New Prodrugs of the Antiprotozoal Drug Pentamidine

Article abstract of DOI:10.1002/cmdc.201100422

Pentamidine is an effective antimicrobial agent that is approved for the treatment of African trypanosomiasis but suffers from poor oral bioavailability and central nervous system (CNS) penetration. This work deals with the development and systematic char

Full text of DOI:10.1002/cmdc.201100422

DOI: 10.1002/cmdc.201100422
New Prodrugs of the Antiprotozoal Drug Pentamidine
Joscha Kotthaus,[a] Jꢀrke Kotthaus,[a] Dennis Schade,[a] Ulrike Schwering,[a]
Helen Hungeling,[a] Helge Mꢀller-Fielitz,[b] Walter Raasch,[b] and Bernd Clement*[a]
Pentamidine is an effective antimicrobial agent that is ap-
proved for the treatment of African trypanosomiasis but suffers
from poor oral bioavailability and central nervous system (CNS)
penetration. This work deals with the development and sys-
tematic characterisation of new prodrugs of pentamidine. For
this reason, numerous prodrugs that use different prodrug
principles were synthesised and examined in vitro and in vivo.
Another objective of the study was the determination of per-
meability of the different pentamidine prodrugs. While some
of the prodrug principles applied in this study are known, such
as the conversion of the amidine functions into amidoximes or
the O-alkylation of amidoximes with a carboxymethyl residue,
others were developed more recently and are described here
for the first time. These newly developed methods aim to in-
crease the affinity of the prodrug for the transporters and me-
diate an active uptake via carrier systems by conjugation of
amidoximes with compounds that improve the overall solubili-
ty of the prodrug. The different principles chosen resulted in
several pentamidine prodrugs with various advantages. The
objective of this investigation was the systematic characterisa-
tion and evaluation of eight pentamidine prodrugs in order to
identify the most appropriate strategy to improve the proper-
ties of the parent drug. For this reason, all prodrugs were ex-
amined with respect to their solubility, stability, enzymatic acti-
vation, distribution, CNS delivery, and oral bioavailability. The
results of this work have allowed reliable conclusions to be
drawn regarding the best prodrug principle for the antiproto-
zoal drug pentamidine.
Introduction
Pentamidine (1) is an antimicrobial drug used to treat trypano-
somiasis, leishmaniasis, and Pneumocystis carinii pneumonia
(PcP). Due to two strongly basic amidine moieties, pentami-
dine suffers from poor oral bioavailability and thus has to be
given by intramuscular (i.m.) or intravenous (i.v.) administra-
tion. Unfortunately, most of the infections mentioned above
occur in tropical or subtropical countries that usually have
poor medical care systems. Consequently, the need for i.m. or
i.v. administration limits the clinical use of pentamidine (1) in
most regions and shows the need for a pentamidine derivative
that can be administered orally. Furthermore, pentamidine (1)
lacks sufficient central nervous system (CNS) penetration and
so is only effective in the treatment of early stages of African
trypanosomiasis, and not in the meningoencephalitic state,
where the pathogens have infiltrated the CNS. Consequently,
several attempts have been made to improve both the oral
bioavailability and CNS penetration of pentamidine (1). Some
research efforts involved altering the structure of pentamidine
(1) by modifying different residues or by cyclising the amidine
function.[1–3] The most successful derivatives that arose from
these efforts were DB75 (furamidine) and its prodrug DB285
(pafuramidine), which showed good efficacy both in vitro and
in vivo.[4] However, clinical development of DB289 was ceased
in 2008 due to unexplained idiosyncratic drug-induced organ
toxicity.
amidine moieties into less basic amidoximes resulting in more
lipophilic molecules. Since amidoximes are uncharged at phys-
iological pH, the gastrointestinal absorption rates of these
pharmacologically inactive amidoximes are significantly im-
proved.[5] The extensive reduction of amidoximes into their
pharmacologically active amidines was first demonstrated in
1988 for the model compound benzamidoxime and was later
shown in vivo for pentamidine–monoamidoxime 2 and pen-
tamidine–diamidoxime 3.[6,7] Recently, the enzyme system re-
sponsible for this reduction was identified as a previously un-
known molybdenum-containing system, which was named mi-
tochondrial amidoxime reducing component (mARC).[8–10] How-
ever, studies in rats revealed that both pentamidine–monoami-
doxime 2 and -diamidoxime 3 have low oral bioavailability.[6]
In order to optimise the pharmacokinetic properties and
CNS delivery of these compounds, further modifications were
made that resulted in N,N’-bis(acetoxy)pentamidine 4, which
has greatly improved lipophilicity compared with the amidox-
ime prodrugs 2 and 3, and showed oral bioavailability in stud-
ies with rats and pigs. Unfortunately, this drug has very limited
solubility, the bioavailability detected was very low, and no
[a] Dr. J. Kotthaus, Dr. J. Kotthaus, Dr. D. Schade, Dr. U. Schwering,
Dr. H. Hungeling, Prof. Dr. B. Clement
Pharmaceutical Institute, Department of Pharmaceutical and Medicinal
Chemistry, Christian-Albrechts-University of Kiel
Gutenbergstraße 76-78, 24118 Kiel (Germany)
Further approaches have dealt with the development of
pentamidine prodrugs. In order to overcome both poor oral
bioavailability and insufficient CNS penetration, we previously
synthesised several prodrugs that rely on different prodrug
principles.[5–7] Earlier studies dealt with the conversion of the
E-mail: bclement@pharmazie.uni-kiel.de
[b] Dr. H. Mꢀller-Fielitz, Prof. Dr. W. Raasch
Institute of Experimental and Clinical Pharmacology and Toxicology
University of Lꢀbeck, Ratzeburger Allee 160, 23538 Lꢀbeck (Germany)
ChemMedChem 2011, 6, 2233 – 2242
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2233

B. Clement et al.
MED
CNS penetration was demonstrated in pigs.[11] Further attempts
dealing with the development of prodrugs with improved sol-
ubility led to the identification of N,N’-bis(valoxy)pentamidine
9. This prodrug was designed in analogy to the valin conjugate
valacyclovir that is absorbed by peptide transporter 1
(PepT1).[12] However, prodrug 9 showed adequate oral bioavail-
ability in rats, but only small amounts were detected in the
brain.[13]
porters are known to possess broad substrate specificity and
to be involved in the uptake of several drugs, such as metho-
trexate and pravastatin.[17,18]
Within this work, we synthesised eight prodrugs for system-
atic characterisation and evaluation with respect to their solu-
bility, stability, bioactivation, permeability, and bioavailability to
draw a reliable conclusion on the most appropriate prodrug of
pentamidine.
Taken together, none of these previously described prodrugs
fulfill all or at least most of the favoured criteria, such as high
oral bioavailability, CNS penetration, and good solubility. Con-
sequently, this study aimed to develop pentamidine prodrugs
with significantly improved solubility, oral absorption, blood–
brain barrier permeability, and presystemic metabolism. Thus,
we successfully designed and examined four new pentamidine
prodrugs, which are based on different prodrug principles.
