Development and characterization of glutamyl-protected N-hydroxyguanidines as reno-active nitric oxide donor drugs with therapeutic potential in acute renal failure

Article abstract of DOI:10.1021/jm400146r

Acute renal failure (ARF) has high mortality and no effective treatment. Nitric oxide (NO) delivery represents a credible means of preventing the damaging effects of vasoconstriction, central to ARF, but design of drugs with the necessary renoselectivity is challenging. Here, we developed N-hydroxyguanidine NO donor drugs that were protected against spontaneous NO release by linkage to glutamyl adducts that could be cleaved by γ-glutamyl transpeptidase (γ-GT), found predominantly in renal tissue. Parent NO donor drug activity was optimized in advance of glutamyl adduct prodrug design. A lead compound that was a suitable substrate for γ-GT-mediated deprotection was identified. Metabolism of this prodrug to the active parent compound was confirmed in rat kidney homogenates, and the prodrug was shown to be an active vasodilator in rat isolated perfused kidneys (EC₅₀ ～50 μM). The data confirm that glutamate protection of N-hydroxyguanidines is an approach that might hold promise in ARF.

Full text of DOI:10.1021/jm400146r

Article
pubs.acs.org/jmc
Development and Characterization of Glutamyl-Protected
N‑Hydroxyguanidines as Reno-Active Nitric Oxide Donor Drugs with
Therapeutic Potential in Acute Renal Failure
Qingzhi Zhang,‡ Philip Milliken,§ Agnieszka Kulczynska,‡ Alex M. Z. Slawin,‡ Adele Gordon,§
Nicholas S. Kirkby,§ David J. Webb,§ Nigel P. Botting,‡,† and Ian L. Megson∥,*
‡EASTChem, School of Chemistry and Centre for Biomolecular Sciences, The University of St. Andrews, North Haugh, St. Andrews
KY16 9ST, U.K.
§Centre for Cardiovascular Science, The Queen’s Medical Research Institute, The University of Edinburgh, 47 Little France Crescent,
Edinburgh EH16 4TJ, U.K.
∥Free Radical Research Facility, Department of Diabetes and Cardiovascular Science, The University of The Highlands & Islands,
Centre for Health Science, Old Perth Road, Inverness IV2 3JH, U.K.
S
* Supporting Information
ABSTRACT: Acute renal failure (ARF) has high mortality and no
effective treatment. Nitric oxide (NO) delivery represents a credible
means of preventing the damaging effects of vasoconstriction, central
to ARF, but design of drugs with the necessary renoselectivity is
challenging. Here, we developed N-hydroxyguanidine NO donor
drugs that were protected against spontaneous NO release by linkage
to glutamyl adducts that could be cleaved by γ-glutamyl transpeptidase
(γ-GT), found predominantly in renal tissue. Parent NO donor drug
activity was optimized in advance of glutamyl adduct prodrug design.
A lead compound that was a suitable substrate for γ-GT-mediated
deprotection was identified. Metabolism of this prodrug to the active
parent compound was confirmed in rat kidney homogenates, and the
prodrug was shown to be an active vasodilator in rat isolated perfused
kidneys (EC50 ∼50 μM). The data confirm that glutamate protection
of N-hydroxyguanidines is an approach that might hold promise in ARF.
aftermath of ischemia-reperfusion, exacerbated by reactive
oxygen species (ROS) production from infiltrating inflamma-
tory cells, is key to the etiology of ARF.20 ROS have multiple
deleterious effects, but rapid reaction with the endogenous
vasodilator, antithrombotic, and anti-inflammatory agent, nitric
oxide (NO),21,22 is a crucial component, not only because of
the loss of the highly beneficial effects of NO20,23,24 but also
through generation of nephrotoxic peroxynitrite (ONOO−) as
the product of the reaction.25 Endogenous NO is synthesized
from L-arginine (L-Arg) and oxygen by NO synthase (NOS)
enzymes and is critical in regulating renal perfusion and
glomerular hemodynamics under basal conditions.26 NO is
protective against ARF and increased NO synthesis from
endothelial and neuronal NOS (eNOS, nNOS) is beneficial in
ischemia:27 NOS inhibition decreases NO production, further
reducing renal function in ischemic ARF in rats.28,29 These
effects are reversed by both L-Arg28 and tetrahydrobiopterin
(NOS cofactor);24 L-Arg also abolishes the associated increase
in ROS.20 NO donor pretreatment has been shown to enhance
INTRODUCTION
■
Acute renal failure (ARF) affects ∼5% of all hospitalized
patients and 10−30% of intensive care patients.1 Prognosis is
poor, with a mortality rate exceeding 50%2 and only 60−70% of
survivors regaining full independence.3 Ischemia-reperfusion
injury is the most common cause of ARF, but sepsis,
hypotension, and nephrotoxic drugs are also important.4 Aortic
cross-clamping during cardiopulmonary bypass surgery is a
major cause of ischemic ARF, although acute changes in cardiac
output due to myocardial infarction or ventricular fibrillation
can also underpin poor renal perfusion, leading to organ
failure.5 Renal replacement therapy is the only recognized
treatment for ARF;6 pharmacological interventions, including
low dose dopamine,7−9 dopexamine (synthetic dopamine
analogue),10−12 mannitol/loop diuretics,13−15 calcium entry
blockers,16 anaritide (atrial natriuretic peptide),17 and N-
acetylcysteine,18,19 have all failed to show reproducible clinical
benefits. This highlights the need for effective therapies to treat
ARF and to reduce the reliance on costly renal replacement
therapy, and the challenges linked to achieving this goal.
Oxidative stress due to mitochondrial dysfunction and
xanthine oxidase (XO; EC 1.17.3.2) activation in the immediate
Received: January 29, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Structures and Associated Codes for Key NOHA and Synthesized Test Compounds
renal function in ischemic models30 and to suppress expression
of endothelin-1,31 a potent vasoconstrictor that is increased in
postischemic kidneys.32 Delivery of exogenous NO may,
therefore, offer several levels of protection in ARF.
Progress toward novel NO donor drugs has been
disappointingly slow, primarily on account of the difficulty in
designing drugs that reproduce the site-specific effects of
endogenous NO, which is driven by the highly regulated nature
of eNOS and the limited reach of NO from its source of
generation.33 The N-hydroxyguanidines (NHGs) include the
intermediate in NOS-mediated synthesis of NO, Nω-hydroxy-L-
arginine (NOHA), and might offer a fresh approach to tissue-
selective NO donor drugs.34 Some NHGs are known to be
metabolized by NOS,35 as well as a range of other abundant
enzymes, such as cytochrome P450s,36 peroxidases,37 and
XO,38 offering the potential to generate NO and possibly to
reduce enzyme-derived ROS generation,38 in a nonselective
manner throughout the body. NHGs have two accessible
functional groups (−OH and −NH2), which allow manipu-
lation into prodrugs through esterification or amidation. With a
view to imparting renoselectivity to NHGs, we sought to
design, synthesize, and test glutamyl-protected drugs, a strategy
B
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Reagents and conditions: (i) BrCN, NaOAc, MeOH, 0 °C to rt, 2 h; (ii) NH2OH·HCl, KOH (cat.), MeOH, reflux, 5 h.
that has previously been employed in protecting the active
moiety of L-DOPA for use in ARF.39−42 This approach exploits
the relatively high expression of γ-glutamyl transpeptidase (γ-
GT; EC 2.3.2.2) in the kidney (5−10 times higher than in liver
or pancreas).43,44 The enzyme acts to cleave the glutamate
adduct, with a view to promoting renoselective activation of the
protected compound and avoiding renal “steal” that would
accompany nonselective vasodilatation. The ideal glutamyl-
protected product, therefore, would be biologically inert, except
in the presence of renal tissue, where it would be activated to
cause tissue-specific, NO-mediated vasodilatation that was not
necessarily dependent on a functional endothelium.
endothelium-dependent vasodilator, acetylcholine (ACh), was
used to test the functional integrity of the endothelium and as
an indicator of the effectiveness of NOS inhibition by L-NAME
(Figure 2). NOHA (100 μM) was found to be an effective
vasodilator that was only partially inhibited by L-NAME.
Product 1a (100 μM), on the other hand, had weak vasodilator
properties that were abolished by L-NAME.
Prototype Prodrugs. To impart potential for renoselec-
tivity to 1a, a γ-glutamyl group was conjugated onto the
guanidino N-OH to generate a protected molecule with the
potential to be a substrate for γ-GT. The glutamate moiety was
derived from commercially available difluoren-9-ylmethyl
protected glutamic acid (Fmoc-Glu-OFm), the protecting
group of which can be removed easily by organic base under
mild conditions. To make the side-chain of 1a match the
protecting protocol, the terminal carboxylic group was masked
with fluoren-9-ylmethyl before it was reacted with cyanogen
bromide and, subsequently, with NH2OH·HCl, to generate the
NHG, 1b (Scheme 2).
RESULTS
■
Proof of Principle: Synthesis of Biologically Active
Parent and Glutamyl-Protected NHG prototypes. Proto-
type NHG Parent Drug and Equivalent Urea (Negative
Control). A prototype parent drug (1a; Table 1) was
synthesized. 1a had similarities to NOHA, but without the α-
amino group, which was considered to present problems for
downstream modification of the molecule in prodrug develop-
ment; other deviations from NOHA were the addition of an N-
methyl group and a shorter carbon chain length. Synthesis
adopted the common two-step reaction45−50 involving
cyanization of N-methyl-4-aminobutanoic acid with cyanogen
bromide, followed by nucleophilic addition of hydroxylamine to
the resultant cyanamide (Scheme 1). This compound exists as a
zwitterion and is soluble in water.
Reaction of 1b with Fmoc-Glu-OFm in the presence of N-(3-
dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride
(EDC) initially formed the linear O-glutamate 2, as confirmed
1
by TLC and H NMR spectra. However, with time and during
work up and purification, it gradually transformed into the
heterocyclic derivative 3 with the loss of a molecule of water:
fractions collected as the linear product 2 from column
chromatography became a mixture of products 2 and 3 after
solvent removal. This cyclization process continued during the
process of Fm and Fmoc removal under basic conditions and
was accompanied by expulsion of pyroglutamic acid and
regeneration of the parent NHG (Scheme 3).
An alternative approach was to introduce the glutamyl group
into the amino moiety of the parent compound (Scheme 3).
