.96 (t, 2H),
1.87-2.01 (m, 2H), 1.58-1.86 (m, 6H), 1.35-1.50 (m, 2H)
4.31 (t, 1H), 4.07 (t, 1H), 3.30 (t, 2H), 2.96 (t, 2H),
1.87-2.01 (m, 2H), 1.58-1.86 (m, 6H), 1.35-1.50 (m, 2H)
2
29 D-Lys-D-ArgNO -NH2
2
Sch em e 1a
give L-ArgNO -L-Dbu-NH2 (23) resulted in approximately
3-fold increases in the potency of both nNOS and iNOS
inhibition. As the number of methylenes in the side
chain decreased from four (21) to two (23), the inhibition
of nNOS and iNOS increased, but that to eNOS de-
creased. This shows up in the dramatic increases in the
selectivity of nNOS over eNOS (>1500-fold in the case
of 23), but there was little change in the relative
selectivity of nNOS over iNOS, which was >190-fold in
the case of 23. However, by shortening the side chain
by one more methylene to Dpr (24), there was a sharp
reduction in inhibition of all isoforms (greater than a
2
factor of 8 for nNOS). L-ArgNO has a longer imine side
2
NO2
chain than Lys, and L-ArgNO -L-Arg
(20) produces a
2
a
Reagents: (a) CuCO3, Cu(OH)2, Boc2O, 1 N NaHCO3, dioxane;
decreased nNOS inhibition and increased iNOS and
eNOS inhibition, which results in significant reduction
in selectivity of nNOS over iNOS and eNOS. The
observed correlation between the length of the amine-
containing side chain and the inhibitory potency and
selectivity of nNOS suggests that both steric and
electronic effects play critical roles in the binding of
these dipeptide inhibitors to the enzyme. The basic
nitrogen of these amine or imine side chains could act
as ionic or hydrogen bond donors for potential interac-
tion with the active site of nNOS. There may be an
electrostatic interaction between the protonated basic
nitrogen and nNOS, but not with eNOS. If the group is
too close to (20, 21, 22) or too far away from (24) the
hydrogen bond or ionic donor of the enzyme, it could
result in a weaker interaction and, consequently, poorer
binding.
(b) benzyl chloroformate, 1 N NaHCO3, H2O; (c) p-nitrophenol,
DCC, EtOAc; (d) dry NH3, THF; (e) H2, 10% Pd/C, EtOH; (f) Boc-
L-ArgNO , EDC, HOBt, DIEA, DMF; (g) 50% TFA in CH2Cl2.
2
amides did not exhibit time-dependent inhibition, even
though nitroarginine is a time-dependent inactiva-
tor.16,29 The standard assay conditions for all three
isoforms were used at a fixed L-arginine concentration
(10 µM) as described previously.29 Although the nNOS
and eNOS used were derived from bovine tissue and
the iNOS from murine tissue, the interspecies similarity
for each isoform is quite high (81-93% identical), so the
source of the enzyme should not be much of a factor in
the evaluation of the inhibitors. The results of the screen
for all 152 dipeptides are shown in the Supporting
Information. By visual comparison of these results, the
10 most nNOS-selective inhibitors (20-29) were chosen
for purification and more accurate Ki data measure-
ments, which are listed in Table 4.
All of the potent and selective dipeptide inhibitors
mentioned above have L-ArgNO at the N-terminus,
2
The initial screen with L-ArgNO at the N-terminus
2
except 26 (vide infra). When L-ArgNO is at the N-
2
revealed that the dipeptide amides containing amino
acids with a nitrogen-containing side chain, such as Lys,
terminus, the favored stereochemistry of the C-terminal
amino acid also is L, except for L-ArgNO -D-Asn-NH2 (25;
2
ArgNO , and His, were relatively potent and selective
2
vide infra) and L-ArgNO -D-Orn-NH2 (27), a diastereomer
2
inhibitors of nNOS. In particular, L-ArgNO -L-Lys-NH2
2
of 22, which shows a 6-fold reduction in both nNOS and
eNOS inhibition while having minimal influence on
iNOS inhibition.