Some prodrug principles are described within this work for the
first time; others were originally established for the model
compound benzamidoxime and have now been applied to
pentamidine (1).[14,15]
There are several rationales for choosing the different pro-
drug principles applied here. Conjugation with polar moieties
resulted in significantly improved solubility (prodrugs 5 and 8),
which is an essential advantage compared with previously de-
scribed pentamidine prodrugs. Prodrugs 6 and 7 show elevat-
ed metabolic stability due to the more complex mode of acti-
vation requiring O-dealkylation and/or multiple reduction
steps. This mode of activation contributes to a reduced presys-
temic activation and as a conse-
Results and Discussion
Synthesis of pentamidine prodrugs
In this work, we successfully synthesised eight prodrugs of the
antiprotozoal drug pentamidine that rely on different prodrug
principles. N,N’-Bis(methoxy)pentamidine 6 was obtained by al-
kylation of pentamidine–diamidoxime 3 with dimethylsulfate
as described in the Experimental Section. Formation of N,N’-
bis(dihydroxy)pentamidine 7 was achieved by addition of 4,4’-
pentamethylendioxy-bis-(N-hydroxybenzencarboximidoylchlor-
ide) to hydroxylamine and subsequent stirring overnight ac-
cording to the conditions described in the Experimental Sec-
tion. N,N’-Bis(succinyloxy)pentamidine 8 was prepared by treat-
ing pentamidine–diamidoxime 3 with succinic anhydride and
subsequent recrystallisation from toluene. All other pentami-
dine prodrugs were synthesised according to the litera-
ture.[11,13,14,19] The chemical structures are summarised in
Table 1.
quence might increase oral bio-
availability and CNS penetration.
Table 1. Structures of all pentamidine derivatives examined.
However, both solubility and
presystemic activation are
known to limit the oral bioavaila-
bility of the pentamidine pro-
drugs developed previously.[11]
Compd[a]
R1,R2
R3
R4
Moreover, the activation of pro-
drug 5 depends on peptidylgly-
cine a-amidating monooxyge-
Pentamidine (1)
Pentamidine–monoamidoxime (2)
Pentamidine–diamidoxime (3)
ÀH
ÀH
ÀH
ÀH
ÀH
ÀH
ÀOH
ÀOH
ÀOH
nase (PAM), an enzyme with
highest activities in the CNS.[14]
N,N’-Bis(acetoxy)pentamidine (4)
ÀH
Thus, this mode of activation
N,N’-Bis(carboxymethoxy)pentamidine (5)
ÀH
could contribute to brain target-
ing.[16] Prodrug 8 is assumed to
be a substrate for transporters
N,N’-Bis(methoxy)pentamidine (6)
N,N’-Bis(dihydroxy)pentamidine (7)
ÀH
ÀOCH3
ÀOH
ÀOCH3
ÀOH
ÀOH
N,N’-Bis(succinyloxy)pentamidine (8)
ÀH
ÀH
resulting in an increased oral
bioavailability by active uptake
mechanisms. Despite many ef-
N,N’-Bis(valoxy)pentamidine (9)
forts to understand the exact
substrate specificity of human
transport systems, the knowl-
edge is still very restricted and
Evaluation of their drug-likeness in silico
does not allow reliable predictions as to whether a newly dis-
covered compound is a substrate or not. Thus, on the basis of
knowledge we currently have, we speculate that prodrug 8
might be a substrate of human organic anion transporters
(OATs) or monocarboxylate transporters (MCTs). Those trans-
The drug-likeness of all prodrugs was evaluated according to
Lipinski’s rule of five to get an initial idea of their potential oral
bioavailability via passive absorption from the intestine.[20] All
essential parameters are summarised in Table 2, and the results
2234
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2233 – 2242

Pentamidine Prodrugs
Table 2. Evaluation of all compounds according to Lipinski’s rule of five.[a]
Table 3. Solubility of pentamidine derivatives 1–9 at different pH levels
and their protein binding.[a]
Compd
MW [gmolÀ1
]
n [O/N]
n [OH/NH]
clogP[b]
Violations
Compd
Solubility [mm]
Protein
binding [%]
1
2
3
4
5
6
7
8
9
340.6
356.4
372.4
456.5
488.5
400.5
404.4
572.6
570.7
6
7
8
10
12
8
10
14
12
6
6
6
4
5
4
6
6
8
2.31
2.54
2.78
3.87
3.10
4.29
1.92
3.22
4.26
1
1
1
0
1
0
1
3
3
pH 2.0
>35000
pH 7.4
pH 9.0
>35000
1
2
3
4
5
6
7
8
9
>35000
1370Æ291
12Æ1
54.6Æ5.3
74.4Æ2.6
92.8Æ1.6
n.d.
22285Æ1244
4211Æ231
14Æ8
1257Æ40
4Æ1
2Æ1
3Æ2
924Æ152
1304Æ28
11626Æ71
8Æ1
10530Æ1572 98.2Æ1.2
10Æ2
21Æ3
n.d.
>35000
95Æ8
96.5Æ2.1
97.1Æ1.2
90.6Æ4.5
unstable
>35000
7500Æ340
157Æ19
10780Æ70
84Æ18
[a] The rule of five predicts good oral bioavailability if not more than one
of the following criteria is violated: not more than 10 hydrogen-bond ac-
ceptors; not more than 5 hydrogen-bond donors; a molecular weight not
greater than 500 Da; a clogP value not greater than 5.[20] Violations of the
rule of five are illustrated in bold style. [b] Lipophilicity (clogP) was calcu-
lated using the software ChemBioDraw Ultra (version 11).
[a] Solubility and protein binding were determined as described in the Ex-
perimental Section. n.d.=not determined due to solubility issues.
prodrugs show significant advantages or disadvantages re-
garding protein binding.
show that, with the exception of N,N’-bis(succinyloxy)pentami-
dine 8 and N,N’-bis(valoxy)pentamidine 9, all prodrugs are as-
sumed to possess adequate oral bioavailability according to
the rule of five (i.e., one or less violations). In contrast to the
other prodrugs examined, prodrugs 8 and 9 are designed to
target transporters and as such their availability would not be
dependent on passive uptake. Consequently, the negative esti-
mations for these prodrugs predicted by the rule of five (three
violations) can be disregarded. These results prove the necessi-
ty to further investigate all prodrugs to identify the most ap-
propriate application for pentamidine (1).