Although this is not the site for NO generation from the NHG,
it was considered likely that the bulky glutamyl group could
sterically hinder enzyme-mediated release of NO. For this
purpose, Boc protection was employed because the N-OH
group is more nucleophilic and reactive than −NH2. This
resulted in a stable intermediate 5, but the same 1,2,4-
oxadiazole 3 formed slowly again once the O-Boc group was
removed in CF3CO2H/DCM. The process continued through-
out deprotection of Fm/Fmoc in Et3N/DCM and during
workup and purification, although the formation of pyrogluta-
mic acid was found to be much slower in this case because the
amide bond is stronger than that of the ester.
Because of this complication, we obtained 1,2,4-oxadiazoles 3
in ∼50% yield and pure 4 (Table 1). The expected product 7
could not be isolated on account of its rapid decomposition;
compound 8 (Table 1) was highly hygroscopic and found to
readily degrade to 4 as monitored by NMR spectroscopy
(Scheme 3).
Bioactivation of Prototype Prodrugs by γ-GT. The
cyclic glutamyl-protected adduct of 1a, compound 4 (100 μM),
failed to induce vasodilatation at this concentration, irrespective
of preincubation with γ-GT (Figure 3). Prodrug 8 was tested
The equivalent urea (1a(u)) was synthesized via the same
intermediate cyanamide but followed by acidic hydrolysis.
[Nitrite + Nitrate] (NOx) Generation and Vasodilator
Activity of NHG Parent Drug Prototypes. The endogenous
NHG, NOHA (positive control; 100 μM), was found to
generate NOx (a surrogate marker for NO) in substantial
quantities (5−11 μM) after incubation with xanthine (X)/XO,
and in rat liver, renal cortex, and renal medullar tissue extracts
(1 h, 37 °C). The specific XO inhibitor, allopurinol (100 μM),
inhibited X/XO-mediated NOx generation, while the NOS
inhibitor, Nω-nitro-L-arginine methyl ester (L-NAME; 200 μM)
abolished the NOx generation associated with liver extracts and
substantially inhibited that in both cortical and medullary renal
tissue (Figure 1A). γ-GT did not induce NOx generation from
NOHA (Figure 1A).
Product 1a (100 μM) was found to have a similar NOx-
generating profile to NOHA in the presence of X/XO, but the
NOx yield in tissue extracts (1−4 μM) was lower than that for
NOHA (Figure 1B).
X/XO failed to induce NOx generation from the urea
counterpart (1a(u)) of 1a, while low NOx concentrations (∼1
μM) measured in tissue extracts treated with 1a(u) were not
inhibited by L-NAME and were not different from those found
in tissue extracts alone (Figure 1C).
To assess the vasodilator properties of product 1a in
comparison to NOHA, precontracted, endothelium-intact rat
aortic rings ( L-NAME, 200 μM) were employed. The
C
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 2. Vasodilator activity of ACh, NOHA and product 1a in
endothelium-intact, precontracted rings of rat aorta in the presence
and absence of the NOS inhibitor, L-NAME (200 μM). ACh and
product 1a-mediated vasodilatation was deemed to be almost
exclusively NOS-dependent, while that induced by NOHA was
considered to be partially NOS-independent. *P < 0.05, ***P <
0.001, paired Student’s t-test L-NAME.
for vasodilatation and found to be inactive in the absence of γ-
GT but produced a modest (∼18%) relaxation after
preincubation with γ-GT that was abolished by the NO
scavenger, oxyhemoglobin (Hb; 10 μM).
Design and Synthesis of Parent NHG Drug Library. On
the basis of the preliminary results from the proof of principle
study, 15 NHGs were synthesized (Table 1: 12 novel, 3 known
1c−s) bearing different alkyl and aryl side-chains, following the
same two-step procedure45−50 from commercial or in-house
amines,51 with a view to optimizing NO-donor parent drug
efficacy and establishing a structure−activity relationship.
All NHGs have a typical resonance at ca. δC 158 ppm for the
guanidino carbon in the 13C NMR spectrum. Using DMSO-d6
as solvent, all 1H NMR spectra for the N-monosubstituted final
products showed four distinctive proton signals for the
protonated hydroxyguanidine group (ca. δH 7.7 for NH2, δH
8.0 for alkyl or aryl NH, δH for N′H, δH 10.4 for OH). The
assignments were based on integration/coupling pattern
(Supporting Information (SI) Figure S1) and 2D 1H−13C
HMBC spectrum (Figure S2) in SI represented by 1g.
The synthetic NHG yields (Table 2) were affected by
repeated recrystallization steps necessary to achieve sufficiently
high purity of the product and satisfactory microanalysis results.
In the case of 1s, even repeated recrystallization failed to
generate product of sufficient purity. Instead, Boc protection
followed by column chromatography and removal of the
protecting group by 4N HCl in dioxane was used to generate
pure product.
The vasodilator effects of the synthesized compounds were
tested in vitro using endothelium-intact rat isolated aortic rings.
The principal aim of the screen was to find compounds that
were not entirely dependent on NOS for activation but for
which the majority of the vasodilator activity was nevertheless
dependent on the NO:sGC pathway. Experiments were
therefore conducted in the presence of the NOS inhibitor, L-
NAME (200 μM), and the reversibility of responses was
assessed using the sGC inhibitor, 1H-[1,2,4]oxadiazolo[4,3-
a]quinoxalin-1-one (ODQ; 50 μM). Promising compounds
were tested further using Hb (10 μM), an acknowledged
scavenger of NO, and L-Arg (100 μM; approximate plasma
levels), which might compete with the drug under endogenous
conditions, not only at the active site for NOS but also at the
system y+/CAT amino acid transporters,52,53 should NHGs
Figure 1. Release of NOx from (A) the endogenous NHG, NOHA (n
= 4−5), (B) the novel NHG, 1a (n = 4−5), and (C) the
corresponding urea 1a(u) (negative control, n = 4−5; all 100 μM)
in the presence of γ-GT (100 mU/5 mM gly/gly), xanthine/xanthine
oxidase (X/XO; 100 μM/100 mU), and rat tissue homogenates.
Parallel experiments were conducted in the presence of specific
inhibitors (5-sulfosalicylate (100 μM) for γ-GT, allopurinol (100 μM)
for XO, and the NOS inhibitor, L-NAME (200 μM) for the tissue
homogenates) to help confirm enzyme specificity. *P < 0.05, **P <
0.01; paired t-test in presence vs absence of specific inhibitor.
D
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 2
a
Reagents and conditions: (i) FmOH, TsOH, reflux; (ii) BrCN, NaOAc, MeOH; (iii) NH2OH·HCl, K2CO3, MeOH.
a
Scheme 3
a
Reagents and conditions: (i) Fmoc-Glu-OFm, HOBt, EDC, DCM, DMF, rt, 3 d; (ii) Et3N, DCM; (iii) (Boc)2O, DMF, H2O, K2CO3, 0 °C, 2 h,
100%; (iv) CF3COOH, DCM, 0 to rt, 2 h.
contribution of NOS to vasodilatation. The conventional
endothelium-dependent vasodilator, ACh, was used to both
confirm the function of the endothelium in test rings and to
validate the model with respect to the inhibitors used (Figure
4A). The results are summarized in Table 2 and shown in their
entirety for selected test compounds in Figure 4B,C.
The results revealed that drug efficacy was improved by
substitution of the terminal carboxylic acid moiety of product
1a with a methyl group (1c) or an aromatic ring (1e) and that
the length of the carbon chain backbone of the molecule (1d−
g) also affected efficacy, with a chain length of 2C (1e)
appearing optimal for both vasodilator activity and dependence
on sGC. Substitution on the aromatic ring with F (1h), Cl (1i),
or Br (1j) reduced the potency of the NHG with respect to
vasodilatation and also reduced the dependence on sGC. A
similar effect was seen with −CH3 substitution (1k) at this
position and, while −OH (1m) substitution apparently
increased vasodilator activity, there was a concomitant loss in
sGC-dependence. Substitution with a methoxy group at three
different positions on the aromatic ring (1n−q) had a
detrimental effect in terms of dependence on sGC. Substitution
with −SO2NH2 (1s) at the same position abolished the
vasodilator effects of the compound. Finally, addition of a
second aromatic ring (1r) also reduced efficacy.
Figure 3. Vasodilator activity of two prototype prodrugs (4, 8) with
and without preincubation with γ-GT. Prodrug 4 failed to cause
vasodilatation, irrespective of the presence of γ-GT. However, prodrug
8 was induced to cause modest vasodilatation after incubation with γ-
GT (100 mU/mL, 60 min, 37 °C), and the effect was abolished by the
NO scavenger, Hb (10 μM). *P < 0.05, **P < 0.01 (Dunnet’s post-
test following one-way ANOVA for product 8 γ-GT and [γ-GT +
Hb]; n = 5−9). NS: no statistical difference between response to 4
γ-GT (n = 5; paired Student’s t test).
Figure 4 illustrates the extended testing that was conducted
utilize these transporters to access cells. Parallel experiments in
the absence of L-NAME confirmed the extent of the
on 1e as the lead parent drug identified from the screen; the
E
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 2. NHG Parent Compound Synthetic Yield and Bioactivity in Rat Aortic Rings (Shaded Box Indicates Lead Selection)
endothelium-dependent vasodilator, ACh, and NOHA were
tested in parallel for reference purposes. ACh-mediated
vasodilatation in precontracted, endothelium-intact rat aortic
rings was abolished by L-NAME, Hb, and ODQ. However, L-
Arg did not significantly affect the vasodilator properties of
ACh (Figure 4A). Treatment of aortic rings with NOHA
induced concentration-dependent vasodilatation that was
abolished by ODQ and substantially inhibited by both L-
NAME and Hb (Figure 4B). The effect was also inhibited by
coincubation with L-Arg. Product 1e had weaker vasodilator
activity than NOHA in the absence of the NO:sGC pathway
inhibitors and L-Arg, while its effects were all but abolished by
ODQ. However, the inhibitory effects of L-NAME and Hb
were substantially weaker than they were for NOHA, while
responses to product 1e were slightly augmented by L-Arg. The
crystal structure of the chosen lead parent drug, 1e, is shown in
Figure 5.