(21) showed excellent inhibitory potency of nNOS over
eNOS, which prompted us to investigate the optimal
spatial preference of the free amino group with lysine
analogues. Deletion of one methylene unit from the
In contrast to the above results, when ArgNO is at
2
amino side chain of lysine, giving L-ArgNO -L-Orn-NH2
2
the C-terminus, it is more selective as the D-isomer (L-
2
2
(22), increased the inhibitory potency of 21, as well as
the selectivity. Removal of one more methylene unit to
Lys-D-ArgNO -NH2 (28) and D-Lys-D-ArgNO -NH2 (29)),
except for L-ArgNO -L-Arg -NH2 (20), which is mildly
NO2
2

3150 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
Huang et al.
Ta ble 4. Nitric Oxide Synthase Inhibition by the Selected Nω-Nitroarginine-Containing Dipeptide Amidesa
a
The enzymes used for the Ki determinations are bovine brain nNOS, recombinant murine iNOS, and recombinant bovine eNOS. See
b
Experimental Section. In most cases, the Ki values represent single measurements, but with 6 data points and correlation coefficients
0.980-0.996. When duplicate runs were made, standard deviations of (8-12% were observed. c The ratio of Ki (eNOS or iNOS) to Ki
(nNOS); all are nNOS selective. Side chains of the amino acids other than ArgNO
.
e Data taken from ref 29.
d
2
compounds. However, this is not the case; L-ArgNO -D-
Asn-NH2 (25) is the diastereomer that is the most potent
inhibitor of nNOS with high selectivity over eNOS
(>1200-fold). Since the amide amino group of the
asparagine side chain is not basic and, therefore, not
protonated, the prevalent interaction of the side chain
with the enzyme is a hydrogen bond interaction, not an
electrostatic one. However, it is unlikely that this isomer
binds at the same site as 23 since the stereochemistry
of the C-terminal amino acid is epimeric with 23.
2
selective. This may be the result of the binding of the
nitroarginine residue to the same binding site, regard-
less of its position in the dipeptide. For example, when
D-nitroarginine is at the C-terminus, it may flip over
180° to assume an L-nitroarginine-like configuration at
the N-terminus for binding (Figure 1).30 If that is the
case, then D-ArgNO should be favored at the C-terminus.
2
NO2
Both L-Lys-D-ArgNO -NH2 (28) and D-Lys-D-Arg -NH2
2
(29) have excellent nNOS/iNOS selectivity because of
their dramatic decreases in iNOS inhibition. Note that
D-Phe-D-ArgNO -OMe (19), which is an extremely poor
D-ArgNO -L-Ser-NH2 (26) also is a selective nNOS
2
2
inhibitor of iNOS but a much better inhibitor of nNOS,
inhibitor, but, as in the case of 25, the stereochemistry
is epimeric compared with that for 23. The side chain
of serine is the same as that of Dpr, except for a
hydroxyl group instead of the amino group of Dpr. The
hydroxyl group of serine is a good hydrogen bond donor,
but is not charged like the amino group of Dpr would
be. Nonetheless, 24 and 26 have similar Ki values for
both nNOS and eNOS and, therefore, similar nNOS/
eNOS selectivities. However, binding of 26 to iNOS
decreases significantly compared with 24, which results
in the sharp increase in selectivity of nNOS over iNOS
for 26 compared with 24. Possibly, the different stereo-
chemistry at the N-terminus prevents the appropriate
interaction with iNOS.
also has D-ArgNO at the C-terminus. Since both L-Lys-
2
D-ArgNO -NH2 (28) and D-Lys-D-ArgNO -NH2 (29) have
similar selectivities, the stereochemistry of the amine-
containing residue at the N-terminus does not appear
to be highly critical. These observations indicate that
the geometry of these dipeptide inhibitors at the active
site have profound effects on their binding to the
different isoforms of NOS; in particular, the position of
2
2
ArgNO in the sequence has a larger effect on the binding
2
to iNOS than to nNOS or eNOS.
Only two nonamine side chain containing dipeptide
amides were found to be highly selective, namely,
NO2
L-ArgNO -D-Asn-NH2 (25) and D-Arg -L-Ser-NH2 (26).