Stability
Incubations in phosphate buffer at pH 2.0, 9.0 and 7.4 simulate
conditions in the stomach, intestine and blood circulation, re-
spectively. They were performed to provide evidence of possi-
ble chemical instability that might limit the development of
these prodrugs as successful drug. Particular interest was set
on the behaviour of the prodrugs at pH 2.0 and 9.0 to assess
the possible extent of hydrolysis or inactivation in the stomach
and intestine. Our experiments demonstrated that most pro-
drugs are highly stable in the different media tested. However,
O-acylated derivatives 4, 8 and 9 were hydrolysed—as expect-
ed—at basic and acidic pH (Figure 1). In particular, the hydroly-
sis of prodrug 8 in acidic media is very pronounced. Thus, a
gastro-resistant formulation would be required to prevent pro-
drugs 4 and 8 from undergoing extensive hydrolysis prior to
absorption. Interestingly, valine derivative 9 is rather stable at
pH 2.0 and can be administered orally without the need for a
gastro-resistant formulation. Similar findings were obtained
from incubations in murine and human plasma, in which rapid
hydrolysis of prodrugs 4 and 9 was observed. Interestingly, in
comparison to these derivatives, prodrug 8 is rather slowly ac-
tivated by plasma enzymes. Moreover, there were no further
metabolites detectable except those resulting from ester hy-
drolysis. Prodrug 6 showed a slight degradation at pH 7.4 and
9.0. All other prodrugs showed no degradation (data not
shown). Hydrolysis of O-acylated prodrugs 4, 8, and 9 in
plasma is desired because ester cleavage is required for activa-
tion. Further incubations performed with prodrugs 4, 8, and 9
and unspecific carboxylic esterase proved good cleavage of
the ester and thus activation of these prodrugs (data not
shown).
Evaluation of the drug-likeness in vitro
Solubility and protein binding
Solubility and protein binding are important properties that
have to be considered when developing new drugs. Both can
limit the clinical use by causing difficulties concerning bioavail-
ability or drug–drug interactions. In general, good solubility is
considered to have a positive influence on oral bioavailability.
Thus, we investigated the solubility at three essential pH levels
simulating conditions in the stomach (pH 2.0), intestine
(pH 9.0), and blood circulation (pH 7.4). Solubility data are sum-
marised in Table 3 and demonstrate that most prodrugs have
good or fair solubility, with the exception of prodrug 4 at
pH 2.0. At pH 7.4 and 9.0, only prodrugs 2, 5 and 8 show good
solubility, whereas prodrugs 3, 4 and 6 are nearly insoluble.
The protein binding of all prodrugs was determined by ul-
trafiltration. As can be seen in Table 3, all prodrugs possess
protein bindings of at least 74%. In general, high protein bind-
ing of more than 90% is considered to be critical in regard to
drug–drug interactions.[21] Most prodrugs examined have pro-
tein binding levels within this critical range and thus, might
cause drug–drug interactions, when co-administered with
other drugs that possess high protein binding. However, rapid
activation of the prodrug into the active form considerably
lowers the risk of those side effects. In summary, none of the
Activation
The pentamidine prodrugs described in this work are based on
diverse prodrug principles that require different enzyme sys-
ChemMedChem 2011, 6, 2233 – 2242
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2235

B. Clement et al.
MED
~
!
&
^
.
Figure 1. Stability of prodrugs 4, 6, 8, and 9 in diverse media. pH 2.0: ; pH 7.4: ; pH 9.0: ; human plasma: N; rat plasma:
tems, such as mARC, cytochrome P450s (CYPs) and PAM for ac-
tivation (Scheme 1).[8,14,22,23] For that reason, the incubation
mixtures varied in their compositions (see the Experimental
Section). Thus, a direct comparison of activation rates cannot
be made. However, we investigated the activation of each pro-
drug by human and porcine enzyme preparations in an opti-
mised assay. Porcine enzyme preparations were used because
of their suitability for the prediction of physiological conditions
in human.[24,25] The results of our experiments demonstrate
considerable rates of conversion from the prodrug into the
active drug pentamidine (1) for all prodrugs examined, proving
the suitability of the principles used for the development of
our prodrugs (Table 4). Incubations with human enzyme prepa-
rations resulted in slightly lower conversion rates compared
with those obtained with porcine enzyme sources (data not
shown). Human tissues, from which microsomes and mito-
chondria were extracted, were obtained from patients suffering
from cancer or hepatitis, possibly resulting in lower or modi-
fied enzyme levels when compared with healthy individuals;
this might be the reason for lower turnover rates in these ex-
periments. Due to difficulties simulating the complex activation
of prodrug 5 in a single in vitro assay containing PAM and re-
ductive enzyme preparations, we examined the activation of 5
by conversion into 3, which is intensively investigated and
known to be excellently activated.
Prodrugs 6 and 7 were designed to increase the metabolic
stability of pentamidine derivatives and consequently reduce
their presystemic activation. As can be seen in Table 4, both
prodrugs possess relatively low turnover rates and thus dem-
onstrate improved metabolic stability compared with other
prodrugs examined.
The different principles result in several advantages and dis-
advantages when considering drug–drug-interactions and
tissue distribution of activating enzymes. For example, N,N’-bis-
(methoxy)pentamidine 6 is activated by oxidative dealkylation
by CYPs to give pentamidine–diamidoxime 3, and therefore it
bears the risk of mediating drug–drug interactions. Similar ob-
servations were made for the pentamidine analogue DB289
(pafuramidine), which relies on the same prodrug principle
and is O-demethylated by CYP1A2, CYP3A4, and CYP4F.[23,26] All
other pentamidine derivatives are activated independently of
CYP enzymes by esterases, PAM or the mARC enzyme
system.[5,8,9,11,14] Interactions mediated by these enzymes have
not yet been described in the literature. PAM shows the high-
2236
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2233 – 2242

Pentamidine Prodrugs
Scheme 1. Activation of pentamidine prodrugs. Compound names: pentamidine (1); pentamidine–monoamidoxime (2); pentamidine–diamidoxime (3); N,N’-
bis(acetoxy)pentamidine (4); N,N’-bis(carboxymethoxy)pentamidine (5); N,N’-bis(methoxy)pentamidine (6); N,N’-bis(dihydroxy)pentamidine (7); N,N’-bis(succi-
nyloxy)pentamidine (8); N,N’-bis(valoxy)pentamidine (9). Abbreviations: mitochondrial amidoxime reducing component (mARC); peptidylglycine a-amidating
monooxygenase (PAM).
Table 4. Prodrug activation by different enzyme preparations.[a]
studies are often carried out before performing in vivo experi-
ments. Thus, we evaluated our pentamidine prodrugs and
Conversion rates
compared their permeability with that of pentamidine (1). Pen-
tamidine (1) and N,N’-bis(carboxymethoxy)pentamidine 5 show
only low permeability. For prodrug 5, this might be caused by
high efflux rates at the apical membrane due to active trans-
port mechanisms, since we observed increased transport rates
from basolateral to apical media for 5 in comparison to the
rates observed for apical to basolateral. All other prodrugs
showed permeability values greater than the parent drug pen-
tamidine (1) by factor two to three (Figure 2).
In summary, we were able to show increased permeability
for our prodrugs in comparison to the active drug pentamidine
(1). The highest permeability increase was observed for N,N’-
bis(succinyloxy)pentamidine 8, which could be seen as the
most suitable prodrug of pentamidine so far (Figure 2).
