Synthesis and biological evaluation of l-valine-amidoximeesters as double prodrugs of amidines

Article abstract of DOI:10.1016/j.bmc.2011.01.066

In general, drugs containing amidines suffer from poor oral bioavailability and are often converted into amidoxime prodrugs to overcome low uptake from the gastrointestinal tract. The esterification of amidoximes with amino acids represents a newly developed double prodrug principle creating derivatives of amidines with both improved oral availability and water solubility. N-valoxybenzamidine (1) is a model compound for this principle, which has been transferred to the antiprotozoic drug pentamidine (8). Prodrug activation depends on esterases and mARC and is thus independent from activation by P450 enzymes. Therefore, drug-drug interactions or side effects will be minimized. The synthesis of these two compounds was established, and their biotransformation was studied in vitro and in vivo. Bioactivation of N-valoxybenzamidine (1) and N,N′-bis(valoxy)pentamidine (7) via hydrolysis and reduction has been demonstrated in vitro with porcine and human subcellular enzyme preparations and the mitochondrial Amidoxime Reducing Component (mARC). Moreover, activation of N-valoxybenzamidine (1) by porcine hepatocytes was studied. In vivo, the bioavailability in rats after oral application of N-valoxybenzamidine (1) was about 88%. Similarly, N,N′-bis(valoxy) pentamidine (7) showed oral bioavailability. Analysis of tissue samples revealed high concentrations of pentamidine (8) in liver and kidney.

Full text of DOI:10.1016/j.bmc.2011.01.066

Bioorganic & Medicinal Chemistry 19 (2011) 1907–1914
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and biological evaluation of L-valine-amidoximeesters as double
prodrugs of amidines
Joscha Kotthaus a, Helen Hungeling a, Christiane Reeh a, Jürke Kotthaus a, Dennis Schade a, Silvia Wein b,
Siegfried Wolffram b, Bernd Clement a,
⇑
a Department of Pharmaceutical and Medicinal Chemistry, Christian-Albrechts-University of Kiel, Gutenbergstr. 76-78, D-24118 Kiel, Germany
b Institute of Animal Nutrition and Physiology, Christian-Albrechts-University of Kiel, Hermann-Rodewald-Str. 9, D-24118 Kiel, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
In general, drugs containing amidines suffer from poor oral bioavailability and are often converted into
amidoxime prodrugs to overcome low uptake from the gastrointestinal tract. The esterification of amidox-
imes with amino acids represents a newly developed double prodrug principle creating derivatives of
amidines with both improved oral availability and water solubility. N-valoxybenzamidine (1) is a model
compound for this principle, which has been transferred to the antiprotozoic drug pentamidine (8).
Prodrug activation depends on esterases and mARC and is thus independent from activation by P450
enzymes. Therefore, drug-drug interactions or side effects will be minimized. The synthesis of these two
compounds was established, and their biotransformation was studied in vitro and in vivo. Bioactivation
of N-valoxybenzamidine (1) and N,N0-bis(valoxy)pentamidine (7) via hydrolysis and reduction has been
demonstrated in vitro with porcine and human subcellular enzyme preparations and the mitochondrial
Amidoxime Reducing Component (mARC). Moreover, activation of N-valoxybenzamidine (1) by porcine
hepatocytes was studied. In vivo, the bioavailability in rats after oral application of N-valoxybenzamidine
(1) was about 88%. Similarly, N,N0-bis(valoxy)pentamidine (7) showed oral bioavailability. Analysis of
tissue samples revealed high concentrations of pentamidine (8) in liver and kidney.
Received 26 December 2010
Revised 28 January 2011
Accepted 31 January 2011
Available online 3 February 2011
Keywords:
Benzamidine
Biotransformation
Pentamidine
Prodrug
Valine
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
has been studied in greater detail afterwards.8,9 So far, no drug–
drug interactions or toxic effects related to the reduction of ami-
Prodrug principles are needed to overcome pharmaceutical,
pharmacokinetic, and pharmacodynamic barriers such as poor sol-
ubility, insufficient oral absorption, inadequate blood–brain barrier
permeability, or presystemic metabolism. Numerous drugs and
drug candidates contain strong basic moieties, such as guanidines
or amidines.1 Due to their strong basicity,2 amidines are proton-
ated under physiological conditions forming highly mesomerically
stabilized amidinium-ions. Since they predominantly exist in their
protonated form in the gastrointestinal tract, they are not signifi-
cantly absorbed via passive transport mechanisms. In most cases
this results in extremely low bioavailability after oral applica-
tion.3,4 Amidoximes can be used as bioprecursors of amidines.
Hydroxylation of an amidine to the corresponding amidoxime
leads to a considerably decreased basicity and to an increase in
lipophilicity.5 Amidoximes are uncharged at physiological pH, in
general pharmacologically inactive, and their absorbance from
the gastrointestinal tract is significantly improved.4 The extensive
reduction of benzamidoxime (2) to the pharmacologically active
benzamidine (3) has already been demonstrated in 19886,7 and
doximes to amidines have been identified.10 In mitochondria the
N-reduction of amidoximes is mediated by a newly discovered
molybdenum containing enzyme, the mitochondrial Amidoxime
Reducing Component (mARC).11,12 The amidoxime prodrug
principle was first applied to pentamidine (8) by our laboratory
in 1985.13 In the meantime, it has been transferred to several other
amidines in order to increase oral bioavailability.4
Pentamidine (8) is effective in therapy for the hemolymphatic
stage of trypanosomiasis and antimony-resistant leishmania-
sis.14–16 Furthermore, it is established in the treatment of Pneumo-
cystis carinii Pneumonia (PcP).17 As an aromatic diamidine,
pentamidine requires intravenous or inhalative application. Unfor-
tunately, most of the infections mentioned above occur in tropical
or subtropical countries that usually have poor medical care sys-
tems. Consequently, this way of application limits the medicinal
use of pentamidine 1 in most regions and shows the need for a
pentamidine derivative that can be administered orally. Moreover,
intravenous application is associated with toxic side effects includ-
ing neutropenia, abnormal function, rash, nausea, and vomiting,
which may seriously limit the ability of patients to continue their
treatment.18 In order to reduce toxicity other research groups
developed pentamidine analogues with modified linkers between
⇑
Corresponding author. Tel.: +49 431 8801126; fax: +49 431 8801352.
E-mail address: bclement@pharmazie.uni-kiel.de (B. Clement).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.01.066

1908
J. Kotthaus et al. / Bioorg. Med. Chem. 19 (2011) 1907–1914
the benzamidoxime moieties.19,20 The most established com-
pounds are DB75 (Furamidine) and its prodrug DB289 (Pafurami-
dine), which showed both oral bioavailability and efficacy in
in vivo experiments.21 However, development was discontinued
in 2008 due to unexplained idiosyncratic drug induced organ
toxicity.