2
Asparagine is similar in structure to 2,4-diaminobutyric
acid (Dbu), except for a carbonyl instead of a methylene
adjacent to the terminal amino group. If the similarity
between asparagine and Dbu is relevant, then the L-L
Some of the dipeptide amides with an alkyl side chain
show eNOS inhibition preference. In these cases, D-D
isomers with D-ArgNO at the C-terminus, such as Gly-
2
NO2
NO2
D-ArgNO -NH2, D-Val-D-Arg -NH2, and D-Leu-D-Arg
-
2
isomer should be the most selective, since L-ArgNO -L-
NH2, are favored. The acidic amino acid (Asp or Glu)
containing dipeptide amides exhibited little or no inhi-
2
Dbu-NH2 (23) is the most selective of that series of

Nω-Nitroarginine-Containing Dipeptide Amides
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3151
electronic factors greatly influence inhibitory potency
and isoform selectivity of these dipeptide inhibitors.
Among the compounds tested, maximal inhibition po-
tency (130 nM) and selectivity (1538-fold) for nNOS over
eNOS, with high selectivity (192-fold) for nNOS over
iNOS, was achieved with L-ArgNO -L-Dbu-NH2 (23);
2
maximal selectivity (2765-fold) for nNOS over iNOS
with high selectivity (135-fold) for nNOS over eNOS was
realized with L-Lys-D-ArgNO -NH2. This investigation
2
opens the door to the design of potent and highly
selective inhibitors of nNOS. Further peptidomimetic
modifications are currently under investigation for
potential therapeutically useful compounds to treat
neuronal disorders caused by excess NO production.
Exp er im en ta l Section
Ch em istr y. All Fmoc protected amino acids and coupling
reagents were purchased from Advanced ChemTech, Inc. The
dipeptide amides were synthesized on an automated peptide
synthesizer (ACT model 357, Advanced ChemTech, Inc.) using
established protocols31 and purified on a Whatman Partisil C18
semi-prep HPLC column (9.4 × 125 mm). The dipeptides were
eluted from the column using a gradient of 100% solvent A
(0.1% TFA in H2O) to 100% of solvent B (0.08% TFA in 60%
CH3CN/H2O) over 30 min at a flow rate of 2 mL/min. 1H NMR
spectra were recorded on a Varian Gemini-300 spectrometer
in the solvent indicated. Chemical shifts are reported in parts
per million (δ) relative to tetramethylsilane (TMS, in CDCl3)
or sodium 3-(trimethylsilyl)-1-propanesulfonate (DSS, in D2O).
Electrospray mass spectra were performed on a Micromass
Quattro II spectrometer. Elemental analyses were obtained
from Oneida Research Services, Inc., Whiteboro, NY. Thin-
layer chromatography was carried out on E. Merck precoated
silica gel 60 F254 (0.25 mm thickness) plates. Amino acids were
visualized with a ninhydrin spray reagent or a UV/vis lamp.
ICN silica gel 60 (230-400 mesh) was used for flash chroma-
tography.
F igu r e 1. Proposed model to rationalize the binding of
epimers to the active site of nNOS.
bition of any of the NOS isoforms, which supports the
involvement of an anionic group at the active site with
which 20-24 can interact. The histidine-containing
dipeptide amides are not as potent as the other amine-
containing dipeptide amides, but a trend similar to that
observed with the amine-containing dipeptide amides
4-[N-(ter t-Bu toxyca r bon yl)a m in o]-L-2,4-d ia m in obu tyr -
ic Acid (L-Dbu (Boc)-OH). This compound was prepared in
a 40% yield as a white solid from L-2,4-diaminobutyric acid
(2.5 g, 21.2 mmol) according to the procedure of Scott et al.:32
1H NMR (D2O) δ 3.69 (dd, 1H), 3.20 (m, 2H), 2.01 (m, 2H),
1.43 (s, 9H).
is seen, namely, that L-ArgNO -L-His-NH2 and L-His-D-
2-[N-(Ben zoxyca r b on yl)a m in o]-4-[N-(ter t-b u t oxyca r -
b on yl)a m in o]-L-2,4-d ia m in ob u t yr ic Acid (Cb z-L-Db u -
(Boc)-OH). L-Dbu(BOC)-OH (0.5 g, 2.3 mmol) was dissolved
in 10 mL of 1 M Na2CO3 at 0 °C. To the stirred solution was
added dropwise 0.55 mL of benzyl chloroformate (4.6 mmol).
