Heterocyclic analogues of L-citrulline as inhibitors of the isoforms of nitric oxide synthase (NOS) and identification of N(δ)-(4,5-dihydrothiazol-2-yl)ornithine as a potent inhibitor

Article abstract of DOI:10.1016/S0968-0896(99)00136-4

L-Thiocitrulline is a known potent inhibitor of several isoforms of nitric oxide synthase (NOS). To explore the structure-activity relationships (SARs) for this molecule in more depth than has previously been reported, three analogues substituted at the sulphur of the isothioureas have been synthesised. In two of these, the S-substituent was 'tied back' sterically by cyclisation to the nitrogen remote from the amino-acid unit. N(δ)-(4,5-Dihydrothiazol-2-yl)ornithine was identified as an inhibitor of rat inducible and constitutive isoforms of NOS and of a constitutive NOS derived from a human tumour xenograft. Analogous N(δ)-(thiazol-2-yl)ornithines were less active, whereas the corresponding N(δ)-(oxazol-2-yl)ornithine and N(δ)-(pyrimidin-2-yl)ornithine failed completely to inhibit NOS. A new efficient preparation of the critical synthetic intermediate, N(α)-Boc-thiocitrulline t-butyl ester, has been developed. Further exploration of the SAR with 2-amino-5-(heterocyclylthio)pentanoic acids (synthesised from 2-(Boc-amino)-5-bromopentanoic acid t-butyl ester), with N-(4-aminobutyl)thiourea and with 2-(4-aminobutylamino)-4,5-dihydrothiazole enabled refinement of our previous model for binding of the substrate, L-arginine, and the inhibitors to NOS.

Full text of DOI:10.1016/S0968-0896(99)00136-4

Bioorganic & Medicinal Chemistry 7 (1999) 1787±1796
Heterocyclic Analogues of L-Citrulline as Inhibitors of the
Isoforms of Nitric Oxide Synthase (NOS) and Identi®cation of
Nꢀ-(4,5-Dihydrothiazol-2-yl)ornithine as a Potent Inhibitor
Saraj Ulhaq, a,b,y Edwin C. Chinje, b Matthew A. Naylor, c Mohammed Ja€ar, b
Ian J. Stratford b and Michael D. Threadgill a,*
aDepartment of Pharmacy and Pharmacology, University of Bath, Claverton Down, Bath BA2 7AY, UK
bMRC Experimental Oncology Laboratory, Department of Pharmacy, University of Manchester, Oxford Road,
Manchester M13 9PL, UK
cGray Laboratory Cancer Research Trust, PO Box 100, Mount Vernon Hospital, Northwood, Middlesex HA6 2JR, UK
Received 23 July 1998; accepted 25 February 1999
AbstractÐl-Thiocitrulline is a known potent inhibitor of several isoforms of nitric oxide synthase (NOS). To explore the structure±
activity relationships (SARs) for this molecule in more depth than has previously been reported, three analogues substituted at the
sulphur of the isothioureas have been synthesised. In two of these, the S-substituent was `tied back' sterically by cyclisation to the
nitrogen remote from the amino-acid unit. Nd-(4,5-Dihydrothiazol-2-yl)ornithine was identi®ed as an inhibitor of rat inducible and
constitutive isoforms of NOS and of a constitutive NOS derived from a human tumour xenograft. Analogous Nd-(thiazol-2-yl)or-
nithines were less active, whereas the corresponding Nd-(oxazol-2-yl)ornithine and Nd-(pyrimidin-2-yl)ornithine failed completely to
inhibit NOS. A new ecient preparation of the critical synthetic intermediate, Na-Boc-thiocitrulline t-butyl ester, has been devel-
oped. Further exploration of the SAR with 2-amino-5-(heterocyclylthio)pentanoic acids (synthesised from 2-(Boc-amino)-5-bro-
mopentanoic acid t-butyl ester), with N-(4-aminobutyl)thiourea and with 2-(4-aminobutylamino)-4,5-dihydrothiazole enabled
re®nement of our previous model for binding of the substrate, l-arginine, and the inhibitors to NOS. # 1999 Elsevier Science Ltd.
All rights reserved.
Introduction
the neuronal form (nNOS) and the endothelial form
(eNOS). Several known inhibitors of the isoforms of the
enzyme are analogues of the substrate 1 or of the co-
product 3 (Scheme 1); they include NG-monomethyl-l-
arginine (NMMA, 4),3 NG-nitro-l-arginine (NOARG,
5a)3,4 and its methyl ester (NAME, 5b), Nd-imino-
methyl-l-ornithine (NIO, 6) and l-thiocitrulline 7. The
Ki values reported for these inhibitors are comparable
with the Km for the substrate 1.5
*
Nitric oxide ( NO) is the smallest known messenger
molecule in biological systems and inter alia is respon-
sible for maintaining cardiovascular homeostasis.1 It is
biosynthesised from l-arginine 1 by the various iso-
forms of nitric oxide synthase (NOS), yielding l-citrul-
line 3 as a co-product. As shown in Scheme 1, the
process comprises two separate mono-oxygenation steps
and involves NG-hydroxyarginine 2 as an intermediate.
Both steps require molecular oxygen (O2) and NADPH.
There are two main groups of isoforms of NOS, the
constitutive Ca2+/calmodulin-dependent types (cNOS)
and an inducible Ca2+/calmodulin-independent form
(iNOS). The cNOS types can be further divided2 into
NOS inhibitors that have particular tissue or isozyme
speci®cities open up a variety of therapeutic possibi-
lities.6 For example, N-(3-(aminomethyl)benzyl)ace-
tamidine is a highly selective inhibitor of murine
iNOS.7,8 Recently, NOS inhibitors have been used
selectively to modulate tumour blood ¯ow, oxygenation
and redox status.9±11 In our previous paper,12 we
reported our identi®cation of S-2-amino-5-(imidazol-1-
yl)pentanoic acid 8 as an inhibitor of two rat and one
human isoforms of NOS, and our approaches to
designing prodrugs which would be bioreduced selec-
tively to inhibitors in hypoxic tumour tissue. In the
Key words: Nitric oxide synthase; thiocitrulline; Nd-(4,5-dihydro-
thiazol-2-yl)ornithine; structure±activity relationship.
* Corresponding author. Tel.: +44-(0)1225-826-840; fax: +44-
(0)1225-826-114; e-mail: prsmdt@bath.ac.uk
Present address: Department of Chemistry, University of Regina,
Regina, Saskatchewan S4S 0A2, Canada.
y
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00136-4

1788
S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
Scheme 1. Steps in the NOS-catalysed oxidation of l-arginine 1 to l-citrulline 3, via NG-hydroxyarginine 3 and structures of known inhibitors 4±8
of NOS.
present paper, we report our ®ndings on the develop-
ment of new, more potent inhibitors, the contribution of
these ®ndings to the establishment of models for bind-
ing of amino-acid inhibitors to NOS, and our identi®-
cation of a new strong inhibitor of the isoforms of the
enzyme.
the benzoyl group from the intermediate N-benzoyl-
thiourea, gave poor yields (<25%) of 13.
The simple S-alkyl thiocitrulline derivative 15 was pre-
pared eciently by alkylation of the thiourea 13 at sul-
phur with 2-iodopropane under basic conditions,
followed by acidolytic cleavage of the Boc and But ester
protecting groups of the intermediate 14. The ®rst
example in which the isothiourea is incorporated into a
heterocycle is the dihydrothiazine 17. Synthesis of this
target molecule was achieved by alkylation of 13 at sul-
phur with 1,3-dibromopropane, with cyclisation being
completed in one pot through intramolecular alkylation
at the less sterically hindered nitrogen. Deprotection of
the intermediate 16 gave 17 in 17% overall yield from
13. The ®ve-membered ring analogue 19 was prepared,
via 18, in slightly better yield (23%), using 1,2-dibromo-
ethane as the bifunctional electrophile. For comparison
of biological activity of aromatic with non-aromatic
heterocycles, the thiazole-amino-acids 21 and 23 were
prepared. Hantzsch condensations of the protected
thiocitrulline 13 with chloroacetaldehyde and with chloro-
acetone under the standard neutral conditions gave the
thiazole 20 and the 4-methylthiazole 22, respectively, in
satisfactory yields. The usual acidolytic deprotection then
provided the target thiazolyl amino acids 21 and 23.