cyclization were 2-fold (Scheme 3): the N-OH or −NH2
groups are ideally positioned with respect to the γ-glutamyl
for generation of a highly stable five-membered aromatic ring
and the tendency for the formation of pyroglutamic acid due to
the instability of the N-hydroxyguanido ester/amide. A detailed
description of a range of strategies used to disrupt the potential
for cyclization and the evolution of the prodrugs has been
reported elsewhere.54
Dipeptidyl Protected NHG Prodrug. A dipeptidyl
prodrug O,N-bis(GABAGlu)-NHG (prodrug 9; Table 1) was
successfully synthesized using two equivalents of dipeptide to
protect both the N-OH and the −NH2 groups of the NHG54 in
order to deprive these groups of the opportunity to attack the
nearby CO group, causing cyclization (Scheme 4). A γ-
aminobutyric acid (GABA) fragment55 was employed as a
simplified version of Glu, which as evidenced above, has a
strong suicidal tendency through formation of lactam. The
glutamyl residue itself serves as the recognizable residue for γ-
GT activity. The amidic bond between the two amino acids was
considered to be as stable as a normal peptide bond and able to
survive the workup. Upon incubation with γ-GT, it was
envisaged that the enzyme would cleave the terminal glutamyl
group; the newly exposed linkers on both the O and N limbs
Design and Synthesis of NHG Prodrug, Based on
Parent Lead. Having selected product 1e as the lead parent
drug, experimental procedures were initiated to synthesize a
range of glutamyl protected prodrugs. The task was colored by
the issues of cyclization that had come to light in the proof of
principle phase of the study. It was deduced that the reasons for
F
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
NHG Prodrugs with an Aromatic Linkage. The inability
of γ-GT to activate the GABA-linked prodrug (9) prompted
the application of another self-immolative spacer, the amino-
benzyl ether (compound 10 in Scheme 5; Table 1). γ-Glutamyl
anilines, especially p-nitro glutamylanilide, are well-known
substrates for γ-GT.56−58 Prodrugs bearing a glutamylaniline
moiety would be more likely to represent suitable substrates for
γ-GT. In addition, similar linkage has been employed previously
in anticancer prodrugs design.59−69
Prodrug 10 was successfully synthesized from p-glutamyla-
minobenzyloxyamine and amino(phenethyliminio)-
methanesulfonate in modest yield after removal of the
protecting group. Details of the synthetic procedure for 10
are published elsewhere.54 Prodrug 10 induced modest
relaxation of rat aortic rings that was enhanced after incubation
with γ-GT (Figure 6A). Responses in both the presence and
absence of γ-GT showed substantial reversal upon addition of
ODQ.
On the back of these findings, the vasodilator activity of
prodrug 10 was tested in the isolated perfused rat kidney,
alongside the parent compound, 1e; prodrug 9 was included as
a negative control on the basis that it did not induce relaxation
in rat aortic rings, irrespective of the presence of γ-GT. In
keeping with the findings in aortic rings, prodrug 9 failed to
induce vasodilatation in isolated perfused kidneys, even at
concentrations as high as 250 μM (Figure 6B). Prodrug 10, on
the other hand, induced concentration-dependent vasodilata-
tion in the kidney (EC50 ∼50 μM). Renal metabolism profiles
for prodrug 10 in rat kidney homogenates illustrate that
compound 1e is generated from prodrug 10 in a time and
concentration-dependent manner over a 45 min incubation
period (Figure 6C).
DISCUSSION AND CONCLUSIONS
■
This research program adopted a multidisciplinary approach to
first prove principle and then design and screen for a glutamyl-
protected NHG prodrug that might ultimately offer renose-
lectivity for use in ARF. Prototype parent and prodrugs
successfully proved principle, but also highlighted a propensity
for cyclization of simple glutamyl-protected NHG prodrugs,
which prevented effective cleavage by γ-GT. A library of parent
molecules was subsequently designed and screened for
NO:sGC-mediated vasodilator effects and a lead compound
selected for glutamyl protection. The cyclization problem was
overcome by incorporation of an aromatic linker between the
glutamyl and NHG moieties, which was spontaneously lost
after γ-GT-mediated cleavage of the glutamyl group. Using this
technique, a suitable prodrug (10) was synthesized and tested
in several models to confirm sensitivity to γ-GT-mediated
activation, decomposition to the parent drug in the presence of
rat renal tissue and vasodilator activity in isolated rat perfused
kidneys. The results confirm that glutamyl-mediated protection
of NHGs represents a valid approach to NO-induced, reno-
active vasodilatation that could have therapeutic advantages in
areas of unmet clinical need, and ARF in particular.
Figure 4. Log concentration−response curves for (A) ACh, (B)
NOHA, and (C) product 1e in the absence (control) and presence of
the NOS inhibitor, L-NAME (200 μM), the endogenous substrate for
NOS, L-Arg, the NO scavenger, Hb (10 μM), and the sGC inhibitor,
ODQ (50 μM). *P < 0.05, ***P < 0.001, 2-way ANOVA with
Bonferroni post-test (n = 12−15 for ACh; n = 4−9 for NOHA and n =
6−8 for 1e).
Figure 5. X-ray structure of product 1e, the lead parent drug chosen
The proof of principle study confirmed that activity of the
parent drug required the guanidino moiety (the corresponding
urea was inactive) and was induced by X/XO enzyme
preparations and NOS in tissue extracts. In addition, it was
established that it is possible to generate γ-GT-dependent
NHG-derived prodrugs, providing the molecule does not
cyclize. The modest vasodilatory efficacy of both the parent
molecule and the prodrug indicated that there was substantial
for prodrug development.
would spontaneously drop off as an effect of thermodynami-
cally favored cyclization, releasing the parent NHG in the
process.
However, prodrug 9 failed to induce relaxation, irrespective
of preincubation with γ-GT (data not shown).
G
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Scheme 4. Design of O,N-Bis(GAbAGlu)-NHG Prodrug
Scheme 5. Design of a γ-Glutamyl Prodrug of NHG with an Aromatic Linker
room for improvement in terms of optimization of chemical
structure. While prodrug 8 confirmed that the concept of
glutamyl protection was valid, the propensity toward cyclization
of this product also served notice that optimization of the
structure was necessary and that a more sophisticated approach
would be required to develop stable prodrugs that would resist
cyclization and retain the potential for γ-GT-mediated
activation. Similarly, there was room to optimize the parent
drug to generate more NO and to be less NOS-dependent in
case NOS was dysfunctional in the target disease state.
actions of which were found to be almost entirely dependent on
NOS activation, NO generation, and stimulation of smooth
muscle sGC. Interestingly, however, the vasodilator activity of
NOHA was inhibited by L-Arg, suggesting that L-Arg effectively
competes with NOHA at the active site of eNOS and/or at the
amino acid transporter system y+.52,74 The polar characteristics
of NOHA suggest that it is unlikely to be able to cross the lipid
bilayer and its structural similarity to L-Arg could allow it to
hijack the amino acid transporter systems. It is also apparent
that, while the vasodilator properties of NOHA are dependent
on NOS, they do not require the enzyme to be Ca2+-activated
(e.g., by ACh, shear stress, hypoxia).
In stark contrast to NOHA, 1e was considerably less sensitive
to L-NAME and Hb-mediated inhibition while retaining almost
exclusive reliance on sGC activation. These results imply that
not only is the activation of this drug largely independent of
NOS but that the NO generated to stimulate sGC is released at
a site relatively inaccessible to Hb and in close proximity of
sGC (i.e., inside smooth muscle cells, as opposed to
endothelium). Crucially, the effects of product 1e were not
inhibited, but were perhaps even enhanced, by L-Arg. We argue
that the increased hydrophobicity of 1e enables it to bypass the
amino acid transporter systems, eliminating the risk of L-Arg-
mediated inhibition in vivo, a distinct advantage over NOHA
for in vivo activity in the face of abundant L-Arg.
Effective prodrug design proved challenging. It was obvious
from the outset that, while generation of glutamyl adducts
undoubtedly protected the resultant molecule from tissue-
mediated NO generation, retaining characteristics necessary for
γ-GT-mediated activation was complicated by the propensity of
the products to spontaneously cyclize. It was very difficult to
isolate pure prototype prodrug 8 because it spontaneously
transformed into 4. The solution to the problem came in the
form of an aromatic linker54 that both prevented cyclization
and spontaneously dropped out of the molecule after cleavage
of the glutamyl group by γ-GT. The resulting compound
(prodrug 10) had weak vasodilator properties in rat aortic rings
that were significantly enhanced by addition of γ-GT and were
mainly sGC-dependent. Release of the corresponding parent
drug (1e) was confirmed in rat kidney extracts, and prodrug 10
was found to have vasodilator effects in rat isolated perfused
kidneys that were not evident for a GABA-linked equivalent
(prodrug 9).
NOHA, and a growing literature relating to synthetic NHG
generation,34,70−72 provided the basis for design and synthesis
of novel NHGs that might have improved pharmacological
profiles. The principal difference between the prototype
developed for the proof of principle study (product 1a) and
NOHA was the absence of the α-amino group, which was
deemed necessary to increase hydrophobicity and to avoid
protection/deprotection issues that were likely to arise at the
conjugation stage of prodrug design. However, this molecule
was found to only have weak NOx-generating capabilities and
vasodilator properties compared to the endogenous NHG,
NOHA. The first step in the evolution of the molecule was to
render it less polar by replacement of the terminal carboxylic
group with first, a methyl group (product 1c), and then an
aromatic ring (products 1d−s). In addition, the methyl group
on the central N was considered redundant and was removed.
Together, these modifications markedly increased the vaso-
dilatory effect of the compounds in comparison to the
prototype 1a. Subsequent modifications to the structure
revealed that the optimal chain length for the carbon backbone
was ∼2C (1e); increasing or decreasing the carbon chain length
reduced the dependence of the response on sGC. A similar
structure−activity relationship had previously been shown for
analogues of NOHA.73 Finally, it was established that
substitution on the aromatic ring invariably reduced drug
efficacy and/or dependence on sGC. Of the conformations
investigated, the desirable molecular characteristics to confer
vasodilator effects that were predominantly (>75%) mediated
by stimulation of sGC were a nonpolar aromatic terminal group
linked to the guanidino moiety by a short carbon chain (2C).
The lead compound from this screen was therefore deemed to
be product 1e, and we undertook to explore the vasodilator
characteristics of this particular parent drug in detail.
To fully appreciate the results obtained with 1e, it is
important to first consider those obtained with NOHA, the
In summary, we have successfully developed a novel reno-
active NO donor prodrug 10. The key features of the molecule
H
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
generating NO or possibly HNO that would rapidly isomerize
to NO at physiological pH.
While we acknowledge that prodrug 10 does not possess
perfect characteristics for renoselectivity, its properties certainly
indicate that glutamyl protection of NHGs is a valid approach
to development of a new generation of drugs for use in ARF,
the primary target of unmet need, but also potentially in other
renal conditions that might benefit from increased renal blood
flow, including hypertension. Further research with prodrug 10
is merited in animal models of cross-clamping and sepsis to test
the viability of this approach in clinically relevant applications.