The solution was stirred at 0 °C for 30 min and at room
temperature for 1 h. TLC (n-butanol:acetic acid:water, 5:1:1)
showed that the reaction was complete. Citric acid (20%) was
used to acidify the mixture, and then the white precipitate
was extracted twice with EtOAc. The organic layer was washed
with 10% Na2CO3 solution, which was acidified with 20% citric
acid again, and extracted with EtOAc twice. The EtOAc layer
was washed with water and saturated NaCl and dried over
MgSO4. After evaporation of the solvent, the product was
obtained in a 90% yield as a white solid: 1H NMR (CDCl3) δ
7.35 (m, 5H), 5.58 (d, 1H), 5.12 (s, 2H), 4.43 (m, 1H), 3.20 (m,
2H), 2.05 (m, 1H), 1.83 (m, 1H), 1.45 (s, 9H).
2-[N-(Ben zoxyca r b on yl)a m in o]-4-[N-(ter t-b u t oxyca r -
bon yl)a m in o]-L-2,4-d ia m in obu tyr ic p-Nitr op h en yl Ester
(Cbz-L-Dbu (Boc)-ONp ). This compound was prepared from
Cbz-L-Dbu (Boc)-OH (1.6 g, 4.5 mmol) according to the
procedure of Tesser et al.33 and was used in the next step
without purification.
2-[N-(Ben zoxyca r b on yl)a m in o]-4-[N-(ter t-b u t oxyca r -
bon yl)a m in o]-L-2,4-d ia m in obu tyr ic Am id e (Cbz-L-Dbu -
(Boc)-NH2). Cbz-L-Dbu (Boc)-ONp was dissolved in THF (20
mL, freshly distilled), and dry NH3 was passed near the surface
of the stirred solution in an ice bath. The precipitate was
collected after 1 h and was recrystallized from ethanol-
2
ArgNO -NH2 are the most potent and selective of the
2
eight isomers. Similar to D-Phe-D-ArgNO methyl ester
2
(19), D-Tyr-D-ArgNO -NH2 and D-Trp-D-ArgNO -NH2
showed modest inhibitory potency toward both nNOS
and eNOS.
2
2
Con clu sion
A series of dipeptide amides containing ArgNO were
2
synthesized and screened for biological activity in vitro
against the three isoforms of NOS. Several conclusions
can be drawn from these studies: (1) the active sites of
NOS are large enough to accommodate inhibitors as
large as dipeptides; (2) there are very subtle differences
in the active sites of the three isoforms of NOS, as
evidenced by the observation that relatively minor
structural modifications can produce highly divergent
inhibition results; (3) the dipeptides containing an
amino group containing side chain show high inhibitory
potency with nNOS and excellent selectivity over both
eNOS and iNOS, which supports a possible electrostatic
interaction or hydrogen bonding interaction between the
dipeptide side chain and nNOS, but not eNOS and
iNOS; (4) there may be multiple hydrogen bond donors
at the active site of nNOS; and (5) both steric and

3152 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16
ether, giving the product as a white solid in a 40% yield:34 [R]D
Huang et al.
(6) Thomsen, L. L.; Iversen, H. K.; Lassen, L. H.; Olesen, J . The
role of nitric oxide in migraine pain. CNS Drugs 1994, 2, 417-
22.
(7) Dorheim, M. A.; Tracey, W. R.; Pollock, J . S.; Grammas, P. Nitric
oxide synthase activity is elevated in brain microvessels in
Alzheimer’s disease. Biochem. Biophys. Res. Commun. 1994, 205,
659-65.