Prodrug
[nmol pentamidine (1) minÀ1 mgÀ1 protein]
PL Ms
PL Mt
PK Ms
PK Mt
2
3
4
6
7
8
9
6.67Æ0.48
0.97Æ0.07
1.16Æ0.01
0.12Æ0.07
0.14Æ0.01
0.79Æ0.24
0.47Æ0.19
13.57Æ1.81
5.08Æ0.17
0.54Æ0.10
0.20Æ0.11
0.25Æ0.01
1.63Æ0.03
0.97Æ0.08
20.24Æ3.04
6.43Æ0.40
2.05Æ0.13
0.15Æ0.05
0.34Æ0.06
1.51Æ0.46
0.65Æ0.22
22.86Æ4.92
4.78Æ0.16
0.10Æ0.02
0.24Æ0.14
0.27Æ0.02
0.41Æ0.09
0.81Æ0.05
[nmol pentamidine–diamidoxime (3) minÀ1 mgÀ1 protein]
PAM
5
16.63Æ0.87
[a] Pig liver microsomes (PL Ms); pig liver mitochondria (PL Mt); pig
kidney microsomes (PK Ms); pig kidney mitochondria (PK Mt); peptidyl-
glycine a-amidating monooxygenase (PAM).
Evaluation of the drug-likeness in vivo
est expression level in the CNS and is the enzyme responsible
for the activation of prodrug 5.[14,16] This selective activation
pathway is assumed to contribute to a brain targeting.
In summary, all pentamidine derivatives except N,N’-bis-
(methoxy)pentamidine 6 appear to be appropriate candidates
for use as prodrugs of pentamidine.
Initial bioavailability studies with 3, 4, 5, 6, 7, 8 and 9 in 2 rats
showed that N,N’-bis(dihydroxy)pentamidine 7 and N,N’-bis(-
succinyloxy)pentamidine 8 are the most suitable prodrug can-
didates for pentamidine (data not shown). Thus, we focused
on these derivatives in an enlarged study with 33 rats. For de-
tails see the Experimental Section.
Permeability and transport studies
Oral bioavailability and distribution
Determination of permeability is important for the prediction
of oral bioavailability of potential drug candidates. Permeability
Analysis of plasma samples obtained after i.v. administration of
pentamidine (1) resulted in plasma profiles with a rapid initial
ChemMedChem 2011, 6, 2233 – 2242
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2237

B. Clement et al.
MED
concentrations of 1 in the brain were quite low, ranging from
62Æ34 nggÀ1 (prodrug 7) to 25Æ14 nggÀ1 (prodrug 8). Con-
trol studies with 1 given orally were performed, and the results
confirmed the poor oral bioavailability of the unmodified drug
as detectable amounts of 1 were only found in the liver and
kidney. Table 5 shows the calculated relative bioavailability of
all compounds tested.
Table 5. Relative bioavailability of both pentamidine prodrugs examined.
Bioavailability[a] [%]
Tissue
Compd 1
10 mgkgÀ1
(i.v.)
Compd 1
50 mgkgÀ1
(p.o.)
Prodrug 7
50 mgkgÀ1
(p.o.)
Prodrug 8
50 mgkgÀ1
(p.o.)
Liver
Kidney
Lung
Spleen
Heart
Brain
100
100
100
100
100
100
4.5Æ011
1.1Æ0.9
n.d.
n.d.
n.d.
71.7Æ60.1
3.8Æ2.2
1.4Æ1.2
3.1Æ2.2
5.0Æ2.3
1.2Æ2.5
97.8Æ73.7
6.2Æ2.8
4.9Æ2.7
1.0Æ1.6
3.5Æ1.3
5.3Æ4.4
Figure 2. Caco-2 permeability of all compounds examined.
distribution phase and a long terminal half-life of approximate-
ly 300 min, which is in accordance with literature data.[27] Un-
fortunately, analysis of plasma samples after oral administration
of either pentamidine prodrugs 7 or 8 at a dose of 50 mgkgÀ1
revealed no detectable concentrations of pentamidine (1).
Within this study, prodrug 8 was administered in 100 mm
phosphate buffer (pH 9.0) to limit the hydrolysis of 8 prior to
absorption in the acidic environment of the stomach.
n.d.
[a] The relative bioavailability was calculated as described in the Experi-
mental Section. n.d.=not detected (<0.2%).
While we were able to demonstrate oral bioavailability, due
to our study design, we have not examined efficacy so far.
Thus, to gain more data, we compared the IC50 values of 1
against trypanosomes (0.8–3.2 nm), leishmania (820–2590 nm)
and plasmodia (35–129 nm) with the concentrations of 1 de-
tected in the tissues examined.[29–32] To compare our data
(nggÀ1) with the IC50 values from the literature (nm), we con-
verted tissue concentrations into micromolar (mm) values using
a hypothetical water content in the tissue of 70%. In doing so,
we obtained tissue concentrations in the micromolar range
except in the brain. Thus, oral application of both prodrugs re-
sulted in efficacious concentra-
Pentamidine (1) is known to accumulate in tissues—pre-
dominantly in the liver and kidney.[28] For this reason, we har-
vested six organs (liver, kidney, spleen, lung, heart, and brain)
and analysed them to determine the tissue content of 1. We
were able to detect considerable concentrations of 1 in every
tissue indicating an adequate oral bioavailability of both 7 and
8. As expected, the highest concentrations were found in the
liver and kidney, whereas concentrations in spleen, lung, heart,
and brain were significantly lower (Figure 3). Unfortunately, the
tions of 1 in all tissues.
In summary, we were able to
demonstrate that the prodrug
principles used are suitable to
increase the oral bioavailability
of pentamidine (1). Both N,N’-
bis(dihydroxy)pentamidine
7
and N,N’-bis(succinyloxy)penta-
midine 8 are absorbed and acti-
vated, yielding efficacious con-
centrations of pentamidine (1).
Conclusions
Based on the results obtained
from the in silico, in vitro, and
in vivo experiments, we were
able to evaluate eight pentami-
dine prodrugs systematically.
These results demonstrate that
all prodrug principles evaluated
Figure 3. Distribution of pentamidine 1 after oral administration of prodrugs 7 (&), 8 (&) and pentamidine 1 (&).
Dosage: prodrugs 7 and 8: 50 mgkgÀ1; pentamidine 1: 10 mgkgÀ1
.
2238
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2233 – 2242

Pentamidine Prodrugs
After addition of dimethyl sulfate (0.75 mL) in dioxane (1 mL) at 0–
58C, the mixture was stirred for 2 h at RT. The compound was ex-
tracted with ethyl acetate (150 mL) and purified by column chro-
matography (SiO2; hexane/ethyl acetate, 4:6) to give compound 6
as a white crystalline powder (96 mg, 30%); mp: 2158C; 1H NMR
([D6]DMSO): d=1.57 (m, 2H, CH2), 1.78 (qn, 4H, 3J=6.7 Hz, CH2),
are suitable for application to pentamidine (1). Some prodrugs
show certain drawbacks that might limit their therapeutic use.