In vivo experiments would test whether the activity seen in rat
isolated kidneys translates into true renoselectivity, as defined
by increased blood flow without an impact on systemic blood
pressure and would identify optimal concentrations required to
prevent ARF. While the acute nature of the primary disease
target ensures that toxicity on account of chronic dosing is not
a consideration, acute toxicity of NHGs in this clinical setting is
not known. An issue that would require special attention in this
regard is the potential for enhanced generation of ONOO− in
disease states.
This study represents a significant step forward in the quest
for an effective preventative agent for use in ARF to address an
urgent clinical need. The results point to successful develop-
ment of a prodrug that is active in renal tissue and has the
potential for renoselectivity. Optimization of the prototype and
testing in models of ARF are merited.
EXPERIMENTAL SECTION
■
NMR spectra were recorded on a Bruker Advance 300 (1H 300 MHz,
13C 75.4 MHz) or 400 (1H 400 MHz, 13C 100.6 MHz) spectrometer.
Chemical shifts (δ) in ppm are given relative to Me4Si, coupling
constants (J) are given in Hz. Low resolution and high resolution
electrospray mass spectra were recorded on a Micromass LC-T (time-
of-flight). Analytical TLC was carried out on either Merck Silica 60
F254 or RP-18 F254S plates. Purity of the tested compounds was >95%
based on either elemental analysis or HPLC analysis.
General Procedure for the Synthesis of NHGs. A 3 M solution
of cyanogen bromide in DCM (1 equiv) was added dropwise to an ice-
cooled suspension of the amine or corresponding salt (commercially
available or in house-made) and sodium acetate (3 equiv) in MeOH
with stirring. The mixture was stirred in an ice bath for 2 h, and then
the ice bath was removed and the reaction mixture was allowed to stir
at room temperature overnight. After removal of the solvent in
vacuum, the residue was partitioned between DCM and water. The
aqueous layer was extracted with DCM three times. The combined
organic phases were dried over MgSO4. Removal of the solvent under
reduced pressure and purification using column chromatography
(silica gel, Pet ether/Et2O 1:2) gave the intermediate cyanamide. The
cyanamide was mixed with hydroxylamine hydrochloride (1 equiv)
and potassium carbonate (0.1 equiv) in anhydrous MeOH. The
mixture was heated at reflux for 6 h under an argon atmosphere. After
cooling, the solvent was removed at reduced pressure and the desired
product was crystallized from CH2Cl2 and diethyl ether. Further
purification was achieved by redissolving the crude product in
minimum amount of methanol followed by filtration to remove any
insoluble material. Titration of the filtrate with diethyl ether usually led
to the precipitation of pure product; if microanalysis was not
satisfactory, repeated recrystallization from MeOH/Et2O was
performed.
Figure 6. (A) Effect of preincubation of prodrug 10 (100 μM) with γ-
GT (100 mU/mL) on vasodilator responses in precontracted rat aortic
rings in the presence and absence of the sGC inhibitor, ODQ (50
μM). NS: no significant difference, *P < 0.05, paired Student’s t-test vs
response in absence of γ-GT; n = 5. (B) Vasodilator effect of the lead
prodrug 10, its parent counterpart, 1e, and a glutamyl-protected NHG
that is not a substrate for γ-GT (prodrug 9; negative control) in the
isolated perfused rat kidney model. (C) Appearance of parent
compound 1e in rat kidney homogenates treated with prodrug 10
(1, 10, 100 μM) with time.
are afforded by the nonselective NO-generating properties of
the parent molecule, coupled with the protective effects
prescribed by the glutamyl adduct. Its vasodilator properties
are enhanced by the enzyme γ-GT, which is most highly
expressed in the renal medulla and, unlike the enzyme-resistant
prodrug 9, it successfully induces vasodilation in isolated
perfused kidneys. The specific enzyme(s) involved in the
activation of the cleaved parent NHG in renal tissue is not yet
known, but the literature would suggest that a range of enzymes
with peroxidase activity could effect the activation step,
N-3-Carboxypropyl-N-methyl-N′-hydroxyguanidine 1a. Colorless
solid (70%). Found: C, 40.96; H, 7.61; N, 23.76%. C6H13N3O3
requires C, 41.14; H, 7.48; N, 23.99%. δH (300 MHz, D2O) 1.79
(quintet, 2H, J 7.2 Hz, CH2-2), 2.12 (t, 2H, J 7.2 Hz, CH2-3), 2.91(s,
3H, NCH3), 3.26 (t, 2H, J 7.3 Hz, CH2-1). δC (75.46 MHz, D2O)
I
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
23.2, 33.7, 35.7, 50.1, 158.9 (CNOH), 182.2. m/z (ESI+) 176 [M +
7.2 Hz, CH2-1), 7.31 (2H, m, J 8.4 Hz, H-2′, 6′), 7.36 (2H, d, J 8.4 Hz,
H-3′, 5′), 7.72 (2H, br s, NH2), 7.95 (1H, br t, J 5.3 Hz, NH), 9.85
(1H, br s, NH), 10.44 (br s, 1H, OH). δC (75.46 MHz, DMSO-d6)
33.8, 41.7, 128.2, 130.8, 131.0, 137.33, 158.1 (CNOH). m/z (ESI+)
214 [M + H]+.
H]+, 198 [M + Na]+, 214 [M + K]+.
N-3-Carboxypropyl-N-methylurea 1a(u). N-3-Carboxypropyl-N-
methylcyanamide (284 mg, 2 mmol) was dissolved in 4N HCl
solution. The mixture was heated at reflux for 6 h, cooled to room
temperature, and extracted with DCM (three times). The combined
extracts were dried (MgSO4) and solvent removed under reduced
pressure to give a gum-like residue. The residue was purified by
column chromatography (silica gel, EtOAc:DCM 6:1) to give the
product as a white solid (256 mg, 80%). Found: C, 44.87; H, 7.38; N,
17.54%. C6H12N2O3 requires C, 44.99; H, 7.55; N, 17.49%. δH (300
MHz, D2O) 1.57 (quintet, 2H, J 7.2 Hz, CH2-2), 2.11 (t, 2H, J 7.2 Hz,
CH2-3), 2.60 (s, 3H, CH3), 3.01 (t, 2H, J 7.2 Hz, CH2-1). δC (75.46
MHz, D2O) 22.7, 31.4, 34.5, 47.9, 161.4, 178.4. m/z (ESI+) 161 [M +
H]+, 183 [M + Na+].
N-2-(4′-Bromophenyl)ethyl-N′-hydroxyguanidine Hydrochloride
1j. Colorless solid (67%). Found: C, 36.78; H, 4.16; N, 14.28%.
C9H13BrClN3O requires C, 36.70; H, 4.45; N, 14.26%. δH (300 MHz,
DMSO-d6) 2.77 (2H, t, J 7.2 Hz, CH2-2), 3.35−3.43 (2H, m, CH2-1),
7.22−7.27 (2H, m, H-2′, 6′), 7.47−7.52 (2H, m, H-3′,5′), 7.70 (2H,
br s, NH2), 7.91 (1H, t, J 6.3 Hz, NH), 9.81 (1H, br s, NH), 10.42
(1H, br s, OH). δC (75.46 MHz, DMSO-d6) 33.8, 41.6, 119.5, 131.1,
137.7, 158.1 (CNOH). m/z (ESI+) 260 [M + H]+.
N-2-(4′-Methylphenyl)ethyl-N′-hydroxyguanidine Hydrochloride
1k. Colorless solid (30%). Found: C, 52.46; H, 6.94; N, 18.45%.
C10H16ClN3O requires C, 52.29; H, 7.02; N, 18.29%. δH (400 MHz,
DMSO-d6) 2.26 (3H, s, CH3), 2.74 (2H, t, J 7.2 Hz, CH2-2), 7.13 (4H,
2 × d, J 8.4 Hz, Ar−H), 7.69 (2H, br s, NH2), 7.89 (1H, t, J 6.5 Hz,
NH), 9.82 (1H, br s, NH), 10.36 (1H, br s, OH). δC (100 MHz,
DMSO-d6) 21.1, 34.5, 42.5, 129.2, 129.4, 135.6, 135.8, 158.6 (C
NOH). m/z (ESI+) 194 [M + H]+.
N-n-Butyl-N′-hydroxyguanidine 1c.49 Colorless crystal (48%).
Found: C, 36.09; H, 8.89; N, 25.34%. C6H14ClN3O3 requires C,
35.82; H, 8.42; N, 25.07%. δH (300 MHz, DMSO-d6) 0.86 (t, 3H, J 7.4
Hz, CH3), 1.26 (hexatet, 2H, J 7.4 Hz, CH2-3), 1.48−1.39 (m, 2H,
CH2-2), 3.13 (dt, 2H, J 5.8, 7.4 Hz, CH2-1), 7.68 (br s, 2H, NH2), 7.96
(t, 1H, J 5.8 Hz, NH), 9.82 (br s, 1H, NH), 10.43 (br s, OH). δC
(75.46 MHz, DMSO-d6) 158.7, 40.7, 30.9, 19.6, 13.9. m/z (ESI+) 132
[M−Cl−].
N-2-(4′-Hydroxyphenyl)ethyl-N′-hydroxyguanidine Hydrochlor-
ide 1m. Colorless solid (40%). Found: C, 46.49; H, 5.94; N,
18.05%. C9H14ClN3O2 requires C, 46.66; H, 6.09; N, 18.14%. δH (300
MHz, DMSO-d6) 2.67 (2H, t, J 7.2 Hz, CH2-2), 3.26−3.73 (2H, m,
CH2-1), 6.66−6.73 (2H, m, H-3′, 5′), 7.01−7.08 (2H, m, H-2′,6′),
7.65 (2H, br s, NH2), 7.81 (1H, t, J 6.4 Hz, NH), 9.29 (1H, s, phenyl-
OH), 9.77 (1H, br s, NH), 10.38 (1H, br s, OH). δC (75.46 MHz,
DMSO-d6) 33.6, 42.3, 115.1, 128.2, 129.6, 155.9, 158.1 (CNOH).
m/z (ESI+) 196 [M + H]+.