Chemical Synthesis
S-2-Amino-5-(imidazol-1-yl)pentanoic acid 812 and
thiocitrulline 7 are inhibitors of NOS that may be con-
sidered to bind to the substrate (arginine) binding site
through the a-amino acid zwitterion moiety, presenting
the imidazole 3-N and the sulphur of the (iso)thiourea
as strong ligands for the haem iron. The approach
adopted in the present study was to investigate the
e€ects on NOS inhibitory activity of incorporating the
(iso)thiourea unit of 7 into aromatic and partly satu-
rated heterocycles and to test the requirement for the
CH2NH group, which may form a hydrogen bond in the
active site of the enzymes and thus contribute to bind-
ing. The common synthetic intermediate for the target
azathiaheterocycles 17, 19, 21 and 23 and for the S-
alkylisothiourea 15 was the protected thiocitrulline 13
(Scheme 2). Ornithine 9 was protected at the d-amine
with Cbz, selectivity being achieved by temporary com-
plexation of the carboxylate and the a-amine with
Cu2+, in a modi®cation of the method employed by us13
and by Yajima et al.14 for the selective e-acylation of
lysine; Clarke and Waight15 have prepared Orn(Cbz)
OH by this method. Boc and t-butyl ester protection
were then introduced at the exposed amine and car-
boxylic acid of Orn(Cbz)OH 10, giving BocOrn(Cbz)
OBut 11. Hydrogenolysis exposed the e-amine in 12.
The thiourea unit in 13 was introduced in very high
yield in two steps, formation of the isothiocyanate with
thiophosgene and reaction with ammonia; the inter-
mediate was not isolated. An alternative sequence,
reaction with benzoyl isothiocyanate and hydrolysis of
Each of the target amino acids 15, 17, 19, 21 and 23
shown in Scheme 2 contains the sulphur atom and both
of the nitrogen atoms of the thiourea present in the lead
inhibitor thiocitrulline 7. To test the requirement for
these particular atoms for NOS inhibitory activity,
amino acids analogous to the active (dihydro)thiazoles
19 and 21 and to the dihydrothiazine 17 were synthe-
sised, in which one or more of these atoms have been
replaced by an alternative heteroatom. 2-Substituted
4,5-dihydrooxazoles are usually prepared by cyclisation
of N-(2-hydroxyethyl)amides with thionyl chloride or
similar dehydrating agents16 but 2-alkylamino-4,5-dihy-
drooxazoles are reported relatively infrequently in the

S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
1789
Scheme 2. Synthesis of S-isopropylthiocitrulline 15, the potent inhibitor 19 and related compounds from the critical intermediate Na-Boc-thioci-
trulline t-butyl ester 13. Reagents: (i) CuCO3, BnOCOCl, Na2CO3; (ii) EDTA Na+2 salt; (iii) Me2CˆCH2, concd H2SO4, 1,4-dioxane; (iv) Boc2O,
Et3N, H2O; (v) H2, Pd/C, EtOH; (vi) CSCl2, CaCO3, CHCl3; (vii) NH3, MeOH; (viii) Me2CHI, Et3N, MeCN; (ix) HCl, CH2Cl2; (x) Br(CH2)3Br,
KOBut, THF; (xi) Br(CH2)2Br, K2CO3, THF, Á; (xii) ClCH2Cl, THF; (xiii) ClCH2COMe, THF.
literature.17,18 As shown in Scheme 3, BocOrnOBut 12
was carbamoylated eciently by 2-chloroethyl isocya-
nate to give the urea 24. Cyclisation was e€ected by the
method developed by Wong et al.19 using potassium
¯uoride on alumina as the base, to give the 4,5-dihy-
drooxazole 24 in high yield, as shown in Scheme 3.
Acidolytic deprotection gave the dihydrooxazolyl amino
acid 26. The analogous pyrimidine 29 (Scheme 3) was
synthesised from BocOrnOH 27,20 protection of the
acid function being unnecessary in this case during
the SNAr reaction with 2-chloropyrimidine. Removal of
the Boc group of 28 was followed by neutralisation
during chromatography to a€ord 29.
S-alkylated products 31 and 33 in moderate and
excellent yields, respectively. Acidolytic deprotection
removed the Boc and But ester groups to a€ord the
targets 32 and 34 as their dihydrochloride salts.
Scheme 5 shows the synthetic approaches to the target
compounds 39 and 41, which represent the lead inhibi-
tors thiocitrulline 7 and the 4,5-dihydrothiazolyl amino
acid 19, respectively, but which lack the carboxylic acid
moiety. These were designed to test the requirement for
electrostatic or hydrogen-bonding interactions between
the a-carboxylate of the inhibitors and the enzyme. Such
compounds, if potent in their NOS-inhibitory activity,
may be easier to formulate as drugs and may penetrate
cell membranes more readily than their zwitterionic
parents. Reaction of the mono-Boc protected putrescine
3523 with carbon disulphide and methylation without
isolation of the intermediate dithiocarbamate salt gave
the dithiocarbamate ester 36. The thiourea 38 was
formed in good yield by displacement of methanethio-
late by ammonia. Cyclisation to the 4,5-dihydrothiazole
40 with 1,2-dibromoethane was e€ected under reaction
conditions analogous to those used for the protected
dihydrothiazolyl amino-acid 18. Removal of the Boc
protection from the aminobutyl chains of 36, 38 and 40
under the usual conditions furnished the target amino-
butyl compounds 37, 39 and 41, respectively, as their
hydrochloride salts.
Target compounds 32 and 34 (Scheme 4) lack the
potential hydrogen-bond donor N±H at the position
corresponding to the CH2NH of arginine, the substrate
of NOS. Thus synthesis and evaluation of these com-
pounds will test the contribution of this hydrogen bond
to the binding of the more potent inhibitors and may
contribute to understanding the mode of binding of the
natural substrate. The synthetic sequence to both targets
was di€erent to that used for the other heterocyclic amino
acids above, in that the (protected) amino acid unit was
introduced as an electrophile, the 5-bromopentanoate
ester 30.12,21,22 Treatment of 30 with thiazole-2-thiolate
and with the anion derived from imidazole-2-thione/
2-mercaptoimidazole lead to the formation of the

1790
S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
Scheme 4. Synthesis of sulphide-linked analogues 32 and 34.
Reagents: (i) 4,5-dihydrothiazole-2-thiol, NaHCO3, MeOH; (ii) HCl,
CH2Cl2; (iii) imidazole-2-thiol, NaHCO3, MeOH.
the e€ect of introducing a bulky alkyl substituent at
sulphur was tested with the S-isopropyl isothiourea 15.
This compound retained signi®cant activity but was ca.
tenfold less potent than 7. In the corresponding dihy-
drothiazine 17, the conformation of the S-substituent is
constrained by linkage back to the nitrogen of the iso-
thiourea; the inhibitory activity is maintained, relative
to 15. However, imposition of further conformational
constraint by incorporating the isothiourea into the 5-
membered dihydrothiazole ring of 19 led to a threefold
increase in potency, as judged by IC50 against the two
rat enzymes (against rat nNOS: 17: IC50=15 mM; 19:
IC50=4.3 mM). However, when these (constrained) iso-
thioureas were examined for their inhibitory activity
using the cNOS of human origin, the dihydrothiazole 19
appeared to be slightly more potent than the lead com-
pound thiocitrulline 7 (7: IC50=2.0 mM; 19: IC50=
1.3 mM). As found for the NOS of rodent origin, 15 and
17 were ca. 10-fold less potent than 7 against the human
enzyme. Interestingly, the isoform selectivity of 19 is
di€erent to that of 7; whereas 7 is threefold more active
against rat iNOS than against rat nNOS, 19 is twofold
more active against rat nNOS than against rat iNOS.