For example, prodrugs 3, 4 and 6 suffer from poor solubility,
whereas prodrug 5 shows only low permeability and did not
show improved CNS delivery in preliminary tests in rats. In ad-
dition, prodrug 6 bears the risk that it may mediate drug–drug
interactions due to its activation by cytochrome P450 en-
zymes.
3
3.72 (s, 6H, CH3), 3.99 (t, 4H, J=6.4 Hz, O-CH2), 5.93 (s, 4H, NH2),
6.92 (m, 4H, AA’BB’, ArH), 7.58 ppm (m, 4H, AA’BB’, Ar-H); 13C NMR
([D6]DMSO): d=22.1 (CH2), 28.3 (CH2), 60.4 (CH3), 67.4 (O-CH2),
113.9 (ArCH), 124.6 (ArC), 127.0 (ArCH), 150.8 (ArC), 159.5 ppm (C=
N); IR (KBr):n˜ =3446, 2936, 1636, 1610, 1518, 1398, 1246, 1050,
836 cmÀ1; MS (ESI) m/z: 401 [M+H]+, 235 [C13H18N2O2 +H]+, 201
The N,N’-bis(succinyloxy)pentamidine 8 was identified as the
most appropriate prodrug. It displays the best characteristics
with respect to solubility, activation, permeability, and conse-
quently oral bioavailability. The instability at acidic pH is not
considered to be a critical factor, because this hydrolysis does
not limit the clinical use when administered in a gastro-resist-
ant formulation. However, further studies concerning efficiency
against trypanosomiasis and leishmaniasis are needed to
obtain more detailed information about the suitability of this
prodrug. In this respect, particular interest is set at the ability
of N,N’-bis(succinyloxy)pentamidine 8 to cross the blood–brain
barrier, a property that is needed to be efficacious against the
second state of African trypanosomiasis. For this reason, effica-
cy studies of 8 are needed and will be performed in the future.
Currently, in vivo imaging studies with radiolabelled pentami-
dine prodrugs and detection by means of single photon emis-
sion computed tomography (SPECT) are currently in progress.
[M+2H]2+
,
185, 177, 136, 119 [C7H7N2]+; Anal. calcd for
C21H28N4O4·1.1H2O (MW=420.3): C 60.01, H 7.24, N 13.33, found: C
60.02, H 6.99, N 13.14.
4,4’-Pentamethylendioxy-bis-(N,N’-dihydroxy)benzencarboximi-
damide (N,N’-bis(dihydroxy)pentamidine) (7): Hydroxylamine
(0.2 g, 6.1 mmol) was dissolved in dry EtOH (15 mL) and treated
dropwise with a solution of 4,4’-pentamethylendioxy-bis-(N-hydrox-
ybenzencarboximidoylchloride) (0.25 g, 0.6 mmol) in dry Et2O
(15 mL). After stirring overnight at RT, the mixture was filtrated,
and the filtrate was concentrated to approximately 5 mL and treat-
ed with petroleum ether (5 mL). The first precipitation was discard-
ed. After storage in a refrigerator overnight, the desired compound
was isolated as a white, crystalline solid (121 mg, 50%); mp:
1298C; 1H NMR ([D6]DMSO): d=1.57 (m, 2H, CH2), 1.79 (qn, 4H,
3
3J=7.3 Hz, CH2), 4.02 (t, 4H, J=6.4 Hz, CH2), 6.91 (m, 4H, AA’BB’,
Ar-H), 7.49 (m, 4H, AA’BB’, Ar-H), 8.09 (s, 2H, OH), 8.32 (s, 2H, OH),
10.16 ppm (s, 2H, NH); 13C NMR ([D6]DMSO): d=22.2 (CH2), 28.3
(CH2), 67.4 (O-CH2), 113.7 (ArCH), 124.0 (ArC), 129.0 (ArCH), 156.6
(ArC), 159.1 (C=N); IR (KBr): n˜ =3324, 2938, 1652, 1520, 1470, 1250,
1176, 986 cmÀ1; MS (ESI) m/z: 405 [M+H]+, 389 [C19H24N4O5 +H]+;
Anal. calcd for C19H24N4O6·0.4H2O (MW=411.63): C 55.44, H 6.07, N
13.61, found: C 55.55, H 6.27, N 13.32.
Experimental Section
Synthesis
Reagents: Pentamidine–monoamidoxime 2, pentamidine–diami-
doxime 3, N,N’-bis(acetoxy)pentamidine 4, N,N’-bis(carboxymethox-
y)pentamidine 5 were synthesised according to previously pub-
lished procedures.[11,14,19] Pentamidine was purchased as the diise-
thionate salt Pentacarinat 300 (Sanofi–Aventis, Frankfurt, Germany).
All other reagents used were obtained commercially in the highest
purity available.
4,4’-Pentamethylendioxy-bis-(N-carboxypropionyloxy)benzami-
dine (N,N’-bis(succinyloxy)pentamidine) (8): Pentamidine–diami-
doxime 3 (1 g, 2.7 mmol) was dissolved in acetone (250 mL) and
succinic anhydride (540 mg, 5.4 mmol) was added. The mixture
was stirred under reflux for 4 h. The solvent was evaporated in va-
cuo and the precipitate was recrystallised from toluene to give the
desired compound as a white crystalline powder (1 g, 68%); mp:
1418C; 1H NMR ([D6]DMSO): d=1.59 (m, 2H, CH2), 1.79 (qn, 4H,
3J=6.7 Hz, CH2), 2.52 (t, 4H, 3J=6.6 Hz, CH2), 2.68 (t, 4H, 3J=
General: Melting points were measured on a Bꢀchi 510 Melting
Point apparatus and are uncorrected. Infrared (IR) spectra were ob-
1
tained on a Perkin–Elmer FTIR 1600 PC spectrophotometer. H and
3
6.6 Hz, CH2), 4.04 (t, 4H, J=6.5 Hz, O-CH2), 6.63 (s, 4H, NH2), 6.99
13C NMR spectra were recorded on a Bruker ARX 300 NMR spec-
(m, 4H, AA’BB’, Ar-H), 7.65 (m, 4H, AA’BB’, Ar-H), 12.18 ppm (br s,
2H, COOH); 13C NMR ([D6]DMSO): d=22.1 (CH2), 27.9 (CH2), 28.3
(CH2), 28.8 (CH2), 67.5 (O-CH2), 113.9 (ArCH), 123.5 (ArC), 128.1
(ArCH), 156.2 (ArC), 160.3 (C-NH2), 170.2 (COOR), 173.5 ppm
(COOH); IR (KBr): n˜ =3478, 3348, 2940, 2870, 1732, 1698, 1612,
1472, 1250 cmÀ1; MS (ESI) m/z: 573 [M+H]+, 555 [MÀH2O+H]+,
1
trometer using the following frequencies: H: 300.13 MHz and 13C:
75.47 MHz. Chemical shifts (d) are reported in ppm relative to tet-
ramethylsilane (TMS) as an internal standard. All coupling constants
(J) were obtained by first-order analysis of the multiplets and are
quoted in Hz. The following NMR abbreviations are used: broad
(br), singlet (s), doublet (d), triplet (t), quintet (qn), unresolved mul-
tiplet (m). Low-resolution electrospray ionisation (ESI) mass spectra
(MS) were recorded on a Bruker Esquire LC mass spectrometer.