The dicationic structure of pentamidine is essential for selective
and rapid accumulation in the parasite and for its pharmacological
effect against trypanosomes.22–26 Owing to the highly ionic nature
of the diamidines, these drugs are not absorbed from the gastroin-
testinal tract and do not readily pass the blood–brain barrier to
achieve therapeutic levels in the brain. In contrast to the active
principle pentamidine (8), the pentamidine diamidoxime (5)
shows oral bioavailablility.5,27 However, for therapy of late-stage
sleeping sickness, where the infection is established in the central
nervous system, only eflornithine and the arsenic containing drug
melarsoprol are effective.28 Arsenical drugs are highly toxic and
increasing occurrence of resistance is becoming a major problem.
Therefore, it is essential to develop drugs that can penetrate the
blood–brain barrier and thus replace those highly toxic drugs.29
In this respect
Previously, our group improved the lipophilicity and oral bio-
availability of the pentamidine diamidoxime prodrug (5) by O-
acetylation forming N,N0-bis(acetoxy)pentamidine. It was antici-
pated that N,N0-bis(acetoxy)pentamidine would also be able to
pass the blood–brain barrier via passive diffusion.30 However, a
major problem of N,N0-bis(acetoxy)pentamidine is its low water
solubility, which limits its usefulness for therapeutic application
and has to be considered critically regarding absorption mecha-
nisms in the gastrointestinal tract.
To date, several prodrug approaches to improve water solubility
have been described, including salt formation and incorporation of
ionizable and polar groups into molecules.31,32 Especially esters
such as phosphates or hemisuccinates are extensively used to in-
crease the water solubility of drugs.1,33,34
mediated transporters and thus cross cell membranes. Based on
the results of previous screening studies, esterification with the
amino acid L-valine appeared to be the most promising candi-
date.35 The transport of valaciclovir through membranes was med-
iated predominantly by the PEPT1 transporter although this
prodrug did not possess a peptide bond and resulted in a higher
bioavailability.36 This novel double prodrug principle allows the
development of prodrugs that can be injected in case of emergency
or may be applied orally during long-term therapy. Furthermore,
these prodrugs may enable penetration through the blood–brain
barrier. In case of pentamidine (8), this new approach might im-
prove therapy, eliminating toxic effects caused by injection. With
the double prodrug principle, tissue targeting of the drug may be
possible.
The first aim of the study was the synthesis of the model com-
pound N-valoxybenzamidine (1) and the investigation of its bio-
transformation in vitro and in vivo. Of special interest were the
oral bioavailability and the tissue distribution after application of
N-valoxybenzamidine (1). Afterwards, the prodrug principle was
transferred to N,N0-bis(valoxy)pentamidine (7) and we examined
water solubility, stability, in vitro bioactivation, and organ distri-
bution of 7 including postabsorptive conversion to the active
metabolite pentamidine (8) in rats.
2. Materials and methods
2.1. Chemicals and reagents
Benzamidoxime (2) was synthesized according to a prior pub-
lished method.37 Pentamidine diamidoxime (5) was synthesized
according to the method described by Clement and Raether.13 Re-
combinant human mARC1 and 2 were obtained as described by
Kotthaus et al.38 All other chemicals were commercially available
and of analytical grade, except acetonitrile and methanol, which
were of HPLC grade.
A new double prodrug principle was developed for amidines in
order to maintain their water solubility and at the same time im-
prove bioavailability. This will be achieved by O-substitution of
2.2. Synthesis
the amidoxime prodrug with the amino acid L-valine. Activation
Uncorrected melting points were measured on a Büchi 510
Melting Point apparatus. IR spectra were obtained on a Perkin El-
mer FTIR 1600 PC spectrophotometer. 1H and 13C NMR spectra
were recorded on a Bruker ARX 300 NMR spectrometer using the
following frequencies: 1H: 300.13 MHz and 13C: 75.47 MHz. Chem-
ical shifts (d values) are reported in ppm relative to TMS as an
internal standard. All coupling constants (J values) were obtained
of the double prodrug proceeds in two steps: first the prodrug is
hydrolyzed to its amidoxime, and subsequent is reduced to the ac-
tive compound (Scheme 1).
Another aim was enabling the drug to penetrate through the
blood–brain barrier by peptide transporters. Hence, the prodrug
contains an amino acid moiety that can potentially target carrier-
Scheme 1. Metabolic activation of N-valoxybenzamidine (1) and N,N0-bis(valoxy)pentamidine (7).

J. Kotthaus et al. / Bioorg. Med. Chem. 19 (2011) 1907–1914
1909
by first order analysis of the multiplets and are quoted in Hertz.
The following NMR abbreviations are used: br (broad), s (singlet),
d (doublet), t (triplet), qn (quintet), m (unresolved multiplet).
Low resolution electrospray ionisation mass spectra were recorded
on a Bruker Esquire-LC mass spectrometer. The substances were
dissolved in acetonitrile or methanol. Recording of the high mass
resolution spectrum was on a Bruker FT-ICR APEX II spectrometer,
electrospray ionization. The substance was dissolved in methanol.
Elemental analysis was performed on a CHNS analysator (HEKA-
tech GmbH) at the department of inorganic chemistry (Christian-
Albrechts-University of Kiel, Germany).
[DAOꢀ2H2O+H]+; Anal. Calcd for C39H58N6O10 (770.93): C, 60.76;
H, 7.58; N, 10.90. Found: C, 60.98; H, 7.87; N, 11.21.
2.2.4. 4,40-(Pentamethylenedioxy)-bis-(2-amino-3-methyl-O-
butyryl)benzamidoxime tetrahydrochloride.
N,N0-bis(valoxy)pentamidine (7)
Boc-protected 6 was dissolved in anhydrous Et2O and gaseous
HCl was bubbled through the solution for 5 min. The mixture
was stirred overnight at room temperature and the isolated precip-
itate washed several times with Et2O. Yield 59%; IR (KBr):
m
~ 3396,
2942, 1804, 1752, 1636, 1608, 1506, 1308, 1260 cmꢀ1
;
1H NMR
(DMSO-d6): d 1.02 (d, 6H, 3J = 6.8 Hz, CH3), 1.03 (d, 6H, 3J = 6.8 Hz,
CH3), 1.58 (qn, 2H, 3J = 6.6 Hz, CH2), 1.80 (qn, 4H, 3J = 7.0 Hz,
CH2), 2.32 (m, 2H, CH(CH3)2), 3.92 (t, 2H, 3J = 5.3 Hz, CH–NH),
4.05 (t, 4H, 3J = 6.3 Hz, O–CH2), 7.00 (d, 3J = 8.9 Hz, 4H, Ar-H), 7.11
(s, 4H, NH2), 7.67 (d, 3J = 8.8 Hz, Ar-H), 8.34 (br s, 2H, NH), 8.81
(s, 6H, NH3); 13C NMR (DMSO-d6): d 17.9, 18.4, 22.1, 28.3, 29.4,
57.3, 67.6, 114.2, 122.9, 128.3, 157.2, 160.6, 165.5; HRMS (ESI)
m/z calcd for C29H43N6O6 (MH+) 571.32386, found 571.32385;
Anal. Calcd for C29H46N6O6Cl4 ꢁ 1.9H2O (750.77): C, 46.40; H,
6.69; N, 11.19. Found: C, 46.52; H, 6.93; N, 10.98.