N-Benzyl-N′-hydroxyguanidine Hydrochloride 1d.34 Colorless
solid (45%). Found: C, 47.50; H, 5.85; N, 20.97%. C8H12ClN3O
requires C, 47.65; H, 6.00; N, 20.84%. δH (300 MHz, DMSO-d6) 4.42
(2H, d, J 6.4 Hz, CH2-1), 7.25−7.41 (5H, m, Ar−H), 7.81 (2H, br s,
NH2), 8.40 (1H, t, J 6.8 Hz, NH), 9.88 (1H, br s, NH), 10.57 (1H, br
s, OH). δC (75.46 MHz, DMSO-d6) 43.5, 127.1, 127.4, 128.4, 137.3,
158.3 (CNOH); m/z (ESI+) 166 [M + H]+.
N-(4-Phenylethyl)-N′-hydroxyguanidine Hydrochloride 1e.34 Col-
orless flake crystals (70%). Found: C, 50.12; H, 6.08; N, 19.48%.
C9H14ClN3O requires C, 50.12; H, 6.38; N, 19.71%. δH (300 MHz,
DMSO-d6) 2.79 (2H, t, J 7.2 Hz, CH2-2), 3.43−3.67 (2H, m, CH2-1)
7.18−7.24 (3H, m, H-2′, 4′, 6′), 7.27−7.32 (2H, m, H-3′, 5′), 7.75
(2H, br s, NH2), 8.00 (1H, t, J 6.4 Hz, NH), 9.88 (1H, br s, N′H),
10.53 (1H, br s, OH). δC (75.46 MHz, DMSO-d6) 34.5, 42.5, 127.2,
N-2-(2′-Methoxyphenyl)ethyl-N′-hydroxyguanidine Hydrochlor-
ide 1n. Colorless crystals (46%). Found: C, 49.02; H, 6.62; N,
17.11%. C10H16ClN3O2 requires C, 48.88 H, 6.56; N, 17.10%. δH (300
MHz, DMSO-d6) 2.78 (2H, t, J 7.2 Hz, CH2-2), 3.28−3.36 (2H, m,
CH2-1), 3.78 (3H, s, OCH3), 6.84−6.99 (2H, m, H-3′, 5′), 7.17−7.25
(2H, m, H-4′,6′), 7.65 (2H, br s, NH2), 7.85 (1H, t, J 5.8 Hz, NH),
9.79 (1H, br s, NH), 10.38 (1H, br s, OH). δC (75.46 MHz, DMSO-
d6) 28.9, 40.7, 55.3, 110.6, 120.2, 126.0, 127.9, 130.1, 157.1, 158.2
(CNOH). m/z (ESI+), 210 [M + H]+.
+
129.1, 129.3, 138.6, 159.0 (CNOH). m/z (ESI+) 180 [M + H] .
N-(3-Phenylpropyl)-N′-hydroxyguanidine Hydrochloride 1f. Col-
orless flake crystals (58%). Found: C, 52.01; H, 6.90; N, 18.41%.
C10H16ClN3O requires C, 52.29; H, 7.02; N, 18.29%. δH (300 MHz,
DMSO-d6) 1.77 (2H, q, J 7.2 Hz, CH2-2), 2.59 (2H, t, J 7.2 Hz, CH2-
3), 3.15 (2H, q, J 7.2 Hz, CH2-1), 7.14−7.23 (3H, m, H-2′, 4′, 6′),
7.25−7.31 (2H, m, H-3′, 5′), 7.71 (2H, br s, NH2), 8.05 (1H, t, J 7.2
Hz, NH), 9.82 (1H, br s, NH), 10.43 (1H, br s, OH). δC (75.46 MHz,
D2O) 29.4, 31.9, 40.4, 126.2, 128.5, 128.7, 141.5, 158.6 (CNOH).
m/z (ESI+), 194 [M + H+].
N-2-(3′-Methoxyphenyl)ethyl-N′-hydroxyguanidine Hydrochlor-
ide 1p. Colorless solid (53%). Found: C, 48.10; H, 6.65; N,
17.01%. C10H16ClN3O2·0.2H2O requires C, 48.18; H, 6.63; N, 16.85%.
δH (300 MHz, DMSO-d6) 2.76 (2H, t, J 7.2 Hz, H-2), 3.38−3.43 (2H,
m, H-1), 3.74 (3H, s, OCH3), 6.75−6.89 (3H, m, H-2′, 4′, 6′), 7.21
(1H, t, J 8.4 Hz, H-5′), 7.72 (2H, br s, NH2), 7.92 (1H, t, J 5.9 Hz,
NH), 9.84 (1H, br s, NH), 10.29 (1H, br s, OH). δC (75.46 MHz,
DMSO-d6) 34.9, 42.2, 55.3, 112.3, 114.8, 121.4, 129.7, 140.21, 158.5
(CNOH), 159.6. m/z (ESI+) 210 [M + H]+.
N-(4-Phenylbutyl)-N′-hydroxyguanidine Hydrochloride 1g. Col-
orless flake crystals (55%). Found: C, 53.77; H, 7.82; N, 17.18%.
C11H18ClN3O requires C, 54.21; H, 7.40; N, 17.24%. δH (300 MHz,
DMSO-d6) 1.42−1.64 (4H, m, CH2-2,3), 2.58 (2H, t, J 7.2 Hz, CH2-
4) 3.16 (2H, dt J 7.2 Hz, J 6.5 Hz, CH2-1), 7.14−7.21 (3H, m, H-2′,
4′, 6′), 7.24−7.29 (2H, m, 3′, 5′), 7.66 (2H, br s, NH2), 7.92 (1H, t, J
6.5 Hz, NH), 9.76 (1H, br s, NH), 10.34 (1H, br s, OH). δC (75.46
MHz, DMSO-d6) 27.9, 28.1, 34.6, 38.6, 125.7, 128.3, 141.9, 158.2
(CNOH). m/z (ESI+) 208 [M + H]+.
N-2-(4′-Methoxyphenyl)ethyl-N′-hydroxyguanidine Hydrochlor-
ide 1q. Colorless crystals (58%). Found: C, 49.16; H, 6.24; N,
16.89%. C10H16ClN3O2 requires C, 48.88; H, 6.56; N, 17.10%. δH (300
MHz, DMSO-d6) 2.72 (2H, t, J 7.2 Hz, CH2-2), 3.30−3.38 (2H, m,
CH2-1), 3.72 (3H, s, OCH3), 6.83−6.89 (2H, m, H-3′, 5′), 7.16−7.22
(2H, m, H-2′,6′), 7.69 (2H, br s, NH2), 7.89 (1H, t, J 5.8 Hz, NH),
9.83 (1H, br s, NH), 10.45 (1H, br s, OH). δC (75.46 MHz, DMSO-
d6) 34.1, 42.6, 55.4, 114.1, 130.2, 130.7, 157.8, 158.1 (CNOH); m/z
(ESI+), 210 [M + H]+.
N-2-(4′-Fluorophenyl)ethyl-N′-hydroxyguanidine Hydrochloride
1h. Colorless solid (42%). Found: C, 45.93; H, 5.59; N, 18.28%.
C9H13ClFN3O requires C, 46.26; H, 5.61; N, 17.98%. δH (300 MHz,
DMSO-d6) 2.77 (2H, t, J 7.1 Hz, CH2-2), 3.36 (2H, quartet, J 7.1 Hz,
CH2-1), 7.07−7.16 (2H, m, H-2′, 6′), 7.26−7.35 (2H, m, H-3′, 5′),
7.72 (2H, br s, NH2), 7.95 (1H, br t, J 5.9 Hz, NH), 9.85(1H, br s,
NH), 10.44 (br s, 1H, OH). δC (75.46 MHz, DMSO-d6) 34.0, 42.4,
115.4 (d, 2JCF 20.9 Hz, C-3′, 5′), 131.1 (d, 3JCF 8.1 Hz, C-2′, 6′), 134.8
N-2-(1′-Naphthyl)ethyl-N′-hydroxyguanidine Hydrochloride 1r.
Off-white solid (79%). Found: C, 58.47; H, 6.14; N, 15.86%.
C13H16ClN3O requires C, 58.76; H, 6.07; N, 15.81%. δH (300 MHz,
DMSO-d6) 3.24−3.32 (2H, m, CH2-1), 3.45−3.55 (2H, m, CH2-2),
7.36−7.46 (2H, m, H-6′, 7′), 7.43−7.56 (2H, m, H-2′, 3′), 7.73 (2H,
br s, NH2), 7.82 (1H, m, H-5′), 7.94 (1H, d, J 8.4 Hz, H-8′), 8.05 (1H,
br t, J 6.2 Hz, NH), 9.85 (1H, br s, NH), 10.46 (1H, br s, OH). δC
(75.46 MHz, DMSO-d6) 31.5, 41.3, 123.6, 125.5, 125.7, 126.2, 126.9,
127.1, 128.6, 131.4, 133.4, 134.2, 158.1 (CNOH). m/z (ESI+) 230
[M + H]+.
4
1
(d, JCF J 2.9 Hz, C-1′), 158.5 (CNOH), 161.5 (d, JCF 242 Hz, C-
4′). δF 117.2; m/z (ESI+) 198 [M + H]+.
N-2-(4′-Chlorophenyl)ethyl-N′-hydroxyguanidine Hydrochloride
1i. Colorless solid (46%). Found: C, 42.42; H, 5.08; N, 16.74%.
C9H13Cl2N3O·0.2H2O requires C, 42.60; H, 5.32; N, 16.56%. δH (300
MHz, DMSO-d6) 2.79 (2H, t, J 7.2 Hz, CH2-2), 3.38 (2H, quartet, J
N-2-(4′-Sulfamoylphenyl)ethyl-N′-hydroxyguanidine Hydrochlor-
ide 1s. Colorless solid (21%); Found: C, 37.00; H, 5.32; N, 18.69%.
C9H15ClN4O3S requires C, 36.67; H, 5.13; N, 19.01%. δH (300 MHz,
J
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
DMSO-d6) 2.87 (2H, t, J 7 Hz, CH2-2), 3.36−3.47 (2H, m, CH2-1),
7.32 (2H, br s, NH2), 7.42−7.48 (2H, m, H-2′, 6′), 7.69 (2H, br s,
NH2), 7.73−7.79 (2H, m, H-3′, 5′), 7.87 (1H, br s, NH), 9.75 (1H, br
s, NH/OH), 10.34 (1H, br s, NH/OH). δC (75.46 MHz, DMSO-d6)
34.6, 42.0, 126.1, 129.7, 142.8, 142.9, 158.6 (CNOH). m/z (ESI+)
259 [M + H]+.