1
) -13.8 (c ) 1.45, methanol, 24 °C); H NMR (DMSO-d6) δ
7.35 (m, 7H), 7.02 (s, 1H), 6.77 (brs, 1H), 5.02 (s, 2H), 3.92 (m,
1H), 2.95 (m, 2H), 1.52-1.83 (m, 2H), 1.36 (s, 9H).
L-Nω-Nit r oa r gin in e-2,4-L-d ia m in ob u t yr ic Am id e (L-
Ar gNO -L-Dbu -NH2‚2TF A, 23). The above amide (400 mg) was
2
dissolved in methanol and hydrogenated in the usual manner.
The catalyst was filtered through a Celite pellet, and the
methanol was evaporated in vacuo. The resulting oil, L-Dbu
(Boc)-NH2 (244 mg, 1.12 mmol) was treated with EDC (236
(8) Bhargava, H. N. Attenuation of tolerance to, and physical
dependence on, morphine in the rat by inhibition of nitric oxide
synthase. Gen. Pharmacol. 1995, 26, 1049-53.
(9) Seo, H. G.; Takata, I.; Nakamura, M.; Tatsumi, H.; Suzuki, K.;
Fujii, J .; Taniguchi, N. Induction of nitric oxide synthase and
comcommitant suppression of superoxide dismutases in experi-
mental colitis in rats. Arch. Biochem. Biophys. 1995, 324, 41-
7.
mg, 1.23 mmol), HOBt (166 mg, 1.23 mmol), Boc-L-ArgNO -OH
2
(394 mg, 1.23 mmol), and DIEA (215 µL, 1.23 mmol) in 20 mL
of cold DMF. After 1 h of stirring in an ice bath, the reaction
mixture was warmed to room temperature. DMF was evapo-
rated under vacuum for another hour, and the residue was
treated with 20 mL of EtOAc. The precipitate was dissolved
in water, and the aqueous solution was extracted three times
with EtOAc. The combined extracts were washed with 5%
NaHCO3, H2O, and brine and dried over MgSO4. The product
was purified by silica gel chromatography (12:1 CHCl3:CH3-
OH, Rf 0.25). The dipeptide was treated with 5 mL of TFA/
CH2Cl2 (1:1 v/v) to remove the Boc group. About 100 mg of
pure dipeptide TFA salt was obtained: 1H NMR (D2O) δ 4.47
(dd, 1H), 4.09 (t, 1H), 3.30 (t, 2H), 3.08 (m, 2H), 2.01-2.25
(m, 2H), 1.89-2.01 (m, 2H), 1.60-1.75 (m, 2H). HRMS (ES)
Calcd: 319.1837 (MH+). Found: 319.1819. Anal. (C10H22N8O4‚
2TFA‚1.2H2O) Calcd: C, 29.58; H, 4.64; N, 19.71. Found: C,
29.58; H, 4.39; N, 19.61.
(10) Kubes, P.; Suzuki, M.; Granger, D. N. Nitric oxide: an endog-
enous modulator of leukocyte adhesion. Proc. Natl. Acad. Sci.
U.S.A. 1991, 88, 4651-5.
(11) Marletta, M. A. Approaches toward selective inhibition of nitric
oxide synthase. J . Med. Chem. 1994, 37, 1899-1907.
(12) (a) Collins, J . L.; Shearer, B. G.; Oplinger, J . A.; Lee, S.; Garvey,
E. P.; Salter, M.; Duffy, C.; Burnette, T. C.; Furfine, E. S.
N-Phenylamidines as selective inhibitors of human neuronal
nitric oxide synthase: structure-activity studies and demon-
stration of in vivo activity. J . Med. Chem. 1998, 41, 2858-2871.
(b) Cowart, M.; Kowaluk, E. A.; Daanen, J . F.; Kohlhaas, K. L.;
Alexander, K. M.; Wagenaar, F. L.; Kerwin, J . F., J r. Nitroaro-
matic amino acids as inhibitors of neuronal nitric oxide synthase.
J . Med. Chem. 1998, 41, 2636-2642.
(13) Wright, C. W.; Rees, D. D.; Moncada, S. Protective and patho-
logical roles of nitric oxide in endotoxin shock. Cardiovasc. Res.
