Isoform-selective substrates of nitric oxide synthase

Article abstract of DOI:10.1021/jm0340703

Because of the double-edged nature of NO, the development of isoform-selective NOS substrates is a highly desirable goal. Given the striking similarity in the heme active sites of the three NOS isoforms, it presents an challenging problem. Several N-aryl-N′-hydroxyguanidines have recently been shown as substrates that are selective for iNOS over nNOS. Here, we report the first success that 3 is a good substrate for nNOS (70% activity of NOHA, K_m ≈ 40 ± 6 μM) over iNOS.

Full text of DOI:10.1021/jm0340703

J . Med. Chem. 2003, 46, 2271-2274
2271
Letters
Sch em e 1. NO Formation and Differences in Three
Isoforms
Isofor m -Selective Su bstr a tes of Nitr ic
Oxid e Syn th a se
Qiang J ia,† Tingwei Cai,† Mingchuan Huang,†
Huiying Li,‡ Ming Xian,† Thomas L. Poulos,‡ and
Peng G. Wang*,†
Department of Chemistry, Wayne State University,
Detroit, Michigan 48202, and Department of Molecular
Biology & Biochemistry and the Program in
Macromolecular Structure, University of California,
Irvine, California 92697
Received March 26, 2003
Abstr a ct: Because of the double-edged nature of NO, the
development of isoform-selective NOS substrates is a highly
desirable goal. Given the striking similarity in the heme active
sites of the three NOS isoforms, it presents an challenging
problem. Several N-aryl-N′-hydroxyguanidines have recently
been shown as substrates that are selective for iNOS over
nNOS. Here, we report the first success that 3 is a good
substrate for nNOS (70% activity of NOHA, Km ≈ 40 ( 6 µM)
over iNOS.
Ch a r t 1
In tr od u ction . NO is involved in a number of physi-
ological and pathophysiological processes1 such as smooth
muscle relaxation,2 inhibition of platelet aggregation,
macrophage cytotoxicity, inflammatory diseases,3,4 iron
metabolism, neurotransmission, and neurotoxicities.5,6
The endogenous synthesis of NO is catalyzed by three
isoforms of NO synthase (NOS) (iNOS, nNOS, and
eNOS), using L-arginine as substrates via a two-step
reaction (Scheme 1).7 In the first step, an intermediate,
N-hydroxyarginine (NOHA), is formed by the incorpora-
tion of an oxygen atom into the guanidine function of
L-arginine. In the second step, NOHA is further oxidized
to form L-citrulline and NO.
Three different NOS isoforms are expressed in dif-
ferent tissues and are highly regulated transcriptionally
or posttranscriptionally (Scheme 1).1,8 Because of the
double-edged nature of NO in both basic physiological
functions and various pathological conditions,9 the de-
velopment of isoform-selective NOS inhibitors and sub-
strates is a highly desirable goal. Many compounds have
been found to be potent, selective NOS inhibitors.10,11
However, only several R-amino acids closely related to
L-Arg or NOHA, such as homo-L-Arg, homo-NOHA, and
E-dehydro-L-Arg are shown to be NOS substrates.12 The
very limited number of substrates for NOSs suggests
that the R-amino acid portion and highly specific
structural features are required for substrates of NOS.13
Very recently, some N-alkyl/aryl substituted hydroxy-
guanidine compounds were found to be novel NOS
substrates that act as substrates for the second step of
the NOS catalytic cycle and that “short-circuit” the first
oxidation step.12-14 Among them, N-butyl-N′-hydroxy-
guanidine 1 and N-isopropyl-N′-hydroxyguanidine 2
(Chart 1) are the most active substrates for all three
NOS isoforms. Crystal structures show that 1 has the
same binding mode as that of NOHA while 2 showed a
novel binding mode that shed light on substrate binding
and the catalytic mechanism.15
In an effort to better understand the structure-act-
ivity relationship and to find new isoform-selective NOS
substrates, we have synthesized a series of compounds
based on known structure-activity relationships. Com-
pound 3 is so far the first selective substrate for nNOS
over iNOS. A new mechanism is proposed that could
also aid in further understanding NOS-selective sub-
strates.