Compounds were dissolved in MeCN or MeOH. High-resolution
mass spectrometry (HRMS) was performed on a Bruker FT-ICR
APEX II spectrometer using electrospray ionisation. Here, the com-
pounds were dissolved in MeOH. Elemental analyses were per-
formed on a CHNS analyser (HEKAtech GmbH, Wegberg, Germany)
at the Department of Inorganic Chemistry, University of Kiel, Ger-
many.
473
[MÀC4H4O3 +H]+,
455
[MÀC4H4O3ÀH2O+H]+,
373
[C19H24N4O4 +H]+, Anal. calcd for C27H32N4O10 (MW=572.56): C
56.64, H 5.63, N 9.79, found: C 56.85, H 6.01, N 9.60.
Biological evaluation
Calculation of lipophilicity (clogP): The lipophilicity of all com-
pounds was calculated for their uncharged state using ChemBio-
Draw Ultra (version 11) (CambridgeSoft, Cambridge, MA, USA).
4,4’-Pentamethylendioxy-bis-(N-methoxy)benzamidine (N,N’-bis-
(methoxy)pentamidine) (6): Pentamidine–diamidoxime 2 (0.3 g,
0.8 mmol) was dissolved in dioxane (1 mL) and 2n NaOH (10 mL).
Stability in diverse media: The stability of all prodrugs was tested at
a concentration of 200 mm in 50 mm potassium phosphate buffer
at pH 2.0, pH 7.4, and pH 9.0 at 208C over a time period of
ChemMedChem 2011, 6, 2233 – 2242
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2239

B. Clement et al.
MED
360 min. Every 15 min, a sample (100 mL) was analysed by HPLC to
quantify the concentrations of prodrug and degradation products.
Additional incubations were performed with all prodrugs at a con-
centration of 200 mm in human and murine plasma for 150 min at
378C. Every 15 min, a sample (100 mL) was taken and treated with
MeCN (100 mL). After centrifugation at 10000 g for 10 min, the su-
pernatant was analysed by HPLC and the concentrations of the
prodrug and metabolites were measured. Furthermore, prodrugs
4, 8, and 9 were incubated at a concentration of 100 mm with un-
specific carboxylic esterase (1 U) from pig liver (Sigma–Aldrich,
Taufkirchen, Germany) at 378C in 50 mm potassium phosphate
buffer (pH 7.4) for 60 min. Every 15 min, samples (100 mL) were an-
alysed by HPLC.
The integrity of the monolayers was evaluated by means of transe-
pithelial electrical resistance (TEER) measurements. Only monolay-
ers with a TEER value of at least 250 WꢂcmÀ2 were used for trans-
port experiments. The transport of lucifer yellow (100 mgmLÀ1
)
across the monolayers was measured after the experiment for 1 h.
Only monolayers with lucifer yellow passage ꢀ1% were analysed.
Transport experiments were carried out in apical medium (2-(N-
morpholino)ethanesulfonic acid (MES), 10 mm; Hank’s buffered salt
solution (HBSS, 10ꢂ), 100 mL; twice-distilled H2O, 900 mL) and ba-
solateral medium (4-(2-hydroxyethyl)-1-piperazineethanesulfonic
acid (HEPES), 10 mm; d-(+)-glucose, 25 mm; HBSS (10ꢂ), 100 mL;
twice-distilled H2O, 900 mL L. The monolayers were washed twice
and then preincubated for 15 min. After TEER measurement, the
experiment was started by adding 600 mL of medium containing
the test compound. The inserts were then transferred to a new cell
culture plate containing 1500 mL medium per well. An aliquot of
apical medium (100 mL) was removed immediately to determine
the t0 concentration. After 10, 25, 45, 70, 120 min, the inserts were
moved to cell culture plates containing fresh medium. The
medium was analysed by HPLC. Permeability was calculated by the
following equation: Papp =(dc/dt)/Aꢂc0(donor). Compounds 2, 4, and
7 were not tested in these studies.
Determination of solubility: An insoluble amount (5 mg) of every
prodrug was shaken in 50 mm potassium phosphate buffer
(300 mL) at pH 7.4 or pH 9.0, or 0.01n HCl (pH 2.0) for 60 min. After
centrifugation at 13000 g for 10 min, the supernatant was analysed
by HPLC to quantify the concentration of prodrug dissolved.
Incubations with liver enzyme preparations: Human and porcine mi-
tochondria and microsomes from liver and kidney were incubated
at 378C in a mixture containing prodrug (500 mm), NADH (1 mm)
and protein (0.3 mg) in 50 mm potassium phosphate buffer
(150 mL) at pH 6.3 over a period of 30 min. Moreover, incubations
of prodrugs 4, 8, and 9 contained carboxylic esterase (1 U) from
pig liver (Sigma–Aldrich, Taufkirchen, Germany), and those per-
formed with prodrug 6 contained an addition amount of NADPH
(1 mm). Afterwards, reactions were terminated by addition of
MeCN (150 mL). Samples were shaken for 10 min, proteins were
sedimented by centrifugation (10000 g, 10 min), and supernatants
were analysed by HPLC.
Animals: Male Sprague Dawley (SD) rats (Charles River Laboratories,
Sulzfeld, Germany) were used. The study was conducted according
to the US National Institutes of Health Guidelines for the Care and
Use of Laboratory Animals and authorised by the local regulatory
authority (Ministerium fꢀr Landwirtschaft, Umwelt and lꢃndliche
Rꢃume des Bundeslandes Schleswig–Holstein, Germany). The ani-
mals were kept at room temperature with a 12/12 h dark (8 AM–
8 PM)/light (8 PM–8 AM) cycle. They received a standard diet and
water ad libitum. Rats (398Æ20 g) were habituated to research as-
sistants and vice versa two weeks before drug treatment was initi-
ated. Rats were randomised into groups receiving either pentami-
dine 1 i.v. (n=10), pentamidine 1 p.o. (n=3), prodrug 7 or 8 both
p.o. (n=10).