Scheme 3. Synthesis of Nd-(4,5-dihydrooxazol-2-yl)ornithine 26 and
Nd-(pyrimidin-2-yl)ornithine 29. Reagents: (i) Cl(CH2)2NCO, THF;
(ii) KF/alumina, MeCN; (iii) HCl, CH2Cl2; (iv) 2-chloropyrimidine,
Et3N, MeOH, Á; (v) HCl, H2O, EtOAc; (vi) NH3, H2O, MeOH.
Evaluation as Inhibitors of NOS
All the candidate inhibitors synthesised as described
above were evaluated for their inhibitory activity
against iNOS and nNOS derived from rat lung and
from rat brain, respectively. As described in the Experi-
mental section, the assay was based on the conversion
of [14C]-1 to [14C]-3.
As an initial screen, most compounds were tested at
1.0 mM against the two isoforms of the rat enzyme; the
results are shown in Table 1. The initial screen was
conducted at 2.0 mM for 29 and at 100 mM for 37, 39
and 41. IC50 values were also obtained for selected com-
pounds. Taking thiocitrulline 7 as the lead compound,
With the potent activity of 19 having been established,
the structure±activity relationship around the ®ve-
membered ring was explored. Conversion of the ring to
the aromatic thiazole 21 reduced activity markedly, with
IC50=200 mM against the rat iNOS, and introduction of

S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
1791
Scheme 5. Synthesis of methyl N-(4-aminobutyl)dithiocarbamate 37, N-(4-aminobutyl)thiourea 39 and 2-(4-aminobutylamino)-4,5-dihydrothiazole
41. Reagents: (i) CS2, Et3N, THF; (ii) MeI; (iii) HCl, CH2Cl2; (iv) NH3, MeOH; (v) Br(CH2)2Br, K2CO3, THF, Á.
Table 1. Inhibition of the isoforms of NOS by know inhibitors and by heterocyclic analogues of l-citrulline 3
Compound
% Inhibition of rat iNOS
(initial screen)
% Inhibition of rat nNOS
(initial screen)
IC50 (-M) versus
rat iNOS
IC50 (-M) versus
rat nNOS
IC50 (-M) versus
H647 cNOS
5a (NOARG)
5b (NAME)
6 (NIO)
7
(triocitrulline)
3.0‹0.2
6.0‹0.3
2.3‹0.1
4.6‹0.2
1.7‹0.1
1.9‹0.2
8
99‹2 (1.0 mM)
96‹5 (1.0 mM)
84‹2 (1.0 mM)
100‹7 (1.0 mM)
81‹4 (1.0 mM)
26‹4 (1.0 mM)
18‹6 (1.0 mM)
3‹2 (2.0 mM)
6‹6 (1.0 mM)
15‹1 (1.0 mM)
5‹6 (100 mM)
25‹5 (100 mM)
64‹2 (100 mM)
100‹2 (1.0 mM)
97‹3 (1.0 mM)
86‹2 (1.0 mM)
93‹1 (1.0 mM)
71‹2 (1.0 mM)
14‹10 (1.0 mM)
12‹3 (1.0 mM)
8‹3 (2.0 mM)
2‹2 (1.0 mM)
5‹2 (1.0 mM)
4‹3 (100 mM)
13‹4 (100 mM)
49‹1 (100 mM)
31.6‹1.5
14.5‹0.7
24.5‹1.8
8.1‹2.2
199‹33
18.5‹1.3
13.0‹0.6
23.5‹0.9
27.5‹1.7
1.3‹0.1
15
17
19
21
23
26
29
32
34
37
39
41
13.0‹0.5
4.3‹0.1
100‹10
a methyl substituent on the thiazole virtually abolished
activity in 23. Replacement of the soft ligand sulphur in
19 with oxygen, a potentially harder ligand, in the
dihydrooxazole 26 also abolished inhibitory activity.
The pyrimidine 29 was also inactive, even at 2.0 mM.
Whereas evaluation of the sulphides 32 and 34 tested
the requirement for a hydrogen-bonding N H link
between the heterocycle and the amino acid, the impor-
tance of the a-carboxylate to binding and inhibition was
examined using 37, 39 and 41. These aminobutylamino
compounds were evaluated at a standard test con-
centration 100 mM. As expected, the dithiocarbamate 37
was inactive but 39, the `decarboxylated analogue' of
thiocitrulline 7 showed weak activity, being >100-fold
less active than the lead compound. Interestingly,
removal of the carboxylate had a lesser deleterious e€ect
on the inhibitory activity in the 4,5-dihydrothiazole ser-
ies, in that 41 was some 20-fold less active than 19
against the rat isoforms of NOS, as judged by IC50
values.
In compounds 32 and 34, the heterocycle is linked to the
amino-acid unit through a lipophilic, non-hydrogen-
bonding sulphur, rather than through the potential
hydrogen-bond donor N H so far examined. The
thiazolyl sulphide 32 is the strict analogue of the
4,5-dihydrothiazolylamino compound 19; however,
although 19 was a potent inhibitor of the rat enzymes
(IC50=8.1 mM versus rat iNOS and IC50=4.3 mM ver-
sus rat nNOS), the S-linked analogue 32 did not inhibit
either rat isoform at the standard initial test concentra-
tion, 1.0 mM. The corresponding imidazole 34 was also
inactive.
In the present study, the compounds synthesised and
evaluated have a wide range of inhibitory potencies

1792
S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
against the isoforms of NOS. Some structure±activity
concepts can be gleaned; these can be rationalised in
terms of a model for the binding of these agents to
the substrate binding site of the enzymes, as shown in
Figure 1. In the simpler model we developed in our
previous study12 using 2-amino-5-azolylpentanoic acids,
it was evident that ligation of the heterocyclic nitrogen
to the haem iron atom was important for binding. We
also speculated that there may be binding sites for the
anionic a-carboxylate and cationic a-ammonium groups
of the amino acid unit, located at appropriate distances
from the haem to allow presentation of the heterocycle
to the iron. The ®ndings of the present study allow us to
present a re®ned model for the binding of the sulphur-
containing inhibitors (Fig. 1). The known inhibitor,
thiocitrulline 7, could bind at the substrate binding site
as shown, with electrostatic and/or hydrogen-bonding
recognition of the a-ammonium and a-carboxylate
functions and ligation of the soft sulphur centre to the
haem iron. In addition, it may be speculated that there
is a further hydrogen-bond from the isothiourea N H
to residues in the NOS protein, as shown in Figure 1.
The most e€ective inhibitor amongst the 2-amino-5-
azolylpentanoic acids, the imidazole 8, ®ts this model
but lacks the possibility of this hydrogen bond. The S-
isopropyl group in 15 also ®ts the model but it is evident
that this addition of an alkyl group to the sulphur of
thiocitrulline reduces anity either by sterically
obstructing approach of the ligand to the iron atom or
by preventing it being negatively charged. Methylation
of thiocitrulline at sulphur has been reported24 to reduce
inhibitory activity by ca. 10-fold, which is consistent
with our ®ndings. However, when the steric restriction
of the approach of the sulphur to the iron is lessened
by `tying back' the S-substituent to the nitrogen in the
4,5-dihydrothiazole 19, the inhibitory activity is restored
to the potency of thiocitrulline 7. Thus we propose that
19 binds as shown in Figure 1 and that the diminution
of activity upon alkylation at sulphur in thiocitrulline is
a steric e€ect, rather than prevention of formation of a
thiolate anion.