2.2.1. 2-(N-tert-Butoxycarbonylamino)-3-methyl-O-
butyrylbenzamidoxime (4)
Benzamidoxime (2) (1.21 g, 8.9 mmol) was dissolved in anhy-
drous acetonitrile. After addition of N-t-butoxycarbonyl-L-valine
(2.60 g, 12.0 mmol), 4-DMAP (0.15 g, 0.80 mmol) and DCC the solu-
tion was stirred for 24 h at room temperature. The solvent was
evaporated under reduced pressure and the crude product was
recrystallized from H2O/EtOH (1/3, v/v). Yield 86%; mp: 147 °C;
IR (KBr): m ;
~ 3496, 2928, 1744, 1686, 1612, 1392, 1374 cmꢀ1 1H
NMR (DMSO-d6): d 0.83 (d, 6H, 3J = 6.4, CH3), 1.40 (s, 9H, CH3),
2.08 (m, 1H, CH(CH3)2), 4.05 (t, 1H, 3J = 8.0, CH–NH), 6.85 (s, 2H,
NH2), 7.31 (d, 1H, 3J = 9.2, NH), 7.44 (m, 3H, Ar-H), 7.52 (m, 2H,
Ar-H); 13C NMR (DMSO-d6): d 18.3, 19.0, 28.2, 30.1, 59.1, 78.3,
126.7, 128.3, 130.5, 131.4, 155.7, 157.0, 168.9; LRMS (ESI) m/z:
693 [2M+Na]+, 336 [M+H]+, 280 [Mꢀ(CH3)2C@CH2+H]+, 137
[BAO+H]+, 118 [Val+H]+; Anal. Calcd for C17H25N3O4 (335.41): C,
60.88; H, 7.51; N, 12.53. Found: C, 60.96; H, 7.71; N, 12.63.
2.3. In vitro stability of N-valoxybenzamidine (1) and N,N0-
bis(valoxy)pentamidine (7)
Both compounds were incubated in concentrations of 100 lM
at 37 °C in 100 mM phosphate buffer pH 2.0, 6.3, and 7.4. Samples
were taken at different timepoints between 0 and 120 min and
analyzed by HPLC.
2.2.2. 2-Amino-3-methyl-O-butyrylbenzamidoxime
dihydrochloride. N-valoxybenzamidine (1)
Boc-protected 4 was dissolved in anhydrous dioxane. Gaseous
HCl was flushed through the solution for 5 min and the mixture
was stirred overnight at room temperature. The product was pre-
cipitated by addition of EtOAc and afterwards washed several
2.4. In vitro biotransformation of N-valoxybenzamidine (1) and
N,N0-bis(valoxy)pentamidine (7) by subcellular enzyme
preparations and mARC
In vitro assays were performed with human and porcine en-
zyme preparations, such as microsomes, mitochondria, 9000g
supernatants, and cytosolic fractions. Human liver and kidney
samples for enzyme preparations were obtained from the Univer-
sity Medical Center Schleswig-Holstein (Kiel, Germany). Donors
were patients who were subjected to a partial hemi-hepatectomy
because of secondary liver tumors. These studies were accredited
from the local medical ethics committee and from the patients. A
standard incubation mixture contained 0.3 mg protein and 1 mM
prodrug (N-valoxybenzamidine (1) or N,N0-bis(valoxy)pentamidine
(7)) in 100 mM potassium phosphate buffer pH 6.3. After preincu-
bating the mixture for 5 min at 37 °C under aerobic conditions, the
reaction was started by the addition of NADH (final concentration
times with Et2O. Yield 30%; IR (KBr):
m~ 3238, 3052, 2910, 2774,
1800, 1682, 1600, 1374 cmꢀ1 1H NMR (DMSO-d6): d 1.02 (d, 3H,
;
3J = 4.7 Hz, CH3), 1.04 (d, 3H, 3J = 4.7 Hz, CH3), 2.02 (m, 1H,
CH(CH3)2), 3.94 (t, 1H, 3J = 5.4, CH–NH2), 7.22 (s, 2H, NH2), 7.48
(m, 3H, Ar-H), 7.72 (m, 2H, Ar-H), 8.80 (s, 3H, NH3), 9.03 (br s,
1H, NH); 13C NMR (DMSO-d6): 18.1, 29.4, 57.2, 126.8, 128.3,
130.6, 130.1, 157.6, 165.40; LRMS (ESI) m/z: 493 [2M+Na]+, 471
[2M+H]+, 236 [M+H]+, 137 [BAO+H]+, 121 [BA+H]+, 119
[BAOꢀH2O+H]+, 118 [Val+H]+; Anal. Calcd for C12H19N3O2Cl2
(308.21): C, 46.76; H, 6.21; N, 13.63. Found: C, 46.54; H, 6.25; N,
13.12.
1 mM) to a final volume of 250
30 min at 37 °C under aerobic conditions. The reaction was termi-
nated by the addition of 250 L cold acetonitrile and vortexing.
After centrifugation at 10000g (Mikrozentrifuge Hettich, Tuttlin-
gen, Germany), 10 L of the supernatant were analyzed by HPLC.
lL. The mixture was incubated for
2.2.3. 4,40-(Pentamethylenedioxy)-bis-[2-(N,N0-tert-
butoxycarbonyl-amino)-3-methyl-O-butyryl]benzamidoxime
(6)
l
Pentamidine diamidoxime (5) (1.0 g, 2.7 mmol) was dissolved
l
in anhydrous acetone. After addition of N-t-butoxycarbonyl-
L
-va-
Additionally, incubations of both prodrugs with recombinant
human mARC1 and mARC2 were performed according to a previ-
ously published method.38 Those incubations contained 0.5 mM
substrate (1 or 7), 1 mM NADH, 2 U carboxyl-esterases from pig li-
ver (Sigma–Aldrich, Taufkirchen, Germany), 76 pmol cytochrome
b5, 7.6 pmol NADH cytochrome b5 reductase, and 76 pmol recombi-
line (1.77 g, 8.2 mmol), 4-DMAP (0.1 g, 0.8 mmol) and DCC
(1.67 g, 8.1 mmol) the solution was stirred for 24 h at room tem-
perature. The solvent was evaporated in vacuo and the crude prod-
uct was recrystallized from H2O/EtOH/acetone (1/3/1, v/v/v). Yield
35%; mp: 174 °C; IR (KBr):
m~ 3494, 2964, 1744, 1690, 1616, 1392,
1366 cmꢀ1 1H NMR (DMSO-d6): d 0.91 (d, 12H, 3J = 6.6 Hz, CH3),
;
nant hmARC1 or hmARC2 in 150
After incubation at 37 °C for 30 min the reactions were terminated
by the addition of 150 L acetonitrile. Precipitated proteins were
lL 20 mM MES buffer pH 6.0.