Ch em istr y. A series of N-substituted N′-hydroxyguan-
idines developed from NOHA, 1, and 2 have been syn-
thesized by using the general procedures previously re-
ported for the preparation of hydroxyguanidines starting
from the corresponding amines (Scheme 2).13,16 N-sub-
stituted guanidines were synthesized according to ref
17 (Scheme 2). Compound 12 was synthesized according
* To whom correspondence should be addressed. Phone: 313-993-
6759. Fax: 313-577-2941. E-mail: pwang@chem.wayne.edu.
† Wayne State University.
‡ University of California, Irvine.
10.1021/jm0340703 CCC: $25.00 © 2003 American Chemical Society
Published on Web 05/03/2003

2272 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 12
Letters
Sch em e 2
Ta ble 1. NO Formation from Compounds 1-16 by NOSa
NO formation, %
compd
iNOS
100
nNOS
eNOS
NOHA
100
100
F igu r e 1. Two different orientations of 1 (A) and 2 (B) at the
NOS substrate-binding site. The hydrophobic patch is marked
by an arc.
L-Arg
112
212
3
4
5
6
7
8
9
10
11
12
13
14
15
16
50
42
15
72
38
20
38
64b
70c
<0.5
28 ( 8
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
6% to Arg18
70 ( 8d
37 ( 4
5e
26 ( 4
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
owing to steric crowding by surrounding protein side
chains, while smaller alkyl groups cannot be held tightly
by the enzyme. A double or triple bond (4 and 5) that
introduced rigidity of the butyl group also reduced
activity (nNOS: 37% and 5% to NOHA, respectively).
The terminal branched alkyl group, such as 6, tert-butyl,
isobutyl, and 3-methylbutyl, reduced the activity dra-
matically (nNOS: <0.5%, <0.5%, 2%, and 6% to NOHA,
respectively).13 The structure-activity relationship for
iNOS and eNOS is similar to that of nNOS.12,13
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
Substitution on the hydroxyl group of 1 (giving 7 and
8) abolished its NO formation activity for all three NOS
isoforms. Silverman and co-workers also reported two
NOHA derivatives with the hydrogen of the hydroxyl
group substituted by allyl or tert-butyl groups. The Km
values of the two NOHA derivatives for iNOS were 100
times higher than that of NOHA.19 Of the plain guani-
dine compounds, only N-(3,3,3-trifluoropropyl)guanidine
was reported to have 35% NO formation activity com-
pared to Arg for iNOS.18 All of the others tested were
very week substrates or nonsubstrates. It is obvious that
the hydroxyl group is important for guanidine-derived
compounds to be NOS substrates.
Unsymmetric N-disubstituted 9-11 were not sub-
strates. Symmetric N,N′-disubstituted 12 was also not
a substrate. N-Methyl-substituted NOHA is even an
NOS inhibitor.20 Therefore, we conclude that the N-
monosubstituted hydroxyguanidine moiety is the key
structure for NOS substrates.
A very exciting discovery is that 3 exhibits 70%
NOHA activity for nNOS but no activity for iNOS. This
is the first highly active substrate selective for nNOS
over iNOS. According to the crystal structure of nNOS
complexed with 2,15 the binding mode of 2 (Figure 1B)
is totally different from that of NOHA and 1 (Figure
1A). The hydroxyguanidine moiety rotates 120°, and the
isopropyl-attached N atom takes the place of the OH-
substituted terminal N atom. The isopropyl group fits
in a small hydrophobic pocket formed by Pro565,
Val567, and Phe584. Given the structural similarity
between 2 and 3, 3 is very likely bound to the active
site in mode B as well. The molecular modeling results
also support mode B binding (Table 2).