In compound 32, a sulphide links the dihydrothiazole
and the amino-acid units. This link is not capable of
being a hydrogen-bond donor and provides a direct test
for the putative hydrogen-bond shown for the binding
of the substrate l-arginine and the inhibitors 7 and 19.
The inactivity of 32 attests to the importance of this
attractive interaction in this sulphur-ligand series. Feld-
man et al.25 observed a similar requirement in that l-
arginine 1 is the substrate for the isoforms of NOS but
the analogue l-indospicine (in which the guanidine is
replaced by CH2C(ˆN+H2)NH2) does not bind. In
contrast, this hydrogen bond appears to be unnecessary
in the strong binding12 of the imidazole 8 and of 2-
amino-5-(5-methyl-2-nitrophenylthio)pentanoic acid.26
Finally, removal of the anionic a-carboxylate dimin-
ishes but does not abolish inhibitory activity. Thus
comparison of the interactions of 39 with those of 7 and
of the aminobutylaminodihydrothiazole 41 with the
potent thiazole-amino acid 19 (Fig. 1) suggest that the
binding of this carboxylate anion is not one of the major
determinants in recognition and binding of the substrate
and the inhibitors by the isoforms of NOS. This obser-
vation provides a potential lead for the development of
new inhibitors with uncharged side-chains, based on the
1-substituted imidazole ligand 8 and the 2-(substituted
amino)-4,5-dihydrothiazole ligand 19. This re®ned
model is consistent with those previously proposed26±28
for binding of 1, of isothiourea NOS inhibitors and of
Figure 1. Models for binding of the substrate 1, the inhibitors thiocitrulline 7, 2-amino-5-(imidazol-1-yl)pentanoic acid 8, Nd-(4,5-dihydrothiazol-2-
yl)ornithine 19 and the `decarboxylated' analogues 39 and 41 to the substrate binding site of NOS.

S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
1793
some known amino-acid NOS inhibitors to the sub-
strate binding site.
Ph-H5); 13C NMR d 27.78, 28.31, 28.49, 30.32, 53.36,
66.44, 79.77, 82.17, 128.42, 128.57, 128.89, 135.83,
155.39, 171.32, 172.67; MS (FAB) m/z 423 (M+H).
Nꢁ-(1,1-Dimethylethoxycarbonyl)ornithine 1,1-dimethyl-
ethyl ester (12). Compound 11 (20.3 g, 48 mmol), in
EtOH (120 mL), was stirred with H2 in the presence of
Pd/C (10%, 700 mg) for 48 h. Filtration (Celite1), eva-
poration and chromatography (CH2Cl2:MeOH, 19:1)
Conclusion
In this paper, we have described our exploration of the
structure±activity relationships for inhibition of the iso-
forms of NOS, using thiocitrulline 7 as a known lead
compound. Synthesis and evaluation of series of Nd-
(heterocyclyl)ornithines, 2-amino-5-(heterocyclylthio)-
pentanoic acids and analogues lacking the carboxylate
have led to identi®cation of Nd-(4,5-dihydrothiazol-2-
yl)ornithine 19 as a potent new inhibitor of the isoforms
of NOS with potency similar to those of the known
inhibitors NOARG 5a, NAME 5b, NIO 6 and thioci-
trulline 7. Our enzyme-inhibition data have led us to
reinforce and re®ne our previous model12 for binding of
inhibitors to the l-arginine substrate binding site of the
enzymes (Fig. 1). In particular, we have shown that the
a-carboxylate anion in the new lead 19 is desirable but
not essential for inhibitory activity. The results of our
further exploration of the structure±activity relation-
ships in this region of 2-substituted 4,5-dihydrothiazoles
(derived from 19) and 1-substituted imidazoles (derived
from 8) will be reported separately, as will our use of 19
as a `warhead' delivered by tissue-selective prodrugs.
1
gave 12 (15.3 g, 77%) as an oil, (lit.29 oil); H NMR d
1.43 (9H, s, But), 1.46 (9H, s, But), 1.88±2.32 (4H, m,
b,g-H2), 4.18 (1H, br, a-H), 4.38 (2H, t, J=7 Hz, d-H2),
5.13 (2H, br, NH2).
Nꢁ-(1,1-Dimethylethoxycarbonyl)-Nꢀ-aminothiocarbonyl-
ornithine 1,1-dimethylethyl ester (13). Thiophosgene
(2.07 g, 18 mmol) was stirred vigorously with 12 (2.0 g,
6.6 mmol) and CaCO3 (740 mg, 6.8 mmol) in CHCl3
(45 mL) and water (5 mL) for 16 h. The suspension was
®ltered. The aqueous phase of the ®ltrate was separated
and was extracted thrice with CHCl3. The combined
organic ®ltrate and extracts were dried and the solvent
was evaporated. The residue was taken up in MeOH
(30 mL) and cooled to 4ꢀC. NH3 was passed through
the solution for 20 min and the mixture was stirred for
3 h at 0ꢀC. Evaporation and chromatography (EtOAc:
hexane, 4:1) gave 13 (1.68 g, 89%) as a white solid: mp
45±47ꢀC (lit.30 oil); H NMR d 1.44 (9H, s, But), 1.47
1
(9H, s, But), 1.68±1.90 (4H, m, b, g-H4), 3.51 (2H, m, d-
H2), 4.13 (1H, m, a-H), 5.28 (1H, d, J=7 Hz, a-NH),
6.21 (1H, br, NH), 6.60 (1H, br, NH); 13C NMR d
26.34, 27.94, 28.30, 32.33, 49.34, 51.53, 77.47, 80.29,
155.45, 166.82, 171.54.
Experimental
General methods
NMR spectra were recorded on samples in CDCl3,
unless otherwise stated. IR spectra were recorded on
KBr discs, unless otherwise stated. Mass spectra were
obtained by electron-impact (EI), chemical-ionisation
(CI) or fast atom bombardment (FAB) techniques in
the positive ion mode, unless otherwise stated. The sta-
tionary phase for chromatography was silica gel. Melt-
ing points are uncorrected. Solutions in organic solvents
were dried with MgSO4. Solvents were evaporated
under reduced pressure. All chiral amino acids are of l
con®guration, unless otherwise stated. The brine was
saturated. Nd-Cbz-ornithine 10 was prepared essentially
by the method of Clarke and Waight,15 as modi®ed by
us13 for the synthesis of Ne-Cbz-lysine.
Nꢁ-(1,1-Dimethylethoxycarbonyl)-Nꢀ-(imino(1-methyl-
ethylthio)methyl)ornithine 1,1-dimethylethyl ester (14).
Compound 13 (500 mg, 1.4 mmol) was stirred with 2-
iodopropane (367 mg, 2.2 mmol) and Et3N (450 mg,
4.4 mmol) in dry MeCN (5 mL) at 50ꢀC for 16 h. Eva-
poration and chromatography (EtOAc:Me2CO, 1:1)
1
gave 14 (378 mg, 68%) as a colourless oil. H NMR d
1.25 (6H, d, J=7 Hz, 2ÂMe), 1.84 (18H, s, 2ÂBut),
1.89±2.20 (4H, m, b,g-H4), 3.49 (2H, t, J=7 Hz, d-H2),
3.60 (1H, m, CHMe2), 4.23 (1H, t, J=6 Hz, a-H), 5.26
(1H, m, NH); MS (FAB) m/z 390.2425 (M+H) (C18H36
N3O4S requires 390.2427).