1992, 26, 48-57.
(14) Hansen, D. W., J r.; Peterson, K. B.; Trivedi, M.; Kramer, S. W.;
Webber, R. K.; Tjoeng, F. S.; Moore, W. M.; J erome, G. M.;
Kornmeier, C. M.; Manning, P. T.; Connor, J . R.; Misko, T. P.;
Currie, M. G.; Pitzele, B. S. 2-Iminohomopiperidineium salts as
selective inhibitors of inducible nitric oxide synthase (iNOS). J .
Med. Chem. 1998, 41, 1361-1366. (b) Garvey, E. P.; Oplinger,
J . A.; Furfine, E. S.; Kiff, R. J .; Laszlo, F.; Whittle, B. J . R.;
Knowles, R. G. 1400W is a slow, tight binding, and highly
selective inhibitor of inducible nitric oxide synthase in vitro and
in vivo. J . Biol. Chem. 1997, 272, 4959-4963. (c) Nakane, M.;
Klinghofer, V.; Kuk, J . E.; Donnelly, J . L.; Budzik, G. P.; Pollock,
J . S.; Basha, F.; Carter, G. W. Novel potent and selective
inhibitors of inducible nitric oxide synthase. Mol. Pharmacol.
1995, 47, 831-834.
Biology. All of the NOS isoforms used in the initial screens
are recombinant enzymes overexpressed in Escherichia coli
from different sources. The murine macrophage iNOS was
expressed and isolated according to the procedures of Hevel
et al.35 The rat neuronal nNOS was expressed36 and purified37
as described. The bovine endothelial eNOS was isolated and
purified as reported.38 However, the Ki values of the most
selective inhibitors were determined with bovine brain nNOS39
instead of the recombinant nNOS. Nitric oxide formation was
monitored by the hemoglobin capture assay as described
previously.40 The Ki values were measured from Dixon plots41
with various L-arginine and inhibitor concentrations.
(15) Olken, N. M.; Marletta, M. A. NG-Methyl-L-arginine functions
as an alternate substrate and mechanism-based inhibitor of
nitric oxide synthase. Biochemistry 1993, 32, 9677-85.
(16) Furfine, E. S.; Harmon, M. F.; Paith, J . E.; Garvey, E. P.
Selective inhibition of constitutive nitric oxide synthase by NG-
nitroarginine. Biochemistry 1993, 32, 8512-17.
Su p p or tin g In for m a tion Ava ila ble: Bar graphs of the
effect of each of the 152 dipeptide amides on nNOS, eNOS,
and iNOS activity, presented as the percent inhibition com-
pared to control (0% inhibition). This information is available
free of charge via the Internet at http://pubs.acs.org.
(17) McCall, T. B.; Feelisch, M.; Palmer, R. M. J .; Moncada, S.
Identification of Nδ-iminoethyl-L-ornithine as an irreversible
inhibitor of nitric oxide synthase in phagocytic cells. Br. J .
Pharmacol. 1991, 102, 234-8.
Ack n ow led gm en t. We thank Professor Michael A.
Marletta (University of Michigan) for the recombinant
E. coli cells which express murine macrophage iNOS,
and Prof. D. Martin Watterson and Dr. Thomas Lukas
(Northwestern University Medical School) for kindly
providing the automated peptide synthesizer and for
their generous technical support during the solid-phase
synthesis. We also thank Dr. J ose´ Antonio Go´mez-Vidal
for helpful discussions. We are grateful to the National
Institutes of Health for financial support of this work
to R.B.S. (GM 49725), M.A.M. (CA 50414), and B.S.S.M.-
(GM 52419) and to the Robert A. Welch Foundation (AQ
1192) for financial support to B.S.S.M.
(18) Babu, B. R.; Griffith, O. W. N5-(1-Imino-3-butenyl)-L-ornithine.
A neuronal isoform selective mechanism-based inactivator of
nitric oxide synthase. J . Biol. Chem. 1998, 273, 8882-8889.