(60 mg/kg body weight) prior to heparinization (500 U). A midline
laparotomy incision was made from pelvis to sternum, exposing the
abdominal cavity. A cannula was inserted into the renal artery via the
superior mesenteric artery and tied in place. Perfusion with a modified
Krebs−Henseleit solution (containing 6.7% bovine serum albumin and
20 amino acids) commenced immediately, with a view to minimizing
renal hypoxia. The right kidney and ureter were dissected free and
transferred to a temperature controlled moisture chamber (Harvard
834/8; Harvard Apparatus) and perfused with 40 mL of recirculating
perfusate at a constant rate of 10 mL/min via a peristaltic roller pump
(Gilson miniplus 3). Perfusion pressure was measured via a side port
and transducer, with 10 mL/min perfusion inducing a pressure of 60.6
2.9 mmHg. Following equilibration (20 min), the kidney vasculature
was constricted by addition of the α1-adrenoceptor agonist, PE (0.3
μM incremental additions), until perfusion pressure reached 120−140
mmHg. Cumulative concentrations of 1e, 9, and 10 (1 μM to 1 mM)
were added to the perfusate to assess the impact of these compounds
on renal perfusion.
Renal Prodrug Activation. The selected candidate prodrug 10 was
incubated at three concentrations (1, 10, and 100 μM) with freshly
pooled (n = 3) male rat (Wistar) kidney homogenates and samples
taken at 15 min intervals over a 45 min incubation period for LC-MS/
MS analysis (Quattro Micro Mass Spectrometer, Micromass, UK;
positive ion mode, Phenomenex Luna 2.5 μm C18 column, mobile
phase 0.1% formic acid in 95% water, 5% acetonitrile). Total protein
concentrations were determined by Lowry assay.
N-Phenethyl-N′-GlugGABAoxy-N″-GluGABAguanidine 9.54 Color-
less powder (47%). Found: C, 53.43; H, 6.65; N, 15.97%. C27H41N7O9
requires C, 53.37; H, 6.80; N, 16.14%. δH (300 MHz, D2O) 1.79−1.64
(m, 4H, 2 × GABA-β), 2.05−1.94 (m, 4H, 2 × Glu-β), 2.45−2.22 (m,
8H, 2 × GABA-α + 2 × Glu-γ), 2.69 (t, 1H, J 6.5 Hz, 1/2 CH2-2), 2.81
(t, 1H, J 6.5 Hz, 1/2 CH2-2), 3.15 −3.06 (m, 4H, 2 × GABA-γ CH2),
3.29 (dt, 2H, J 9.4, 6.6 Hz, CH2-1), 3.66−3.60 (m, 2H, 2 × Glu-α),
7.30−7.07 (m, 5H, PhH). δC (CDCl3, 100 MHz) 23.8, 23.9, 24.1, 26.4,
29.8, 31.5, 33.7, 36.3, 38.5, 42.0, 42.4, 43.6, 54.1, 126.6, 128.7, 129.0,
138.8, 173.2 (overlapped), 173.9, 174.6. m/z (ESI+) 608 [M + H]+. m/
z (ESI−) 606 [M − H]−.
N-Phenethyl-N′-(4-(γ-glutamylamino))benzyloxyguanidine10.54
Colorless solid (0.150 g, 74%). Purity was 98% according to HPLC
analysis. δH (D2O, 300 MHz) 2.16 (q, 2H, J 7.4 Hz, Glu-β), 2.60−2.54
(m, 2H, Glu-γ), 2.69 (t, 2H, J 6.6 Hz, CH2-2), 3.32 (t, 2H, J 6.6 Hz,
CH2-1), 3.74 (t, 1H, J 6.2 Hz, Glu-α), 4.75 (s, 2H, CH2), 7.11 (d, 2H,
J 8.2 Hz, Ar H), 7.48−7.24 (m, 7H, Ar H). δC (62.5 MHz, DMSO-d6),
27.6, 33.5, 35.8, 42.9, 54.1, 74.4, 119.2, 126.5, 128.8, 129.0, 129.2,
133.7, 138.5, 140.0, 154.8, 170.0, 170.8. m/z (ESI+) 414 [M + H]+.
Biological Assessment. Animals. The animal care and exper-
imental procedures were in accordance with the United Kingdom
Animal (Scientific Procedures) Act, 1986. Male Wistar rats weighing
between 250 and 450 g were used in all animal experiments.
In Vitro NOx Analysis. Stock solutions of X/XO, γ-GT/gly-gly were
made fresh on each experimental day in KPE buffer. Crude
homogenates of rat liver, renal medulla, and renal cortex were
prepared in ice-cold HBSS containing 1% Triton X from freshly killed
adult male Wistar rats using a mechanical homogenizer and sonicator
bath. NHGs (100 μM) were added to enzyme preparations (100 μM
X/100 mU XO and 100 mU γ-GT/5 mM gly-gly) or tissue
homogenates with and without the relevant inhibitors (5-sulfosalicylic
acid (100 μM) for γ-GT; allopurinol for XO (100 μM) and L-NAME
(200 μM) for NOS in homogenates for (1 h, 37 °C), before samples
were centrifuged (1000g; 4 °C) and NOx assayed using the Griess test
(R&D Systems). This test does not measure NO directly, but
Statistical Analysis. All data are expressed as mean
SEM.
Statistical analysis of NOx data was conducted using paired Student t-
tests. Comparisons of vasodilator effects were conducted using
Student’s t-tests or one-way ANOVAs followed by Dunnett’s post-
tests, as appropriate, while those for concentration−response curves
was performed using 2-way ANOVA, followed by Bonferroni post-
tests. P < 0.05 was considered to be significant (analysis conducted
using Graphpad Prism 5.00).
ASSOCIATED CONTENT
■
S
* Supporting Information
Preparation and characterization of compound 1b, 3, 4, 5, 8; 1H
1
NMR and 2D H−13C HMBC spectra of N-(4-phenylbutyl)-
N′-hydroxyguanidine 1g in DMSO-d6; HPLC traces of 9 and
10. This material is available free of charge via the Internet at
http://pubs.acs.org
−
−
measures [NO2 + NO3 ] as a surrogate of NO.
Vasodilator Experiments. Rings (∼3 mm) of rat aorta were
mounted in a myograph (Danish Myo model 703, 37 °C) containing
Krebs buffer solution (composition mM: NaCl 118, KCl 4.7, CaCl2
2.5, MgSO4 1.2, KH2PO4 1.2, NaHCO3 35, glucose 10), oxygenated
with 95% O2, 5% CO2. Vascular rings were stretched to 15 mN over a
1 h period, and the maximal contraction determined by replacing
standard Krebs buffer solution with Krebs containing 60 mM KCl.
Subsequently, aortic rings were precontracted to ∼80% of the KCl-
induced maximal contraction with phenylephrine (PE 10−300 nM)
prior to confirmation of endothelial integrity using the endothelium-
dependent vasodilator, acetylcholine (ACh; 10 μM). Relaxation of
>80% was considered to confirm functional integrity of the
endothelium and rings that failed to relax to this extent were rejected.
Following washout of ACh, precontracted rings were either treated
with NHGs (100 μM, single exposure in preliminary experiments or
cumulative concentrations of 1−100 μM in subsequent character-
ization experiments). Parallel rings were incubated with L-NAME (200
μM), L-Arg (100 μM), hemoglobin (10 μM), or ODQ (50 μM) prior
to and during exposure to ACh, NOHA, and NHG analogues to
determine whether the compounds acted independently of NOS,
compete with L-Arg for cellular uptake, generate NO that is scavenged
by hemoglobin, or act via sGC activation, respectively. In some
experiments involving prodrugs, glutamyl-linked NHGs (10 mM)
were preincubated with γ-GT (100 mU/5 mM gly-gly; 1h, 37 °C)
prior to addition to the myograph.
AUTHOR INFORMATION
Corresponding Author
*Phone: +44 (0)1463 279562. Fax: +44 (0)1463 711245. E-
mail: ian.megson@uhi.ac.uk.
■
Author Contributions
All authors contributed to the writing of this manuscript and
approved the final version.
Notes
The authors declare no competing financial interest.
†Nigel P. Botting died 4 June 2011.
ACKNOWLEDGMENTS
We thank the Wellcome Trust for supporting this research
(Catalyst Biomedica, Translational Award 063729/Z/01/Z).
■
a
ABBREVIATIONS USED
■
ACh, acetylcholine; ARF, acute renal failure; γ-GT, γ-glutamyl
transpeptidase; Hb, hemoglobin; NHG, N-hydroxyguanidine;
NOHA, N-hydroxyarginine; NO, nitric oxide; NOx, [nitrite +
nitrate]; NOS, NO synthase; L-NAME, Nω-nitro-L-arginine
methyl ester; ODQ, 1H-[1,2,4]oxadiazolo[4, 3-a]quinoxalin-1-
one; ROS, reactive oxygen species; X, xanthine; XO, xanthine
Isolated Perfused Kidney. The isolated perfused kidney preparation
was performed as previously described.75 Briefly, rats were
anaesthetized by intraperitoneal injection with sodium pentobarbital
K
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(16) Ladefoged, S. D.; Andersen, C. B. Calcium channel blockers in
kidney transplantation. Clin. Transplant. 1994, 8, 128−133.
(17) Allgren, R. L.; Marbury, T. C.; Rahman, S. N.; Weisberg, L. S.;
Fenves, A. Z.; Lafayette, R. A.; Sweet, R. M.; Genter, F. C.; Kurnik, B.
R.; Conger, J. D.; Sayegh, M. H. Anaritide in acute tubular necrosis.
Auriculin Anaritide Acute Renal Failure Study Group. N. Engl. J. Med.
1997, 336, 828−834.
(18) Briguori, C.; Manganelli, F.; Scarpato, P.; Elia, P. P.; Golia, B.;
Riviezzo, G.; Lepore, S.; Librera, M.; Villari, B.; Colombo, A.;
Ricciardelli, B. Acetylcysteine and contrast agent-associated neph-
rotoxicity. J. Am. Coll. Cardiol. 2002, 40, 298−303.
oxidase; Fm, (9-fluorenyl)methyl; Fmoc, (9-fluorenyl)-
methoxycarbonyl; GABA, γ-aminobutanoyl; Glu, γ-glutamyl;
glygly, glycylglycine; EDC, 1-ethyl-3-(3-dimethylaminopropyl)-
carbodiimide; HOBt, N-hydroxybenzotriazole hydrate;
(Boc)2O, di-tert-butyl dicarbonate; Boc, tert-butoxycarbonyl;
TsOH, p-toluenesulfonic acid; DMSO, dimethyl sulfoxide;
DMF, dimethylformamide; Et3N, triethylamine; DCM, di-
chloromethane; HBSS, Hanks-balanced saline solution; KPE,
potassium phosphate-EDTA; LC-MS/MS, liquid chromatog-
raphy mass spectrometry; PE, phenylephrine hydrochloride
(19) Durham, J. D.; Caputo, C.; Dokko, J.; Zaharakis, T.; Pahlavan,
M.; Keltz, J.; Dutka, P.; Marzo, K.; Maesaka, J. K.; Fishbane, S. A
randomized controlled trial of N-acetylcysteine to prevent contrast
nephropathy in cardiac angiography. Kidney Int. 2002, 62, 2202−2207.