1.40 (s, 18H, CH3), 1.58 (qn, 2H, 3J = 6.9 Hz, CH2), 1.81 (qn, 4H,
3J = 7.6 Hz, CH2), 2.07 (m, 2H, CH(CH3)2), 4.06 (m, 6H, O–CH2,
CH–NH), 6.71 (s, 4H, NH2), 7.00 (d, 4H, 3J = 8.8 Hz, Ar-H), 7.30 (d,
2H, 3J = 8.8 Hz, NH), 7.66 (d, 4H, 3J = 8.7 Hz, Ar-H); 13C NMR
(DMSO-d6): d 18.2, 19.0, 22.1, 28.1, 28.3, 30.1, 59.1, 67.5, 78.2,
114.1, 123.3, 128.1, 155.7, 156.5, 160.4, 168.9; LRMS (ESI) m/z:
771 [M+H]+, 572 [MꢀC10H17NO3+H]+, 373 [DAO+H]+, 337
l
sedimented by centrifugation and the supernatant was analyzed
by HPLC.
Further incubations were performed wit both prodrugs and 0.5
U carboxyl-esterases from pig liver (Sigma–Aldrich, Taufkirchen,
Germany) in 100 mM potassium buffer pH 7.4.

1910
J. Kotthaus et al. / Bioorg. Med. Chem. 19 (2011) 1907–1914
2.5. In vitro biotransformation of N-valoxybenzamidine (1) by
cultured pig hepatocytes
methanol (90/10, v/v) and concentrations of N-valoxybenzamidine
(1), benzamidoxime (2), and benzamidine (3) were determined by
HPLC analysis.
The isolation and cultivation of porcine hepatocytes was carried
out according to the method published by Fröhlich et al.39 Incuba-
tions were performed on day three after culturing. Hepatocytes
were incubated with 1 mM substrate in Krebs–Henseleit buffer
pH 7.4 for 1 h at 37 °C and 5% CO2 in humidified air. The incuba-
tions were stopped by aspirating the supernatant from the culture
dishes. The supernatant was frozen at ꢀ20 °C and freeze-dried.
2.8. Tissue preparation
2.8.1. Application of N-valoxybenzamidine (1)
The organs were stored at ꢀ80 °C. After thawing they were
washed and homogenized. The homogenate was resuspended in
600
tion of 600
the supernatant was concentrated to dryness with nitrogen, resus-
pended in 100 water/methanol (90/10, v/v), centrifuged
(10000g, 10 min), and 10 L were analyzed by HPLC.
lL water, centrifuged (5000g, 5 min), and precipitated by addi-
Afterwards, the residue was taken up in 150
lL of 100 mM phos-
l
L acetonitrile. After centrifugation (10000g, 10 min),
phate buffer (pH 6.3)/methanol (90/10, v/v). 10
lL were analyzed
by HPLC.
lL
l
2.6. Animal study
2.8.2. Application of N,N0-bis(valoxy)pentamidine (7)
Animal studies were approved by the department for agricul-
ture, environment and rural areas of the german state Schleswig
Holstein. Male Wistar rats (n = 2 or 3 for each substance, initial
body weight 230 10 g, Charles River Laboratories, Germany) were
housed in cages with sawdust-covered solid flooring in a con-
trolled environment (22 2 °C, humidity 65%) with a 12-h light/
dark cycle. Animals had free access to feed (maintenance labora-
tory chow Altromin, Lage) and tap water. The rats were anesthe-
The work up procedure for the organs is similar to N-
valoxybenzamidine (1). After concentration of the supernatant the
residue was resuspended in 150
lL water/methanol (50/50, v/v).
2.9. HPLC methods
2.9.1. In vitro investigations
2.9.1.1. Analysis of N-valoxybenzamidine (1). Analysis were car-
ried out on a Waters Alliance System consisting of a Waters e2695
XC separations module and a Waters 2995 photodiode array detec-
tor (260 nm) with column heater according to a prior published
method.41 Separations were carried out at room temperature on
tized by intraperitoneal injection of
a mixture of xylazine
hydrochloride (10 mg kgꢀ1
)
and ketamine hydrochloride
(75 mg kgꢀ1) and permanent canulas were surgically inserted into
the carotid artery and the jugular vein. The catheters were externa-
lised in the nape of the neck, and filled with physiological saline
containing heparin (50 I.E. mlꢀ1). The animals were allowed to re-
cover from surgery until complete compensation of body weight
loss due to surgery. Bioavailability was tested in animals starved
for 16 h and substances were applied either by oral gavage or intra-
venous injection. Blood withdrawal was performed via arterial
catheter. Eight hours after prodrug application, animals were anes-
thetized by intraarterial pentobarbital injection and euthanized by
complete blood withdrawal from the abdominal aorta. Tissues
(lung, brain, heart, liver, spleen, and kidney) were removed and
sampled for subsequent analyses. Principles of laboratory animal
care (NIH Publication No. 85-23, revised 1985) were followed
and all animal experiments were approved by the Ethics Commit-
tee according to German Animal Protection Law. Each rat received
an oral dose of N-valoxybenzamidine (1) of 50 mg kgꢀ1 or an oral
a LiChropsher 60 RP-select B column (125 ꢁ 4 mm; 5
lm, Merck,
Darmstadt, Germany) with a RP-select B (4 ꢁ 4 mm) guard column.
The mobile phase consisted of 0.1% trifluoroacetic acid with 20 mM
potassium phosphate buffer (pH 6.5)/acetonitrile (65/35, v/v). The
flow rate was set to 1 mL minꢀ1 and the wavelength to 229 nm.
Retention times were 1.9 0.1 min for benzamidoxime (2),
3.5 0.1 min for benzamidine (3), and 5.2 0.2 min for N-
valoxybenzamidine (1). The limit of quantification was 0.5 lM
for all metabolites.
2.9.1.2. Analysis of N,N0-bis(valoxy)pentamidine (7). Analysis
were performed on a LiChrospher 60 RP-select B (125 ꢁ 4 mm,
5
l
m; Merck, Darmstadt, Germany) with
a
RP-select
B
,
(4 ꢁ 4 mm) guard column. The flow rate was set to 1 mL minꢀ1
the column temperature 25 °C, the wavelength to 260 nm, and
dose of N,N0-bis(valoxy)pentamidine (7) of 100 mg kgꢀ1
.
the injection volume was 10 lL. The mobile phase for stability
Compounds 1 and 7 were dissolved in 100 mM phosphate buffer
pH 6.3. Benzamidine (3) was dissolved in water for injection and
investigations consisted of 0.1% TFA pH 2.5 and methanol (42/58,
v/v). N,N0-bis(valoxy)pentamidine (7) was eluted after 2.5
0.1 min. The mobile phase for metabolism studies consisted of
30 mM octanesulfonic acid, 20 mM tetramethylammoniumchlo-
ride, pH 3.0 and methanol (48/52, v/v). The retention times were
7.1 0.3 min for pentamidine diamidoxime (5), 8.3 0.3 min for
pentamidine monoamidoxime and 9.7 0.4 min for pentamidine
administered intravenously in a concentration of 10 mg kgꢀ1
.
Blood samples (300 lL) were taken at 30, 60, 90, 120, 240, and
480 min after oral application and at 10, 20, 40, 60, 120, and
360 min after injection. Due to the high tendency of pentamidine
(8) to accumulate in tissues,40 no plasma samples were taken after
oral application of N,N0-bis(valoxy)pentamidine (7).