(19) Zhang, H. Q.; Fast, W.; Marletta, M. A.; Martasek, P.; Silverman,
R. B. Potent and selective inhibition of neuronal nitric oxide
synthase by NG-propyl-L-arginine. J . Med. Chem. 1997, 40,
3869-70.
(20) Narayanan, K.; Spack, L.; McMillan, K.; Kilbourn, R. G.;
Hayward, M. A.; Masters, B. S. S.; Griffith, O. W. S-Alkyl-L-
thiocitrullines. J . Biol. Chem. 1995, 270, 11103-10.
(21) These kinetic experiments have been repeated, and the Ki values
are a little different than those published: nNOS ) 0.11 µM;
iNOS ) 80 µM; eNOS ) 10 µM; selectivities are nNOS/iNOS )
727 and nNOS/eNOS ) 91.
(22) Wolff, D. J .; Lubeskie, A. Aminoguanidine is an isoform-selective,
mechanism-based inactivater of nitric oxide synthase. Arch.
Biochem. Biophys. 1995, 316, 290-301.
(23) Garvey, E. P.; Oplinger, J .; Tanoury, G. J .; Sherman, P. A.;
Fowler, M.; Marshall, S.; Harmon, M. F.; Paith, J . E.; Furfine,
E. S. Potent and selective inhibition of human nitric oxide
synthase. J . Biol. Chem. 1994, 269, 26669-26676.
(24) Shearer, B. G.; Lee, S.; Oplinger, J . A.; Frick, L. W.; Garvey, E.
P.; Furfine, E. S. Substituted N-phenylisothioureas: potent
inhibitors of human nitric oxide synthase with neuronal isoform
selectivity. J . Med. Chem. 1997, 40, 1901-05.
Refer en ces
(1) Kerwin, J . F., J r.; Lancaster, J . R., J r.; Feldman, P. L. Nitric
oxide: a new paradigm for second messengers. J . Med. Chem.
1995, 38, 4342-62.
(2) Kerwin, J . F., J r.; Heller, M. The arginine-nitric oxide path-
way: A target for new drugs. Med. Res. Rev. 1994, 14, 23-74.
(3) Stuehr, D. J .; Griffith, O. W. Mammalian nitric oxide synthases.
Adv. Enzymol. Relat. Areas Mol. Biol. 1992, 65, 287-346.
(4) Knowles, R. G.; Moncada, S. Nitric oxide synthases in Mammals.
Biochem. J . 1994, 198, 249-258.
(5) (a) Choi, D. W.; Rothman, S. M. The role of glutamate neuro-
toxicity in hypoxic-ischemic neuronal death. Annu. Rev. Neuro-
sci. 1990, 13, 171-182. (b) Garthwaite, J . In The NMDA
Receptor; Watkins, J . C., Collingridge, G. L., Eds.; Oxford
University Press: Oxford, England, 1989; pp 187-205.
(25) (a) Moore, W. M.; Webber, R. K.; Fok, K. F.; J erome, G. M.;
Conner, J . R.; Manning, P. T.; Wyatt, P. S.; Misko, T. P.; Tjoeng,
F. S.; Currie, M. G. 2-Iminopiperidine and other 2-imonoaza-
heterocycles as potent inhibitors of human nitric oxide synthase
isoforms. J . Med. Chem. 1996, 39, 669-672. (b) Moore, W. M.;

Nω-Nitroarginine-Containing Dipeptide Amides
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 16 3153
Webber, R. K.; Fok, K. F.; J erome, G. M.; Kornmeier, C. M.;
Tjoeng, F. S.; Currie, M. G. Inhibitors of human nitric oxide
synthase isoforms with the carbamidine moiety as a common
structural element. Bioorg. Med. Chem. 1996, 4, 1559-64.
(26) Wolff, D. J .; Gribin, B. J . The inhibition of the constitutive and
inducible nitric oxide synthase isoforms by indazole agents. Arch.
Biochem. Biophys. 1994, 311, 300-306.