a
The initial rate of NO synthesis was determined at 37 °C using
spectrophotometric oxyhemoglobin assay for NO. Briefly, 0.5-1
unit of enzyme, 5 µM BH4, and 2-5 mM DTT were added to a
prewarmed cuvette that contained 50 mM HEPES (pH 7.4),
supplemented with 15 µM oxyhemoglobin, 100 units/mL SOD, 100
units/mL catalase, 200 µM NADPH, 4 µM FAD, 4 µM FMN, 5 µM
BH4, and 0.5 mM substrate at the desired concentration, to give
a final volume of 0.9 mL. In the case of nNOS and eNOS, 1 mM
CaCl2 and 10 µg/mL CaM were present. For iNOS, 1 mM
magnesium acetate was added. The reference cuvette had the same
composition except that 50 mM HEPES, 5 µM BH4, and 2-5 mM
DTT were added instead of NOS-containing solutions. The NO-
mediated conversion of oxyhemoglobin to methemoglobin was
monitored over time as an increase in absorbance at 401 nm and
quantitated using an extinction coefficient of 38 mM-1 cm-1. The
rates were expressed as a percentage of those found for NOHA
b
(mean value ( SD from at least three experiments). Km ≈ 67
µM. c Km ≈ 56 µM. Km ≈ 40 ( 6 µM. e The highest value from
d
three experiments.
to ref 15. All new compounds were fully characterized
by 1H and 13C NMR and high-resolution mass spectrom-
etry.
Resu lts a n d Discu ssion . The activity of these N′-
hydroxyguanidine compounds as substrates of NOS was
evaluated by a hemoglobin assay that is based on the
conversion of oxyhemoglobin to methemoglobin by
NO.12,13 For evaluation, NOHA was used as a control
indicating 100% activity. The results are summarized
in Table 1.
According to the crystal structure of nNOS complexed
with 1,15 the hydroxyguanidino moiety is anchored by
multiple H-bonds and is oriented in a manner similar
to what is observed for NOHA at the NOS active site
(Figure 1A). Such orientation allows the n-butyl group
to be accommodated in the space surrounded by Gln478
and Val567 of nNOS. According to the structure-
activity-relationship assay,12,13 the alkyl group size
should be between three and five carbons for good
substrates. Large alkyl groups cannot fit into the space
Data in Table 2 indicate that all the good substrates
have high affinity in mode A except for 3. Compound 2
has high affinity in both modes. The crystal structure
cannot fully exclude the possibility that 2 adopts mode

Letters
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 12 2273
Sch em e 3
Ta ble 2. FlexiDock Binding Affinity of
N-Alkyl-N′-hydroxyguanidines with nNOSa
ture-function relationships of isoform-selective NOS
substrates and the mechanism of NO synthase.
compd
mode A
mode B
nNOS activity, %
Ack n ow led gm en t. This work is supported by Na-
tional Institutes of Health, Grants GM54074 (P.G.W.)
and GM33688 (T.L.P.).
1
2
3
4
-74.12
-74.24
-46.72
-71.48
-60.18
-51.48
-47.47
64
70
70
33
12
6
-62.68
-65.58
N-sec-butyl
N-(3-methylbutyl)
N-isobutyl
Su p p or tin g In for m a tion Ava ila ble: Experimental sec-
tion. This material is available free of charge via the Internet
at http://pubs.acs.org.
2
a
FlexiDock on Sybyl 6.7 is used to calculate the binding affinity.
All of the docking scores have units of energy (kcal/mol). These
values are considered to be relative docking energies and do not
represent the actual binding energies. However, for a series of
compounds, the docking energies are expected to have the same
rank-order correlation as the true binding energies.
Refer en ces
(1) Moncada, S.; Palmer, R. M.; Higgs, E. A. Nitric oxide: physiol-
ogy, pathophysiology, and pharmacology. Pharmacol. Rev. 1991,
43, 109-142.
(2) Tomasian, D.; Keaney, J . F.; Vita, J . A. Antioxidants and the
bioactivity of endothelium-derived nitric oxide. Cardiovasc. Res.
2000, 47, 426-435.
(3) Brennan, P. A. The actions and interactions of nitric oxide in
solid tumours. Eur. J . Surg. Oncol. 2000, 26, 434-437.
(4) Grisham, M. B.; J ourd’Heuil, D.; Wink, D. A. Nitric oxide. I.
Physiological chemistry of nitric oxide and its metabolites:
implications in inflammation. Am. J . Physiol. 1999, 276, G315-
G321.
(5) Christopherson, K. S.; Bredt, D. S. Nitric oxide in excitable
tissues: physiological roles and disease. J . Clin. Invest. 1997,
100, 2424-2429.