(8). The limit of quantification was 1 lM for all compounds.
2.7. Processing of plasma samples
2.9.2. In vivo investigations
2.9.2.1. Analysis of N-valoxybenzamidine (1). Separation was
For the application of N-valoxybenzamidine (1) and benzami-
dine (3) blood samples were drawn into heparinized containers,
immediately centrifuged (1500g, 10 min, 4 °C) and stored at
ꢀ80 °C. For analysis of N-valoxybenzamidine (1) and benzamidine
(3), samples were worked up by solid phase extraction (Strata X,
carried out at room temperature by isocratic elution on a Synergy
Max RP 80A column (250 ꢁ 4.6 mm; 5
lm, Phenomenex, Aschaff-
enburg, Germany) with a C 12 (4 ꢁ 2 mm) guard column. The
eluate was monitored at 229 nm. The mobile phase consisted of
octanesulfonic acid 10 mM, pH 2.5 and acetonitrile (82.5/17.5,
v/v). The retention times were 23.5 0.5 min for benzamidoxime
(2) and 25.6 0.5 min for benzamidine (3). Standard curves for
each metabolite were generated that were linear with correlation
coefficients >0.99. Precision of the assays and accuracy were
33
equilibration and conditioning of the column, 150
luted with 450 L aqua bidest were added. The column was
washed with 600 L water and the substances were eluted with
1000 L water (pH 3)/methanol (40/60, v/v). The samples were
concentrated to dryness. The residues were taken up in water/
lm
column, Phenomenex, Aschaffenburg, Germany). After
lL plasma di-
l
l
l
assessed by adding different concentrations 0–50
lM of ben-
zamidoxime (2) and benzamidine (3) to plasma. Standard curves

J. Kotthaus et al. / Bioorg. Med. Chem. 19 (2011) 1907–1914
1911
were obtained by linear regression analysis. The recoveries were
and samples was 50
lL. The retention times were 10.6 0.3 min
100.8 13.2% and 96.6 9.3% for benzamidoxime (2) and
benzamidine (3), respectively. The limit of quantification was
for pentamidine diamidoxime (5), 11.5 0.3 min for pentamidine
monoamidoxime, and 12.0 0.4 min for pentamidine (8). The limit
about 0.25
lM for both, benzamidoxime (2) and benzamidine (3).
of quantification was 0.1 lM for all compounds. Electrospray ioni-
For analysis of N-valoxybenzamidine (1) in different organs, the
HPLC method described for in vitro investigations was used. The
limit of quantification after working up the organs was about
zation (ESI) mass spectra were recorded in the positive-ion mode
on an Esquire-LC mass spectrometer (Bruker, Bremen, Germany).
Synthetic standards were analyzed in comparison to metabolites
from the in vivo samples. For diamidoxime, ions with m/z 373
[M+H]+, 355 [MꢀH2O]+, and 187 [M+2H]2+ were found. For the
monoamidoxime, ions with m/z 357 [M+H]+ and 179 [M+2H]2+
were characteristic. Pentamidine (8) showed characteristic ions
with m/z 341 [M+H]+, 205 [MꢀC7H8N2O]Å+, 171 [M+2H]2+, and
137 [C7H8N2O+H]+. The samples were assessed in full scan mode
in the range of m/z 50–650 to detect other possible metabolites.
Peak areas of ion chromatograms with the sum of ions at m/z were
obtained by Data Analysis™ integration software (Data Analysis™,
Version 3.0, Bruker Daltonics). Standard curves were linear over
this range with correlation coefficients >0.995.
0.375 lM for both, N-valoxybenzamidine (1) and benzamidine (3).
2.9.2.2. Analysis of N,N0-bis(valoxy)pentamidine (7). Separations
were performed analogously to those described for the in vitro
investigations with the slight modification that the concentration
of octanesulfonic acid was increased to 30 mM. The retention times
were 12.9 0.6 min for pentamidine diamidoxime (5), 15.4
0.5 min for pentamidine monoamidoxime and 18.5 0.5 min for
pentamidine (8). The limit of quantification was 1 lM for all com-
pounds. Standard curves for each metabolite were generated and
found to be linear with correlation coefficients >0.999. Precision
of the assays and accuracy were assessed by adding various
concentrations (i.e., 0–50 lM) of the diamidoxime (5), mono-
3. Results and discussion
3.1. Chemistry
amidoxime, and pentamidine (8) to plasma. Recovery of pentami-
dine (3) from plasma and samples, which contained the same
amount of each metabolite dissolved in water was 107.3 9.8%
and recovery of the monoamidoxime was 125.7 11.2%. The limit
Synthesis of the target compounds (i.e., N-valoxybenzamidine
(1) and N,N0-bis(valoxy)pentamidine (7) was carried out as shown
in Scheme 2.
of quantification was about 0.5
lM for all detected metabolites.
2.9.2.3. LC/MS method. An Agilent model 1100 HPLC system
In the first step, Boc-protected L-valine was esterified with ben-
equipped with a binary pump G1312 A and a Hewlett Packard Deg-
asser G1322 A was used for mass spectrometric analysis. Separa-
tions were carried out at room temperature on a LiChrospher 100
zamidoxime (1) using N,N0-dicyclohexylcarbodiimide (DCC) and 4-
(N,N-dimethylamino)pyridine (4-DMAP) in anhydrous acetonitrile.
Afterwards, the protecting group was cleaved with gaseous HCl in
dry dioxane. N,N0-bis(valoxy)pentamidine (7) was synthesized
analogously, only differing in the solvents used. The new prodrug
of pentamidine possessed excellent solubility in aqueous media.
5 mM of 7 were soluble in 50 mM phosphate buffer pH 6.3. In
RP-18 column (250 ꢁ 4 mm; 5
lm, Merck, Darmstadt, Germany)
and a RP-select B guard column (4 ꢁ 4 mm). The detection wave-
length was 260 nm. The mobile phase consisted of 0.1% trifluoro-
acetic acid in aqua bidest and acetonitrile (76/24, v/v). The flow
rate was 0.7 mL minꢀ1. The injection volume for both standards
Scheme 2. (A) Synthesis of N-valoxybenzamidine (1) and (B) N,N0-bis(valoxy)pentamidine (7). Reagents: (a) tert-butoxycarbonyl-
HCl(g), dioxane; (c) tert-butoxycarbonyl- -valine, DCC, 4-DMAP, acetone; (d) HCl(g), ether.
L-valine, DCC, 4-DMAP, acetonitrile; (b)
L

1912
J. Kotthaus et al. / Bioorg. Med. Chem. 19 (2011) 1907–1914
comparison, only 50 lM
of N,N0-bis(acetoxy)pentamidine and
Table 1
In vitro activation of N-valoxybenzamidine (1) and N,N0-bis(valoxy)pentamidine (7)
by human and porcine liver and kidney microsomes, mitochondria, 9000g superna-
tant, and cytosola
100 lM pentamidine diamidoxime (5) were soluble in that media.