Nꢀ-(Imino(1-methylethylthio)methyl)ornithine dihydro-
chloride (15). Compound 14 (250 mg, 640 mmol) was
treated with HCl, as for the synthesis of 21 except that
the solvent was CH2Cl2 and the material was recrys-
tallised (EtOH:CH2Cl2), to give 15 (62%) as a highly
Nꢁ-(1,1-Dimethylethoxycarbonyl)-Nꢀ-(phenylmethoxy-
carbonyl)-L-ornithine 1,1-dimethylethyl ester (11). Orn
(Cbz)OH 1015 (54.0 g, 202 mmol) was stirred vigorously
with 2-methylpropene (160 mL) and H2SO4 (15 mL) in
1,4-dioxan (200 mL) at 0ꢀC for 6 h. The solution was
added during 20 min to Et3N (120 mL) and water
(200 mL) at 10ꢀC. Di-t-butyl dicarbonate (44.0 g,
202 mmol) was added. The mixture was stirred for 5 h.
The evaporation residue, in EtOAc, was washed (aq
KHSO4, water, brine) and was dried. Evaporation and
chromatography (hexane:ethyl acetate, 2:1) gave 11
1
hygroscopic white solid: H NMR d (D2O) 1.38 (6H, d,
J=7 Hz, 2 (Me), 1.73±2.05 (4H, m, b,g-H4), 3.44 (2H, t,
J=7 Hz, d-H2), 3.79 (1H, septet, J=7 Hz, CHMe2), 4.03
(1H, t, J=6 Hz, a-H); MS (FAB) m/z 235.1351 (M+H)
(C9H21N3O2S requires 235.1356).
Nꢀ-(4,5-Dihydro-1,3-thiazin-2-yl)-Nꢁ-(1,1-dimethylethoxy-
carbonyl)ornithine 1,1-dimethylethyl ester (16). Com-
pound 13 (440mg, 1.3 mmol) was stirred with Br(CH2)3Br
(572 mg, 2.8 mmol) and KOBut (140 mg, 1.3 mmol)
in THF (3 mL) at 35ꢀC for 16 h. Evaporation and
1
(28.0 g, 82%) as a colourless oil, (lit.29, oil); H NMR d
1.43 (9H, s, But), 1.46 (9H, s, But), 1.83±2.60 (4H, m, b,
g-H4), 3.20 (2H, m, d-H2), 4.18 (1H, m, a-H), 4.85 (1H,
br, NH), 5.02±5.20 (3H, m, PhCH2+NH), 7.35 (5H, s,

1794
S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
chromatography (EtOAc!EtOAc:EtOH, 2:1) gave 16
(125 mg, 25%) as a colourless oil: IR (®lm) n 3300, 1710,
1650, 1605 cm 1; 1H NMR d 1.44 (9H, s, But), 1.47 (9H,
s, But), 1.70±1.95 (4H, m, b,g-H4), 2.20±2.40 (2H, m,
thiazine 5-H2), 3.15±3.25 (2H, m, d-H2), 3.60±4.00 (4H,
m, thiazine 4,6-H4), 4.22 (1H, m, a-H), 5.42 (1H, d,
J=7.9 Hz, a-NH), 9.05 (1H, br, d-NH); 13C NMR d
26.81, 28.04, 28.99, 30.56, 31.77, 40.80, 43.29, 49.52,
53.66, 82.17, 82.33, 156.02, 164.15, 171.58; MS (EI) m/z
387 (M), 277 (M Boc).
Nꢀ-(Thiazol-2-yl)ornithine hydrochloride (21). A solution
of 20 (300 mg, 810 mmol) in THF (5 mL) was saturated
with HCl at 0ꢀC and was stirred at 20ꢀC for 30 min. The
precipitate was collected by ®ltration under N2, washed
(THF) and dried to give 21 (130 mg, 65%) as a white
solid: mp 213±215ꢀC; 1H NMR (DMSO-d6) d 1.87±2.10
(4H, m, b,g-H4), 3.49 (2H, t, J=7.0 Hz, d-H2), 4.10 (1H,
brt, J=6 Hz, a-H), 6.90 (1H, d, J=3 Hz, thiazole 4-H),
7.23 (1H, d, J=3 Hz, thiazole 3-H); MS (FAB) m/z
216.0805 (M+H) (C8H14N3O2S requires 216.0808).
Nꢀ-(4,5-Dihydro-1,3-thiazin-2-yl)ornithine hydrochloride
(17). A solution of 16 was treated with HCl, as for the
synthesis of 21, to give 17 (67%) as a highly hygroscopic
Nꢁ-(1,1-Dimethylethoxycarbonyl)-Nꢀ-(4-methylthiazol-
2-yl)ornithine 1,1-dimethylethyl ester (22). Compound
13 (350 mg, 1.0 mmol) was stirred with chloroacetone
(187 mg, 2.0 mmol) in THF (6 mL) for 1 h. Evaporation
and chromatography (CH2Cl2:hexane, 1:1) gave 22
1
white solid: H NMR ((CD3)2SO) d 1.62±1.98 (4H, m,
b,g-H4), 2.02 (2H, quintet, J=6 Hz, thiazine 5-H2), 3.18
(2H, t, J=6 Hz, d-H2), 3.40±3.60 (4H, m, thiazine 4,6-
H4), 3.89 (1H, m, a-H), 8.4 (3H, br, N+H3), 8.6 (1H,
br, NH), 9.2 (1H, br, NH); 13C NMR ((CD3)2SO) d
21.70, 22.34, 26.68, 49.08, 51.67, 51.93, 62.08, 163.67,
170.80; MS (FAB) m/z 232.1137 (M+H) (C9H18N3O2S
requires 232.1120).
1
(230 mg, 61%) as an oil: H NMR d 1.44 (9H, s, But),
1.46 (9H, s, But), 1.75±1.95 (4H, m, b,g-H4), 2.20 (3H, s,
thiazole-Me), 3.28 (2H, t, J=6 Hz, d-H2), 4.15±4.23
(1H, m, a-H), 5.10 (1H, d, J=7 Hz, a-NH), 5.19±5.26
(1H, br, d-NH), 6.02 (1H, s, thiazole 5-H); 13C NMR d
16.65, 24.78, 28.33, 28.53, 31.91, 46.03, 53.47, 79.81,
82.19, 99.87, 144.32, 155.49, 169.91, 171.56; MS
(FAB) m/z 386.2112 (M+H) (C18H32N3O4S requires
386.2114).
Nꢀ-(4,5-Dihydrothiazol-2-yl)-Nꢁ-(1,1-dimethylethoxy-
carbonyl)ornithine 1,1-dimethylethyl ester (18). Com-
pound 13 (500 mg, 1.4 mmol) was heated under re¯ux
with Br(CH2)2Br (541 mg, 2.9 mmol) and K2CO3
(199 mg, 1.4 mmol) in THF (10 mL) for 16 h. Filtration,
evaporation and chromatography (EtOAc:hexane,
1:1!EtOAc:CH2Cl2:Et3N, 10:4:1) gave 18 (200 mg,
Nꢀ-(4-Methylthiazol-2-yl)ornithine hydrochloride (23).
Compound 22 was treated with HCl, as for the syn-
thesis of 21, to give 23 (65%) as a white solid: mp 250ꢀC;
1H NMR (D2O) d 1.43±1.82 (4H, m, b,g-H4), 2.01 (3H,
s, thiazole-Me), 3.32 (2H, t, J=7 Hz, 5-H2), 3.65 (1H, t,
J=6 Hz, a-H), 6.07 (1H, s, thiazole 5-H); 13C NMR
(D2O) d 17.81, 24.78, 29.94, 46.21, 56.46, 102.83, 139.20,
162.30, 172.10; MS (FAB) m/z 230.0962 (M+H)
(C9H16N3O2S requires 230.0965).