(6) Bredt, D. S. Endogenous nitric oxide synthesis: biological
functions and pathophysiology. Free Radical Res. 1999, 31, 577-
596.
A. According to the calculations, 3 prefers binding mode
B. This is very likely the catalytically active binding
mode because the low calculated binding affinity in
mode A is inconsistent with the observed high activity
and affinity (Km ≈ 40 ( 6 µM for this compound; Km
≈
13 ( 2 µM for NOHA21).
Since all the known mechanisms are based on mode
A (Scheme 3),15,19,22-27 we propose a new mechanism to
accommodate the novel binding mode B. The OH group
of hydroxyguanidine in mode B is located too far away
from the heme iron and lends support to the mechanism
wherein the N atom is the source of hydrogen atom
supplied to the ferric superoxy species.15 The binding
mode B can be enforced only if a small N-substituted
alkyl group, such as isopropyl and cyclopropyl, fits into
the small hydrophobic pocket next to the substrate-
binding site. Any smaller or larger N-alkyl group on
hydroxyguanidine will lead to the preferred binding
mode A. It is very likely that binding mode B contributes
to the high NOS selectivity observed for 3.
(7) Alderton, W. K.; Cooper, C. E.; Knowles, R. G. Nitric oxide
synthases: structure, function and inhibition. Biochem. J . 2001,
357, 593-615.
(8) Roman, L. J .; Martasek, P.; Masters, B. S. Intrinsic and extrinsic
modulation of nitric oxide synthase activity. Chem. Rev. 2002,
102, 1179-1190.
(9) J ia, Q.; J anczuk, A. J .; Cai, T.; Xian, M.; Wen, Z.; et al. NO
donors with anticancer activity. Expert Opin. Ther. Pat. 2002,
12, 819-826.
(10) Auvin, S.; Auguet, M.; Navet, E.; Harnett, J . J .; Viossat, I.; et
al. Novel inhibitors of neuronal nitric oxide synthase with potent
antioxidant properties. Bioorg. Med. Chem. Lett. 2003, 13, 209-
212.
(11) Marletta, M. A. Approaches toward selective inhibition of nitric
oxide synthase. J . Med. Chem. 1994, 37, 1899-1907.
(12) Dijols, S.; Perollier, C.; Lefevre-Groboillot, D.; Pethe, S.; Attias,
R.; et al. Oxidation of N(ω)-hydroxyarginine analogues by NO-
synthase: the simple, non amino acid N-butyl N′-hydroxyguani-
dine is almost as efficient an NO precursor as N(ω)-hydroxyargi-
nine. J . Med. Chem. 2001, 44, 3199-3202.
(13) Xian, M.; Fujiwara, N.; Wen, Z.; Cai, T.; Kazuma, S.; et al. Novel
substrates for nitric oxide synthases. Bioorg. Med. Chem. 2002,
10, 3049-3055.
(14) Mansuy, D.; Boucher, J . L. Oxidation of N-hydroxyguanidines
by cytochromes P450 and NO-synthases and formation of nitric
oxide. Drug Metab. Rev. 2002, 34, 593-606.
Con clu sion . Given the striking similarity in dimeric
quaternary structure and especially in the heme active
sites of the three NOS isoforms,28,29 the discovery of
isoform-selective substrates presents an especially chal-
lenging problem. We have reported N-aryl-N′-hydroxy-
guanidines as substrates selective for iNOS over nNOS.13
Here, we showed the first reverse case that 3 is a good
substrate selective for nNOS over iNOS. These results
open the way toward a better understanding of struc-

2274 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 12
Letters
(15) Li, H.; Shimizu, H.; Flinspach, M.; J amal, J .; Yang, W.; et al.
The novel binding mode of N-alkyl-N′-hydroxyguanidine to
neuronal nitric oxide synthase provides mechanistic insights into
NO biosynthesis. Biochemistry 2002, 41, 13868-13875.
(16) Roberts, J . D.; Mazur, R. H. Small-ring compounds. IV. Inter-
conversion reactions of cyclobutyl, cyclopropylcarbinyl, and
allylcarbinyl derivatives. J . Am. Chem. Soc. 1951, 73, 2509-
2520.