3.2. In vitro stability of N-valoxybenzamidine (1) and N,N0-
bis(valoxy)pentamidine (7)
Enzyme source
Activation of 1 to
benzamidine (3)
Activation of 7 to
pentamidine (8)
b
[nmol minꢀ1 (mg protein)ꢀ1
]
Stability studies revealed that cleavage of the ester can occur by
chemical hydrolysis (Fig. 1). N-valoxybenzamidine (1) was rela-
tively stable at pH 6.3 and pH 7.4, whereas the degradation to-
wards benzamidoxime (2) increased rigorously at pH 2.0
resulting in an immediate and complete hydrolysis to 2. In con-
trast, hydrolysis of N,N0-bis(valoxy)pentamidine (7) is quite inde-
pendent from pH-value. Interestingly, stability of compound 7 is
increased compared to 1, which might be a result from the substi-
tution with the O-alkyl-residue. Consequently, in subsequent ani-
mal studies, prodrug 1 was administered in a buffered solution
to prevent hydrolysis during gastrointestinal transit. Furthermore,
complete hydrolysis of both prodrugs was observed in the pres-
Pig liver microsomes
Pig liver mitochondria
Pig liver cytosole
Pig liver 9000g supernatant
Pig kidney mitochondria
Pig kidney microsomes
Pig kidney 9000g supernatant
7.0 0.3
8.6 0.3
1.2 0.2
3.4 0.5
8.8 0.9
10.4 0.3
5.6 1.0
0.5 0.2
1.0 0.1
ND
0.4 0.1
0.8 0.1
0.7 0.2
0.2 0.1
Human liver microsomes
Human liver mitochondria
Human liver cytosol
Human liver 9000g supernatant
Human kidney microsomes
Human kidney 9000g supernatant
Human kidney cytosol
0.9 0.1
1.8 0.1
ND
0.2 0.1
1.3 0.1
ND
0.3 0.1
0.2 0.1
ND
0.2 0.1
0.3 0.1
ND
ND
2.0 0.1
ND
0.4 0.2
ence of 0.5 U esterases/250 lL (data not shown).
Human kidney mitochondria
3.3. In vitro metabolism of N-valoxybenzamidine (1) and N,N0-
bis(valoxy)pentamidine (7)
ND, not detectable <0.06 nmol minꢀ1 (mg protein)ꢀ1
.
a
For detailed protocols see Section 2.4. An incubation mixture consisted of
0.3 mg protein, 1 mM prodrug, 1 mM NADH as cosubstrate in 100 mM potassium
phosphate buffer pH 6.3.
In vitro metabolism of both N-valoxybenzamidine (1) into ben-
zamidine (3) and N,N0-bis(valoxy)pentamidine (7) into pentami-
dine (8) was demonstrated by porcine and human microsomes,
mitochondria, 9000g supernatants, and cytosolic fractions. The
turnover rates determined are shown in Table 1. The highest
amounts of benzamidine (3) and pentamidine (8) were detected
when using microsomal or mitochondrial preparations. The cyto-
solic fractions were mainly able to cleave the ester but no further
reduction of the amidoxime was observed. The enzyme responsible
for reduction is bound to membranes and therefore not present in
the cytosolic fractions. Human tissues were obtained from patients
suffering from cancer or hepatitis, possibly resulting in lower or
modified enzyme levels in comparison with healthy persons. Thus,
a lack of quality of human tissues, from which microsomes and
mitochondria were extracted, might be a reason for lower turnover
rates in these experiments.
As expected, turnover rates of the pentamidine prodrug 7 are
considerably lower than those of N-valoxybenzamidine (1). This
observance can easily be explained by the more complex activation
of 7 that requires hydrolysis and reduction of two N-valoxyamidine
groups. mARC has recently been identified as a 35 kDa molybdo-
protein and has been demonstrated to participate predominantly
in the reduction of amidoximes or guanidines.11,12,38 Our results
b
Turnover rates are means SD of two independent experiments measured
twice.
demonstrate an excellent activation of both N-valoxybenzamidine
´
(1) and N,N-bis(valoxy)pentamidine (7) by hmARC1 and hmARC2
in the presence of cytochrome b5 and NADH cytochrome b5 reduc-
tase (Table 2).
Hepatocytes represent a well-accepted model for the simula-
tion of in vivo biotransformation. In our laboratory, the biotrans-
formation of several amidoximes has already been studied with
human and porcine hepatocytes.39,42 The present study demon-
strates that N-valoxybenzamidine (1) was activated by porcine
hepatocytes. Turnover rates of 650 227 and 175 39 pmol minꢀ1
(mg protein)ꢀ1 were obtained for the biotransformation of N-
valoxybenzamidine (1) to benzamidoxime (2) and benzamidine
(3), respectively, and indicate the perfect suitability of this prodrug
principle.
3.4. In vivo metabolism of N-valoxybenzamidine (1)
N-Valoxybenzamidine was applied orally at 50 mg kgꢀ1 body
weight to three male Wistar rats in 100 mM phosphate buffer pH
Figure 1. Effect of different pH values on the stability of N-valoxybenzamidine (1) and N,N0-bis(valoxy)pentamidine (7). The reaction contained 100
lM prodrug in 100 mM
potassium phosphate buffer pH 2.0, pH 6.3, and pH 7.4.

J. Kotthaus et al. / Bioorg. Med. Chem. 19 (2011) 1907–1914
1913
Table 2
In plasma, mainly benzamidine (3) was detected. Maximum
plasma concentrations of up to 24 M were observed (Table 3).
In vitro activation of N-valoxybenzamidine (1) and N,N0-bis(valoxy)pentamidine (7)
by the recombinant human mitochondrial Amidoxime Reducing Component
(hmARC)a
l
In contrast, benzamidoxime (2) was found in only one sample,
which was taken 30 min after oral application of 1. The course of
the plasma concentration–time curves shows that the prodrug
was quickly absorbed from the gastrointestinal tract into the blood
and subsequently activated to 3. Maximum plasma concentration
values of benzamidine (3) were already detected 30 min after oral
application of the prodrug. Mean oral bioavailability of benzami-
dine (3) after application of N-valoxybenzamidine (1) was about
88% (Table 3). Previous studies revealed that oral bioavailability
of benzamidine after application of benzamidoxime to rats was
about 75%.43
Activation of 1 to
benzamidine (3)
Activation of 7 to
pentamidine (8)
b
[nmol minꢀ1 (mg protein)ꢀ1
]
Standard incubation mixture
With hmARC1
With hmARC2
67.4 14.7
43.7 4.2
48.8 10.9
10.9 0.9
a
For detailed protocols see Section 2.4. An incubation mixture consisted of
0.5 mM prodrug, 1 mM NADH, 2 U carboxyl-esterase, 76 pmol cytochrome b5,
7.6 pmol NADH cytochrom b5 reductase, and 76 pmol of hmARC1 or 2 in 20 mM
MES buffer pH 6.0.
b
In order to investigate distribution of benzamidine (3) we
examined six organs (i.e., lung, brain, heart, liver, spleen, and
kidney) that were removed 8 h after oral application of N-
valoxybenzamidine 6 (Fig. 3).