1
35%) as a colourless oil: H NMR d 1.46 (9H, s, But),
1.48 (9H, s, But), 1.65±1.93 (4H, m, b,g-H4), 3.14 (2H,
t, J=7 Hz, thiazole 5-H2), 3.35 (2H, m, d-H2), 3.61 (2H,
t, J=7 Hz, thiazole 4-H2), 4.17 (1H, m, a-H), 5.38
(1H, d, J=8 Hz, a-NH), 5.79 (1H, br, d-NH); MS
(FAB) m/z 372.1962 (M+H) (C17H30N3O4S requires
372.1957).
Nꢀ - (2 - Chloroethylaminocarbonyl) - Nꢁ - (1,1 - dimethyl -
ethoxycarbonyl)ornithine 1,1-dimethylethyl ester (24). 2-
Chloroethylisocyanate (211 mg, 2.0 mmol) was added
during 5 min to 12 (500 mg, 1.7 mmol) in THF (5 mL)
at 0ꢀC. The mixture was stirred at 20ꢀC for 16 h.
Evaporation and chromatography (EtOAc) gave 24
(530 mg, 78%) as a white solid: mp 103±105ꢀC; 1H
NMR d 1.44 (9 H, s, But), 1.46 (9 H, s, But), 1.62±1.96
(4H, m, b,g-H4), 3.21 (2H, t, J=6.4 Hz, NCH2), 3.33
(2H, t, J=6.4 Hz, NCH2), 3.61 (2H, t, J=6.4 Hz,
CH2Cl), 4.15 (1H, m, a-H), 4.90 (1H, br, NH), 4.98
(1H, br, NH), 5.18 (1H, d, J=7 Hz, a-NH); 13C NMR d
25.82, 28.00, 28.35, 30.60, 39.87, 42.12, 44.88, 53.62,
79.96, 82.19, 155.76, 158.30, 171.82; MS (FAB) m/z 791/
789/787 (2M+H), 396/394 (M+H), 358 (M HCl),
340/338 (M Me2CˆCH2), 296/294 (M Boc).
Nꢀ-(4,5-Dihydrothiazol-2-yl)ornithine dihydrochloride
(19). Compound 18 was treated with HCl, as for the
synthesis of 21, to give 19 (67%) as a highly hygroscopic
1
white solid: H NMR (D2O) d 1.67±1.86 (4H, m, b,g-
H4), 3.53±3.65 (2H, m, d-H2), 3.24 (2H, t, J=7 Hz,
thiazoline 4-H2), 3.93±4.07 (1H, m, 2-H), 4.24 (2H, t,
J=7 Hz, thiazoline 5-H2); 13C NMR (D2O) d 26.72,
28.91, 30.32, 31.69, 48.21, 58.34, 167.65, 172.36; MS
(FAB) m/z 218.0979 (M+H) (C8H16N3O2S requires
218.0963).
Nꢁ-(1,1-Dimethylethoxycarbonyl)-Nꢀ-(thiazol-2-yl)orni-
thine 1,1-dimethylethyl ester (20). Compound 13
(350 mg, 1.0 mmol) was stirred with chloroacetaldehyde
(100 mg, 1.3 mmol) in THF (5 mL) for 2 h. Evaporation
and chromatography (EtOAc:hexane, 1:1) gave 20
Nꢀ-(4,5-Dihydrooxazol-2-yl)-Nꢁ-(1,1-dimethylethoxy-
carbonyl)ornithine 1,1-dimethylethyl ester (25). Com-
pound 24 (516 mg, 3.6 mmol) was heated under re¯ux
with KF on alumina (40%, 516 mg, 3.6 mmol) in MeCN
(10 mL) for 16 h. The mixture was cooled and ®ltered
(Celite1). Evaporation and chromatography (EtOAc:
hexane:Et3N, 20:40:3) gave 25 (390 mg, 83%) as a col-
(230 mg, 61%) as a white solid: mp 69±71ꢀC; H NMR
1
d 1.44 (9H, s, But), 1.45 (9H, s, But), 1.71±1.91 (4H, m,
b,g-H4), 3.36 (2H, m, d-H2), 4.25 (1H, m, a-H), 5.18
(1H, d, J=7 Hz, a-NH), 5.72±5.83 (1H, br, d-NH), 6.48
(1H, d, J=3.7 Hz, thiazole 4-H), 7.11 (1H, d, J=3.7 Hz,
thiazole 5-H); 13C NMR d 24.04, 28.00, 28.68, 31.60,
45.47, 53.51, 79.81, 82.15, 106.38, 139.10, 155.47,
170.32, 171.67; MS (FAB) m/z 743 (2M+H), 372.1966
(M+H) (C17H30N3O4S requires 372.1957).
1
ourless oil: H NMR d 1.45 (9H, s, But), 1.46 (9H, s,
But), 1.55±1.90 (4H, m, b,g-H4), 3.20 (2H, m, d-H2),
3.75 (2H, t, J=7 Hz, oxazole 4-H2), 4.17 (1H, m, a-H),

S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
1795
1
4.25 (2H, t, J=7.1 Hz, oxazole 5-H2), 4.43 (1H, m, d-
NH), 5.32 (1H, d, J=7.9 Hz, a-NH); 13C NMR d 25.68,
28.32, 28.90, 30.34, 42.63, 52.45, 53.69, 67.85, 79.57,
81.84, 155.50, 161.47, 171.85; MS (FAB) m/z 358
(M+H).
H, 5.73; N, 8.38%; H NMR ((CD3)2SO) d 1.70±1.91
(4H, m, 3,4-H4), 3.13±3.20 (2H, m, 5-H2), 3.17 (2H, t,
J=7 Hz, thiazole 4-H2), 3.95 (1H, m, 2-H), 4.16 (2H, t,
J=7 Hz, thiazole 5-H2), 8.42±8.60 (3H, br, N+H3). 13
NMR ((CD3)2SO) d 25.02, 28.98, 31.86, 31.97, 51.50,
62.10, 168.78, 170.83; MS (FAB) m/z 235 (M+H).
C
Nꢀ-(4,5-Dihydrooxazol-2-yl)ornithine hydrochloride (26).
Compound 25 was treated with HCl, as for the syn-
thesis of 21 except that the solvent was CH2Cl2, to give
1,1-Dimethylethyl S-2-(1,1-dimethylethoxycarbonyl-
amino)-5-(imidazol-2-ylthio)pentanoate (33). Imidazole-
2-thiol (492 mg, 4.9 mmol) was stirred with NaHCO3
(412 mg, 4.9 mmol) in MeOH (10 mL) for 30 min. The
solvent was evaporated. The residual Na salt was sus-
pended in dry EtOH (5 mL) and was cooled to 0ꢀC.
Compound 3012 (750 mg, 2.1 mmol) was added and the
mixture was stirred at 20ꢀC for 6 h. Water (10 mL) was
added and the mixture was extracted with EtOAc
(3Â10 mL). Washing (water, brine), drying, evaporation
and chromatography (EtOAc) gave 33 (700 mg, 89%) as
1
26 (67%) as a highly hygroscopic white solid: H NMR
(D2O) d 1.62±1.98 (4H, m, b,g-H4), 3.58 (2H, m, 5-H2),
3.75 (2H, t, J=7 Hz, oxazole 4-H2), 4.00 (1H, m, a-H),
4.23 (2H, t, J=7 Hz, oxazole 5-H2); 13C NMR (D2O)
24.98, 27.69, 29.25, 34.15, 54.62, 66.23, 158.23, 175.85;
MS (FAB) m/z 202.1185 (M+H) (C8H16N3O3 requires
202.1193).