(27) Hagen, T. J .; Bergmanis, A. A.; Kramer, S. W.; Fok, K. F.;
Schmelzer, A. E.; Pitzele, B. S.; Swenton, L.; J erome, G. M.;
Kornmeier, C. M.; Moore, W. M.; Branson, L. F.; Conner, J . R.;
Manning, P. T.; Currie, M. G.; Hallinan, E. A. 2-Iminopyrro-
lidines as potent and selective inhibitors of human inducible
nitric oxide synthase. J . Med. Chem. 1998, 41, 3675-83.
(28) Thiemermann, C.; Mustafa, M.; Mester, P. A.; Mitchell, J . A.;
Hecker, M.; Vane, J . R. Inhibition of the release of endothelium-
derived relaxing factor in vitro and in vivo by dipeptides
containing NG-nitro-L-arginine. Br. J . Pharmacol. 1991, 104, 31-
38.
(29) Silverman, R. B.; Huang, H.; Marletta, M. A.; Martasek, P.
Selective inhibition of neuronal nitric oxide synthase by Nω-
nitroarginine- and phenylalanine-containing dipeptides and
dipeptide esters. J . Med. Chem. 1997, 40, 2813-17.
(30) Recently, we synthesized the N-acetylated N-terminal amine and
found that it was a much poorer inhibitor of all of the isozymes
(at least 200-fold; Huang, H., and Silverman, R. B., unpublished
results). If steric hindrance is not a problem, then the electro-
static interaction shown in Figure 1 may be important.
(31) Lukas, T. J .; Mirzoeva, S.; Slomczynska, U.; Watterson, D. M.
Identification of novel classes of protein kinase inhibitors using
combinatorial peptide chemistry based on functional genomics
knowledge. J . Med. Chem. 1999, in press.
(33) Tesser, G. I.; Schwyzer, R. Synthesis of 17, 18-diornithine-δ-
corticotropin-(1-24)-tetracosapeptide, a biologically active ana-
logue of adrenocorticotropic hormone. Helv. Chim. Acta 1966,
49, 1013-22.
(34) Bodanszky, M.; Klausner, Y. S.; Mutt, V. Synthesis of the
vasoactive intestinal peptide (VIP) I. the C-terminal cyanogen
bromide fragment. Bioorg. Chem. 1972, 2, 30-38.
(35) Hevel, J . M.; White, K. A.; Marletta, M. A. Purification of the
inducible murine macrophage nitric oxide synthase. J . Biol.
Chem. 1991, 266, 22789-91.
(36) Roman, L. J .; Sheta, E. A.; Martasek, P.; Gross, S. S.; Liu, Q.;
Masters, B. S. S. High-level expression of functional rat neuronal
nitric oxide synthase in Escherichia coli. Proc. Natl. Acad. Sci.
U.S.A. 1995, 92, 8428-32.
(37) Gerber, N. C.; Montellano, P. R. Neuronal nitric oxide syn-
thase: expression in Escherichia coli, irreversible inhibition by
phenyldiazene, and active site topology. J . Biol. Chem. 1995,
270, 17791-96.
(38) Martasek, P.; Liu, Q.; Roman, L. J .; Gross, S. S.; Sessa, W. C.;
Masters, B. S. S. Characterization of bovine endothelial nitric
oxide synthase expressed in Escherichia coli. Biochem. Biophys.
Res. Commun. 1996, 219, 359-65.
(39) Furfine, E. S.; Harmon, M. F.; Paith, J . E.; Garvey, E. P.
Selective Inhibition of Constitutive Nitric Oxide Synthase by
L-NG-Nitroarginine. Biochemistry 1993, 32, 8512-17.
(40) Hevel, J . M.; Marletta, M. A. Nitric oxide synthase assays
Methods Enzymol. 1994, 133, 250-8.
(41) Dixon, M. The determination of enzyme inhibitor constants.
(32) Scott, J . W.; Parker, D.; Parrish, D. R. Improved syntheses of
Nꢀ-tert-butyloxycarbonyl-L-lysine and NR-benzyloxycarbonyl-Nꢀ-
tert-butyloxycarbonyl-L-lysine. Synth. Comm. 1981, 11, 303-
314.
Biochem. J . 1953, 55, 170-1.
J M990111C