Turnover rates are means SD of three incubations.
No N-valoxybenzamidine (1) or benzamidoxime (2) was de-
tected in any organ, whereas benzamidine (3) was detected in
lung, heart, spleen, liver, and kidney with highest concentrations
of 3 in the metabolically most active organs, liver and kidney.
Unfortunately, no metabolite was found in the brain, probably
due to extensive metabolism of N-valoxybenzamidine (1) to ben-
zamidine (3) so that the prodrug did not reach the blood–brain
barrier in sufficiently high concentrations (detection limit of ben-
zamidine P 0.375 lM). Furthermore, efflux mechanisms might
have prevented accumulation in the brain.44 Insufficient bleeding
might be responsible for the high concentrations in spleen.
3.5. In vivo metabolism of N,N0-bis(valoxy)pentamidine (7)
Figure 2. Plasma concentration–time curves of benzamidine (3) after oral appli-
cation N-valoxybenzamidine (1) to three rats at 50 mg kgꢀ1
.
N,N0-bis(valoxy)pentamidine (7) was given orally at 100 mg kgꢀ1
to two rats. Analysis of tissue samples revealed a rapid and
almost complete activation of the prodrug 7 into the drug pentami-
dine (8). Thus, no concentrations of the prodrug 7 and the pentami-
dine diamidoxime (5) were detected in the tissues examined,
whereas pentamidine (8) and the intermediate pentamidine mono-
amidoxime were found in considerable concentrations. Pentami-
dine monoamidoxime was detected in the lung in concentrations
of 200 10 ng gꢀ1 tissue and in the kidney in concentrations of
300 20 ng gꢀ1 tissue. The active form pentamidine (8) was found
in every organ, but mainly in kidney and liver (Fig. 3). The identity
of pentamidine monoamidoxime and pentamidine (8) in all organs
was confirmed by HPLC and LC/MS.
Table 3
Pharmacokinetic parameters of benzamidine (3) after oral application of N-valoxy-
benzamidine (1) to three rats at 50 mg kgꢀ1
Rat 1
Rat 2
Rat 3
Mean SD
tmax (min)
cmax M)
t1/2 (min)
30.0
24.2
267.9
356.5
154.5
0.4
90.0
23.6
132.9
219.3
122.8
0.6
30.0
22.1
125.1
202.3
107.9
0.6
50 35
23.3 1.1
175.3 80.3
259.4 84.5
128.4 23.8
0.5 0.1
(l
MRT (min)
Vd (dL kgꢀ1
)
Cl (dL minꢀ1 kgꢀ1
)
Bioavailability (%)
118.8
74.5
69.8
87.7 27.0
Interestingly, pentamidine (8) was detected in the brain in con-
centrations of 200 20 ng gꢀ1 organ. The higher lipophilicity of the
pentamidine-prodrug 7 compared to the benzamidine-prodrug 1
could be the reason pentamidine (8) being able to pass the
6.3 that should elevate the gastric pH and minimize hydrolysis of 1
prior to absorption. The corresponding plasma concentration–time
curves of benzamidine (3) are shown in Figure 2.
Figure 3. Detected concentrations of benzamidine (3) and pentamidine (8) in lung, brain, heart, liver, spleen, and kidney after oral application of N-valoxybenzamidine (1)
(50 mg kgꢀ1, n = 3) and N,N0-bis(valoxy)pentamidine (7) (100 mg kgꢀ1, n = 2), respectively.

1914
J. Kotthaus et al. / Bioorg. Med. Chem. 19 (2011) 1907–1914
blood–brain barrier in small amounts. These results indicate that
N,N0-bis(valoxy)pentamidine (7) might possibly be efficacious in
the treatment of the late-stage African sleeping sickness. However,
further studies are necessary to prove this assumption.
Due to the fact that pentamidine (1) is known to accumulate in
tissues, we determined tissue contents and no plasma concentra-
tions. Unfortunately, oral bioavailability is calculated by means of
plasma concentrations. Thus we cannot declare a precise value
for the oral bioavailability of N,N0-bis(valoxy)pentamidine (7).
However, concentrations detected in the tissues indicate an ade-
quate oral bioavailability the pentamidine prodrug 7.
References and notes
1. Ettmayer, P.; Amidon, G. L.; Clement, B.; Testa, B. J. Med. Chem. 2004, 47, 2393–
2404.
2. Albert, A. G.; Phillips, R. J. Chem. Soc. 1948, 2240–2249.
3. Young, R. C.; Mitchell, R. C.; Brown, T. H.; Ganellin, C. R.; Griffiths, R.; Jones, M.;
Rana, K.; Saunders, D.; Smith, I. R.; Sore, N. E. J. Med. Chem. 1988, 31, 656–671.
4. Peterlin-Masic, L.; Cesar, J.; Zega, A. Curr. Pharm. Des. 2006, 12, 73–91.
5. Clement, B. WO 9,501,168, 1995.
6. Clement, B.; Schmitt, S.; Zimmermann, M. Arch. Pharm. (Weinheim, Ger.) 1988,
321, 955–956.
7. Hauptmann, J.; Paintz, M.; Kaiser, B.; Richter, M. Pharmazie 1988, 43, 559–560.
8. Clement, B.; Mau, S.; Deters, S.; Havemeyer, A. Drug Metab. Dispos. 2005, 33,
1740–1747.
9. Clement, B. Drug Metab. Rev. 2002, 34, 565–579.
10. Bürenheide, A.; Kunze, T.; Clement, B. Basic Clin. Pharmacol. Toxicol. 2008, 103,
61–65.
4. Conclusions
11. Havemeyer, A.; Bittner, F.; Wollers, S.; Mendel, R.; Kunze, T.; Clement, B. J. Biol.
Chem. 2006, 281, 34796–34802.
Conjugation of amidoximes with the amino acid L-valine repre-
12. Grünewald, S.; Wahl, B.; Bittner, F.; Hungeling, H.; Kanzow, S.; Kotthaus, J.;
Schwering, U.; Mendel, R. R.; Clement, B. J. Med. Chem. 2008, 51, 8173–8177.
13. Clement, B.; Raether, W. Arzneim.-Forsch. 1985, 35, 1009–1014.
14. Denise, H.; Barrett, M. P. Biochem. Pharmacol. 2001, 61, 1–5.
15. Apted, F. I. Pharmacol. Ther. 1980, 11, 391–413.
16. Bryceson, A. D.; Chulay, J. D.; Mugambi, M.; Were, J. B.; Gachihi, G.; Chunge, C.
N.; Muigai, R.; Bhatt, S. M.; Ho, M.; Spencer, H. C. Trans. R. Soc. Trop. Med. Hyg.
1985, 79, 705–714.
17. Newman, S. P.; Simonds, A. K. Lung 1990, 168, 685–691.
18. Sands, M.; Kron, M. A.; Brown, R. B. Rev. Infect. Dis. 1985, 7, 625–634.
19. Jones, H. E.; Bundel, G. K.; Tidwell, R. R.; Hall, J. E.; Farr, S. J.; Richards, R. J.
Toxicology 1993, 80, 1–12.