Nꢁ-(1,1-Dimethylethoxycarbonyl)-Nꢀ-(pyrimidin-2-yl)-
ornithine (28). BocOrnOH 2720 (2.00 g, 8.6 mmol) was
heated under re¯ux with 2-chloropyrimidine (500 mg,
4.3 mmol) and Et3N (430 mg, 4.2 mmol) in MeOH
(60 mL) for 3 days. Evaporation and chromatography
(acetone:MeOH, 1:1) gave 28 (347 mg, 26%) as a white
a colourless oil: H NMR d 1.44 (9H, s, But), 1.45 (9H,
1
s, But), 1.61±2.00 (4H, m, 3,4-H4), 3.09 (2H, m, 5-H2),
4.15 (1H, m, 2-H), 5.25 (1H, br, NH), 7.20 (2H, s, imi-
dazole 4,5-H2); MS (FAB) m/z 372 (M+H).
solid: mp 164±166ꢀC; H NMR ((CD3)2SO) d 1.35 (9H,
S-2-Amino-5-(imidazol-2-ylthio)pentanoic acid dihydro-
chloride (34). Compound 33 was treated with HCl, as
for the synthesis of 21, to give 34 (63%) as a highly
1
s, But), 1.87 (4H, m, 3,4-H4), 3.35 (2H, m, 5-H2), 4.35
(1H, m, 2-H), 7.03 (1H, t, J=7.9 Hz, pyrimidine 5-H),
8.30 (2 H, d, J=7.9 Hz, pyrimidine 4,6-H2); MS (FAB)
m/z 312 (M+H).
1
hygroscopic white solid: H NMR (D2O) d 1.40±1.58
(2H, m, 4-H2), 1.72±1.88 (2H, m, 3-H2), 2.83 (2H, t,
J=7 Hz, 5-H2), 3.55 (1H, t, J=6 Hz, 2-H), 7.01 (2H, s,
imidazole 4,5-H2); 13C NMR (D2O) d 25.87, 29.84,
34.70, 55.18, 124.81, 140.15, 175.30; MS (FAB) m/z
216.0820 (M+H) (C8H14N3O2S requires 216.0867).
Nꢀ-(Pyrimidin-2-yl)ornithine (29). Compound 28 (710mg,
2.2 mmol) was stirred with hydrochloric acid (4 M,
15 mL) and EtOAc (25 mL) for 4 h. Evaporation and
chromatography (MeOH:35% aq NH3, 49:1) gave 29
(300 mg, 63%) as a white hygroscopic solid: mp 195±
Methyl N-(4-(1,1-dimethylethoxycarbonylamino)butyl)-
dithiocarbamate (36). CS2 (1.37 g, 18 mmol) was added
to 3523 (1.7 g, 9 mmol) in THF (10 mL) at 0ꢀC, followed
by Et3N (900 mg, 9 mmol). The mixture was stirred at
0ꢀC for 2 h. MeI (1.3 g, 9 mmol) was added and the
reaction was stirred at 0ꢀC for 10 min and at 20ꢀC for
16 h. Filtration, evaporation and chromatography
(CH2Cl2:EtOAc, 1:1) gave 36 (2.0 g, 80%) as colourless
oil: Found: C, 47.70; H, 7.93; N, 9.95. C11H22N2O2S2
197ꢀC; H NMR (D2O) d 1.75±2.00 (4H, m, 3,4-H4),
1
3.04 (2H, t, J=7.1 Hz, 5-H2), 4.27 (1H, m, 2-H), 6.77
(1H, t, J=5.0 Hz, pyrimidine 5-H), 8.32 (2H, d, J=
5.0 Hz, pyrimidine 4,6-H2); MS (FAB) m/z 211.1183
(M+H) (C9H15N4O2 requires 211.1195).
1,1-Dimethylethyl S-5-(4,5-dihydrothiazol-2-ylthio)-2-
(1,1-dimethylethoxycarbonylamino)pentanoate (31). 4,5-
Dihydrothiazole-2-thiol (500 mg, 4.3 mmol) was stirred
with NaHCO3 (354 mg, 4.3 mmol) in MeOH (10 mL) for
30 min. The solvent was evaporated. The residual Na
salt was suspended in dry EtOH (5 mL) and was cooled
to 0ꢀC. Compound 3012 (400 mg, 1.3 mmol) was added
and the mixture was stirred at 20ꢀC for 6 h. Water
(10 mL) was added and the mixture was extracted with
EtOAc (3Â10 mL). Washing (water, brine), drying,
evaporation and chromatography (EtOAc:hexane, 1:2)
1
requires C, 47.45; H, 7.96; N, 10.05%; H NMR d 1.94
(9H, s, But), 1.52±176 (4H, m, 2,3-H4), 2.62 (3H, s,
SMe), 3.15 (2H, m, 1-H2), 3.55 (2H, m, 4-H2), 4.96 (1H,
br, NH), 8.15 (1H, br, NH); 13C NMR d 25.56, 27.31,
28.12, 39.67, 46.59, 78.95, 132.45, 156.06; MS (CI) m/z
279 (M+H).
Methyl N-(4-(aminobutyl)dithiocarbamate hydrochloride
(37). Compound 36 was treated with HCl, as for the
synthesis of 21, to give 37 (71%) as a white solid: mp
119±121ꢀC; H NMR (D2O) d 1.62±1.80 (4H, m, 2,3-
H4), 2.58 (3H, s, SMe), 3.04 (2H, t, J=6 Hz, 4-H2), 3.78
(2H, t, J=6 Hz, 1-H2); MS (FAB) m/z 179.0696
(M+H) (C6H15N2S2 requires 179.0677).
1
gave 31 (230 mg, 44%) as a colourless oil: H NMR d
1.44 (9H, s, But), 1.46 (9H, s, But), 1.65±1.96 (4H, m,
3,4-H4), 3.11 (1H, dt, J=11, 7 Hz, 5-H), 3.14 (1H, dt,
J=11, 7 Hz, 5-H), 3.40 (2H, t, J=8.0 Hz, thiazole 5-H2),
4.20 (1H, m, 2-H), 4.22 (2H, t, J=8.0 Hz, thiazole 4-H2),
5.31 (1H, d, J=8 Hz, a-NH); m/z (FAB) 390 (M+H).
1
1,1-Dimethylethoxycarbonyl N-(4-(thioureido)butyl)-
carbamate (38). NH3 was passed through 36 (780 mg,
2.8 mmol) in MeOH (10 mL) at 0ꢀC for 10 min. The
solution was stirred at 0ꢀC for 2 h and at 20ꢀC for 16 h.
Evaporation and chromatography (CH2Cl2:EtOAc:
Et2O, 1:2:2) gave 38 (350 mg, 51%) as a white solid: mp
S-2-Amino-5-(4,5-dihydrothiazol-2-ylthio)pentanoic acid
dihydrochloride (32). Compound 31 was treated with
HCl, as for the synthesis of 21, to give 32 (63%) as a
white solid: mp 160±162ꢀC; Found: C, 28.70; H, 5.30;
N, 8.13. C8H14N2O2S2.2HCl. 2H2O requires C, 28.75;

1796
S. Ulhaq et al. / Bioorg. Med. Chem. 7 (1999) 1787±1796
157±159ꢀC; 1H NMR d 1.43 (9H, s, But), 1.60±1.71 (4H,
m, 2,3-H4), 3.20 (2H, m, 4-H2), 3.52 (2H, m, 1-H2), 4.93
(1H, br, NH), 6.51 (1H, br, SH), 7.07 (1H, br, NH); 13C
NMR d 27.81, 28.39, 29.63, 43.35, 45.05, 79.71, 128.77,
156.77; MS (CI) m/z 248 (M+H).
References
1. Moncada, S.; Palmer, R. M. J.; Higgs, E. A. Pharmacol.
Rev. 1991, 43, 109.
2. Forstermann, U.; Schmidt, H. H. H. W.; Pollock, J. S;
Sheng, H.; Mitchell, J. A.; Warner, T. D.; Nakane, M.;
Murad, F. Biochem. Pharmacol. 1991, 42, 1849.