Incubations with peptidylglycine a-amidating monooxygenase (PAM):
Incubations were performed in analogy to those described previ-
ously by Schade et al.[14]
Determination of protein binding: The protein binding of each pro-
drug was examined at three different concentrations (10 mm,
25 mm, and 50 mm) by ultrafiltration. The prodrugs were solved in
50 mm potassium phosphate buffer (pH 7.4) with 4% albumin, in-
cubated for 15 min and centrifuged using Microcon centrifugal
filter units YM 30 (Millipore, Billerica, MA, USA) at 13000 g for
10 min. Control samples were treated analogously but contained
no albumin. The filtrate was analysed by HPLC. Protein binding
data reported in Table 3 represent the mean value of the determi-
nations at the three different concentrations. Due to poor solubili-
ty, prodrug 3 was examined at a concentration of 10 mm only.
Surgery: Two days before the pharmacokinetic study, chronic poly-
ethylene catheters were inserted into the right femoral vein and
artery while the animals were under pentobarbitone anaesthesia
(67 mgkgÀ1). Catheters were tunnelled under the back skin, exteri-
orised in the region of the cervical vertebra, and fixed at the skin.
Thereafter, rats were housed individually in cages (heightꢂwidthꢂ
length: 20ꢂ22ꢂ25 cm) until the end of the study.
Animal study: The oral bioavailability of prodrugs 7 and 8 was in-
vestigated in a study with 33 rats. Both compounds were suspend-
ed in a solution of gum arabic (10%) dissolved in 100 mm phos-
phate buffer and orally administered by gavage to rats at a dose
of 50 mgkgÀ1 (n=10). Due to the rapid hydrolysis of compound 8,
the solution was adjusted to pH 9.0 to increase the gastric pH and
minimise the fraction of hydrolysed prodrug. Moreover, pentami-
dine 1 (50 mgkgÀ1) was given orally by gavage to a control group
of rats (n=3). Blood samples were taken with microvettes (Sar-
stedt, Nꢀmbrecht, Germany) after 20, 40, 60, 90, 120, 240 and
360 min via an arterial implanted catheter. Pentamidine 1 was
given intravenously to rats at a dose of 10 mgkgÀ1 (n=10). Blood
samples of approximately 300 mL were collected 5, 10, 20, 40, 75,
150, 300 min after intravenous dosing. Plasma samples were ob-
tained by centrifugation at 10000 g for 5 min and were immediate-
ly frozen at À808C. Animals were euthanised by decapitation
360 min after application of the test compounds, and the following
organs were harvested: liver, kidney, lung, spleen, heart, and brain.
The tissues were frozen immediately and stored at À808C.
Cell culture and transport studies: The human colon adenocarcino-
ma cell line, Caco-2 was purchased at passage 40 from the Europe-
an Collection of Cell Cultures (ECACC), Salisbury, UK. Cultured mon-
olayers of Caco-2 cells were used in transport experiments be-
tween passages 45–60. The cells were cultured at 378C in 5% CO2
and maintained in Dulbecco’s modified Eagle’s medium (DMEM)
with GlutaMAXTM (Life Technologies, Darmstadt, Germany), d-glu-
cose (4500 mgLÀ1), foetal calf serum (FCS; 10%), penicillin
(100 UmLÀ1),
streptomycin
(100 mgmLÀ1
)
and
glutamine
(1.5 mgmLÀ1) but without sodium pyruvate. Stock cultures were
grown in 75 cm2 tissue culture flasks and were split 1:6 at 80–90%
confluency using ethylenediaminetetraacetic acid (EDTA; 0.02%)
and trypsin (0.05%). Cells were seeded at 200000 cells per cm2 on
polyethylene terephtalate membrane inserts (1.13 cm2, 2ꢂ
106 pores per cm2, 0.4 mm pore size) and cultured for 21 days. The
medium was changed every two to three days and on the day
before the experiment.
2240
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2233 – 2242

Pentamidine Prodrugs
Preparation of plasma and tissue samples: Plasma samples were
treated with equal volumes of MeCN. Samples were then shaken
for 45 min and centrifuged at 10000 g for 15 min. The supernatant
was analysed by HPLC. Tissue samples were thawed to room tem-
perature and cut into pieces weighing approximately 1000 mg
(liver) or 500 mg (other tissues). The tissue samples were homoge-
nised in water (2 mL) and then treated with MeCN (2 mL) to pre-
cipitate proteins. After shaking for 45 min, the samples were centri-
fuged at 12000 g for 15 min. The supernatants were dried with
compressed air and lyophilised overnight. The residues were resus-
pended in MeOH/water (50:50, 400 mL), shaken for 1.5 h, and cen-
trifuged at 12000 g for 15 min. The supernatant was analysed by
HPLC.
Acknowledgements
We thank Franziska Hohm (Pharmaceutical Institute, University
of Kiel) for excellent technical assistance. The study described
here arose from a larger project within the scope of a patent
value fund. In this context, we acknowledge the financial support
provided by the Dritte Patentportfolio Beteiligungsgesellschaft
GmbH & Co. KG.
Keywords: amidoximes · oral bioavailability · pentamidines ·
prodrugs · trypanosomiasis
Determination of the relative bioavailability: In general, oral bioavail-
ability is calculated from the plasma concentration versus time plot
for a drug after both intravenous and oral administration. Due to
the high protein binding of pentamidine 1 and its tendency to ac-
cumulate in tissues, no plasma concentrations of pentamidine 1
were detectable after oral administration of either prodrug exam-
ined. In order to compare the prodrugs, we used the pentamidine
concentrations detected in the different tissues and defined the
concentration obtained after intravenous application of pentami-
dine as 100%. The relative bioavailability was calculated by means
of the pentamidine concentrations measured after oral treatment
with prodrug and that obtained after intravenous application of
pentamidine.
[1] D. W. Boykin, A. Kumar, G. Xiao, W. D. Wilson, B. C. Bender, D. R. McCur-
dy, J. E. Hall, R. R. Tidwell, J. Med. Chem. 1998, 41, 124–129.
[2] S. K. Jones, J. E. Hall, M. A. Allen, S. D. Morrison, K. A. Ohemeng, V. V.
Reddy, J. D. Geratz, R. R. Tidwell, Antimicrob. Agents Chemother. 1990,
34, 1026–1030.
[3] S. M. Rahmathullah, J. E. Hall, B. C. Bender, D. R. McCurdy, R. R. Tidwell,
D. W. Boykin, J. Med. Chem. 1999, 42, 3994–4000.
[4] T. Wenzler, D. W. Boykin, M. A. Ismail, J. E. Hall, R. R. Tidwell, R. Brun, An-
timicrob. Agents Chemother. 2009, 53, 4185–4192.
[5] B. Clement, Drug Metab. Rev. 2002, 34, 565–579.
[6] B. Clement, M. Immel, R. Terlinden, F. J. Wingen, Arch. Pharm. 1992, 325,
61–62.
[7] B. Clement, S. Schmitt, M. Zimmermann, Arch. Pharm. 1988, 321, 955–
956.
[8] S. Gruenewald, B. Wahl, F. Bittner, H. Hungeling, S. Kanzow, J. Kotthaus,
U. Schwering, R. R. Mendel, B. Clement, J. Med. Chem. 2008, 51, 8173–
8177.