20. Boykin, D. W.; Kumar, A.; Spychala, J.; Zhou, M.; Lombardy, R. J.; Wilson, W. D.;
Dykstra, C. C.; Jones, S. K.; Hall, J. E.; Tidwell, R. R. J. Med. Chem. 1995, 38, 912–
916.
21. Mdachi, R. E.; Thuita, J. K.; Kagira, J. M.; Ngotho, J. M.; Murilla, G. A.; Ndung’u, J.
M.; Tidwell, R. R.; Hall, J. E.; Brun, R. Antimicrob. Agents Chemother. 2009, 53,
953–957.
22. Matovu, E.; Seebeck, T.; Enyaru, J. C.; Kaminsky, R. Microbes Infect. 2001, 3, 763–
770.
23. Bray, P. G.; Barrett, M. P.; Ward, S. A.; de Koning, H. P. Trends Parasitol. 2003, 19,
232–239.
24. Das, B. P.; Boykin, D. W. J. Med. Chem. 1977, 20, 1219–1221.
25. Wilson, B. N.; Tanious, F. A.; Mathis, A.; Hall, J. E.; Stephens, C. E.; Boykin, D. W.
Curr. Med. Chem. 2005, 5, 389–408.
26. Neidle, S.; Pearl, L. H.; Skelly, J. V. Biochem. J. 1987, 243, 1–13.
27. Clement, B.; Immel, M.; Terlinden, R.; Wingen, F. J. Arch. Pharm. (Weinheim,
Ger.) 1992, 325, 61–62.
28. Barrett, M. P.; Fairlamb, A. H. Parasitol. Today 1999, 15, 136–140.
29. Soeiro, M. N.; De Souza, E. M.; Stephens, C. E.; Boykin, D. W. Expert Opin. Invest.
Drugs 2005, 14, 957–972.
30. Clement, B.; Bürenheide, A.; Rieckert, W.; Schwarz, J. ChemMedChem 2006, 1,
1260–1267.
31. Stella, V. J.; Martodihardjo, S.; Terada, K.; Rao, V. M. J. Pharm. Sci. 1998, 87,
1235–1241.
32. Varma, M. V.; Khandavilli, S.; Ashokraj, Y.; Jain, A.; Dhanikula, A.; Sood, A.;
Thomas, N. S.; Pillai, O.; Sharma, P.; Gandhi, R.; Agrawal, S.; Nair, V.;
Panchagnula, R. Curr. Drug Metab. 2004, 5, 375–388.
33. Huttunen, K. M.; Kumpulainen, H.; Leppanen, J.; Rautio, J.; Jarvinen, T.;
Vepsalainen, J. Synlett 2006, 701–704.
sents a new double prodrug principle leading to compounds that
are highly water soluble and have improved oral bioavailability
compared to the unmodified amidoximes. Our in vitro and
in vivo studies demonstrated the excellent suitability of this pro-
drug principle for amidines. Both N-valoxybenzamidine (1) and
N,N0-bis(valoxy)pentamidine (7) were activated in vitro by all en-
zyme preparations investigated (i.e., porcine and human subcellu-
lar enzyme fractions, hmARC1, hmARC2, and hepatocytes). The
activation relies on esterases and mARC and is thus independent
from P450 enzymes minimizing the risk for drug-drug interactions
and undesired side effects. Our results demonstrate the increase of
solubility in comparison to the amidoxime prodrugs. Moreover,
both prodrugs show excellent oral bioavailability. In addition to
absorption by diffusion, the transport by amino acid and peptide
transporters in the gastrointestinal tract is feasible and will be
the subject of further investigations. In vivo, an oral bioavailability
of benzamidine (3) of about 88% was observed after oral adminis-
tration of prodrug 1 to rats. The high bioavailability also excludes
the possibility of an ester hydrolysis prior to absorption.
N,N0-bis(valoxy)pentamidine (7) entered the cells of all tissues
investigated. This effect is essential for the antiprotozoic effect of
pentamidine (8) considering that the parasites are spread through-
out the body. Principally, 7 was activated completely to the drug
pentamidine (8). Also, its solubility was improved over 100-fold
in comparison to N,N0-bis(acetoxy)pentamidine and to pentami-
dine diamidoxime (5). These observations indicate that the devel-
opment of this new prodrug principle may also considerably
improve the treatment of the second stage African sleeping sick-
ness. The drug can be applied intravenously in case of emergency
as well as orally during long-term treatment with overall improved
pharmacokinetic properties because of its excellent water solubil-
ity. Furthermore, the active drug pentamidine (8) was detected in
the brain, although only in small amounts. Passing the blood–brain
barrier is essential for curing the second stage of the African
sleeping sickness. Consequently, further studies are necessary to
investigate efficacy of the pentamidine prodrug 7. In this respect,
the development of pentamidine prodrugs based on other amino
acids might be reasonable possibly resulting in prodrugs being
better capable of crossing the blood–brain barrier.
34. Mantyla, A.; Rautio, J.; Nevalainen, T.; Keski-Rahkonen, P.; Vepsalainen, J.;
Jarvinen, T. Eur. J. Pharm. Sci. 2004, 23, 151–158.
35. Acosta, E. P.; Fletcher, C. V. Ann. Pharmacother. 1997, 31, 185–191.
36. Han, H.; de Vrueh, R. L.; Rhie, J. K.; Covitz, K. M.; Smith, P. L.; Lee, C. P.; Oh, D.
M.; Sadee, W.; Amidon, G. L. Pharm. Res. 1998, 15, 1154–1159.
37. Krueger, P. Ber. Dtsch. Chem. Ges. 1885, 18, 1053–1060.
38. Kotthaus, J.; Wahl, B.; Havemeyer, H.; Kotthaus, J.; Schade, D.; Schönberg, D. G.;
Mendel, R.; Bittner, F.; Clement, B. Biochem. J 2010, 433, 383–391.
39. Fröhlich, A. K.; Girreser, U.; Clement, B. Drug Metab. Dispos. 2005, 33, 1532–
1537.
40. Wingen, F.; Brägas, B. Arzneim.-Forsch. 1991, 41, 937–945.
41. Kotthaus, J.; Steinmetzer, T.; van de Locht, A.; Clement, B. J. Enzyme Inhib. Med.
Chem 2011, 26, 115–122.
42. Clement, B.; Christiansen, K.; Girreser, U. Chem. Res. Toxicol. 2001, 14, 319–326.
43. Liu, L.; Ling, Y.; Havel, C.; Bashnick, L.; Young, W.; Rai, R.; Vijaykumar, D.; Riggs,
J. R.; Ton, T.; Shaghafi, M.; Graupe, D.; Mordenti, J.; Sukbuntherng, J. 13th North
America ISSX Meeting; Maui, HI, 2005.
44. Chen, Y.; Dalwadi, G.; Benson, H. A. Curr. Drug Delivery 2004, 1, 361–
376.
Acknowledgements
We would like to thank Melissa Zietz and Sven Wichmann for
technical assistance and Dr. Ulrich Girreser for his advice and assis-
tance with the LC/MS analysis.