(17) Kai, Y.; Knochel, P.; Kwiatkowski, S.; Dunitz, J . D.; Oth, J . F.
M.; et al. Structure, synthesis, and properties of some persub-
stituted 1,2-dinitroethanes. In quest of nitrocyclopropyl-anion
derivatives. Helv. Chim. Acta 1982, 65, 137-161.
(18) Dijols, S.; Boucher, J . L.; Lepoivre, M.; Lefevre-Groboillot, D.;
Moreau, M.; et al. First non-R-amino acid guanidines acting as
efficient NO precursors upon oxidation by NO-synthase II or
activated mouse macrophages. Biochemistry 2002, 41, 9286-
9292.
(23) Korth, H. G.; Sustmann, R.; Thater, C.; Butler, A. R.; Ingold, K.
U. On the mechanism of the nitric oxide synthase-catalyzed
conversion of N omega-hydroxyl-L-arginine to citrulline and
nitric oxide. J . Biol. Chem. 1994, 269, 17776-17779.
(24) Crane, B. R.; Arvai, A. S.; Ghosh, S.; Getzoff, E. D.; Stuehr, D.
J .; et al. Structures of the N(ω)-hydroxy-L-arginine complex of
inducible nitric oxide synthase oxygenase dimer with active and
inactive pterins. Biochemistry 2000, 39, 4608-4621.
(25) Tierney, D. L.; Huang, H.; Martasek, P.; Masters, B. S.; Silver-
man, R. B.; et al. ENDOR spectroscopic evidence for the position
and structure of NG-hydroxy-L-arginine bound to holo-neuronal
nitric oxide synthase. Biochemistry 1999, 38, 3704-3710.
(26) Tierney, D. L.; Huang, H.; Martasek, P.; Roman, L. J .; Silverman,
R. B.; et al. ENDOR Spectroscopic Evidence for the Geometry
of Binding of Retro-Inverso-Nω-Nitroarginine-Containing Dipep-
tide Amides to Neuronal Nitric Oxide Synthase. J . Am. Chem.
Soc. 2000, 122, 7869-7875.
(27) Tantillo, D. J .; Fukuto, J . M.; Hoffman, B. M.; Silverman, R. B.;
Houk, K. N. Theoretical Studies on NG-Hydroxy-L-arginine and
Derived Radicals: Implications for the Mechanism of Nitric
Oxide Synthase. J . Am. Chem. Soc. 2000, 122, 536-537.
(28) Raman, C. S.; Li, H.; Martasek, P.; Babu, B. R.; Griffith, O. W.;
et al. Implications for isoform-selective inhibitor design derived
from the binding mode of bulky isothioureas to the heme domain
of endothelial nitric-oxide synthase. J . Biol. Chem. 2001, 276,
26486-26491.
(19) Huang, H.; Hah, J . M.; Silverman, R. B. Mechanism of nitric
oxide synthase. Evidence that direct hydrogen atom abstraction
from the O-H bond of NG-hydroxyarginine is not relevant to
the mechanism. J . Am. Chem. Soc. 2001, 123, 2674-2676.
(20) Moynihan, H. A.; Roberts, S. M.; Weldon, H.; Allcock, G. H.;
Anggard, E. E.; et al. Synthesis and biological activity of the
novel nitric oxide synthase inhibitor Nw-hydroxy-Nw-methyl-
L-arginine. J . Chem. Soc., Perkin Trans. 1 1994, 769-771.
(21) Moali, C.; Boucher, J . L.; Sari, M. A.; Stuehr, D. J .; Mansuy, D.
Substrate specificity of NO synthases: detailed comparison of
L-arginine, homo-L-arginine, their N ω-hydroxy derivatives, and
(29) Wang, P. G.; Xian, M.; Tang, X.; Wu, X.; Wen, Z.; et al. Nitric
oxide donors: chemical activities and biological applications.
Chem. Rev. 2002, 102, 1091-1134.
N
ω-hydroxynor-L-arginine. Biochemistry 1998, 37, 10453-
10460.
(22) Marletta, M. A. Nitric oxide synthase structure and mechanism.
J . Biol. Chem. 1993, 268, 12231-12234.
J M0340703