3. Rees, D. D.; Palmer, R. M. J.; Schultz, R.; Hodson, H. F.;
Moncada, S. Br. J. Pharmacol. 1990, 101, 746.
4. Moore, P. K.; Al-Swayeh, O. A.; Chong, N. W. S.; Evans,
R. A; Gibson, A. Br. J. Pharmacol. 1990, 99, 408.
5. Grith, O. W.; Kilbourne, R. G. Methods Enzymol. 1996,
268, 375.
6. Marletta, M. A. J. Med. Chem. 1994, 37, 1899.
7. Garvey, E. P.; Oplinger, J. A.; Fur®ne, E. S.; Ki€, R. J.;
Laszlo, F.; Whittle, B. J. R.; Knowles, R. G. J. Biol. Chem.
1997, 272, 4959.
N-(4-Aminobutyl)thiourea dihydrochloride (39). Com-
pound 38 was treated with HCl, as for the synthesis of
21, to give 39 (47%) as a highly hygroscopic white solid:
1H NMR ((CD3)2SO) d 1.49 (2H, quintet, J=6.6 Hz, 2-
H2), 1.72 (2H, quintet, J=6.5 Hz, 3-H2), 2.78 (2H, m, 1-
H2), 3.36 (2H, t, J=6.5 Hz, 4-H2), 4.18 (1H, br, NH),
8.00 (1H, br, NH), 8.07 (3H, s, N+H3), 9.50 (1H, br)
and 10.05 (1H, br) (NH2); 13C NMR ((CD3)2SO) d
29.06, 29.81, 45.85, 45.97, 176.37; MS (FAB) m/z
148.0904 (M+H) (C5H14N3S requires 148.0908).
8. Thomsen, L. L.; Scott, J. M. J.; Topley, P.; Knowles, R. G.;
Keerie, A. J.; Frend, A. J. Cancer Res. 1997, 57, 3300.
9. Andrade, S. P.; Hart, I. R.; Piper, P. J. Br. J. Pharmacol.
1992, 107, 1092.
10. Wood, P. J.; Stratford, I. J.; Adams, G. E.; Szabo, C.;
Thiemermann, C.; Vane, J. R. Biochem. Biophys. Res. Com-
mun. 1993, 192, 505.
11. Wood, P. J.; Sansom, J. M.; Butler, S. A.; Stratford, I. J.;
Cole, S. M.; Szabo, C.; Thiemermann, C.; Adams, G. E. Can-
cer Res. 1994, 54, 6458.
12. Ulhaq, S.; Chinje, E. C.; Naylor, M. A.; Ja€ar, M.; Stratford,
I. J.; Threadgill, M. D. Bioorg. Med. Chem. 1999, 7, 2139.
13. Matthews, S. E.; Pouton, C. W.; Threadgill, M. D. New J.
Chem., submitted for publication.
4,5-Dihydro-2-(4-(1,1-dimethylethoxycarbonylamino)butyl-
amino)thiazole (40). Br(CH2)2Br (1.28 g, 6.8 mmol) was
boiled under re¯ux with 38 (500 mg, 2.0 mmol) and
K2CO3 (276 mg, 2.0 mmol) in THF (10 mL) for 16 h.
Filtration, evaporation and chromatography (EtOAc:
hexane, 1:1!EtOAc:CH2Cl2:Et3N, 10:4:1) gave 40
s, But), 1.55±1.90 (4H, m, butyl 2,3-H2), 3.15 (1H, t,
J=7 Hz, butyl 1-H2), 3.37 (2H, m, butyl 4-H2), 3.50
(2H, t, J=7 Hz, thiazole 5-H2), 3.99 (2H, t, J=7 Hz,
thiazole 4-H2), 4.72 (1H, br, NH); 13C NMR d 27.50,
28.40, 32.30, 41.13, 62.20, 79.19, 156.26, 158.74; MS
(CI) m/z 280 (M+H).
1
(200 mg, 40%) as a colourless oil: H NMR d 1.50 (9H,
14. Yajima, H.; Watanabe, H.; Okamoto, M. Chem. Pharm.
Bull. 1971, 19, 2185.
15. Clarke, D. R.; Waight, E. S. J. Chem. Soc. C 1971, 3743.
16. Frump, J. A. Chem. Rev. 1971, 71, 483.
17. Malen, C.; Desnos, M.; Laubie, M.; Poignant, J. C. Ger-
man Patent 2362754, 1974.
18. Peet, N. P.; Sunder, S. US Patent 4256755, 1981.
19. Wong, W. C.; Wang, D.; Forray, C.; Vaysse, P. J. J.;
Branchek, T. A.; Gluchowski, C. Bioorg. Med. Chem. Lett.
1994, 4, 2317.
20. Arold, H.; Haller, K. J. Prakt. Chem. 1969, 311, 3.
21. Olsen, R. K.; Ramasamy, K.; Emery, T. J. Org. Chem.
1984, 49, 3527.
22. Ulhaq, S.; Chinje, E. C.; Naylor, M. A.; Stratford, I. J.;
Threadgill, M. D. Anti-Cancer Drug Design 1997, 12, 61.
23. Hu, W. Q.; Hesse, M. Helv. Chim. Acta 1996, 79, 548.
24. Narayanan, K.; Spack, L.; McMillan, K.; Kilbourn, R. G.;
Hayward, M. A.; Masters, B. S. S.; Grith, O. W. J. Biol.
Chem. 1995, 270, 11103.
25. Feldman, P. L.; Chi, S.; Sennequier, N.; Stuehr, D. J.
Bioorg. Med. Chem. Lett. 1996, 6, 111.
26. Cowart, M.; Kowaluk, E. A.; Daanen, J. F.; Kohlhaas, K.
L.; Alexander, K. M.; Wagenaar, F. L.; Kerwin, J. F. J. Med.
Chem. 1998, 41, 2636.
2-(4-Aminobutylamino)-4,5-dihydrothiazole dihydro-
chloride (41). Compound 40 was treated with HCl, as
for the synthesis of 21, to give 41 as a white solid: mp
162±164ꢀC; H NMR ((CD3)2SO) d 1.55±1.68 (4H, m,
1
butyl 2,3-H4), 2.79 (2H, m, butyl 4-H2), 3.30 (2H, t,
J=6.5 Hz, butyl 1-H2), 3.56 (2H, t, J=7 Hz, thiazole 5-
H2), 3.89 (2H, t, J=7 Hz, thiazole 4-H2), 8.12 (3H, br,
N+H3), 10.40 (1H, br, NH); 13C NMR ((CD3)2SO) d
24.00, 24.76, 30.63, 38.78, 44.31, 48.64, 169.23; MS
(FAB) m/z 159.0943 (M+H) (C7H15N2S requires
159.0955).
NOS inhibition studies
Measurements of the inhibitory activity of the test
compounds against rat iNOS and rat nNOS and against
cNOS derived from H647 human cells were made as
described previously.12 The results are shown in Table 1 as
the mean of triplicate experiments‹standard deviation.
27. Garvey, E. P.; Oplinger, J. A.; Tanoury, G. J.; Sherman, P.
A.; Fowler, M.; Marshall, S.; Harmon, M. F.; Paith, J. E.;
Fur®ne, E. S. J. Biol. Chem. 1994, 269, 26669.
28. Kerwin, J. F.; Lancaster, J. R.; Feldman, P. L. J. Med.
Chem. 1995, 38, 4343.
29. Feldman, P. L. Tetrahedron Lett. 1991, 32, 875.
30. Narayanan, K.; Grith, O. W. J. Med. Chem. 1994, 37,
885.
Acknowledgements
The authors thank Mr. R. R. Hartell and Mr. D. J.
Wood for the NMR spectra and Mr C. Cryer for the
mass spectra. S.U. thanks the Medical Research Council
for a studentship.