(20) Schneider, R.; Raff, U.; Vornberger, N.; Schmidt, M.; Freund,
R.; Reber, M.; Schramm, L.; Gambaryan, S.; Wanner, C.; Schmidt, H.
H.; Galle, J. L-Arginine counteracts nitric oxide deficiency and
improves the recovery phase of ischemic acute renal failure in rats.
Kidney Int. 2003, 64, 216−225.
ADDITIONAL NOTE
■
a
Abbreviations for common amino acids are in accordance with
the recommendations of IUPAC.
REFERENCES
■
(1) Groeneveld, A. B.; Tran, D. D.; van der Meulen, J.; Nauta, J. J.;
Thijs, L. G. Acute renal failure in the medical intensive care unit:
predisposing, complicating factors and outcome. Nephron 1991, 59,
602−610.
(2) Levy, E. M.; Viscoli, C. M.; Horwitz, R. I. The effect of acute
renal failure on mortality. A cohort analysis. JAMA, J. Am. Med. Assoc.
1996, 275, 1489−1494.
(3) Bagshaw, S. M. The long-term outcome after acute renal failure.
Curr. Opin. Crit. Care 2006, 12, 561−566.
(4) Hilton, R. Acute renal failure. BMJ (Br. Med. J.) 2006, 333, 786−
790.
(21) Gryglewski, R. J.; Palmer, R. M.; Moncada, S. Superoxide anion
is involved in the breakdown of endothelium-derived vascular relaxing
factor. Nature 1986, 320, 454−456.
(22) MacKenzie, A.; Martin, W. Loss of endothelium-derived nitric
oxide in rabbit aorta by oxidant stress: restoration by superoxide
dismutase mimetics. Br. J. Pharmacol. 1998, 124, 719−728.
(23) Conger, J.; Robinette, J.; Villar, A.; Raij, L.; Shultz, P. Increased
nitric oxide synthase activity despite lack of response to endothelium-
dependent vasodilators in postischemic acute renal failure in rats. J.
Clin. Invest. 1995, 96, 631−638.
(24) Kakoki, M.; Hirata, Y.; Hayakawa, H.; Suzuki, E.; Nagata, D.;
Tojo, A.; Nishimatsu, H.; Nakanishi, N.; Hattori, Y.; Kikuchi, K.;
Nagano, T.; Omata, M. Effects of tetrahydrobiopterin on endothelial
dysfunction in rats with ischemic acute renal failure. J. Am. Soc.
Nephrol. 2000, 11, 301−309.
(5) Gelman, S. The pathophysiology of aortic cross-clamping and
unclamping. Anesthesiology 1995, 82, 1026−1060.
(6) Grun, R. P.; Constantinovici, N.; Normand, C.; Lamping, D. L.
Costs of dialysis for elderly people in the UK. Nephrol., Dial.,
Transplant. 2003, 18, 2122−2127.
(7) Chertow, G. M.; Sayegh, M. H.; Allgren, R. L.; Lazarus, J. M. Is
the administration of dopamine associated with adverse or favorable
outcomes in acute renal failure? Auriculin Anaritide Acute Renal
Failure Study Group. Am. J. Med. 1996, 101, 49−53.
(8) Lassnigg, A.; Donner, E.; Grubhofer, G.; Presterl, E.; Druml, W.;
Hiesmayr, M. Lack of renoprotective effects of dopamine and
furosemide during cardiac surgery. J. Am. Soc. Nephrol. 2000, 11,
97−104.
(9) Bellomo, R.; Chapman, M.; Finfer, S.; Hickling, K.; Myburgh, J.
Low-dose dopamine in patients with early renal dysfunction: a
placebo-controlled randomised trial. Australian and New Zealand
Intensive Care Society (ANZICS) Clinical Trials Group. Lancet 2000,
356, 2139−2143.
(10) Takala, J.; Meier-Hellmann, A.; Eddleston, J.; Hulstaert, P.;
Sramek, V. Effect of dopexamine on outcome after major abdominal
surgery: a prospective, randomized, controlled multicenter study.
European Multicenter Study Group on Dopexamine in Major
Abdominal Surgery. Crit. Care Med. 2000, 28, 3417−3423.
(11) Dehne, M. G.; Klein, T. F.; Muhling, J.; Sablotzki, A.; Osmer,
C.; Hempelmann, G. Impairment of renal function after cardiopulmo-
nary bypass is not influenced by dopexamine. Renal Failure 2001, 23,
217−230.
(25) Chirino, Y. I.; Hernandez-Pando, R.; Pedraza-Chaverri, J.
Peroxynitrite decomposition catalyst ameliorates renal damage and
protein nitration in cisplatin-induced nephrotoxicity in rats. BMC
Pharmacol. 2004, 4, 20.
(26) Bech, J. N.; Nielsen, C. B.; Pedersen, E. B. Effects of systemic
NO synthesis inhibition on RPF, GFR, UNa, and vasoactive hormones
in healthy humans. Am. J. Physiol. 1996, 270, F845−F851.
(27) Pararajasingam, R.; Weight, S. C.; Bell, P. R.; Nicholson, M. L.;
Sayers, R. D. Prevention of renal impairment following aortic cross-
clamping by manipulation of the endogenous renal nitric oxide
response. Eur. J. Vasc. Endovasc. Surg. 2000, 19, 396−399.
(28) Chintala, M. S.; Chiu, P. J.; Vemulapalli, S.; Watkins, R. W.;
Sybertz, E. J. Inhibition of endothelial derived relaxing factor (EDRF)
aggravates ischemic acute renal failure in anesthetized rats. Naunyn-
Schmiedeberg’s Arch. Pharmacol. 1993, 348, 305−310.
(29) Mashiach, E.; Sela, S.; Winaver, J.; Shasha, S. M.; Kristal, B.
Renal ischemia-reperfusion injury: contribution of nitric oxide and
renal blood flow. Nephron 1998, 80, 458−467.
(30) Matsumura, Y.; Nishiura, M.; Deguchi, S.; Hashimoto, N.;
Ogawa, T.; Seo, R. Protective effect of FK409, a spontaneous nitric
oxide releaser, on ischemic acute renal failure in rats. J. Pharmacol. Exp
Ther. 1998, 287, 1084−1091.
(12) Ralph, C. J.; Tanser, S. J.; Macnaughton, P. D.; Sinclair, D. G. A
randomised controlled trial investigating the effects of dopexamine on
gastrointestinal function and organ dysfunction in the critically ill.
Intensive Care Med. 2002, 28, 884−890.
(31) Kurata, H.; Takaoka, M.; Kubo, Y.; Katayama, T.; Tsutsui, H.;
Takayama, J.; Matsumura, Y. Nitric Oxide Protects against Ischemic
Acute Renal Failure through the Suppression of Renal Endothelin-1
Overproduction. J. Cardiovasc. Pharmacol. 2004, 44 (Suppl 1), S455−
S458.
(13) Kellum, J. A. Use of diuretics in the acute care setting. Kidney
Int., Suppl. 1998, 66, S67−70.
(32) Wilhelm, S. M.; Simonson, M. S.; Robinson, A. V.; Stowe, N. T.;
Schulak, J. A. Endothelin up-regulation and localization following renal
ischemia and reperfusion. Kidney Int. 1999, 55, 1011−1018.
(33) Miller, M. R.; Megson, I. L. Recent developments in nitric oxide
donor drugs. Br. J. Pharmacol. 2007, 151, 305−321.
(14) Dorman, H. R.; Sondheimer, J. H.; Cadnapaphornchai, P.
Mannitol-induced acute renal failure. Medicine (Baltimore) 1990, 69,
153−159.
(15) Mehta, R. L.; Pascual, M. T.; Soroko, S.; Chertow, G. M.
Diuretics, mortality, and nonrecovery of renal function in acute renal
failure. JAMA, J. Am. Med. Assoc. 2002, 288, 2547−2553.
(34) Renodon-Corniere, A.; Dijols, S.; Perollier, C.; Lefevre-
Groboillot, D.; Boucher, J. L.; Attias, R.; Sari, M. A.; Stuehr, D.;
L
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Mansuy, D. N-Aryl N′-hydroxyguanidines, a new class of NO-donors
after selective oxidation by nitric oxide synthases: structure−activity
relationship. J. Med. Chem. 2002, 45, 944−954.
(35) Mansuy, D.; Boucher, J. L. Alternative nitric oxide-producing
substrates for NO synthases. Free Radical Biol. Med. 2004, 37, 1105−
1121.
transpeptidase with potential for reno-selective vasodilatation. Chem.
Commun. 2012, 49, 1389−1391.
(55) Leenders, R. G. G.; Gerrits, K. A. A.; Ruijtenbeek, R.; Scheeren,
H. W.; Haisma, H. J.; Boven, E. beta-Glucuronyl Carbamate Based
Pro-Moieties Designed for Prodrugs in Adept. Tetrahedron Lett. 1995,
36, 1701−1704.
(56) Menard, A.; Castonguay, R.; Lherbet, C.; Rivard, C.; Roupioz,
Y.; Keillor, J. W. Nonlinear free energy relationship in the general-acid-
catalyzed acylation of rat kidney gamma-glutamyl transpeptidase by a
series of gamma-glutamyl anilide substrate analogues. Biochemistry
2001, 40, 12678−12685.
(57) Thompson, G. A.; Meister, A. Interrelationships between
Binding-Sites for Amino-Acids, Dipeptides, and Gamma-Glutamyl
Donors in Gamma-Glutamyl Transpeptidase. J. Biol. Chem. 1977, 252,
6792−6798.
(58) Tate, S. S.; Meister, A. Gamma-Glutamyl-Transferase Trans-
peptidase from Kidney. Methods Enzymol. 1985, 113, 400−419.