HPLC methods
[9] A. Havemeyer, F. Bittner, S. Wollers, R. Mendel, T. Kunze, B. Clement, J.
Biol. Chem. 2006, 281, 34796–34802.
[10] J. Kotthaus, B. Wahl, A. Havemeyer, D. Schade, D. Garbe-Schonberg, R.
Mendel, F. Bittner, B. Clement, Biochem. J. 2011, 433, 383–391.
[11] B. Clement, A. Burenheide, W. Rieckert, J. Schwarz, ChemMedChem
2006, 1, 1260–1267.
General: Columns were purchased from Merck KGaA, Darmstadt,
Germany. Analyses were carried out on a Waters Alliance system
consisting of
a Waters e2695 XC separations module and a
Waters 2995 photodiode array detector (260 nm) with column
heater.
[12] H. Han, R. L. de Vrueh, J. K. Rhie, K. M. Covitz, P. L. Smith, C. P. Lee, D. M.
Oh, W. Sadee, G. L. Amidon, Pharm. Res. 1998, 15, 1154–1159.
[13] J. Kotthaus, H. Hungeling, C. Reeh, D. Schade, S. Wein, S. Wolffram, B.
Clement, Bioorg. Med. Chem. 2011, 19, 1907–1914.
[14] D. Schade, J. Kotthaus, H. Hungeling, J. Kotthaus, B. Clement, ChemMed-
Chem 2009, 4, 1595–1599.
[15] C. Reeh, J. Wundt, B. Clement, J. Med. Chem. 2007, 50, 6730–6734.
[16] G. S. Wand, R. L. Ney, S. Baylin, B. Eipper, R. E. Mains, Metabolism 1985,
34, 1044–1052.
[17] A. Kalliokoski, M. Niemi, Br. J. Pharmacol. 2009, 158, 693–705.
[18] B. L. Urquhart, R. B. Kim, Eur. J. Clin. Pharmacol. 2009, 65, 1063–1070.
[19] B. Clement, W. Raether, Arzneim. Forsch. 1985, 35, 1009–1014.
[20] C. A. Lipinski, F. Lombardo, B. W. Dominy, P. J. Feeney, Adv. Drug Delivery
Rev. 2001, 46, 3–26.
Stability and solubility: Separations were performed on a LiChro-
Cart LiChrospher 60 RP-select B column (125ꢂ4 mm, 5 mm) with a
mobile phase of H2O+0.1% trifluoroacetic acid (TFA) (pH 2.5) and
MeCN. The flow rate was set to 1 mLminÀ1, the column tempera-
ture was kept at 258C, and the injection volume was 10 mL. The
percentage MeCN varied depending on the particular derivative:
(1) 20% MeCN, tR =3.4Æ0.2 min; (2) 20% MeCN, tR =3.2Æ0.2 min;
(3) 40% MeCN, tR =3.0Æ0.2 min; (4) 40% MeCN, tR =5.1Æ0.2 min;
(5) 25% MeCN, tR =5.8Æ0.2 min; (6) 25% MeCN, tR =3.6Æ0.2 min;
(7) 30% MeCN, tR =3.4Æ0.2 min; (8) 20% MeCN, tR =4.4Æ0.2 min;
(9) 30% MeCN, tR =7.1Æ0.3 min.
Activation: Separations were performed on a LiChroCart LiCHros-
pher 60 Rp-select B column (125ꢂ4 mm, 5 mm) with a mobile
phase of H2O+octylsulfonate (10 mm)+tetramethylammonium
chloride (20 mm) (pH 3.0) and MeOH (52:48). The flow rate was set
to 1 mLminÀ1, the column temperature was maintained at 258C,
and the injection volume was 20 mL. Pentamidine 1 eluted at tR =
10.7Æ0.4 min.
[21] P. E. Rolan, Br. J. Clin. Pharmacol. 1994, 37, 125–128.
[22] M. Z. Wang, J. Q. Wu, A. S. Bridges, D. C. Zeldin, S. Kornbluth, R. R. Tid-
well, J. E. Hall, M. F. Paine, Drug Metab. Dispos. 2007, 35, 2067–2075.
[23] J. Y. Saulter, J. R. Kurian, L. A. Trepanier, R. R. Tidwell, A. S. Bridges, D. W.
Boykin, C. E. Stephens, M. Anbazhagan, J. E. Hall, Drug Metab. Dispos.
2005, 33, 1886–1893.
[24] B. Clement, S. Mau, S. Deters, A. Havemeyer, Drug Metab. Dispos. 2005,
33, 1740–1747.
In vivo study: Pentamidine was detected by RP-HPLC. Separations
were performed on a Superspher 60 RP-select B column (250ꢂ
3 mm, 5 mm) with a mixture of H2O+0.1% TFA (pH 2.5) and MeOH
(60:40). The flow rate was set to 0.32 mLminÀ1, and the injection
volume was 35 mL. Pentamidine 1 eluted at tR =24.7Æ0.5 min.
[25] M. J. Myers, D. E. Farrell, K. D. Howard, J. C. Kawalek, Drug Metab. Dispos.
2001, 29, 908–915.
[26] M. Z. Wang, J. Y. Saulter, E. Usuki, Y. L. Cheung, M. Hall, A. S. Bridges, G.
Loewen, O. T. Parkinson, C. E. Stephens, J. L. Allen, D. C. Zeldin, D. W.
Boykin, R. R. Tidwell, A. Parkinson, M. F. Paine, J. E. Hall, Drug Metab.
Dispos. 2006, 34, 1985–1994.
[27] R. Terlinden, A. Rçmer, Med. Klin. (Muenchen, Ger.) 1990, 85 (Suppl. 2),
245–247.
[28] F. Wingen, B. Brꢃgas, Arzneim. Forsch. 1991, 41, 937–945.
ChemMedChem 2011, 6, 2233 – 2242
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2241

B. Clement et al.
MED
[29] J. J. Brendle, A. Outlaw, A. Kumar, D. W. Boykin, D. A. Patrick, R. R. Tid-
well, K. A. Werbovetz, Antimicrob. Agents Chemother. 2002, 46, 797–
807.
[32] M. A. Ismail, R. Brun, T. Wenzler, F. A. Tanious, W. D. Wilson, D. W. Boykin,
Bioorg. Med. Chem. 2004, 12, 5405–5413.
[30] R. K. Arafa, R. Brun, T. Wenzler, F. A. Tanious, W. D. Wilson, C. E. Stephens,
D. W. Boykin, J. Med. Chem. 2005, 48, 5480–5488.
Received: September 6, 2011
[31] I. O. Donkor, T. L. Huang, B. Tao, D. Rattendi, S. Lane, M. Vargas, B. Gold-
berg, C. Bacchi, J. Med. Chem. 2003, 46, 1041–1048.
Published online on October 7, 2011
2242
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2233 – 2242