(59) Elsadek, B.; Graeser, R.; Warnecke, A.; Unger, C.; Saleem, T.;
El-Melegy, N.; Madkor, H.; Kratz, F. Optimization of an Albumin-
Binding Prodrug of Doxorubicin That Is Cleaved by Prostate-Specific
Antigen. Med. Chem. Lett. 2010, 1, 234−238.
(60) Antczak, C.; Bauvois, B.; Monneret, C.; Florent, J. C. A new
acivicin prodrug designed for tumor-targeted delivery. Bioorg. Med.
Chem. 2001, 9, 2843−2848.
(61) de Groot, F. M. H.; Loos, W. J.; Koekkoek, R.; van Berkom, L.
W. A.; Busscher, G. F.; Seelen, A. E.; Albrecht, C.; de Bruijn, P.;
Scheeren, H. W. Elongated multiple electronic cascade and cyclization
spacer systems in activatible anticancer prodrugs for enhanced drug
release. J. Org. Chem. 2001, 66, 8815−8830.
(62) de Groot, F. M. H.; van Berkom, L. W. A.; Scheeren, H. W.
Synthesis and biological evaluation of 2′-carbamate-linked and 2′-
carbonate-linked prodrugs of paclitaxel: selective activation by the
tumor-associated protease plasmin. J. Med. Chem. 2000, 43, 3093−
3102.
(63) Elsadek, B.; Graeser, R.; Esser, N.; Schafer-Obodozie, C.; Abu
Ajaj, K.; Unger, C.; Warnecke, A.; Saleem, T.; El-Melegy, N.; Madkor,
H.; Kratz, F. Development of a novel prodrug of paclitaxel that is
cleaved by prostate-specific antigen: an in vitro and in vivo evaluation
study. Eur. J. Cancer 2010, 46, 3434−3444.
(64) Fossey, C.; Huynh, N. T.; Vu, A. H.; Vidu, A.; Zarafu, I.;
Laduree, D.; Schmidt, S.; Laumond, G.; Aubertin, A. M. Synthesis and
anti-HIV evaluation of hybrid-type prodrugs conjugating HIV
integrase inhibitors with d4t by self-cleavable spacers containing an
amino acid residue. J. Enzyme Inhib. Med. Chem. 2007, 22, 608−619.
(65) Greenwald, R. B.; Pendri, A.; Conover, C. D.; Zhao, H.; Choe,
Y. H.; Martinez, A.; Shum, K.; Guan, S. Y. Drug delivery systems
employing 1,4-or 1,6-elimination: poly(ethylene glycol) prodrugs of
amine-containing compounds. J. Med. Chem. 1999, 42, 3657−3667.
(66) Hay, M. P.; Wilson, W. R.; Denny, W. A. Nitroarylmethylcarba-
mate prodrugs of doxorubicin for use with nitroreductase gene-
directed enzyme prodrug therapy. Bioorg. Med. Chem. 2005, 13, 4043−
4055.
(67) Muller, I. A.; Kratz, F.; Jung, M.; Warnecke, A. Schiff bases
derived from p-aminobenzyl alcohol as trigger groups for pH-
dependent prodrug activation. Tetrahedron Lett. 2010, 51, 4371−4374.
(68) Niculescu-Duvaz, D.; Heyes, J.; Springer, C. J. Structure−
activity relationship in cationic lipid mediated gene transfection. Curr.
Med. Chem. 2003, 10, 1233−1261.
(69) You, H.; Yoon, H. E.; Yoon, J. H.; Ko, H.; Kim, Y. C. Synthesis
of pheophorbide-a conjugates with anticancer drugs as potential cancer
diagnostic and therapeutic agents. Bioorg. Med. Chem. 2011, 19, 5383−
5391.
(70) Lefevre-Groboillot, D.; Boucher, J. L.; Stuehr, D. J.; Mansuy, D.
Relationship between the structure of guanidines and N-hydroxygua-
nidines, their binding to inducible nitric oxide synthase (iNOS) and
their iNOS-catalysed oxidation to NO. FEBS J. 2005, 272, 3172−3183.
(71) Slama, P.; Boucher, J. L.; Reglier, M. Aromatic N-
hydroxyguanidines as new reduction cosubstrates for dopamine beta-
hydroxylase. Biochem. Biophys. Res. Commun. 2004, 316, 1081−1087.
(36) Boucher, J. L.; Genet, A.; Vadon, S.; Delaforge, M.; Henry, Y.;
Mansuy, D. Cytochrome P450 catalyzes the oxidation of N-omega-
hydroxy-L-arginine by NADPH and O2 to nitric oxide and citrulline.
Biochem. Biophys. Res. Commun. 1992, 187, 880−886.
(37) Boucher, J. L.; Genet, A.; Vadon, S.; Delaforge, M.; Mansuy, D.
Formation of nitrogen oxides and citrulline upon oxidation of N-
omega-hydroxy-L-arginine by hemeproteins. Biochem. Biophys. Res.
Commun. 1992, 184, 1158−1164.
(38) Dambrova, M.; Uhlen, S.; Welch, C. J.; Wikberg, J. E.
Identification of an N-hydroxyguanidine reducing activity of xanthine
oxidase. Eur. J. Biochem. 1998, 257, 178−184.
(39) Barthelmebs, M.; Caillette, A.; Ehrhardt, J. D.; Velly, J.; Imbs, J.
L. Metabolism and vascular effects of gamma-L-glutamyl-L-dopa on the
isolated rat kidney. Kidney Int. 1990, 37, 1414−1422.
(40) Wilk, S.; Mizoguchi, H.; Orlowski, M. gamma-Glutamyl dopa: a
kidney-specific dopamine precursor. J. Pharmacol. Exp. Ther. 1978,
206, 227−232.
(41) Li Kam Wa, T. C.; Freestone, S.; Samson, R. R.; Johnston, N.
R.; Lee, M. R. Renal metabolism and effects of the glutamyl derivatives
of L-dopa and 5-hydroxytryptophan in man. Clin. Sci. 1996, 91, 177−
185.
(42) Worth, D. P.; Harvey, J. N.; Brown, J.; Lee, M. R. gamma-L-
Glutamyl-L-dopa is a dopamine pro-drug, relatively specific for the
kidney in normal subjects. Clin. Sci. 1985, 69, 207−214.
(43) Rosalki, S. B. gamma-Glutamyl transpeptidase. Adv. Clin. Chem.
1975, 17, 53−107.
(44) Tate, S. S.; Meister, A. gamma-Glutamyl transpeptidase:
catalytic, structural and functional aspects. Mol. Cell. Biochem. 1981,
39, 357−368.
(45) Jia, Q.; Cai, T.; Huang, M.; Li, H.; Xian, M.; Poulos, T. L.;
Wang, P. G. Isoform-selective substrates of nitric oxide synthase. J.
Med. Chem. 2003, 46, 2271−2274.
(46) Pufahl, R. A.; Nanjappan, P. G.; Woodard, R. W.; Marletta, M.
A. Mechanistic probes of N-hydroxylation of L-arginine by the
inducible nitric oxide synthase from murine macrophages. Biochemistry
1992, 31, 6822−6828.
(47) Wagenaar, F. L.; Kerwin, J. F. Methodology for the Preparation
of N-Guanidino-Modified Arginines and Related Derivatives. J. Org.
Chem. 1993, 58, 4331−4338.
(48) Wallace, G. C.; Fukuto, J. M. Synthesis and bioactivity of N-
omega-hydroxyarginine: a possible intermediate in the biosynthesis of
nitric oxide from arginine. J. Med. Chem. 1991, 34, 1746−1748.
(49) Xian, M.; Fujiwara, N.; Wen, Z.; Cai, T. W.; Kazuma, S.;
Janczuk, A. J.; Tang, X. P.; Telyatnikov, V. V.; Zhang, Y. X.; Chen, X.
C.; Miyamoto, Y.; Taniguchi, N.; Wang, P. G. Novel substrates for
nitric oxide synthases. Bioorg. Med. Chem. 2002, 10, 3049−3055.
(50) Xian, M.; Li, X. P.; Tang, X. P.; Chen, X. C.; Zheng, X. L.;
Galligan, J. J.; Kreulen, D. L.; Wang, P. G. N-Hydroxyl derivatives of
guanidine based drugs as enzymatic NO donors. Bioorg. Med. Chem.
Lett. 2001, 11, 2377−2380.
(51) Caddick, S.; Judd, D. B.; Lewis, A. K. D.; Reich, M. T.; Williams,
M. R. V. A generic approach for the catalytic reduction of nitriles.
Tetrahedron 2003, 59, 5417−5423.
(52) Zani, B. G.; Bohlen, H. G. Transport of extracellular L-arginine
via cationic amino acid transporter is required during in vivo
endothelial nitric oxide production. American J. Physiol: Heart Circul.
Physiol. 2005, 289, H1381−H1390.
(53) Zharikov, S. I.; Block, E. R. Characterization of L-arginine uptake
by plasma membrane vesicles isolated from cultured pulmonary artery
endothelial cells. Biochim. Biophys. Acta 1998, 1369, 173−183.
(54) Zhang, Q.; Megson, I.; Botting, N. P.; Webb, D. J.; Kulczynska,
A. A new class of NO-donor pro-drugs triggered by [gamma]-glutamyl
M
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(72) Moali, C.; Boucher, J. L.; Renodon-Corniere, A.; Stuehr, D. J.;
Mansuy, D. Oxidations of N(omega)-hydroxyarginine analogues and
various N-hydroxyguanidines by NO synthase II: key role of
tetrahydrobiopterin in the reaction mechanism and substrate
selectivity. Chem. Res. Toxicol. 2001, 14, 202−210.
(73) Beranova, P.; Chalupsky, K.; Kleschyov, A. L.; Schott, C.;
Boucher, J. L.; Mansuy, D.; Munzel, T.; Muller, B.; Stoclet, J. C. N-
omega-hydroxy-L-arginine homologues and hydroxylamine as nitric
oxide-dependent vasorelaxant agents. Eur. J. Pharmacol. 2005, 516,
260−267.
(74) McDonald, K. K.; Rouhani, R.; Handlogten, M. E.; Block, E. R.;
Griffith, O. W.; Allison, R. D.; Kilberg, M. S. Inhibition of endothelial
cell amino acid transport system y+ by arginine analogs that inhibit
nitric oxide synthase. Biochim. Biophys. Acta 1997, 1324, 133−141.
(75) De Mello, G.; Maack, T. Nephron function of the isolated
perfused rat kidney. Am. J. Physiol. 1976, 231, 1699−1707.
N
dx.doi.org/10.1021/jm400146r | J. Med. Chem. XXXX, XXX, XXX−XXX